Degrees of recourse

James Surowiecki has been pushing the idea (first mooted by John Hempton) that since bank financing always involves non-recourse by taxpayer (via government deposit insurance), it’s no big deal that the Geithner Plan is built around generous non-recourse lending. Surowiecki:

There is one detail of the plan, though, that people are particularly bothered by, and that is the fact that the plan involves the FDIC guaranteeing loans to private investors. (The way the plan to buy pools of mortgages is set up, investors will be able to borrow six dollars for every one dollar they invest. If their bets go bad, they lose only the one dollar they invested—the FDIC is responsible for paying back all the borrowed money.) Paul Krugman, for instance, calls this the “central issue,” and argues that because the non-recourse loans are a massive subsidy to investors—which they are—the plan will distort the prices that investors are willing to pay for these assets, and therefore “has nothing to do with letting markets work.” Ezra Klein, similarly, argues that because the plan relies on these “non-recourse” loans, the prices it will produce will be in some way “artificial.” Their point is that the Geithner plan, among other things, is supposed to produce real market prices for these toxic assets, which will then give us a better picture of banks’ balance sheets and allow us to avoid valuing these assets at prices that the government thinks have become unduly low because investors are so risk-averse. But by creating a plan in which investors have only a small downside and a big upside, we’re supposedly creating fake prices.

There’s no doubt that the non-recourse loans constitute a big subsidy: while investors’ downside risk isn’t eliminated, since they can still lose all the money they invest, that risk is significantly limited, while their potential upside is significantly increased (since they’re leveraging every dollar they invest six-to-one). Yet for all the criticism of this subsidy, the truth is that the plan’s reliance on non-recourse loans is not an especially radical idea. In fact, it’s essentially the same kind of subsidy that the entire U.S. banking system has depended on for the last seventy-five years. What are FDIC-insured bank deposits, after all? They’re non-recourse loans to banks. You deposit money with a bank—that is, you lend it your money. The bank can then take that money, and leverage it up nine-to-one to make loans or acquire assets. If the loans are good, they keep all the profits. If the loans go bad, the most the bank can lose is the capital it’s invested. All the rest of the bank’s losses are paid for by the FDIC. This is precisely the same arrangement — down to the loans being guaranteed by the FDIC—that the Geithner plan sets up. In effect, it just extends to outside investors, for the purpose of acquiring toxic assets, the same subsidy that banks have been receiving since 1933.

Surowiecki is right in a superficial way, but he misses crucial details. His analogy breaks down in ways that I think are informative.

Non-recourse financing involves bundling a valuable put option along with a loan. The value of that option hinges on the details of the arrangement: An “at-the-money” option is more valuable than an “out-of-the-money” option. The “moneyness” of the implicit put option in a non-recourse arrangement is determined by the degree of leverage the borrower is allowed. The Geithner plan permits a maximum leverage of 7:1 assets to equity. As Surowiecki points out, that degree of leverage is less than the leverage of a traditional bank.

But notional “moneyness” is not the only thing that determines the value of an option. In particular, options are most valuable when the assets that underlie them are very volatile. In order to limit the degree to which banks maximize the value of the deposit insurance option at the taxpayers’ expense, banks are supposed to submit to onerous, intrusive regulation that limits the riskiness, the volatility of the assets they can purchase.

Under the Geithner plan, the government will extend its non-recourse option to investors without preventing them from “swinging for the fences” on risk in order to extract value from the option. On the contrary, the government is insisting its loans be used to purchase assets that have already proved themselves unsuitable for purchase by a regulated entity, by virtue of being volatile and difficult to price. It’s as if an insurance company that ordinarily refuses to cover homes in hurricane states suddenly offered policies only to purchasers looking to build homes on Gulf-coast barrier islands.

Sometimes an option is costly to exercise. Exercise costs reduce the value of an option to its owner along with the expected liability of the option writer. The traditional deposit insurance option was very costly for banks to exercise. Banks were only allowed to exercise the option by being put out of business. That sharply limited the value of the FDIC option to bank employees and shareholders. Usually the “franchise value” of a bank is greater than its regulatory capital. In order to extract value from the option, bank stakeholders must take bets whose (probability-weighted) payoff in a good outcome exceeds not only the loss of regulatory capital, but also the value of the business as an ongoing concern. Moreover, if we are valuing the “traditional” FDIC option, we should go back just a few decades to when most banks were privately held and run by lifers. A typical bank’s owners and employees had an illiquid, undiversified exposure to the well-being of their institution. Calculating franchise value from the price/book of Citi pre-crisis would dramatically underestimate the value of a traditional bank to its controlling stakeholders. Exercise of the FDIC option used to be costly indeed for the owners and managers of banks. There were, if you’ll excuse the terms, “synergies” between regulation and a high cost of exercise: If triggering a deposit insurance payout is very painful, strategies designed to monetize the option need a fly-to-the-moon upside to be worth the risk. But spaceports are more likely to be flagged by regulators than an extra 100 bps on a “AAA” CPDO.

Unlike deposit insurance, the non-recourse option offered to investors in the Geithner plan is completely costless to exercise, once it is in the money. Surrender of the collateral constitutes fulfillment of the lending contract full stop. Pace Megan McArdle, it is not like the non-recourse option implicit in the typical home mortgage, where exercising the option involves a hit to ones credit. There is no default of any sort involved in surrendering the assets rather than repaying the loans. The right to do so will be written into the deals.

Finally, the FDIC option didn’t used to be free. Banks paid a fee in exchange for the deposit insurance. Under the Geithner plan, borrowers will be charged a fee as well, but then they’ll be borrowing at rates dramatically lower than they could on their own. This works with Surowiecki’s story — banks borrow very cheaply from depositors because of the FDIC guarantee. But, as always, the question is price: Given the volatility and uncertainty surrounding the underlying assets and the near-zero cost of exercise, will the FDIC charge a fee high enough to cover the expected liability of the option? I’ll leave that for readers to decide.

Surowiecki’s analogy does work pretty well if we compare the Geithner plan not to traditional banking thirty years ago, but to recent practices of the industry. Thanks to hands-off government and structured-finance shell games, banks were recently able to amass high risk, high yield investment portfolios despite being ostensibly regulated institutions. Institutional changes, including changing norms about the length and terms of bank employment and dominance of the industry by large, heavily-traded limited-liability corporations, decreased the expected cost of exercise to bank employees and informed shareholders, making volatility maximizing strategies more attractive. The option premium, the FDIC fee, was severely underpriced (it was reduced to zero from 1996 to 2006).

If Surowiecki wants to argue that the non-recourse option embedded in the Geithner plan would basically reproduce the subsidy to the banking system offered circa 2006, I’ll readily agree. But it is not reasonable to argue that non-recourse loans offered on generous terms to unregulated investors for the express purpose of purchasing unusually volatile assets represent the “same subsidy that banks have been receiving since 1933.”

 
 

116 Responses to “Degrees of recourse”

  1. Nemo writes:

    The key question is also, “with recourse to what?”

    There is a huge difference between making one “non-recourse” loan to an investor to purchase a large number of assets, with the loans collateralized by all of the assets combined… And making 1,000 “non-recourse” loans to an investor to purchase 1,000 distinct risky instruments, each loan collateralized by that single instrument.

    The FDIC loans are made with recourse to all of the assets of the bank combined. Which is similar to (but slightly different from) your observation that the bank can only exercise its “option” by going out of business.

    Can a single person or company even own and control multiple FDIC-insured banks? I somehow had the impression that the regulations forbade that, for the obvious reasons.

    A single non-recourse loan to a single entity is not a problem. It’s when you have multiple non-recourse loans to the same entity that you get into trouble, since the entity can make multiple bets with the virtual certainty that a few of them will pay off.

    Thus Surowiecki’s argument is nonsense.

  2. John Hempton writes:

    Thanks for the reference. It is about the numbers and the structure.

    If the Geithner funds are 6.5 times levered holding a diverse range of assets then I am fairly comfortable. Not entirely comfortable.

    If they are 6.5 times levered – run by people with massive conflicts of interest and able to buy undiversified assets I am terrified.

    I really can’t tell from the documentation – though conflict of interest appears to be the norm.

    J

  3. James Surowiecki writes:

    “It’s when you have multiple non-recourse loans to the same entity that you get into trouble, since the entity can make multiple bets with the virtual certainty that a few of them will pay off.”

    I understand this is the same argument you made in your original post, but there’s no substance to it without real numbers to back it up. For the multiple loans to create significant problems, the investors need to be certain not just that some bets will hit, but that the pay off from the bets that hit will outweigh all of the misses. Yes, if you assume, as you did, that the possible values of the assets in a pool are 0 or 100, the option is enormously valuable, and the likely loss to the government is enormously high, because the investor enjoys the huge upside, and the government pays for the huge downside. When the range of values is much narrower — which is far more realistic — then the supposedly nefarious effects of the multiple loans significantly diminish in importance, and the risk of investors massively overpaying is far smaller. It’s much more likely that there will be many more situations in which investors in these partnerships lose all their money, and the FDIC almost none at all (since the value of the asset will be only somewhat below what the partnership pays for it), than situations in which the investors make a profit and the FDIC pays for everything.

  4. James Surowiecki writes:

    “In order to limit the degree to which banks maximize the value of the deposit insurance option at the taxpayers’ expense, banks are supposed to submit to onerous, intrusive regulation that limits the riskiness, the volatility of the assets they can purchase.

    “Under the Geithner plan, the government will extend its non-recourse option to investors without preventing them from “swinging for the fences” on risk in order to extract value from the option. On the contrary, the government is insisting its loans be used to purchase assets that have already proved themselves unsuitable for purchase by a regulated entity, by virtue of being volatile and difficult to price.”

    Really? The plan requires the FDIC to vet all the assets that are going to be offered for sale. It will determine the amount of leverage that’s offered to the partnerships, based on the riskiness of the assets for sale, therefore precisely preventing them from swinging for the fences on very risky assets. And I don’t know how these assets can be considered unsuitable for purchase by a regulated entity, when regulated entities are the ones that made the loans and/or bought the assets in the first place, and are not prohibited from buying them now. (The New York Post today has a story on how Citigroup and BoA are buying up distressed mortgages from other banks.)

    I’m also not sure how we know that these assets are so volatile. Just because the market for them is illiquid doesn’t mean that their underlying value is subject to wide swings. Many (most?) of these assets will be collateralized by property. Unless you think property values are going to oscillate wildly in the next twenty years, I’m not sure the underlying value of the asset (even if it’s a mortgage) would necessarily be volatile. But I’m probably missing something important on this point.

  5. James — I claim they are volatile and/or of uncertain valuation because they have proven themselves to be volatile historically, and their illiquidity suggests (though does not prove) continuing disagreement about their current values. It is possible that the period of uncertainty is over, and now they have clear values but spooked investment for “animal spirit” reasons won’t touch them without sugar. But that strikes me as unlikely. How frequently have we seen assets go from par to bid 30 / ask 80, and then settle nicely and permanently at 78? Again, it could happen. But ordinarily the FDIC wouldn’t let banks lever up on an assets so recently volatile and with such divergent prices until after that volatility had settled for a while. We are very far from AAA land.

    You are right that, in theory, the FDIC could withhold leverage except on the same terms with would extend with respect to conservative bank assets in ordinary times. The FDIC does have the power to say no. But from the FDIC’s perspective, it’s probably on the hook either way, but it faces less trouble if it has to make whole bank creditors than a public-private partnership participant. Honestly, just from a political economy perspective, do you really think the FDIC is going to behave like a stern bank examiner wrt its participation in this program?

    That Citi and BoA are buying up distressed mortgages, and that they are permitted to do so is a scandal, from my perspective. What is likely going on is exactly what many of us predicted with the original TARP. These banks now have a uniquely deep-pocketed buyer, and they can earn a spread purchasing from less privileged banks and selling into the publicly supported program. That will help recapitalize the banks, but only at cost of increasing the risk that might need to be offloaded onto public balance sheets. In ordinary times and with ordinary banks, FDIC would have negotiated a formal action with these banks that, among other things, would certainly prevent them from purchasing illiquid/high disagreement assets.

    The irony here is that if FDIC behaved according to the laws that govern it and the norms that usually apply, it could not participate in this program without effectively blocking it. You can argue (and I think you do) that extraordinary times and its systemic risk exception (which has not been explicitly triggered btw) mean that it should take on assets whose value it otherwise would not insure at 7x leverage to help recapitalize the banks. But you can’t then argue that the current regime is equivalent to run-of-the-mill banking: deposit insurance plus extraordinary regulatory forbearance implies a dramatically different kind of subsidy than deposit insurance under cautious regulation and a commitment prompt corrective action that in theory characterizes the FDIC.

  6. JKH writes:

    I’m not sure why these discussions don’t refer to equity as a call option, in the same way they refer to debt as a put option. They both have the same asset value strike, and they’re complementary.

    The portfolio effect as highlighted by Nemo is important. But banks have portfolios too. And they allocate capital according to risk, or volatility, considering the portfolio effect. They also have winners and losers within and across their portfolios. And for both banks and these investors, the net outcome depends on the assumed or realized distribution of volatilities in the outcome – i.e. how volatile is the realized volatility across the different assets of the portfolio. Because of this, I tend to agree with James Surowiecki.

    The argument that these assets are more volatile than typical bank assets begs the question in several ways. To the degree that there’s no liquid market now, they’re volatile now. That may change some. But more importantly, to the degree that these private investors won’t be pinned down by marked to market accounting for purposes of assessing their own equity capital position, they’ll be effectively less volatile in the future. Volatility is in the eye of the measuring beholder. The investor’s equity option will be less volatile due to the domicile. So will be the put. And the put is less likely to be exercised.

    For the same reason, these equity investors also enjoy the benefit of a franchise value comparable to banking. All that’s involved is ignoring marked to market volatility for equity capital purposes, and taking into account the cash flow coming in from the assets. It’s an entirely different way to value the equity option. And the put is less likely to be exercised.

    Any selectively targeted suspension or adjustment to mark to market accounting will make these assets more valuable to the banks. However, this only evens the playing field with the private investors in terms of considering the effect in pricing the assets for the trade.

  7. JKH — It’s closer to a European than an American put. MTM volatility needn’t matter, but what value it settles into is still a great mystery. Doesn’t change the option pricing story at all.

    I don’t get how equity investors in a special purpose fund have “franchise value” to lose comparable to those with a durable connection to a going concern bank.

    The question of how diversified each guaranteed portfolio will turn out to be is very important, and so far unaddressed. Investors maximize option value by segregatin investments into separately guaranteed undiversified portfolios. Will FDIC permit that? It’s unclear.

    But even a diversified portfolio is quite different here than a bank’s portfolio of comparable assets, because the cost of exercise is much lower for the investment funds. Banks very rarely exercise the FDIC option. Banks occasionally do go underwater to greater or lesser degrees with administratively negotiated “prompt corrective action”, not exercise. I don’t think investment fund behavior is likely to be comparable. Do you disagree?

  8. Steve, can you tell us about your econblogger conference call with the Treasury department.

  9. septizoniom2 writes:

    all this discussion is very good.

    one question i have not seen asked is why is there no fee/warrants/stock issued to treasury by the sellers? are not they getting substantial value too from the arrangements?

  10. JKH writes:

    SRM,

    I haven’t seen anybody else make your point about MTM and the European put. I understand it, but must think more about it.

    Full franchise equivalence is pushing it. But embedded in a franchise value calculation is the hold to maturity accrual value on the existing asset liability portfolio, which has the potential/possibility to substantially or more than offset/reverse any MTM deterioration in that portfolio.

    I agree with your point on diversification – which I described as the volatility of realized volatility across the portfolio. But all the scenarios used on various blogs to date deliberately assume a volatile mix of realized directional volatility. It’s easy to screw the taxpayer on a scenario basis. A little more balance in the blogosphere scenario analysis might be in order.

    I must think more about your put exercise comparison between banks and investment funds. I’m not grasping the fundamental difference yet.

  11. Richard — alas, i am no brad delong. i was up all night working on a project, fell asleep in my office at about 7 am, woke up at about 11 to see that i was invited to a conference call i’d have loved to be on. i don’t understand the rhyme or reason behind who did and didn’t get invites, and lots of people both smarter and more prominent than i am did not, but i do give them a bit of credit for inviting someone as mistrustful and hostile to their initiatives as i am.

    septizoniom2 — good question. my answer would be cynical, although others might argue this is a reasonable financing arrangement and nothing more, so why should there be warrants. again, that’s not my view at all…

    JKH — i think we’re using franchise value differently here. i think you are thinking what i’d call the hold to maturity value of an asset portfolio. i’m thinking about the intangible going concern value of a business, both in an arms-length sense and to people whose human capital is firm specific. banks make money by providing services. they are not solely or even predominantly portfolio managers.

    i agree that we all need more clarity on what kind of portfolios FDIC will guarantee. you are write that people like me tend to assume permissiveness towards the highest subsidy scenarios, and that needn’t be the case (although even diversified portfolios of highly correlated hard-to-price assets will support a significant option value). it’d be nice if we could get hard and fast information about this, but i suspect that the best we’ll get is broad guidelines with an FDIC discretion loophole. like discount window collateral guidelines at the fed, that won’t be enough to calm us cynics, fairly or not.

  12. James Surowiecki writes:

    “one question i have not seen asked is why is there no fee/warrants/stock issued to treasury by the sellers? are not they getting substantial value too from the arrangements?”

    Oddly, the fact sheet says that Treasury is getting warrants in the partnerships that it’s investing in. I don’t really know what this means, unless it means that Treasury has an option to buy more of the partnership if its value rises.

  13. James Surowiecki writes:

    “But you can’t then argue that the current regime is equivalent to run-of-the-mill banking: deposit insurance plus extraordinary regulatory forbearance implies a dramatically different kind of subsidy than deposit insurance under cautious regulation . . .”

    Well, let’s say a “different” kind of subsidy — it’s been a long time since I’d say there was cautious regulation of big-bank investments.

    Two things I’m confused by:

    “The FDIC does have the power to say no. But from the FDIC’s perspective, it’s probably on the hook either way, but it faces less trouble if it has to make whole bank creditors than a public-private partnership participant.” When and how would the FDIC have to make a PPIP investor whole?

    More important, I don’t get your point about how MTM volatility doesn’t matter as long as the asset’s value is a mystery. In the first place, many of these assets are pools of mortgages. We now have a reasonably good sense of what the potential downside on unemployment is, and of how much further housing prices could fall. We (well, I don’t, but I’m sure PIMCO does) have a sense of how likely people are to walk away from their mortgages when they’re underwater, and so on. It shouldn’t be that hard to come up with a reasonable narrow range of projcted future cash flows. That range should determine the volatility, shouldn’t it? Even in cases where we’re talking about non-mortgage assets, your argument seems to be that private investors are going to look at these assets and be willing to pay more for those that are illiquid, precisely because they’re illiquid (which you say indicates a disagreement over value and is therefore a sign of volatility). That just seems crazy to me.

  14. James Surowiecki writes:

    Also, as far as the franchise value question goes, there’s something to the argument that the value of the future existence of a bank constrains bidding in a way that the value of the future existence of a PPIP doesn’t. But surely the fact that the PPIP is going to be stuck with these assets once it buys them (the whole point is that there is no secondary market for them now, and no way to know if it will exist), which means its capital is going to be tied up in them — and it’s going to be paying interest on its loans — for who knows how long, should also restrain it from overbidding. What’s typically left out of scenario models like Nemo’s is that it could easily take years before it’s clear whether an asset pool is a winner or loser.

    Also, as far as your skepticism about the FDIC being a rigorous examiner as far as leverage, this will hardly assuage your concerns, but the evaluation of the asset pool will be done by a third-party firm: “In assessing the supportable leverage of the asset pool, the Third Party Valuation Firm will analyze characteristics including

    expected cash flows based on type of interest rates, risk of underlying assets, expected lifetime losses, geographic exposures, maturity profiles and other relevant factors.” The fact that that analysis will be available should also help reduce the expected volatility of the underlying assets.

  15. James — I think we agree that it’s fair to say that the non-recourse loans are not so different than banking under deposit insurance and without effective regulation, as recently prevailed for large banks.

    FDIC makes PPIP investors whole in the sense that if purchased assets fall in value, FDIC extinguishes investors’ liability in exchange for the collateral. More specifically, FDIC assumes their liability to the bank that financed the sale in exchange for the surrendered assets. If the assets had remained on bank balance sheets, assuming bank creditors are mostly guaranteed, FDIC would have assumed the same liability to bank creditors in exchange for the same assets. In a sense, PPIP engineers a Swedish guarantee via a middle-man, but without the political fallout that might occasion an explicit guarantee without a “tough” reorganization, and without capturing any taxpayer upside wrt future bank franchise value.

    I think you are understating the volatility and uncertainty surrounding bank loans. Mortgages that go into foreclosure now can easily end up with a sub-50% recovery value, especially in troubled exurbs. Other loans may be more speculative, e.g. loans made to finance the private equity boom, which may be payable by near bust firms with very low recovery values. If you buy the Summers/Geithner virtuous cycle story, that increases volatility: there are potential very bad outcomes if the plan doesn’t “take”, potential very good outcomes if it does. Not even the very smart PIMCO people can come up with good point estimates for loan values. They’ll (loosely) estimate expected values and distribution, and the distributions will be wide, which implies high variance and a high option value.

    The illiquidity thing is counterintuitive, but it’s true. Purchasers under PPIF won’t be traders: they’ll care primarily about the hold-to-maturity value and the option value of the loans they are buying. The ability to convert to cash at an interim period is unnecessary, given the generous financing. So, they apply little illiquidity discount per se. If the financing was full recourse (but not subject to margin calls), they’d apply a discount for the uncertainty of valuation that sits behind the illiquidity. But the option in a non-recourse loan turns that logic around: The valuation uncertainty that makes the loans illiquid also makes the embedded put option valuable. It’s not the illiquidity per se that is valuable in this case, but the uncertainty in ultimate values that is the cause of that illiquidity.

  16. Another, perhaps more elegant way of looking at why uncertainty is so good for these assets is to point out that the portfolio a borrower ends up with when she accepts a non-recourse loan amounts to a synthetic call option on the assets (by virtue of put/call parity — see the comment thread to this post for a good discussion, which included JKH from this thread). It is never optimal to exercise a call option early. (Technical caveat: early exercise is never optimal when asset-cash flows accrue to the option holder, as they do in this case.) So MTM valuation or the ability to convert early to cash are irrelevant. The value of the option is determined by the the hold-to-maturity distribution portfolio. Like an ordinary call option (assuming some reasonably symmetric distribution around the expected hold-to-maturity value), the wider range of possible outcomes, the more valuable the option.

    If the illiquidity of the assets is caused by uncertainty in the hold-to-maturity values, then illiquidity and the value of the synthetic call option will be positively related.

    A caveat: If the illiquidity is caused by asymmetric information, that is if a “lemons” problem means all the risk is on the downside (at prices where the bank is willing to sell), this analysis needn’t hold. But I’m presuming the PIMCOs of the world will get as much information as the banks have about the loans they’ll purchase, so genuine uncertainty, not assymmetric information drives the hold-to-maturity distribution.

  17. James Surowiecki writes:

    “If the illiquidity of the assets is caused by uncertainty in the hold-to-maturity values, then illiquidity and the value of the synthetic call option will be positively related.”

    Right. I guess what I’m saying is that I’m unconvinced the illiquidity is primarily the product of a dramatic disagreement about hold-to-maturity values (or even a dramatic disagreement about the uncertainty of the hold-to-maturity values). I think it’s more likely the result of a) investor risk aversion; b) bank unwillingness to mark down; and c) investors waiting for fire sales, including those that would follow in the wake of nationalization.

    As far as the making investors whole line goes, I think it’s just a semantic confusion: if I’m understanding you correctly, I think I would say that the FDIC will make the company that the PPIP investor borrowed from whole.

  18. winterspeak writes:

    Do bank deposits even work the way Surowiecki says?

    In the post, he says “[FDIC insured deposits] are non-recourse loans to banks. You deposit money with a bank—that is, you lend it your money. The bank can then take that money, and leverage it up nine-to-one to make loans or acquire assets.”

    This is the standard–and wrong–multiplier model that has deposits being levered up to make loans. In reality, banks simply make the loan, that then creates the deposit. The constraint on bank lending is capital requirements.

    I should know this, but I want to make sure — are deposits counted as capital for capital requirements? (Note, I do not mean reserve requirements).

  19. winterspeak writes:

    OK, I checked this, and it does not seem like deposits have much impact on capital requirements at all. FDIC does not protect equity holders, it simply protects one class of liability holders.

    I don’t think this maps well to PPIP.

  20. winterspeak — i think the mapping is between bank equity holders and PPIP investors. both groups can lever up, purchase a risky portfolio worth much more than they invested, take all the upside, but force the government to eat much of the downside. it’s reasonable to model both groups as having borrowed funds from the state and been given a put option to boot.

    the details of bank lending, as you say, don’t really depend upon depositors. i don’t think banks are quite allowed to create deposits and lend to themselves directly, but so long as capital and increasingly weak reserve requirements are met, a bank can create deposits and lend to a middleman, who then invests in a portfolio of assets. bank leverage is mostly a function of loan demand (ie demand for loans the banks are willing to fund), subject to the capital constraint. in the traditional reserve multiplier model, total deposits were constrained by the central bank, but as reserve requirements have been diluted and central banks have moved to interest rate rules (which implies accommodating quantitative demand for

    deposits), it’s really only the capital constraint that binds. the reserve stuff just amounts to a game individual banks have to play to avoid regulatory fouls or clearing problems.

    one problem with the analogy is that not all bank liabilities are deposits, so banks’ effective put is in theory not written solely by the federal government, but also by bondholders, etc. but as we’ve seen, for the big banks, all nonequity liabilities may capture a state guarantee. from bank equity holders’ perspective, that not all liabilities are deposits only really matters in terms of the cost of funds (bonds are more expensive, at least if you’re not tbtf). in any case, equityholders get to put a lot of the lower tail to bank creditors or their public guarantors. (this is true of leveraged limited liability firms generally. what distinguishes banks is how they lever — they create their own deposits; how much and how cheaply they can lever up; and that the state is the guarantor of most of their liabilities.)

  21. James —

    “I guess what I’m saying is that I’m unconvinced the illiquidity is primarily the product of a dramatic disagreement about hold-to-maturity values (or even a dramatic disagreement about the uncertainty of the hold-to-maturity values). I think it’s more likely the result of a) investor risk aversion; b) bank unwillingness to mark down; and c) investors waiting for fire sales, including those that would follow in the wake of nationalization.”

    So, for there to be a lot of option value, there has to be a high dispersion of possible outcomes around the expected value. I think I’ve confused the issue by using the word “disagreement” so much. Disagreement, to me, is just one way high dispersion of outcomes reflects itself. I say that loan is worth 20¢. You say it’s worth 80¢. We are equally informed, so a neutral observer attributes a probability distribution with expected value of 50¢, but a lot of uncertainty around that mean. Equivalently, there can be just one investor who, after credit analysis, decides a loan’s expected hold-to-maturity value is 50¢ but with a wide range of plausible realizations around that number. Both scenarios are equivalent in terms of valuing the option. No one actually has to disagree. Investors must simply think that a wide range of hold-to-maturity values are possible.

    Your (a), investor risk aversion, means that investors think this. For the non-recourse arrangement NOT to have option value, investors would have to believe they know the hold-to-maturity value with certainty. If they did, there’d be no reason to offer the non-recourse financing. The implicit option would be worthless, and liquidity-constrained investors would immediately accept any full-recourse loans to purchase the asset at anything below its known hold-to-maturity value. The fact that investors are risk averse means they believe that assets might not turn out to be worth what they expect. Under this belief, the non-recourse option is valuable, even if all investors agree on that belief.

    Bank unwillingness to mark down implies no trade, if investors are certain in their valuations. However, if investors are uncertain, then the nonrecourse option has value, and investors should push their bids higher to capture this value, potentially high enough to satisfy banks. This is the usual complaint about PPIF: That extra push, to the degree investor attributed probability distributions are accurate, represents a probably liability to the state. It’s the nonrecourse subsidy.

    Holding out for nationalization fire sales is an interesting possibility. PPIF only addresses this indirectly, by signaling that nationalization is less likely. But perhaps that’s enough. From a taxpayer perspective, it would be better to signal that nationalized bank assets would be sold off slowly and deliberately, or even held to maturity, so there is no point in waiting for firesale values. Paying to strengthen banks’ bargaining position so that strategic investors stop holding out is an indirect and expensive way to eliminate the possibility of fire sales.

    I suppose the real “fire sellers” would be banks trying not to be nationalized when the state hasn’t credibly signaled that it won’t nationalize. So it might make sense to be quick and forceful one way or another: nationalize early and have the receiver demonstrate a capacity to hold rather than quick-liquidate assets, or commit never to nationalize so banks don’t desperately try to stave off the reaper. That second commitment can never quite be credible, though.

    Still, for your (c), you have to explain how heterogeneous investors manage to tacitly collude to hold out. That is, if there are bargains at marks that banks would accept, but there might be better bargains in a fire sale, why don’t some investors decide a bird in the hand beats a bird in a burning bush? And once the best bargains start getting taken, wouldn’t other investors stop holding out for fear of missing the bargains entirely? I don’t really think that (c) is plausible. Or at least it wouldn’t be plausible if potential investors weren’t liquidity constrained, but the government could eliminate liquidity constraint by offering non marginable full recourse loans. That is, if (c) is the problem, there’s no need for non recourse…

  22. I still do not get it how FDIC, that can “barely” guarantee deposits, just up 250,000 USD per depositor per bank, will guarantee debt under Legacy Loans Program. Why somebody, other than depositors, should get high and closer to 100% guarantee? Why then this kind of discrimination in percentage guranteed and maximum amount, if “This is precisely the same arrangement — down to the loans being guaranteed by the FDIC—that the Geithner plan sets up”?

  23. JKH writes:

    SRW,

    “So MTM valuation or the ability to convert early to cash are irrelevant. The value of the option is determined by the the hold-to-maturity distribution portfolio.”

    Only time for a quick note. I’m not so sure about this. I know what you’re saying. But step back big picture and consider the risk a potentially long term holder is willing to bear when not subject to MTM for capital purposes. The holder will view volatility and the willingness to bear that risk differently under any single scenario that consitutes the option valuation tree. Option pricing doesn’t take this into account, American or European. At least I don’t think it does. And the hold to maturity assumption doesn’t seem to account for the freedom to sell assets when the market encourages it, while being able to tough out the bad markets without selling.

  24. septizoniom2 writes:

    re my q and js’s reply: i am aware of that provision in the term sheet, but the partnership is not the seller of the legacy asset. i would have thought that the seller here is obtaining quite significant value from USG by virtue of the low cost non recourse loan and should issue some value to the USG as compensation. anyway, this point is off topic for this quite engaging discussion here. you are all brilliant writers.

  25. RueTheDay writes:

    Wait, what’s this about an “econblogger conference call with Treasury”? Was this like a “get onboard with the program and start blogging favorable material or end up in Guantanamo” type thing?

  26. winterspeak writes:

    SRW: Thanks for your comments. I see what you’re saying, bank equity holders are in a first loss position and can only lose up to what they put in. Surowiecki was talking about deposits though, and I think we agree, that has nothing to do with this.

    I’m still not sure why having a single class of liability holders backed by FDIC makes it non recourse though. Maybe I’m just being dense, but suppose FDIC did not exist. Bank equity holders would still be a first loss position. Banks could still lever up as they wanted to, subject to capital constraints. Banks can exist just fine without deposits, as we have seen.

    Are you saying that FDIC insurance increases the value to bank equity holders (because they are off the hook for one class of liabilities) and thus is a straight transfer? This I agree with, but I also would argue that in banking, during competitive periods, this transfer is competed away and goes to customers.

    In this sense, I think the PPIP analogy works well. There is a straight transfer in the non-recourse nature of the Fed loans, and they hope it will be competed away by the financial intermediaries, and end up with banks. It’s another mechanism for recapitalization through deception, which included the first TARP program, AIG, etc.

    Viewed from this perspective, I understand the analogy, but it also seems obvious to me that FDIC makes PPIP OK in any way.

  27. James Surowiecki writes:

    “Thanks for your comments. I see what you’re saying, bank equity holders are in a first loss position and can only lose up to what they put in. Surowiecki was talking about deposits though, and I think we agree, that has nothing to do with this.”

    I don’t understand why deposits have nothing to do with this: that’s where, in a simple model of a bank, the bank gets the money to lend to its customers. If I start a bank with $1 billion in capital, and I want to make $10 billion in loans, I have to get the $10 billion to lend out from somewhere. So I borrow it by taking in deposits (non-recourse loans guaranteed by the FDIC), and then re-lending that money. Take IndyMac. When it went under, it had $19 billion in deposit liabilities, $20+ billion in loans, and $800 million in cash on hand. Where had all the money that the depositors had deposited — literally — in IndyMac gone? It had gone out in the form of loans, hadn’t it?

  28. Nemo writes:

    James —

    “I understand this is the same argument you made in your original post, but there’s no substance to it without real numbers to back it up.”

    The same is true for your own argument.

    I made a follow-up post where I presented a more sophisticated model, and I put it online so you can experiment with it yourself. You do not need extreme probability distributions to build scenarios where the FDIC gets shafted.

    I am curious why you think it is more “realistic” for “the range of values to be quite narrower”. The market certainly does not agree with you at present.

    Finally, Bill Gross and all the PE folks seem very excited to have the government truncating their downside tail risk. Why would that be, I wonder?

  29. Gentlemen (I don’t detect any ladies) –

    Steve, thanks for the link to my recent rant (I think).

    At a time when the government is projected to run WWII-sized deficits into the indefinite future, what is the marginal bang per G buck in doing this stupid bailout? Doing nothing would be far better, economically and politically. Cf. John Hempton’s call for a Canadian-style highly concentrated, tightly regulated banking system. And Zero Hedge’s leaked GS memo on the ridiculous marks being applied to this stuff at some of the banks. How can a market possibly make here?

    As I point out in a recent update at my site, bank C&I lending is falling for cyclical reasons and will continue to do so.

    The administration is not showing good judgment in rushing a program as stupid as this one, but… that seems to be par for the course so far.

    –Benign

  30. winterspeak writes:

    James Surowiecki: Aaah, this may be a long discussion (or not). Your model of how banks get the money to make loans is the common one. However, it is completely wrong. Not your fault though — this is very poorly understood.

    What you outline is the standard fractional reserve banking story, and it is not how things actually work.

    What actually happens is that, when a bank makes a loan, it credits the customer’s account with a deposit (a liability) and sets up a receivable (an asset). It expands both sides of its balance sheet at the same time. The equity it had originally remains on the equity entry on the liability side of its balance sheet, but now the balance sheet is larger, so the bank is more levered.

    Notice that the creation of the loan creates the deposit. In the fiat money system we have in the US today, loans create deposits. The private sector increases its balance sheets, driven by bank lending, but it is credit extension that creates the deposits. Deposits do not in any way enable credit extension.

    The banks make all the loans they want, and then see how much in reserves they need to post with the Fed. If they are short, they borrow the extra they need from the interbank market (at the federal funds rate). If the banking system as a whole is short, it borrows from the discount window. The Fed manipulates the interbank lending market through open market operations to reach its target for the federal funds rate. Thus reserves do not constrain bank lending — banks lend first and do what they need to do to make reserve requirements. Canada has no reserve requirements.

    The only constraint on bank lending (from the supply side) is capital requirements, not reserve requirements. On the demand side, of course, you need qualified loan applicants.

    “I don’t understand why deposits have nothing to do with this: that’s where, in a simple model of a bank, the bank gets the money to lend to its customers. If I start a bank with $1 billion in capital, and I want to make $10 billion in loans, I have to get the $10 billion to lend out from somewhere.”

    Nope. As a bank you simply expand both side of your balance sheets at the same time, loaning the money (creating a deposit) and creating a receivable at the same instant.

    “So I borrow it by taking in deposits (non-recourse loans guaranteed by the FDIC), and then re-lending that money.”

    Nope. Banks can lend with zero deposits. We have many financial institutions out there that lend, and have no deposits.

    “Take IndyMac. When it went under, it had $19 billion in deposit liabilities, $20+ billion in loans, and $800 million in cash on hand. Where had all the money that the depositors had deposited — literally — in IndyMac gone? It had gone out in the form of loans, hadn’t it?”

    Not sure I’m fully understanding the balance sheet you are describing above. I’m guessing that by $20+ B in loans you mean a liability (loans that are financing the bank), and not a receivable asset (money expected from loans the bank made). $800M cash is an asset. (Please set me straight if I’m wrong).

    When people deposit money in a bank, it becomes a spreadsheet entry that credits their account, and debits the bank’s (where the deposit exists as a liability). Given FDIC, I find it simplest to think of the money residing as an entry on the Federal Reserve spreadsheet.

    When a bank goes bust, essentially assets collapse (loans don’t get paid back) forcing the bank to right down asset values and equity, shrinking both sides of its balance sheet at once. As equity shrinks faster than the balance sheet overall (leverage) the bank runs afoul of its capital requirements, and FDIC steps in. Without FDIC insurance, depositors would simply be one additional class of unsecured liability holders, who may or may not be made whole, or partially whole, in bankruptcy.

    Deposits don’t get loaned out. Loans create deposits.

  31. James Surowiecki writes:

    “$20+ billion in loans” — these were loans IndyMac had made. They were assets on the balance sheet — but they represented $20 billion+ that IndyMac had paid out.

    I remain mystified by your account. I understand that financial institutions can lend without deposits. But IndyMac was lending its deposits. It took in $19 billion in cash deposits — people gave the bank $19 billion of their money. When it went under, only $800 million of that $19 billion was on hand. Where did the $18.2 billion go, if not into the loans IndyMac made and the assets it bought?

  32. RueTheDay writes:

    “Nope. As a bank you simply expand both side of your balance sheets at the same time, loaning the money (creating a deposit) and creating a receivable at the same instant.”

    This would be true for a bank that made a loan to a borrower that also happened to have a deposit account at the same bank. It does not accurately describe what happens when a bank loans money to a borrower who holds the funds at another bank (or who wishes to immediately transfer the loaned funds to another bank). A bank may be able to create deposits and expand its own balance sheet, however, it cannot create deposits at another institution – it needs to physically transfer funds to the second institution (usually done between the accounts both banks hold at the Fed).

  33. winterspeak writes:

    James Surowiecki: Thank you for the clarification. IndyMac was not lending out its deposits. In fact, no bank lends out its deposits.

    I see your point though: people put $19B of cash into the bank, and it only had $800M when it went under. Where did the $18.2B go?

    It is not accurate, though, to think of deposits as some kind of vault cash that go in the bank and then get lent out again. We aren’t on a gold standard, and money is fungible.

    I find it most helpful to think of FDIC deposits as simply going to the Govt, who returns them when the bank goes under. Certainly if you ask any IndyMac customer (all of whom were made whole on their deposits) who returned their money, that is what they would say. So from that perspective, they became a data entry in a Govt spreadsheet.

    I find it hard to tackle the numbers in your example, the math is misleading. Most banks make out loans far in excess of deposits, and depositors aren’t the only liability holders who fund the bank. So, was there anything else on the asset side? What about liabilities — did the bank have $1.8B in equity? If so, why was it deemed insolvent? &c.

    My point is that banks expand both sides of their balance sheets when they make loans. They do not loan deposits out. Making loans, in fact, creates the deposits (and balances those with the creation of a receivable). Cash, especially vault cash, is best thought of as an inventory management issue, and is not fundamental to credit creation.

    To the degree that FDIC offers a subsidy to banks, the value of the option gets competed to depositors. When thinking about the analogy to PPIP, that same value is being competed to the banks. This is the same deceptive recapitalization we’ve seen all along.

  34. I think that some of the confusion comes from the word “deposit”. That makes it sound like someone came and deposited money into a bank.

    But a better word for deposit would be “Bank-To-Customer IOU”. One way a Bank-To-Customer IOU gets created is that customers deposit money in the bank, and the bank issues IOUs promising to pay the customer back.

    Another way that Bank-To-Customer IOUs get created is a customer asks for a bank loan. Nothing much happens, when a bank loan is granted — the customer writes the bank IOUs, and the bank writes the customer IOUs. The customer promises to pay off her IOU over, say, five years, while the bank promises to pay its new Bank-To-Customer IOUs on demand.

    But since Bank-To-Customer IOUs are guaranteed by the government, they are perfect substitutes for government money. Effectively money is created when banks issue redeemable-on-demand IOUs, which we call “deposits”, but which in aggregate have never been deposited, but have been created when a bank decided it would be profitable to issue IOUs to a customer in exchange for an IOU from a customer for the same amount plus interest.

    Looking at the aggregate banking system, the vast majority of deposits were created by loans (all deposits in excess of banks cash and reserves on deposit with the Fed). Looking at individual banks, sometimes deposits are created by literally depositing money, but usually the deposited money was itself created by a loan at another bank.

    (RTD’s scenario can be viewed as Bank A creates Bank-To-Customer IOU, customer redeems Bank A IOU for cash, which she deposits in Bank B in exchange for Bank B Bank-To-Customer IOU. The net effect is the same — the total number of Bank-To-Customer IOUs — deposits — increases. But some government cash — bank reserves — have been caused to move as well. Banks are supposed to take care that they issue Bank-To-Customer IOUs in a manner such that they are capable of coping with the reserve movements customers may compel by redeming those IOUs.)

  35. winterspeak writes:

    RueTheDay: Yes, sometimes banks transfer deposits, and receivables to other banks. The point is that the deposit, wherever it is held in the banking system, was created as the consequence of a loan, wherever that is held in the banking system.

    When you transfer a loan and a deposit to another institution(s), the balance sheet shrinks but not at the expense of equity. The bank is smaller, but less levered. Everything is fine.

    If you just transfer the deposit (as say, cash) then cash on hand decreases on the asset side, and the deposit on the liability side. Once cash on hand is exhausted, the bank cannot fulfill the legal demands of its deposit (liability) holders. That’s the same as missing a bond payment, and it’s traditionally called bankruptcy.

  36. JKH writes:

    Winterspeak,

    It may be possible to consider a corollary to the paradigm here.

    Banks create new money (deposits) by extending new credit. This works at the micro level and the macro level. It’s the fundamental credit money causality that is generally overlooked, leading to a common misinterpretation of money dynamics.

    That obviously doesn’t preclude banks from competing for existing deposits, where deposits attracted on a competitive basis lead to a central bank surplus, or replace deposits that may have been lost on a competitive basis.

    The second mode doesn’t contradict the first.

    It’s just that the popular tendency is to construct a world view of finance and economics on the basis of the second interpretation, which is literally a derivative one, rather than the first interpretation, which is the true foundation of credit money linkage.

    P.S. this is the commercial banking case that parallels the discussion you and I have had regarding the idea of government surpluses, where taxes operationally can precede expenditure, although only in a derivative sense relative to the true cumulative causality, which is that expenditure precedes taxation.

  37. Alex writes:

    Perhaps I have a basic misunderstanding: I thought the FDIC insures the depositors, not the bank. Yes, the bank pays premia but it’s for the benefit of the depositors rather than the bank. Isn’t that how it works? So if the bank blows itself up w/ leveraged bets that go wrong and no longer has the equity to repay depositors then the FDIC makes the depositors whole. I don’t believe the FDIC makes the bank whole on the levered part of its loss… Rather, the FDIC will take over the bank, wiping out its shareholders, and pay out the depositors up to the insured amount (at least). If I’m right then this situation bears no resemblance to the non-recourse loans that will be on offer to the private participants in the PPIP. These guys will be given the much reviled free put. Not so a bank

  38. James Surowiecki writes:

    “I find it hard to tackle the numbers in your example, the math is misleading. Most banks make out loans far in excess of deposits, and depositors aren’t the only liability holders who fund the bank. So, was there anything else on the asset side? What about liabilities — did the bank have $1.8B in equity? If so, why was it deemed insolvent? &c.”

    IndyMac had a small amount of other debt, including preferred stock. But almost all of its liabilities were insured deposits and advances from the Federal Home Loan Bank. It had about $1 billion in equity (supposedly), just before it was taken over. But its total deposits were just a little less than its total loans. (It also had another $6.6 billion in securities.)

    Back to the issue at hand, I understand that on a system-wide level, loans create deposits. But I don’t understand the argument that on an institutional level, in the banks’ case the FDIC subsidy is competed away to those lending to the individual bank (depositors), while in the PPIPs’ case is competed away to those selling to the individual PPIP.

  39. JKH writes:

    Going back a few comments:

    Winterspeak:

    “I’m still not sure why having a single class of liability holders backed by FDIC makes it non recourse though.”

    I think this question is important to the clarity of the discussion and analysis.

    More generally, there seems to be confusion in distinguishing between the following:

    – The role of PPIP debt in relation to PPIP equity

    – The role of collateral in the case of PPIP debt

    – The role of bank debt and deposits in relation to bank equity

    – The role of bank deposits in relation to bank debt

    – The role of the FDIC in relation to bank deposits

    Each of the above constitutes a distinct put option, with a different connotation for the term “non recourse”, which has been used in quite a general way in the discussions.

  40. JKH writes:

    I also think it’s meaningless to talk about “bank put” or “bank non-recourse”.

    “Bank” is nothing in terms of finance.

    Finance, including embedded options, is asset/liability or equity specific.

    Sorry to be negative here. Just letting it out.

  41. winterspeak writes:

    SRW: Sometimes creating new words simplifies things, sometimes it just makes them more complicated! I’m not going to pass any judgement on your IOU terminology — it works for you I’m sure, and it make work for others!

    The reason I phrased my answer as I did is that it’s critical to understand that banks do not loan out the money that gets put into them, as an IOU or anything else. Banks simply write two-way IOUs, and yes you can think of this as creating money, it depends on what your definition of money is. It is certainly balance sheet expansion.

    Private sector balance sheets can only expand in two ways: 1. banks make loans, 2. Government runs deficits. I did not want to bring in Government deficits, because I fear that would be too much for James at this point.

    Transferring a deposit from one bank to another simply credits one account, and debits another. No new money is created.

    JKH: I agree with you. Whatever subsidy there is from the Government implicitly in FDIC insurance, in a competitive banking environment, will get transferred to depositors and banks compete for deposits. So the value in the non-recourse loan is transferred from the Government to the ultimate liability holder.

    My point is, to the extent that the PPIP auction is competitive, the Government non-recourse subsidy will go to the bank. To the extent it is non-competitive, the subsidy gets split between banks and private equity firms. If we want to capitalize banks and protect unsecured liability holders, shouldn’t we do that honestly, or at the very least efficiently? If we’re lucky PPIP will be as good as a straight transfer from the Govt to unsecured bank liability holders and employees. If we’re unlucky, PPIP will be less good that than, maybe a lot less, leaving banks undercapitalized but enriching some already very wealthy PE folks on the way. It’s like they’ve set up a camouflaged leaky pipe between the Fed/Treasury and the banks, drilled lots of holes in it, and then are auctioning chewing gum to plug up the holes. Or something.

    My point is that the operational analogy that actually holds between FDIC and PPIP *weakens* the argument for PPIP, it does not strengthen it, contra Hempton and Surowiecki.

  42. James Surowiecki writes:

    “Perhaps I have a basic misunderstanding: I thought the FDIC insures the depositors, not the bank. Yes, the bank pays premia but it’s for the benefit of the depositors rather than the bank. Isn’t that how it works? So if the bank blows itself up w/ leveraged bets that go wrong and no longer has the equity to repay depositors then the FDIC makes the depositors whole. I don’t believe the FDIC makes the bank whole on the levered part of its loss… Rather, the FDIC will take over the bank, wiping out its shareholders, and pay out the depositors up to the insured amount (at least).”

    The FDIC isn’t going to make the private investors whole on the levered part of their loss — it’s going to make the bank that lent to the private investors whole. If the bet goes bad, the FDIC will effectively take over the PPIP, wipe out the investors, and seize the assets that have been bought. Similarly, in the event that a bank’s levered bets go bad, the FDIC makes the depositors — who I thought one could say had lent their money to the bank, but apparently not — whole.

  43. James Surowiecki writes:

    “Whatever subsidy there is from the Government implicitly in FDIC insurance, in a competitive banking environment, will get transferred to depositors and banks compete for deposits. So the value in the non-recourse loan is transferred from the Government to the ultimate liability holder.

    “My point is, to the extent that the PPIP auction is competitive, the Government non-recourse subsidy will go to the bank. To the extent it is non-competitive, the subsidy gets split between banks and private equity firms.”

    Just to repeat myself, I don’t get this at all. In the Geithner plan, the ultimate liability holder isn’t the bank that the PPIP is buying from. It’s the bank that’s lending to the PPIP, and is on the hook if the leveraged bet goes bad. You have the subsidy going to the bank that’s selling to the PPIP (that’s the bank that, in your model, needs to be recapitalized). Why?

  44. Nemo writes:

    A final question for James:

    Why are these loans non-recourse?

    If saddling the FDIC with the losers is not the intent, then why does PIMCO not have to repay all of the loans they take out under this program, instead of the just the ones that correspond to winning bets?

  45. JKH writes:

    SRW,

    I’m jumping around here because I returned late to the discussion.

    Regarding our exchange on “franchise value”, I note that of Hempton’s five types of solvency tests, the only one that includes the value of future accrual margins of any sort is the third one, “positive economic value of the banking entity”. To the extent that a PPIF has accrual cash flow, which is distinct from an MTM valuation, that accrual cash flow is a subset of all future accrual cash flows that would be considered to be part of “franchise value” or “economic value”. A similar future cash flow subset can be identified for a bank, based on the run off of the existing asset liability book. That’s not to endorse Hempton’s framework per se, but just to illustrate what I meant by “subset of” in the context of such a comparison.

  46. Alex writes:

    “The FDIC isn’t going to make the private investors whole on the levered part of their loss — it’s going to make the bank that lent to the private investors whole. If the bet goes bad, the FDIC will effectively take over the PPIP, wipe out the investors, and seize the assets that have been bought.”

    James, I’m not sure I fully understand what you’re saying here – only the FDIC is going to be lending to the private investors – but I still don’t believe it’s correct :D. Here’s how I believe the PPIP is intended to work. Consider the purchase by PPIP of $100 face value of MBS from some bank for $84. The private investor and the Tsy invest $42 each and each will receive half of any price appreciation on the bond above $84. If the bond eventually pays off in full then both the private investor and the Treasury make $8 on their $42 investments (ignoring coupons and funding costs, for simplicity). The difference is that the private investor can make his $42 purchase with $6 of his own money and $36 of money borrowed from the FDIC. If the bond doesn’t pay off or declines in price then the private investor can walk away with a maximum loss of $6. That’s the non-recourse feature. Of course, the real beneficiary of the non-recourse feature will be the bank that’s selling the bond into PPIP and not the private investor that co-invests with the Treasury. Remember, the private investor sets the price and the Treasury goes along w/ that price. The various private investors who participate in PPIP will bid the bond up until their total investment, including the put embedded in their loan, is almost a break-even proposition. That means the Tsy will be buying at too high a price – since they don’t have the put – and the FDIC is short puts w/o recompense. Both features screw the taxpayer. So, the bank wins and the taxpayer loses – as intended.

  47. James Surowiecki writes:

    “If saddling the FDIC with the losers is not the intent, then why does PIMCO not have to repay all of the loans they take out under this program, instead of the just the ones that correspond to winning bets?”

    The point of the plan is absolutely to make the FDIC bear most of the downside risk of these investments. I’m not arguing with that assertion. The justification for this from Treasury and the Fed is the supposed massive risk aversion on the part of investors, which would make them unwilling to use serious leverage (which is necessary to turn the $150 billion Treasury has at its disposal into $1 trillion) on a recourse basis. I’m agnostic on whether that’s true. But my point all along has only been that FDIC insurance allows an individual bank, subject to Steve’s caveats above, to do essentially what a PPIP does — use leveraged money to make bad bets without having to repay its creditors. If people think the PPIP model represents the robbery of the taxpayer or will create a massive distortion of prices, then it seems to me that they should be saying the same thing about deposit insurance. (Of course, as you’ve pointed out, at least on the surface it’s other banks, and not the taxpayer, who foots the bill, but that’s obviously complicated.)

    Question about the impact of the loans on pricing: Even with a non-recourse loan, doesn’t overpaying with leveraged money magnify your risk? I mean, if I put up $10 for a $10 asset, it has to fall in value by 100% for me to be wiped out. If I put up $10 for a $70 asset, it only has to drop in value by 14% for me to be wiped out. Does that not mean that in the latter case you have less margin for error? Or is that simply built into, as it must be, the expected-value calculation?

  48. Don N writes:

    Since IndyMac was brought into the conversation…

    They had a bunch of the same sort of assets that the Geithner loan plan is addressing. What did they end up being worth? It took some time to wrap that whole takeover action up. It ended up costing ~$4 billion more than originally expected. Was that because the assets really were worth pennies on the dollar or what?

    Don N

  49. JKH writes:

    The equity investors in PPIF don’t give a damn who holds the debt so long as the debt offers the same sort of put option provided by a bank’s creditors and depositors.

    But this put option is worth more than bank funding, not because its FDIC backed, but because it’s collateralized.

    The expected payoff for the equity holder is higher with a collateralized put than a non-collateralized put.

    Beyond that, the expected payoff depends on the distribution of asset volatilities. As in the many examples provided, realized asset values that are both high and low provide a greater payoff than realized asset values that cluster around an average.

  50. James Surowiecki writes:

    “only the FDIC is going to be lending to the private investors”

    No, the FDIC isn’t lending any money. (How could it? It’s only got like $60 billion in the fund, I think.) It’s guaranteeing loans that will be made by private institutions or whoever to the PPIPs. So if the loans go bad, then the FDIC will seize the PPIP’s assets and pay back the institutions that lent the money to the PPIP. That’s exactly what it does with bank depositors.

    Even so, the math on your example is right, except that Treasury enjoys the same benefits of the put and of the leverage as private investors do. In your example, Treasury is only putting up $6, and it’s getting back $8, and it doesn’t lose anything other than the $6 if the investment goes bad.

  51. Alex writes:

    Aha! Thanks for the clarification of the mechanics, James. Can you explain how the Treasury enjoys the same leverage and put as the private investor? Who’s writing the put to the Treasury, the FDIC? If so, the government as a whole is still short a put to the the private investor. The FDIC is short 2 puts, 1 to the Tsy and 1 to the private investor. So the gov’t as a whole is short 1 put and the earlier conclusions still apply. I guess my point is that, from the perspective of the taxpayer, we can bundle the Treasury and the FDIC together here…

  52. Nemo writes:

    James —

    You continue to think in terms of a single loan, instead of multiple non-recourse loans.

    Yes, FDIC insurance encourages banks to overpay for assets; that is why they are heavily regulated. But the “loan” in this case is with recourse to the entire bank.

    Under P-PIF, a single private equity firm will be able to take out multiple independent non-recourse loans to purchase distinct assets. This makes all the difference in the world.

    Your last question (“Doesn’t overpaying with leveraged money magnify my risk?”) makes the same mistake. Yes, one asset purchased with one loan in your example only has to fall 14% for you to get “wiped out”. It also only has to rise 14% for you to double your money. If you can make similar bets on 100s of assets, what starts as a high probability of getting “wiped out” turns into a virtual certainty of reaping huge profits at somebody else’s expense. This has almost nothing in common with a single bank purchasing large numbers of assets with the help of its single massive “put option” from the FDIC.

    I can no longer tell if you are making this mistake deliberately in order to avoid admitting that you are wrong.

  53. James Surowiecki writes:

    “The FDIC is short 2 puts, 1 to the Tsy and 1 to the private investor.”

    Right. This is part of the plan that hasn’t gotten much notice (which makes me think I may be wrong), but as far as I can tell from the Summary of Terms, the economics of the deal for Treasury and the private investor are. The key is that it’s the PPIP as a whole that borrows at the 6-to-1 rate, and since the PPIP’s equity is split (most of the time) 50-50, effectively the Treasury Department is borrowing half the money, non-recourse. So saying the FDIC is writing 2 puts is, in a way, accurate.

    Whether that means we can bundle Treasury and the FDIC together is complicated. I mean, technically, the FDIC does not use taxpayer money: it raises its money from banks. And in this case, it’ll raise some money by charging the private investors a “guarantee fee” (Treasury, as far as I can tell, doesn’t have to pay this). It’s true that if the FDIC fund runs out, it’ll borrow from Treasury to make up the difference, but in theory it will have to pay that money back by raising fees on banks. I think Nemo, in one of his posts, said that the real subsidy here is going from healthy banks (who will have their FDIC fees raised) to the PPIPs.

  54. surferdude writes:

    “I find it hard to tackle the numbers in your example, the math is misleading. Most banks make out loans far in excess of deposits, and depositors aren’t the only liability holders who fund the bank…”

    for much of history, banks did not loan out more than their deposits. large banks could do this mainly out of access to other funding sources. until commercial banks gained access to fhlb system in 1989 as part of firrea legislation almost all community banks had LtD ratios in the 0.6 to 0.8 range. then a LtD ratio over 0.9 would draw the attention of the regulator. over the past 15 years, loans as a share of balance sheet, deposits or capital have increased which is part of the problem not even looking at the riskiness of loan portfolio composition. fractional reserve banking allows for loans to create deposits but loans are not the sole source of deposits. for example, i have more money on deposit than i have in loans.

  55. Alex writes:

    “It’s true that if the FDIC fund runs out, it’ll borrow from Treasury to make up the difference, but in theory it will have to pay that money back by raising fees on banks.”

    Ok, let’s do a simple thought experiment: Let’s say $1T is eventually spent in PPIP on bonds at an avg price of $80 per $100 face value. Let’s say that 1/2 mature w/o problem and that 1/2 have a $20 recovery value – a scenario considered by many to be reasonable and expected. The dollar loss on the losers will be 1/2 * $1T * 60/80 = $375B. Of the $375B, the investors have to eat 1/2 * $1T * 1/7 = $71B and the FDIC will have to find ~$300B to repay the lenders who provided leverage to the Tsy and the private investors. This assumes, of course, that the loans are associated with the individual bond purchases of each borrower. It may be the case (is it?) that the non-recourse feature will be applied to the aggregate purchases, winners and losers both, of a given borrower. In this case, there are gains of 20/80 on $500B and losses of 60/80 on $500B in our example, for a net loss of $250B. The Tsy and private investors presumably put up 1/7 * $1T = $143B of equity, so the FDIC will still have to cover at least $250B – $143B = $107B in this case. The actual figure in our example is sure to be higher than $107B because there will be more than one borrower and winning and losing investments will be distributed unevenly among the PPIP participants. For example, if one guy buys a bond that pays off in full and another guy buys a bond that ends up being worth $20 then the FDIC will have to cover the loan of the second guy w/o being “subsidized” by the first guy’s gain. Either way, that’s a lot of insurance premia that the FDIC would have to raise in order to pay off the Treasury after the FDIC, in turn, gets bailed out! The banks will legitimately argue that there’s no reason for the FDIC to gouge them in order to cover its losses on writing free PPIP puts. So if PPIP goes badly and the puts get exercised then the FDIC won’t have enough dough. The Treasury will have to bail out the FDIC(!) and the FDIC is unlikely to be able to repay Treasury in the foreseeable future. …We clearly need to bundle the FDIC and the Treasury together when thinking about the consequences to the tax payer of PPIP.

  56. James Surowiecki writes:

    “It may be the case (is it?) that the non-recourse feature will be applied to the aggregate purchases, winners and losers both, of a given borrower.”

    I think this will, unfortunately, not be the case — I think the loans are pegged to the PPIP, not the borrower. As you show here, and as Nemo’s model does too, this is potentially very important in terms of the FDIC’s losses.

  57. James Surowiecki writes:

    “Ok, let’s do a simple thought experiment: Let’s say $1T is eventually spent in PPIP on bonds at an avg price of $80 per $100 face value. Let’s say that 1/2 mature w/o problem and that 1/2 have a $20 recovery value – a scenario considered by many to be reasonable and expected. The dollar loss on the losers will be 1/2 * $1T * 60/80 = $375B. Of the $375B, the investors have to eat 1/2 * $1T * 1/7 = $71B and the FDIC will have to find ~$300B to repay the lenders who provided leverage to the Tsy and the private investors. ”

    One thing I’d say about all these thought experiments that have been done in the last few days, from yours to Nemo’s to Krugman’s to Sachs’: they all prove — no offense — what we already knew, which is that if the investors overpay (that is, pay more than the assets’ future cash flows are really worth), then the government (or the FDIC — like you, Alex, I’m not sure it’s useful to distinguish them) — will end up taking big losses. I mean, if you make non-recourse loans to people who make bad bets, then you’re going to lose a lot of money.

    I’m pretty sure Tim Geithner and Bernanke know this. Either they think the assets are, on the whole, significantly undervalued, in which case this plan makes everyone a winner. Or they think that this is an easier way to recapitalize the banks, and that it might have the added benefit of getting the secondary market for mortgages moving again. But I’m fairly certain that the revelation that the government is on the hook in the event of overvaluation isn’t going to shock them.

  58. mittelwerk writes:

    what private banks are financing the PPIP? where is that coming from?

    people here are getting lost in the con game. the economics of the PPIP don’t seem terribly complicated. fundamentally, non-recourse loans on overvalued collateral are indistinguishable from (gift) capital. as for losses, there’s only one entity that can underwrite losses of such magniutude as will occur, and that’s the treasury: and that’s our toxic legacy.

    schematically, this is nationalization — though not of banks but of crappy assets themselves, and only them.

  59. jimbo writes:

    Or they think that this is an easier way to recapitalize the banks

    If by “recapitalize”, you mean, “give lots of money to their bank executive friends”, yeah, I guess this is the easiest way to do it. The fact is if the “Bank of America” was the “Bank of Paramus”, the FDIC would have come in long ago, wiped out the shareholders, fired the executives, and sold off the “troubled assets” to the top bidder. But that’s not how it happens if you’re a Wall Street hotshot.

    The charitable part of me wants to hope that this is a stalling action to distract the markets while the FDIC staffs up to pull off the nationalization – after all, think about what that would entail: the FDIC would have to walk into every branch of BofA, Citi, etc. (do one, you gotta do ’em all) simultaneously and make sure none of the officers walk off with anything. It would make D-day look like a nice day at the beach. So maybe they’re just getting ready, and all this nonsense is just a smokescreen. But then I think, nah – this is America! Land of the crooks! Home of the screwed!

  60. Alex writes:

    Let me re-post a comment I made elsewhere, and solicit any comments:

    What I fail to understand is why the “trickle up” approach hasn’t been embraced. Consider the following approach:

    1) Identify homeowners who owe more than their homes are worth (necessary but not sufficient for a rational homeowner to walk away from his mortgage). The list of homeowners is easy to put together with info from mortgage servicers and/or application forms filled out by homeowners who want assistance.

    2) Determine a reasonable monthly payment for these homeowners. This would probably be a fixed percentage of their current income, as per their most recent W2.

    3) Have the government pay the difference between that reasonable monthly payment and the required monthly mortgage payment. This could be implemented in a variety of ways, e.g. monthly reimbursement of the homeowner for part of the prior month’s full mortgage payment upon evidence of payment.

    Note that this approach wouldn’t require cram downs or any other changes requiring re-documentation and/or legal wrangling about the securities collateralized by these mortgages…

    Such a program would dramatically decrease both current and anticipated defaults and thereby raise the prices on mortgage-backed securities, accomplishing the desired bail-out of the banks. But, unlike the Geithner plan, it would accomplish this bail out by treating the underlying problem (foreclosures and unaffordable debt) as well as the symptom (depressed security prices). Furthermore, it would do so in a way that is both cost-effective (by requiring those homeowners who receive assistance to continue to make monthly payments that they can afford) and has a natural exit strategy (assistance can end when home prices increase enough to make the home worth more than the remaining loan pricipal OR when income increases enough to make the monthly payments without assistance).

    Sometimes the direct approach is best…

  61. James Surowiecki writes:

    Alex, Michael Lewis wrote on this subject in his recent column on the AIG bonus kerfuffle:

    “Since the beginning of the crisis I’ve wondered why the government has found neither the will nor the way to attack the root of the problem — the people who borrowed money to buy homes they shouldn’t have bought.

    “Now I think I understand. It would be too simple. People would understand a lot of small payments to the guy down the street who doesn’t deserve them, and become outraged. Far better to throw trillions at opaque corporations, the inner workings of which no one still really understands.”

    There may be something to the idea that watching the government pay your neighbor’s mortgage would make someone angrier than an elaborate scheme like the Geithner plan.

  62. mencius writes:

    My take is here.

    Shorter Moldbug: not a subsidy, at least not to the PPIP buyers. Reason: the return profile for PPIP investments is generally the same as in any leveraged investment. The toxic-waste buyer makes a lot of money if the toxic waste goes up; loses all his money if it goes down even slightly. Anyone who buys any financial instrument on margin gets just the same result.

    The PPIP is a government loan, not a true subsidy. A true subsidy would be if the government said: we will give you a dime, outright, for every dollar of toxic waste you buy. Of course one can use the word as one likes, but there is a real qualitative difference between a loan and a grant.

    The infamous “non-recourse loan” does not directly affect the risk-return profile of the PPIP buyer, because once his equity is wiped out, he does not care whether the asset is sold on the open market (as in a conventional margin loan) or directly to USG. Thus, this “option” is of zero value to the PPIP buyer. Wiped out is wiped out.

    Not that there aren’t differences between a PPIP loan and a conventional margin loan. There are, and the differences are important.

    (Since the PPIP loan is more desirable, the differences could be described as a subsidy. Again, however, I think it is best to reserve this term for actual gifts, not loans. If any loan that a free market would not offer is described as a subsidy, of course (to paraphrase Schopenhauer) our financial system is as riddled with subsidies as a streetwalker with syphilis. The indictment is so obvious, inescapable and universal that it ceases to be an indictment.)

    The differences are (a) the term of the loan (zero-term for a conventional margin loan, long-term for the PPIP), and (b) the fact that if the PPIP fails for any investment, said barrel of sludge is not dumped back on the market, but bought by USG.

    The result of (a) is that the PPIP buyer need not post continuous collateral – if his leverage is 7 to 1 and the asset falls by 15%, he is not instantly wiped out, as he would be with a conventional margin loan. He is concerned only with the price of the asset at the end of the (multiyear) PPIP term. Smart.

    The result of (b) is that in case the PPIP buyer is wiped out, his vat of reeking, dioxin-flavored garbage is not dumped back on the market, as with a conventional margin loan, but snapped up by dioxin-craving USG, feeding its savage and inexorable thirst for chlorinated hydrocarbons. Therefore, the systemic feedback loop of a bank run is broken. Also smart – at least, if your concern is saving our present banking system.

    I feel this combination is an effective device for preventing the bank-run death spiral that made the toxic assets toxic in the first place. Unfortunately, it is probably a case of closing the barn door after the horse has departed. If MBS investments had been funded by PPIP-style programs *in the first place*, the problem would not exist. But, of course, USG would never have done this.

    Inasmuch as the PPIP is a device that works to sell bank assets to USG at above their market price, it is indeed a subsidy to the banks. But in order to subsidize the banks as such, PPIP investors need to take a bath. In order to subsidize the banks on a large scale – the scale actually required – either the scheme must actually succeed, or the PPIP investors must take an epic bath. I doubt either of these conditions will come true.

    The game theory of the PPIP is just all wrong. The fewer buyers join the herd, the more likely it is that PIMROCK will take a bath. There is only one real Nash equilibrium here, and my guess is that fund managers are collectively intelligent enough to find it.

    As for self-dealing, I don’t expect it. There is simply too much scrutiny. No one at Citi etc would be crazy enough to think they could get away with it. These firms already have compliance departments the size of Saturn. I’m not suggesting that there are any actual moral scruples in the equation – just people smart enough to only cheat when they can do so successfully. I really hope I am not being naive about this.

  63. wow. teach me to go away for a couple of hours.

    there’s not so much i can odd to this (really good) conversation, ‘but here’re a couple of small points:

    — it’s been brought out pretty well in this conversation that the compartmentalization of the financing is really, really import. assuming correlations less than one, a one-asset, one-loan scheme is likely to trigger much greater losses to Tsy/FDIC and higher subsidy to banks and/or investors than loans against large diversified portfolios only. i haven’t seen that issue discussed very much, and didn’t give it much thought prior to reading these comments. it seems like that’s something FDIC could clarify.

    — it’s worth emphasizing, re thought experiments, that Tsy/FDIC loses under PPIP even if investors on average don’t overpay at all (as long as assets are purchased in many separate deals). if investors price bank assets accurately, on half the deals, they earn a profit, but on half the deals they take an equivalent loss, a part of which is eaten by the government. so investors writ large expects a profit on a dollar invested at fair prices, which suggests a rational investor would try to bid a smidgen more than fair price, because that cuts negligibly into her profits but gives her a larger share of the deal. competition among investors leads to overpaying up to asset + option value, which increases Tsy/FDIC’s losses relative to fair prices, but there would be a public loss / private subsidy even if investors could somehow be forced to bid no more than expected hold-to-maturity value.

    — the above is all in terms of a risk-neutral investor. how would risk averse investors behave? the easiest (for me!) way of grokking that is to divide the PPIP investor’s portfolio into two assets, the asset or portfolio purchased from the bank, and the put option. regardless of leverage, a risk neutral investor would pay up to the sum of the expected value of those items, asset + option, because at that cost, the expected value of the transaction greater than zero with any level of leverage. a risk averse investor will pay something less than this, but (assuming she can value the asset and option with equal certainty), she would not distinguish between a dollar of asset value and a dollar of option value, since both assets are purchased with the same degree of leverage. so if the option is worth 20% of the asset, the risk averse investor will “overpay” for the asset by 20%, with which she will really be buying the asset. Whether she is overpaying for the asset in risk neutral terms depends on her risk aversion: if the combination of asset uncertainty and leverage would have caused her to take a 20% haircut for risk, she’ll end up paying precisely the risk-neutral price of the asset (consistent with James’ comment above about trying to overcome risk aversion, and Tsy’s general justification of its plan). the treasury would still lose in the scenario, even though no assets are overpaid for in risk-neutral, hold-to-maturity terms, because half the assets (on avg) would turn out to be worth less than the risk-neutral expected value, and half more: the risk averse investor captures her extra 20% of value from the Tsy, while banks capture precisely the risk neutral value of assets. in effect, if we (unrealistically) presume this kind of perfect calibration of option value and risk haircut, Tsy is paying risk-averse real investors to behave like risk-neutral investors by covering the haircut they would usually impose for risk.

    — @jkh re franchise value, note that in hempton’s post, the “accrual value” can be split between the cash flows from the existing portfolio and cash flows from new business, which are good because bank borrowing rates are pushed low while credit spreads are high during a crisis. that’s why citi could gloat about the past couple of months. a PPIF entity does have the first kind of “accrual value”, ie cash flows from the preexisting portfolio, but not the second, which is likely to be the more important component if legacy assets are impaired. i think i’m agreeing with you, there is an accrual value to the assets that is some subset of the accrual value of an ongoing bank, but the difference between the subset and the total value represents a large “cost of exercise” to a bank rather than an investment fund that makes use of the non-recourse option.

  64. JKH writes:

    SRW,

    Yes, we’re much closer now.

    It does raise the question of whether or not a well run investment fund could successfully manage some degree of asset turnover using the same funding base – sort of a “franchise-light” approach to emulating the banking effect.

  65. James Surowiecki writes:

    “re thought experiments, that Tsy/FDIC loses under PPIP even if investors on average don’t overpay at all (as long as assets are purchased in many separate deals). if investors price bank assets accurately, on half the deals, they earn a profit, but on half the deals they take an equivalent loss, a part of which is eaten by the government.”

    Right — my point was simply that the scenarios which show the government taking massive losses all also involve the private investors overpaying by a lot. They also (necessarily) imagine scenarios in which the value of the underlying asset is very volatile (in Steve’s sense). I think the extreme values in these scenarios are unrealistic, but Steve made an excellent case — way way back up the thread — that mortgage values are in fact very uncertain.

    “in effect, if we (unrealistically) presume this kind of perfect calibration of option value and risk haircut, Tsy is paying risk-averse real investors to behave like risk-neutral investors by covering the haircut they would usually impose for risk.”

    Exactly. I assume this is how Treasury explains the plan to itself.

  66. Alex writes:

    “it’s worth emphasizing, re thought experiments, that Tsy/FDIC loses under PPIP even if investors on average don’t overpay at all (as long as assets are purchased in many separate deals).”

    It’s absolutely crazy for the FDIC to be short a basket of options on loans (1 on each loan for each purchase) rather than being short an option on a basket of loans (i.e. have the total loan made to a given investor supported by all of that investor’s purchased bonds).

    Unnecessary and ridiculous.

  67. mencius writes:

    BTW, James, some of us do have exactly the same opinion of “deposit insurance.”

    You are right that the PPIP and the regular activities of FDIC are logically parallel. But you are wrong to assume that this makes them both healthy. Have you considered the possibility that they’re both pernicious? If not, I recommend a visit to your local specialist in Austrian economics, toot sweet.

    As the above discussion makes abundantly clear, “deposit insurance” can be modeled as a put option granted to bank “depositors,” ie, creditors. It allows the depositor to sell his “deposit,” ie, IOU, to FDIC at par. This valuable option increases the price of said IOUs, so it can be described as an indirect subsidy to banks.

    However, there is an even simpler way to model “deposit insurance.” In this interpretation, the actual transaction is triangular: the “depositor” lends to FDIC, and FDIC lends to the bank. Each of these is a plain vanilla loan with normal recourse provisions. But FDIC, since it has implicit command of an intaglio press, will never, ever default.

    The purpose of “deposit insurance” (both words are Orwellian distortions, because “deposit” in English implies a warehouse receipt rather than a loan, and “insurance” in English implies an insurable risk without moral hazard), is to induce a transaction that would simply not happen in a free market.

    This is the transaction of maturity transformation, in which banks borrow short and lend long. In a free market without “deposit insurance,” the original lender (“depositor”) would not lend to banks which practiced MT.

    Reason: said lender would notice that there is no actual physical process that can convert a 2019 dollar into a 2009 dollar. Therefore, the borrower (the bank) cannot actually make good on all its 2009 commitments – except by selling the loans on the open market, ie, finding new borrowers. If this reminds you of Bernie Madoff, perhaps it should.

    Without an LLR such as FDIC, maturity transformation is no more than a market-manipulation scheme in the market for future money, which drives the price of 2019 dollars up by injecting that market with demand for 2009 dollars. With the LLR, this scheme is stable. Without it, it isn’t.

    Thus the collapse of the shadow banking system, in which there was no LLR. From an epistemological perspective, the error was in the assumption that MT is a free-market economic practice. It ain’t. It requires an LLR who either can print money, or has one heck of a lot of it, or can pull other nasty tricks such as suspending bank liquidation.

    (In the era of Bagehot, the Bank of England used to serve as an LLR for gold. This always took some serious stones, and it also took some serious gold. After WWI, the stones remained but the gold was not sufficient in the light of war debt. Which is basically why we have paper money today.)

    Thus deposit insurance is a camouflaged government loan which is designed to achieve the usual results of a government loan, namely, lowering interest rates for politically favored borrowers. This corrupt transaction has been around since Jesus was a little boy, and probably well before.

    The fact that it is deeply embedded in the Anglo-American financial system is not, I feel, a good argument for it, since the Anglo-American financial system (dating from the establishment of the Bank of England, 1694) has been producing cycles of boom and bust, panics and manias, etc, for its entire freakin’ lifetime.

    These are easily explained by the yield curve instabilities produced by the creation and collapse of MT schemes – ie, the multiple equilibria of a bank run. (Non-Austrians actually do understand this stuff, sort of. Google “Diamond-Dybvig.”)

    In a free-market financial system, there is one private lender of $X at term T for every private borrower of $X at term T. Rocket science this is not. If we would rather indulge in 17th-century Rube Goldberg financial-engineering schemes, we probably deserve exactly the results we’re getting.

  68. James Surowiecki writes:

    “It’s absolutely crazy for the FDIC to be short a basket of options on loans (1 on each loan for each purchase) rather than being short an option on a basket of loans (i.e. have the total loan made to a given investor supported by all of that investor’s purchased bonds).”

    It does seem like it’s going to be weird, if you’re the FDIC, to watch an investor walk away from a $20 million hit from one of his PPIPs even while he’s making $30 million on another.

    I just went back and read the Summary of Terms again, and it doesn’t explicitly say that investors will be allowed to invest in more than one PPIP. But it does seems as if each eligible Asset Pool will be owned by an individual PPIP, and that it’s the partnership, as I said earlier, that’ll take out the loan. So it sounds like a one-asset/one-loan system, with investors allowed to take out multiple loans. But that’s not explicit.

  69. oops. i made a bit of an embarrassing mistake in my previous comment (although it doesn’t affect the qualitative points). in the risk-averse case, if the option would be worth 20% of the asset’s value to a risk-neutral investor, it would actually be worth much more than that to a risk-averse investor, because the value of the put and the value of the asset are (very) negatively correlated.

    qualitatively this doesn’t matter — the presence of the option still offsets some but not all of the risk aversion haircut relative to what a risk neutral investor would pay for the bundle, but quantitatively it matters a great deal. it reduces the degree to which Tsy would have to provide leverage in a good faith attempt to try to offset risk aversion without provoking overpayment relative to the hold-to-maturity asset values.

  70. JKH writes:

    SRW,

    Isn’t a component of the put option premium paid as a risk premium on the rate paid for debt funding (i.e. the premium over the risk free rate)? This would be present valued of course. If so, the total option premium would include the funding premium plus an asset premium (for the risk neutral case, and so on). And the subsidy from Treasury would only include the asset premium.

  71. JKH writes:

    Or is the debt funding at the risk free rate? Do we know that?

  72. winterspeak writes:

    JKH &SRW: All good points, and excellent discussion. Thank you.

    MENCIUS: I think it is helpful to think of FDIC insured deposits as sitting directly on USG books. Certainly, a fraction of them do sit on Government books as reserves.

    JAMES: You’ve made me think harder about what bank deposits actually do. They don’t fund lending, and they aren’t a big part of the system, so I’ve tended to ignore them. It’s also true that banks compete for them (sometimes) so they must have some value.

    Given that I accept my ignorance in this area, should I back down from my broader assertion: that the FDIC/PPIP parallels does not mean that PPIP is fundamentally more of how banking already runs? Maybe.

    The FDIC guarantee is enjoyed by the depositor when the bank goes bankrupt, but it’s possible that any other actor in the system may enjoy the benefits of it at other times. Just as tax incidence does not always fall on the one who is taxed, the value of guarantees may not always be enjoyed by the one whose money is guaranteed. But bank balance sheets fundamentally are not built on loans guaranteed by the Government (FDIC deposits). Tier 1 and Tier 2 equity is not FDIC insured, and it’s what counts for capital requirements.

    PPIP keeps equity in first loss position, but actually guarantees *all* the debt used for leverage, not just a tiny sliver of it. Leveraged lending vehicles exist *without* Govt insured debt. PPIP would not exist without them. This is a material difference.

  73. JKH writes:

    Mencius,

    “The infamous “non-recourse loan” does not directly affect the risk-return profile of the PPIP buyer, because once his equity is wiped out, he does not care whether the asset is sold on the open market (as in a conventional margin loan) or directly to USG. Thus, this “option” is of zero value to the PPIP buyer. Wiped out is wiped out.”

    I’ve been seeking clarity on the precise meaning of non-recourse since discussions began weeks ago. You’ve provided this in the case of the type of non-recourse people refer to when thinking about FDIC backing. I agree with your point.

    The other type of non-recourse as I understand it is the fact that the debt funding is collateralized. Is there any distinction here on this point with respect to risk analysis, compared to the “other” type of non-recourse in the sense of FDIC? I simply haven’t been able to think it through yet.

  74. James Surowiecki writes:

    “You are right that the PPIP and the regular activities of FDIC are logically parallel. But you are wrong to assume that this makes them both healthy. Have you considered the possibility that they’re both pernicious?”

    I have considered it — Steve advocated getting rid of deposit insurance a week or so ago, I think. And I don’t think there’s any doubt that it subsidizes bank profits and encourages risk-taking. I just think that bad as things are now, they were worse before 1933, at least in the U.S.

    In any case, the reason I have been emphasizing the analogy has mostly been because there are lots of people out there insisting that the PPIPs will radically distort market prices or that they represent “robbery,” even as they seem to have no problem at all with deposit insurance. (Paul Krugman is against the PPIPs, but thinks the government should guarantee all bank debt as part of nationalization –how do those things go together?) I don’t think that makes sense — even accepting the various ways in which banks do have more constraints on their behavior than the PPIPs do.

    I actually think it’s perfectly plausible to look at the analogy I’m making and draw exactly the opposite conclusion — I’m saying, crudely, “the PPIPs are getting FDIC-insured loans, just the banks do, so it’s not that crazy a plan,” but you can just flip it on its head and say “since it’s obviously a terrible idea to give the PPIPs access to FDIC-insured loans, we can see that it’s a terrible idea to give banks access to them, too.” I don’t agree with that, but depending on your assumptions, I think it’s a reasonable conclusion to draw.

  75. mencius — a couple of points (re your first comment):

    the question of whether there’s a subsidy or not is more a quantitative than a qualitative question. i don’t think it’s really debatable that a non-recourse loan is equivalent to a full recourse loan plus a financial option, and that financial option values are always greater than zero. but to be a subsidy it is not sufficient to point out that a borrower gets a valuable option. you also must show that the borrower isn’t paying for it.

    as you suggest, nonrecourse financing is ubiquitous in ordinary affairs. any lending to a limited-liability entity is nonrecourse. formal borrowers price the nonrecourse option into the interest rate they charge, and limit the value of the option by imposing covenants, in effect regulating the firm like a private FDIC. i’d argue that limited liability organization does extract a subsidy, not from taxpayers or organized lenders, but from another dispersed group, informal creditors of a firm in the ordinary course of its business. outside of an economics lecture, i don’t think it’s reasonable to suggest that your mom factored in the expected cost of default when she purchased that circuit city gift card for you a year ago. not all subsidies are bad subsidies — i think offering limited liability forms of organization has benefits that exceed its costs for nonfinancial firms.

    but the point is, it’s not the limited liability or the nonrecourse loan that determines the subsidy, it’s how much creditors charge relative to how capable the borrower is of maximizing option value by creating volatility, and how costly the option is to exercise.

    i do think the option value PPIP nonrecourse financing will represent a subsidy, and a large one. but that’s not because it’s a nonrecourse arrangement. it’s because the uncertainty surrounding the value of the assets lent against is high, the cost of exercise is low, and i don’t think the FDIC will charge an adequate premium for guaranteeing the loans. i could be wrong: it is ultimately a quantitative question.

    (and since i don’t know the quantities involved, PPIP will be a lot of custom tailored deals, and FDIC hasn’t published specific information on how it will determine guarantee fees, i’m really concluding based on a theory of the incentives and goals of the policymakers shaping the program. i think they want to offer a subsidy to recapitalize existing bank organizations, and PPIP came to be as a means of doing so that would be both effective and politically acceptable.)

    it’s an excellent point that non-marginability is an important benefit relative to market-available terms in and of itself. if i were designing PPIP, I would offer full recourse, nonmarginable loans at near Treasury rates. that would eliminates reduced pricing due to illiquidity and mark-to-market path risk. but it would require investors to price the true distribution of hold-to-maturity asset values rather than a truncated distribution. it would imply a moderate subsidy to investors, in the form of artificially low borrowing rates, but that’s something Tsy can easily finance at little cost or risk to taxpayers. (The gov’t would still earn a small positive spread.) Again, the issues become quantitative. (This was really Hempton’s original point.) Are the benefits and liabilities of the arrangement reasonably priced, from taxpayers’ perspective? i think no with PPIP, partly because the nonrecourse option will be underpriced.

    it’s not often that i’d accuse you of being insufficiently cynical, but i think you are insufficiently cynical about gaming. already citi and boa are buying assets that a week ago they wouldn’t touch, now that they think this program is a done deal and they’ll have special access to a deep-pocketed dumping ground.

    i think that both private and public actors in the financial world have not unlearned the view, which has been profitable for them for a long time, that with clever financial and legal engineering, they can work their way out of any problem, and make a lot of money to boot. if serious public scrutiny were a binding constraint, the Fed wouldn’t get away with hiding losses in SPVs, and the administration wouldn’t get away with making large fiscal commitments without Congressional authorization by squirreling them away as contingent liabilities to FDIC. a lot of clever people think that that the way to deal with scrutiny is to get cleverer, and it has worked out very well from them.

    the scenario i presented earlier may be overly conspiratorial, but there is a huge subsidy on offer to the consolidated financial sector if they can find some way of coordinating subject to scrutiny constraints. i think this is exactly the sort of problem the consolidated financial sector is good at solving.

    despite all the talk of a subsidy in the nonrecourse option, i think the private investors in PPIP will on average take a large loss. the larger the losses (by virtue of overpricing) to explicit investors in PPIP, the larger the aggregate payoff to the consolidated financial sector. i wouldn’t touch a PPIP-investing “PIMROCK” mutual fund with a 2.3 trillion foot pole.

    btw i think that margin arrangements usually aren’t formally nonrecourse. that is they are collateralized lending arrangements, but if you ave a really bad day at the market and lose more than your equity, they can still come after you for the rest. like a lot of collateralized lending, it may be implicitly nonrecourse in that the difficulty of extracting the difference between the collateral value and what is owed usually prevents much action, but still that raises the cost of exercise since your credit can get killed. securities margin lending is a pretty great deal for the lender: it’s usually lightly leverd (no more than 1:1), they have control rights to protect their interest (your own personal FDIC), and the cost of exercising a default option is high. nevertheless, retail brokerages often charge obscene rates for the privilege.

  76. James Surowiecki writes:

    “it’s an excellent point that non-marginability is an important benefit relative to market-available terms in and of itself. if i were designing PPIP, I would offer full recourse, nonmarginable loans at near Treasury rates. that would eliminates reduced pricing due to illiquidity and mark-to-market path risk.”

    This gets at one of the more interesting, and least discussed, aspects of the plan, which is that, for those assets that are sold, it effectively replaces MTM accounting with hold-to-maturity accounting.

    “already citi and boa are buying assets that a week ago they wouldn’t touch, now that they think this program is a done deal and they’ll have special access to a deep-pocketed dumping ground.”

    Steve, I’ve read various people suggesting that these purchases are somehow about gaming the Geithner plan, but I can’t figure out how that’s supposed to be happening. In the first place, if you accept — even to a limited extent – the deposit insurance/PPIP loan analogy, then all the purchases (Citi’s on one end, and the PPIP on the other) are being made with subsidized/insured money, so I don’t see where the meaningfully big arbitrage opportunity exists. More important, once Citi and BoA buy these assets, they can’t say that they’re not trading, and therefore need to be sold to a PPIP in order to establish a market price. The assets were bought in a market, which means that for these assets there must be liquidity. So I have a hard time believing the FDIC is going to willingly subsidize the purchase of these assets two weeks after Citi bought them.

  77. James —

    just clarifying — i did argue for getting rid of deposit insurance, but only along with getting rid of (non-“narrow”) banking. architecturally, i’d prefer the US banking system circa 1936 to the US banking system circa 1929 or 2006. but i think we can do better than any of those.

    i have no philosophical problem with subsidizing private risk-taking or capital formation, if the diffuse benefits of those activities outweigh the subsidy cost. i just think that the state-guaranteed-highly-levered-risk-investment-fund model as a means of subsidy is particularly prone to capital allocation error, catastrophic breakdown, and corruption. again, we can do better. but we’d have to overcome a pretty entrenched incumbency problem.

  78. James — Why would Citi and BoA do this if it’s all insured money?

    It’s not all insured money. There are hedge funds holding these assets who cannot sell them. Citi now can buy them at a substantially higher price than they could last week, with some assurance that they will have access to a buyer willing to pay yet more. The people at Citi can model option values pretty well: they can offer say half the PPIP option value over last week’s bid to the hedge find holder, and expect to capture the other half from PPIP buyers.

    Even when the money is insured, the cost of exercising the insurance option creates reason enough for trades to happen. Sure, the depositors of a small bank holding these assets are gonna be okay even if loans sour on the bank’s books. But the employees of the firm may prefer to avoid an FDIC receivership. If half the option value is goes some way towards to fixing a whole in smallbank’s balance sheet, while watching assets decay means a Friday night special, smallbank has an incentive to sell.

    The argument (originally from the Economist’s blog?) that if Citi or BoA purchase them, then obviously they trade, and if they trade, they’d be ineligible, seems a bit obtuse to me. It’s possible that a really unfriendly regulator would claim that any recent private transaction meant functioning secondary markets for an asset, but that’s not actually a reasonable view, and I don’t think our regulators are that unfriendly. Regulators could prevent this by requiring individual assets eligible for sale to have been purchased prior to say Jan 1, 2009. But I don’t think they want to — I think they’re glad to let the public purse be used to recapitalize financials broadly, especially since doing so earns a handy profit for the core banks they are trying to help.

  79. RueTheDay writes:

    winterspeak said: “RueTheDay: Yes, sometimes banks transfer deposits, and receivables to other banks. The point is that the deposit, wherever it is held in the banking system, was created as the consequence of a loan, wherever that is held in the banking system.”

    Your post to which I responded made the claim that banks can arbitrarily expand their balance sheet by creating deposits and loans simulatneously. In other words, assuming a 10% reserve requirement, if someone makes a $100 deposit, the bank can immediately make a $1000 loan. That really is not how it works. The banking system as a whole can transform $100 in base money into $1000 in deposits by making loans, but individual banks do not operate that way precisely becaus loans will be made to customers of other banks necessitating transfers of reserves.

    Also, you said:

    “The banks make all the loans they want, and then see how much in reserves they need to post with the Fed. If they are short, they borrow the extra they need from the interbank market (at the federal funds rate). If the banking system as a whole is short, it borrows from the discount window.”

    While yes, that is broadly true, it is constrained by the Fed’s willingness to inject additional reserves into the banking system.

  80. RueTheDay writes:

    winterspeak – Also, while the majority of deposits are created by banks making loans, technically, any time a bank spends money on anything it is creating a deposit. When a bank pays one of its employees, it credits his account, thus creating a deposit. Unlike a non-bank.

  81. James Surowiecki writes:

    “a really unfriendly regulator would claim that any recent private transaction meant functioning secondary markets for an asset, but that’s not actually a reasonable view”

    Well, maybe. But the Post article — the source for all these speculations — doesn’t really make them sound like private transactions. It talks about competing bidders, says that most of the buyers (not sellers) are hedge funds, and talks about the trajectory of prices in this market over recent months. Now, I have no idea how accurate this is, or where the Post is getting its information. But it doesn’t quite sound like a story about private transactions.

    More to the point, the assets they’re talking about couldn’t, if they were re-offered for sale, be bought with the 6-to-1 non-recourse loans, because they’re MBS. So I really don’t see where the arbitrage opportunity exists, since the banks’ purchases will be far more highly subsidized than the PPIPs’.

  82. RueTheDay writes:

    JS said: “Steve, I’ve read various people suggesting that these purchases are somehow about gaming the Geithner plan, but I can’t figure out how that’s supposed to be happening.”

    If the percentage differential between the price at which the asset is carried on the bank’s books and its true market value exceeds the private investors’ required equity under PPIP, then it makes sense for the bank to set up an SPV to buy the asset from the bank, because by so doing, it limits its potential loss to the equity fronted. E.g., asset is booked at $100 and is really worth $60, bank sets up SPV to buy asset with $5 equity (Treasury also puts up $5 equity, remainder financed by non-recourse loan), bank now takes a loss of $5 instead of $40, taxpayers eat the rest of the loss.

  83. James — I think that “maybe” is about right. There’s always been the form of secondary markets for these assets. There have been low speculative bids and high asks, and few trades crossing. Of course purchasing banks are putting higher (probably substantially higher) bids to win the auctions. The question isn’t so much whether there had in recent memory been sales of comparable assets, but whether the markets are broadly liquid, whether trades are taking place at privately agreed prices with some regularity. Obviously that’s a judgment call. The occasional trade can be written off as a fire sale. But if the assets are trading in some sense all the time, then it’d be hard to justify to including them PPIP if you claim its purpose is to find market prices. In the end, it is Treasury and Fed who will determine what assets are eligible. If the cynics are right about what’s motivating these purchases, it seems likely that the banks will have clarified with those agencies what assets would qualify.

    You make a good point that these assets shouldn’t be eligible for the loan selling program. They would instead be eligible for the securities selling program. That program is much less discussed and analyzed, because the leverage subsidy is most obvious in the loan program. The securities program permits 1:1 leveraging with the Treasury, but it also permits unspecified further leveraging up at the Fed via TALF. Also, the securities program is the one most liable to overpayment due to conflict-of-interest: a few, handpicked fundmanagers will participate who might hold substantial debt (though not substantial equity) of the selling bank.

    It’s certainly not impossible that Citi / BoA participation in these markets is unrelated to these programs. What their legitimate purpose would be seems questionable, though. Suppose Citi and BoA really think they are just good assets, and are trading on their own account. Ordinarily, they’d not be permitted to do that, it’d be taking a risky bet while arguably near insolvent, prototypical “gambling for redemption” that any formal regulatory action would try to prevent. BoA’s claimed justification of breathing life into the mortgage market is unpersuasive: BoA has no legitimate role, especially while on public support, of fixing broken markets just because. If it’s acting in the public interest, it should be in coordination with regulators. If it has a legitimate profit motive besides subsidy extraction or gambling, it has been coy about offering it up. Maybe it is protecting the identity of fully private clients on whose behalf it is making purchases, or something like that. Sometimes when there’s smoke, it’s because someone has blown out a candle. But sometimes something’s burning.

  84. James Surowiecki writes:

    “If the percentage differential between the price at which the asset is carried on the bank’s books and its true market value exceeds the private investors’ required equity under PPIP, then it makes sense for the bank to set up an SPV to buy the asset from the bank, because by so doing, it limits its potential loss to the equity fronted. E.g., asset is booked at $100 and is really worth $60, bank sets up SPV to buy asset with $5 equity (Treasury also puts up $5 equity, remainder financed by non-recourse loan), bank now takes a loss of $5 instead of $40, taxpayers eat the rest of the loss.”

    Can’t be done. No non-bank investor in a PPIP can be affiliated with a selling bank, and no bank can have more than 9.99% of a PPIF that it’s offering assets to.

    “Sometimes when there’s smoke, it’s because someone has blown out a candle. But sometimes something’s burning.”

    No doubt.

  85. RueTheDay writes:

    “Can’t be done. No non-bank investor in a PPIP can be affiliated with a selling bank, and no bank can have more than 9.99% of a PPIF that it’s offering assets to.”

    Right, and there’s no way I was doing 80 on the NJ Turnpike on the way home from work tonight because the sign says “Speed Limit: 65”.

    I was using a simplified illustrative example. If you don’t think the big banks are smart enought to come up with ways around the law you’re extremely naive.

  86. babar writes:

    interesting discussion

    > Disagreement, to me, is just one way high dispersion of outcomes reflects itself. I say that loan is worth 20¢. You say it’s worth 80%

    well, the loans _could_ actually be worth different amounts to each of us. there might be several reasons for that. i think that two are usually left out of the discussion:

    — if i am a TBTF bank, the embedded put option is already there so this asset IS worth more to me than to private equity. this program gives private equity access to that same put option, essentially.

    — we might want to be paid differently for risk capital. when a lot of these securities were first made up, people had easy access to cheap leverage and were fine with a 5% return. that leverage is gone and and so is most of the risk capital in the market, and now they want 20% return. this 5% to 20% difference over 3 years counts a lot for the difference in price.

    note that SRW and others have talked a lot about the VOLATILITY of the asset — i don’t think that is the issue. it is the R, not the sigma, that is causing valuation differences at the moment.

    for me the big problem with PPIP is that some of the buyers are also major bondholders; they would essentially be paying themselves and be incentivized to do so. this ruins what pricing information there is.

    SRW — is there a good reason why the text input window on your site is so blastedly small?

  87. babar writes:

    @SRW: Tsy is paying risk-averse real investors to behave like risk-neutral investors by covering the haircut they would usually impose for risk.

    That’s basically what I just said above. So you got it.

  88. babar writes:

    > despite all the talk of a subsidy in the nonrecourse option, i think the private investors in PPIP will on average take a large loss.

    if this happens (as it might if the underlying loans are worse than we expect, which might happen if the economy is worse than we expect) then this program will help the banks or treasury or both — probably both. why? because private equity is taking the first loss. if they are throwing money in and losing it, that takes away from money that the other two parties will lose. on the other hand, if they make money, the reverse is true.

  89. jimbo writes:

    RTD-

    While yes, that is broadly true, it is constrained by the Fed’s willingness to inject additional reserves into the banking system.

    Not true. The Fed (under normal, non-zirp circumstances) does not have discretion about how many reserves it adds or subtracts from the system, so long as it wants to maintain a FF target. It must supply as many reserves as the system is demanding, or interest rates will climb to infinity (since the system cannot remedy a general shortage itself).

    WS –

    I think you may be getting lost in the wilderness, a bit. (Happens all the time people new to “the paradigm”…). Just because loans create deposits and banks are free to expand their balance sheet in both directions does not really mean that they create money completely out of “thin air” – they borrow money in order to lend it. The trick is, the loan creates, in the deposit, the very money the bank is then borrowing back, so it balances out (as Mosler said once, it’s not some special law that gives banks the right to create money – it’s just accounting). But if I take my new loan and write a check on it and that deposit ends up in another bank, it has to borrow that money from somewhere else to balance it’s books. Because they’re tied into the Fed payments system (and thus must submit to bank regulation about what they can keep on their books as assets, they can always borrow against their assets at the current rate.

    Why do banks compete for deposits? Because it’s cheap money – they pay essentially nothing to borrow deposits, while borrowing on the FF market or from the discount window will cost them whatever the current rate is. If you can fund your loans with a lot of deposits, you can make a better spread on your loans and make more money. That’s pretty much it.

  90. James Surowiecki writes:

    “if this happens (as it might if the underlying loans are worse than we expect, which might happen if the economy is worse than we expect) then this program will help the banks or treasury or both — probably both. why? because private equity is taking the first loss.”

    Actually, they’re splitting the first loss with Treasury. Profits and losses for the PPIP are split proportionally between Treasury and the private investors. Private investors do lose everything before the FDIC loses a cent, though.

  91. babar writes:

    @James S: I think that is correct. I left out a step though. I think that if losses on “legacy” assets and loans are severe it’s likely that banks will be nationalized after this program and the assets thereby will be brought onto the treasury balance sheet. if that happens, the bank’s losses become the treasury’s. (I skipped a step, and assumed bondholders wouldn’t shoulder the complete loss. Sorry.)

  92. winterspeak writes:

    RueTheDay: No, we do not agree. A bank with $100 in deposits may generate any amount of loans, depending on how much capital it has. Reserve requirements *do not* constrain lending, despite popular belief otherwise. Spend time on Warren Mosler’s site to see how this works.

    Banks do not transform “base money” (reserves) into deposits by making loans. It is the opposite. Banks make loans, out of nothing, subject only to capital constraints, creating deposits as they do so. Reserve requirements make them post some of those deposits as reserves at the Fed. If a bank finds itself short, it borrows what it needs on the interbank market from another bank that has excess reserves. If the system as a whole is short, it borrows what it needs from the discount window.

    If the Fed shuts the discount window, and the system is short on reserves as a whole, then the interbank market cannot work and I guess the Federal Funds rate goes to zero as there is no longer a bid.

    The causality you have working all the way through is backwards, I’m afraid. You can have a perfectly normal banking system with zero reserves. Canada has one. jimbo is right.

    JIMBO: Thanks for setting me straight on the role of deposits!

  93. RueTheDay writes:

    jimbo: “Not true. The Fed (under normal, non-zirp circumstances) does not have discretion about how many reserves it adds or subtracts from the system, so long as it wants to maintain a FF target. It must supply as many reserves as the system is demanding, or interest rates will climb to infinity (since the system cannot remedy a general shortage itself).”

    The key phrase being “so long as it wants to maintain a target”. The FF rate is not fixed long term. Increased loan volume and thus increased systemic reserve demand will occur as economic growth increases and then the Fed will tighten monetary policy.

  94. RueTheDay writes:

    “RueTheDay: No, we do not agree. A bank with $100 in deposits may generate any amount of loans, depending on how much capital it has. Reserve requirements *do not* constrain lending, despite popular belief otherwise.”

    Of course reserve requirements constrain lending, UNLESS the Fed is willing to inject additional reserves into the system. Granted, in the short term at least, the Fed will be willing to do that to maintain the FF target rate. But it is incorrect to say that reserve requirements do not constrain lending. In effect you are simply saying that the Fed cannot simultaneously determine both reserve levels and the FF rate, that’s all.

    “Spend time on Warren Mosler’s site to see how this works.”

    Mosler is a crank.

    Yes, money is largely endogenous to the system, but the pure horizontalist view is just as untenable as the other extreme.

    “If the Fed shuts the discount window, and the system is short on reserves as a whole, then the interbank market cannot work and I guess the Federal Funds rate goes to zero as there is no longer a bid.”

    It wouldn’t go to zero, it would actually increase as a result of the scramble for reserves, but at the end of the day some institutions would come up short and effectively be put out of business.

    “The causality you have working all the way through is backwards”

    No, I just recognize that the causality actually runs both ways. Neither supply nor demand has primacy.

  95. jimbo writes:

    RTD-

    You yourself just said: if the Fed refuses to provide reserves, it will force otherwise viable institutions out of business (and drive interest rates to sky-high levels) simply because they were the last one the scrum for reserves. Volcker listened to the crank Friedman back in the 80s and tried to run the Fed this way – he was forced to give up on it, since it’s not really possible as a practical matter.

    Think of it like a electric utility: all power comes from the utility, and in theory, the utility could choose to either provide a set amount of current (reserves), or set the voltage (interest rate) and let the amount of current supplied float. But practically, only setting the voltage works – since the wildly fluctuating voltage would destroy every device connected to the grid.

    So yes, reserve requirements could, theoretically, constrain lending. But they don’t, and can’t in the real world.

  96. RueTheDay writes:

    Jimbo – You are conflating two somewhat orthogonal arguments here. One is whether reserve targeting represents effective central bank policy, the other is whether a statutory reserve requirement imposes a constraint on the amount of loans a bank can make. On the former, I agree that interest rate targeting has proven far more effective that monetary aggregate targeting. On the latter, reserve requirements absolutely impose a constraint on lending activities – it just so happens that the Fed is willing under the present regime to inject reserves into the system to ease that constraint in the interest of pursuing the former policy. Let us not forget that the banking system structure as it exists in the US today is not the only possible structure nor is it the only one in existence or to have existed in the history of the US.

  97. winterspeak writes:

    RTD: You’re wrong here, but you certainly have conventional wisdom on your side.

    The Fed sets the FFR target, and then engages in OMO to meet that target. It issues Treasuries to drain excess reserves, and thus creates (and influences) the interbank market to hit its FFR target. Yes, this is a complicated way to set the FFR, but it is what happens in the US.

    It isn’t that interest rate targeting is more effective than monetary aggregate targeting, it’s that interest rate targeting happens every day, and monetary aggregate targeting doesn’t happen at all. It isn’t how the Fed works. You see *huge* changes in reserves today — vast excesses — but this is all a side product of interest rate targeting (which now extends beyond short term Treasuries as the Fed has adopted QE, or is soon about to).

    Many different banking systems are possible. Gold standards are possible. But let’s remain clear about what actually happens in this one.

  98. RueTheDay writes:

    winterspeak – I am very much familiar with the Moslerite worldview after having argued with his sycophants on Usenet for many years. Their criticism of the classical “quantity theorists” and exogenous money is correct. But then they go to the opposite extreme and put money at the “center of the universe” and ignore the rest of economics.

  99. winterspeak writes:

    RTD: We’re not talking about all of Mosler’s theories here, just the operational reality of reserve banking. I’m not recommending any extremes, just insisting that we recognize how banks actually work today.

  100. mencius writes:

    James,

    I just think that bad as things are now, they were worse before 1933, at least in the U.S.

    Spoken like a good defender of the status quo! The history of the world before 1933, as I’m confident you’re aware, does not begin in 1929. Since what we’re watching here is basically the slow-motion collapse of the New Deal financial system, if not the New Deal period, it is worth examining this hoary creation myth.

    The old regime of what Keynes called “orthodox economics” did not collapse overnight. Even in the “classical gold standard” of the Lombard Street era, there was never enough gold in the system to meet all current claims. Bagehot said in _Lombard Street_ (certainly worth a reread today) that 10% rates would “draw gold out of the ground.” But after WWI the system was just watered with way too much paper, ie, war debt.

    Furthermore, in the ’20s there was a credit bubble rather similar to the experience of 1982-2007, in which even more debt was injected into the system and used to back claims for present gold. Basically, in the first Federal Reserve period (1913-33) the financial system was misregulated to the point where it was attempting to violate the laws of physics.

    The correct government action in 1933 would have been a mega-devaluation of the dollar while retaining the gold standard, say to $100 or $150 an ounce. This would have rendered the financial system solvent by the simple expedient of making it physically possible for it to fulfill its obligations.

    Instead the result of the policy process, which at the time was a tug of war between those who still followed orthodox sound-money economics and those who believed in the new inflationism of Keynes, Fisher, Gesell, Major Douglas, etc – the legitimate heirs of the Attwoods and the free-silver men, the progenitors of a long chain of soft-money theorists leading down to our own dear Moslers and Bernankes, produced a horrendous result. They retained the appearance of a gold standard, while in fact transitioning to paper money. Over the last 75 years, this innovation has produced exactly the same effects that any historian would have predicted.

    So, yes – in a sense, orthodox finance and economics caused the bank failures of 1933. But the only problem with them was that they were not orthodox enough. If you have a balanced lending system which does not mismatch its maturities, you don’t need “deposit insurance.” Backing claims to present money with promises of future money is just a bad idea.

    The FDIC was created in 1933, but it is only a minor variation on the Anglo-American model of state-sponsored banking, dating to 1694. If you’re not an expert in 17th-century British history, I can assure you that it was just as corrupt then as it is now. Moreover, the simplest way for the government to protect maturity transformers is just to suspend the liquidation of a bank that cannot pay its claims, as was done in the Bank Restriction of 1797. Effectively, Britain was on paper money for the whole of the Napoleonic Wars. There is nothing new under the sun.

  101. RueTheDay writes:

    Mencius – While a “maturity matched” banking system is appealing in theory, how do you make it work in practice?

    For example:

    I deposit $100 into a maturity matching bank for an agreed upon duration of 1 year. They issue me a 1 year CD and make a 1 year term loan to you. What then stops you from taking the funds and going to a second bank and placing it in a 10 year CD? Given a positively sloped yield curve, there’s certainly an enormous incentive to do exactly that. You can howl and protest that this would be made illegal, but HOW?

  102. mencius writes:

    SRW,

    as you suggest, nonrecourse financing is ubiquitous in ordinary affairs. any lending to a limited-liability entity is nonrecourse. formal borrowers price the nonrecourse option into the interest rate they charge, and limit the value of the option by imposing covenants, in effect regulating the firm like a private FDIC. i’d argue that limited liability organization does extract a subsidy, not from taxpayers or organized lenders, but from another dispersed group, informal creditors of a firm in the ordinary course of its business. outside of an economics lecture, i don’t think it’s reasonable to suggest that your mom factored in the expected cost of default when she purchased that circuit city gift card for you a year ago. not all subsidies are bad subsidies — i think offering limited liability forms of organization has benefits that exceed its costs for nonfinancial firms.

    This is extending the term “subsidy” to almost intergalactic proportions. The reason it is traditional for an LLC to include that notation in its name is that both parties to the contract are seen to understand the limited liability, which is just one of many choices of debt seniority structure. There is no third-party interference at all in this freely entered contract.

    Those – including many “left-libertarians” – who argue that limited liability represents some kind of government-granted privilege are just wrong. Limited liability is an ineluctable consequence of freedom of contract.

    but the point is, it’s not the limited liability or the nonrecourse loan that determines the subsidy, it’s how much creditors charge relative to how capable the borrower is of maximizing option value by creating volatility, and how costly the option is to exercise.

    As I said in my earlier comment, since PPIP loans are more desirable than anything the private sector would offer, you can certainly construct a financial model of the value of the difference, and that value is positive. If you want to call that a subsidy, you are perfectly within your rights to.

    However, what I mean by “subsidy” is a program like this: if you buy a dollar of toxic waste, the government gives you a dime. If the toxic waste goes down by a nickel, you still make a nickel.

    The most essential fact about PPIP is that (unlike my dollar-dime proposal) it cannot turn losses into profits. For a PPIP fund to make money, the price of its toxic waste must increase.

    btw i think that margin arrangements usually aren’t formally nonrecourse. that is they are collateralized lending arrangements, but if you ave a really bad day at the market and lose more than your equity, they can still come after you for the rest.

    Good point.

    i think that both private and public actors in the financial world have not unlearned the view, which has been profitable for them for a long time, that with clever financial and legal engineering, they can work their way out of any problem, and make a lot of money to boot. if serious public scrutiny were a binding constraint, the Fed wouldn’t get away with hiding losses in SPVs, and the administration wouldn’t get away with making large fiscal commitments without Congressional authorization by squirreling them away as contingent liabilities to FDIC. a lot of clever people think that that the way to deal with scrutiny is to get cleverer, and it has worked out very well from them.

    Yes. I am not sure, however, that these same tactics will serve them well in a politically charged environment. Enron finance flourished in the dark. It is not dark now, which is one good thing we can say.

  103. mencius writes:

    RTD,

    I deposit $100 into a maturity matching bank for an agreed upon duration of 1 year. They issue me a 1 year CD and make a 1 year term loan to you. What then stops you from taking the funds and going to a second bank and placing it in a 10 year CD? Given a positively sloped yield curve, there’s certainly an enormous incentive to do exactly that. You can howl and protest that this would be made illegal, but HOW?

    This specific example is easily answered – after 1 year, you have to return your loan. To do this, you have to abort your 10-year CD. To do this, you have to pay a penalty for early withdrawal, which makes your transaction unprofitable.

    But this answer is too easy. It relies on the fact that CDs are not negotiable securities. If they were, you could just sell your 10-year CD with 9 years remaining to someone else who wanted a 9-year CD, and again (thanks to that upward yield curve) profit.

    So, for example, you don’t need banks to perform MT. You can do it yourself at home. If you are a person with a demand for current money, just buy a 30-year Treasury, for which perfectly liquid markets exist at all times, and sell it when you need the money. Thus you earn the 30-year return, instead of the zero-percent rate of your friendly local full-reserve “giro” bank.

    The key to understanding MT is to understand that (in the absence of official protection) this transaction is not, in fact, rational for the investor. It is not rational because he is buying into a market-manipulation scheme, and what goes up must come down. If you are the only one doing this, fine; if others are doing it, as you must assume they will, you’ll get an interest rate spike in Treasuries when the scheme collapses, and you’ll lose money by trying to sell your Treasuries on a down market.

    In short, you are not in a Nash equilibrium (your strategy is not optimal given the assumption that others are practicing the same strategy). So, you are not acting rationally.

    Rather, the rational investor in a balanced lending environment will discover his own maturity preferences and structure his investments around them, communicating the aggregate structure of maturity preference to the market in the best Hayekian manner.

    Now, this is quite a big change, just mentally, from our present system. I feel that any such change, unimaginable though it is for a million reasons, should probably be associated with some regulatory reinforcement, to get things moving in the right direction. However, it is self-interest rather than regulation that makes the system work.

  104. RueTheDay writes:

    “Limited liability is an ineluctable consequence of freedom of contract.”

    So is murder for hire.

  105. RueTheDay writes:

    “This specific example is easily answered – after 1 year, you have to return your loan. To do this, you have to abort your 10-year CD. To do this, you have to pay a penalty for early withdrawal, which makes your transaction unprofitable.”

    Not necessarily. You can simply plan on rolling over the 1 year loan. 99 times out of 100, this will work just fine. 1 time out of 100, you will not be able to roll over the loan (or it will not pay to do so because interest rates have increased to a level above the original 10 year loan rate). At that point, we have a crisis. We’re right back to square one, just as in the current system.

    “But this answer is too easy. It relies on the fact that CDs are not negotiable securities. If they were, you could just sell your 10-year CD with 9 years remaining to someone else who wanted a 9-year CD, and again (thanks to that upward yield curve) profit.

    So, for example, you don’t need banks to perform MT. You can do it yourself at home. If you are a person with a demand for current money, just buy a 30-year Treasury, for which perfectly liquid markets exist at all times, and sell it when you need the money. Thus you earn the 30-year return, instead of the zero-percent rate of your friendly local full-reserve “giro” bank.”

    Precisely. Thus proving my point. Just because the initial transaction is maturity matched does not prevent subsequent maturity mismatch from entering the system.

    “The key to understanding MT is to understand that (in the absence of official protection) this transaction is not, in fact, rational for the investor. It is not rational because he is buying into a market-manipulation scheme, and what goes up must come down. If you are the only one doing this, fine; if others are doing it, as you must assume they will, you’ll get an interest rate spike in Treasuries when the scheme collapses, and you’ll lose money by trying to sell your Treasuries on a down market.”

    The adjustment does not happen instantly, it builds over time. Just like in the present system. Maturity matching the initial transaction does not in any way break the boom bust cycle.

    Additionally, your attempt to place the blame on depsoit insurance is counterfactual. We had financial panics and economic depressions long before we had deposit insurance.

    “Rather, the rational investor in a balanced lending environment will discover his own maturity preferences and structure his investments around them, communicating the aggregate structure of maturity preference to the market in the best Hayekian manner.”

    This is little more than a standard appeal to the libertarian faith that in the absence of government we have utopia. History has demonstrated that faith to be misplaced.

  106. mencius — re limited liability and contract: i agree, within a frictionless “rational actors” framework. limited liability organization just sets up a default arrangement that can easily be overridden by freely contracting parties. to the degree there is a subsidy, it is due to ways in which a frictionless rational actors model does not describe the world. informationally limited for whom negotiating explicit contracts is costly may fail to fully incorporate the limited liability default-setting into their arrangements. the libertarian ethos basically suggests that it’s everyone’s duty to live up to a reasonable fascile of homo economicus. i adopt a more paternalistic and “behavioral” view — i think people diverge (and want to diverge) in important ways from home economicus, because it is costly and unpleasant to be that way. under this view, setting up a limited liability default in ordinary business arrangements amounts to a tax on informal, constrained creditors (and sometimes on larger, informal stakeholders as well). i don’t object to this tax in general: it actually works pretty well, since the informal interactors with business (e.g. customers) gain from the entrepreneurship limited liability encourages. all things considered, it’s one of the better taxes out there. i do object to limited liability organization of financial firms, not on philosophical terms, but based on informal cost/benefit analyisis.

    i do object to the suggestion that LL organization is essential to freedom of contract. it just defines a default contractual setting. in a frictionless rational actors world, creditors and firms could arraneg contracts recourse or nonrecourse as they saw fit, regardless of that default setting.

  107. mencius writes:

    RTD,

    Not necessarily. You can simply plan on rolling over the 1 year loan. 99 times out of 100, this will work just fine.

    I fear you are not quite following the game theory here.

    Are you familiar with the theory of market manipulation as applied in commodity markets, namely the “burying the corpse” problem? Basically, you can create an apparent profit in any asset or security by bidding up its price. If FCOJ is trading at 12, and you buy all the oranges in Florida, you can push it up to 20. However, to produce an actual profit, you need to sell those oranges. Or all you have is oranges – not actual money.

    MT is doing much the same thing in the market for loans. By “backing” claims to 2009 money with 2019 money, you are pushing up the price of 2019 money. However, the reason people want claims to 2009 money, rather than claims to 2019 money – ie, the whole reason your yield curve slopes upward – is that they actually want, well, 2009 money.

    Your structure is a Nash equilibrium only if the people you are borrowing your 1-year money from are actually in error, and actually turn out to be willing to lend for 10 years. Otherwise, the structure will collapse.

    Thus the error in your “99 times out of 100.” Liquidity risk is not a matter of probability. It is a matter of game theory. Either your 1-year money is going to get out of 10-year securities, causing a yield-curve hiccup in which the devil takes the hindmost, or it won’t.

    If it does, you lose, and your behavior is irrational. If it doesn’t, your lenders are behaving irrationally, and it is irrational of you to depend on irrational behavior in others. Ergo, your scheme is irrational either way, and it will not be practiced by prudent actors in a free market.

    While the State may feel free to prohibit it, as it may feel free to act preemptively against any pattern of imprudent behavior among its subjects, any regulations in this direction reinforce rather than contradict self-interest. This makes them good regulations which are easy to enforce, because there is no motivation for clever people to try to find loopholes in them.

    Prohibiting systematic MT is a lot like prohibiting Ponzi schemes. It is not rational to participate in a Ponzi, but a lot of people fall for it anyway. MT banking is not a Ponzi, although it has some superficial similarities (I have attempted without success to promote the phrase “Bagehot scheme”). But the rational investor should behave in the same way toward all systematic distortions of market pricing: avoid them.

    This is little more than a standard appeal to the libertarian faith that in the absence of government we have utopia. History has demonstrated that faith to be misplaced.

    For one so critical of kneejerk responses, you sure are quick with them!

    The belief that “in the absence of government we have utopia” constitutes anarchism, not libertarianism. In the absence of government we have Haiti – madness, destruction and death. Libertarians, at least orthodox ones, most certainly believe in government.

    Moreover, I am neither an anarchist nor a libertarian. I am a reactionary. I believe in absolute sovereignty, a la Robert Filmer. I am not just a Tory, but a real, live Jacobite. My general perspective of Anglo-American history is that it’s basically all gone to crap since they cut the head off Charles I. My idea of a good political leader: Strafford, Metternich, Frederick II, Cromwell, Elizabeth I. If you try expounding this line over at mises.org, you will find that it’s not very popular.

    My view is simply that a rational absolute sovereign is like the farmer in the pastorals of Hesiod or Virgil. His country is his garden, and he does not tolerate weeds. But he works with the forces of nature, not against them. He does not try to “drive out nature with a pitchfork.” In Chinese terms, He is a practitioner of wu wei.

    This is easily confused with libertarianism, but the two have nothing in common. To a true libertarian, “classical liberalism” is a moral imperative, not a prudent policy. To me it is just a prudent policy, and it is only a prudent policy in those cases where it does not turn out to be imprudent.

    So I’m afraid you’ll have to find some Jedi mind tricks that work on Jacobites, because yours do absolutely nothing for me.

  108. mencius writes:

    SRW,

    i do object to the suggestion that LL organization is essential to freedom of contract. it just defines a default contractual setting. in a frictionless rational actors world, creditors and firms could arrange contracts recourse or nonrecourse as they saw fit, regardless of that default setting.

    I object to the suggestion too – LL organization is derivable from freedom of contract, not the reverse.

    Also, it is the contract between equity holders and corporations that constitutes the “nonrecourse loan” in the LLC structure. Since an equity holder is just the most junior creditor, all LLC tells you is that no creditor is so senior that he can actually claw back money from more junior creditors! If you loan money to a corporation, you most certainly have recourse against all its assets, subsidiaries, office furniture, etc, etc.

    the libertarian ethos basically suggests that it’s everyone’s duty to live up to a reasonable facsimile of homo economicus. i adopt a more paternalistic and “behavioral” view — i think people diverge (and want to diverge) in important ways from homo economicus, because it is costly and unpleasant to be that way.

    Have you ever actually read Human Action? I know it is quite a tome, but I think you would enjoy slogging through it.

    Homo economicus is a strawman. People have preferences. They reveal those preferences through their actions. Their preferences may be volatile, and they may not correspond to anyone’s paternal sense of the actor’s best interest. But by definition, it makes no sense to say a person’s actions are at variance with his preferences.

    People also make mistakes. They act in ways that a more rational observer would conclude are not likely to achieve their goals. Nothing in aprioristic economics conflicts with or is refuted by this fact.

    This is all standard Misesian logic. Mises would be particularly disgusted by the Robert Shillers of the world, who choose to attribute perfectly rational, understandable and predictable engineering disasters to the mysterious sway of “animal spirits.” As if the Tacoma Narrows Bridge fell down because the people who were driving over it were thinking bad thoughts.

    The multiple equilibrium of a bank runs is not a matter of “confidence” versus “hysteria.” It is a matter of game theory. Moreover, this theory is perfectly intelligible both in the terms of 20th-century mathematical economics (Diamond-Dybvig), and those of 19th-century deductive economics (Mill and the Bullion School, the Austrians, etc, etc).

    Shiller is just another inflationist crank who makes a living by infecting the public mind with self-serving nonsense – like Mosler, but more successful. It is always interesting to compare the “insider” monetary cranks (Keynes, Fisher) with the “outsider” cranks (Attwood, Silvio Gesell, C.H. Douglas – the latter being Ezra Pound’s favorite economist). The policies are much the same, but the insiders have the talent of sounding sane. Sadly, this only makes them more dangerous.

    So: I am not accusing you of being a “behavioral economist.” Since homo economicus is indeed a strawman, there is no shame in running a lance through it every now and then, if just to keep your lance clean.

    However, those who (unlike you) spend most of their time tilting at this strawman typically have some set of policies for which the strawman’s death is intended to provide evidence. These policies typically involve some sort of funny-money scheme. This scheme typically amounts in practice to some sort of corporate subsidy (in your sense of the term). One cannot dissociate oneself too much from these charlatans.

    I linked earlier to John Stuart Mill’s dismissal of Attwood and the Birmingham School, but in fact his entire Of An Inconvertible Paper Currency is well worth reading. The scorn and contempt with which Mill regards the monetary quacks of his day are too little seen in ours.

  109. mencius writes:

    RTD, one more note:

    Additionally, your attempt to place the blame on depsoit insurance is counterfactual. We had financial panics and economic depressions long before we had deposit insurance.

    “Deposit insurance” is only one way for the State to guarantee bank liabilities. There are a thousand others, mainly involving zombification as an alternative to liquidation. Again, have a look at the Bank Restriction of 1797.

    This was by no means a unique occurrence, either in English history or in American. America, in fact, is the Mecca of funky banking and paper money. American finances were dodgy before the Revolution, during the Revolution, and after the Revolution. For example, one of the great achievements of my political hero, Thomas Hutchinson, was his eradication of paper money in Massachusetts – in the 1740s.

    Moreover, the feature that causes panic is MT in the absence of protection, not its presence. In the presence of protection, you just get inflation. The typical trigger for a panic is a situation in which people come to believe that unprotected MT is safe – as in our late, great “shadow banking system.”

  110. RueTheDay writes:

    Mencius is contradicting himself with his appeal to rationality. In fact, he proves my point when he mentions that participating in Ponzi schemes is irrational yet people still participate. The reason people participate is that some people (the first couple of tiers) do make money and as the scheme unfolds over time, it is difficult to perceive where one lies within the structure. The same applies to borrowing short to lend long during a boom – yes, it cannot last forever, and everyone knows that. It’s a game of musical chairs, but if one manages the timing correctly, one can make significant amounts of money, and of course everyone believes they can time it correctly even though most won’t. Thus he cannot claim that the current MT banking system is sustainable for a period of time though ultimately unstable in the long run while claiming that a system that requires initial maturity match but is vulnerable to subsequent mismatch will not experience the same pattern.

  111. RueTheDay writes:

    “Deposit insurance” is only one way for the State to guarantee bank liabilities. There are a thousand others, mainly involving zombification as an alternative to liquidation. Again, have a look at the Bank Restriction of 1797.

    Suspending the convertibility of bank notes into specie hardly qualifies as a guarantee of bank liabilities. A depositor who just wants his money would perceive it as the opposite.

    Moreover, the feature that causes panic is MT in the absence of protection, not its presence. In the presence of protection, you just get inflation. The typical trigger for a panic is a situation in which people come to believe that unprotected MT is safe – as in our late, great “shadow banking system.”

    This I actually agree with. The government should never have allowed the growth of money market mutual funds and the attendant belief on the part of the general public that they were as safe as bank savings accounts.

  112. mencius writes:

    RTD,

    In fact, he proves my point when he mentions that participating in Ponzi schemes is irrational yet people still participate. The reason people participate is that some people (the first couple of tiers) do make money and as the scheme unfolds over time, it is difficult to perceive where one lies within the structure. The same applies to borrowing short to lend long during a boom – yes, it cannot last forever, and everyone knows that. It’s a game of musical chairs, but if one manages the timing correctly, one can make significant amounts of money, and of course everyone believes they can time it correctly even though most won’t.

    And you say I’m proving your point? As far as I can tell, you’re trying to prove mine! Perhaps we could just agree to agree on this one.

    Thus he cannot claim that the current MT banking system is sustainable for a period of time though ultimately unstable in the long run while claiming that a system that requires initial maturity match but is vulnerable to subsequent mismatch will not experience the same pattern.

    The additional variable in the current MT banking system is Fed protection. Unfortunately, being largely informal, this protection is extremely hard to predict – n’est ce pas? By definition, no market can predict the unpredictable. Markets are not magic.

    And this type of official uncertainty, too, is not a phenomenon new to history. Consider the story of Hamilton and the bondholders. It hardly appears as a validation of the efficient market hypothesis.

    Suspending the convertibility of bank notes into specie hardly qualifies as a guarantee of bank liabilities. A depositor who just wants his money would perceive it as the opposite.

    Au contraire, mon frere. Historically, when the government suspends convertibility – even on behalf of a nominally “private” firm, such as of course the Bank of England – it is taken as the touch of sovereign grace. And the formerly convertible paper currency becomes an inconvertible fiat currency.

    The pound during the Bank Restriction was exactly what it is now: a piece of paper. Of course, that piece of paper carried an implicit promise to return, at some time in the indefinite future, to gold.

    And lo, it even happened. Britain was still great in those days. The Crown had just defeated Napoleon. What has USG defeated lately? Can you still be a hyperpower if a bunch of hill-bandits in the ass-crack of nowhere are more than your measure? Is anyone in Washington today competent to carry Castlereagh’s underwear, or Wellington’s, or even that scoundrel Canning’s?

    But I digress. Point is: the path from pure metal, a la Amsterdamsche Wisselbank or the Hamburg silver system, to pure paper was a long one. Even the sordid, sorry saga of the New Deal era is just the last chapter in our multi-century epic of financial ruin and decay.

    And if you look at the character of the prime movers behind this great degringolade, and the actual accuracy of their predictions and structure of their motivations, you will see that in each and every case they are cranks, preachers, liars or thieves. This cannot possibly be a coincidence.

  113. RueTheDay writes:

    Mencius:

    “The additional variable in the current MT banking system is Fed protection. Unfortunately, being largely informal, this protection is extremely hard to predict – n’est ce pas? By definition, no market can predict the unpredictable. Markets are not magic.”

    I don’t think you can attribute the instability to Fed protection. Money market mutual funds never had any sort of FDIC protection until crisis erupted last Fall. The sort of system you propose – with initial maturity matching but no way of subsequently matching maturities would run into the same issue. Another way of looking at it is that there will always be some sort of implicit Fed protection for too-big-to-fail institutions. No government will ever adopt a “come what may, let the heavens fall, we will not intervene on principle” stance in the face of a crisis that threatens to tear down society.

  114. mencius writes:

    RTD,

    I don’t think you can attribute the instability to Fed protection. Money market mutual funds never had any sort of FDIC protection until crisis erupted last Fall. The sort of system you propose – with initial maturity matching but no way of subsequently matching maturities would run into the same issue. Another way of looking at it is that there will always be some sort of implicit Fed protection for too-big-to-fail institutions. No government will ever adopt a “come what may, let the heavens fall, we will not intervene on principle” stance in the face of a crisis that threatens to tear down society.

    You’re making my argument for me again! There isn’t protection, but then again there always is. If this isn’t unpredictable, what is? Who could have predicted that Bear would be saved, Lehman fail, and Citigroup be saved?

  115. RueTheDay writes:

    The point is that none of this would be any different under your (initially) maturity matched system.

  116. JKH writes:

    I left the following comment at Hempton’s. It was intended to respond to a suite of 4 posts, wherein Hempton contends that Paul Krugman is being illogical by supporting bank debt guarantees, while opposing the Geithner PPIF plan. My analysis relates to degrees of recourse; recourse in the case of the Geithner put is effectively more restricted than recourse for the bank put; therefore the Geithner put is more valuable:

    “The implicit assumption of Krugman and the others is that collective net losses put to taxpayers via the Geithner funds may exceed collective net gains for private equity investors. The examples they choose lead to such asymmetric outcomes.

    In such outcomes, the taxpayer subsidy via the Geithner funds is greater than via the bank debt guarantee. These outcomes are possible because the granular structure of the Geithner put means it is generally closer to the money in total dollars at risk than the aggregate bank put:

    If the bank sells the assets, the assets will generally end up being distributed or subdivided amongst multiple Geithner funds with multiple debt funding and collateralization arrangements, and multiple equity funding allocations. Based on the kind of outcome distribution assumed by Krugman and the others, the Geithner private equity investors can make money while the taxpayer loses.

    If the bank retains the assets, they will be supported by the bank’s equity funding. But if the bank is insolvent, and the debt is guaranteed, all remaining assets effectively will be put to the taxpayer. Any assets that would have produced net gains for equity investors in the Geithner model would be used instead here to reduce the net put to the taxpayer. Therefore, the net subsidy in the bank case is less than in the Geithner case.

    This structural difference between the Geithner put and the bank put could be described variously as the portfolio effect, the diversification effect, the correlation effect, the collateralization effect, or the granular put effect.

    It should not be described as a volatility effect per se. Volatility affects both the Geithner put and the bank put. It affects the degree of superior value experienced in the Geithner put. But it is not the essence of the structural difference between the Geithner put and the bank put.

    One potential criticism of Krugman and his ilk in this case is that they choose outcomes to fit their conclusion. But it is precisely the outliers that create the problem and the risk to the taxpayer. So it’s not clear that even this criticism is warranted.

    In any event, they’re being consistent. Therefore, they’re not being illogical.

    The use of the terminology “non-recourse” in these discussions is confusing and unnecessary. The fact that both bank debt and Geithner debt are non-recourse at the fundamental level of the equity investor is irrelevant to a comparison of the differences in these funding alternatives. The fact that both bank debt and Geithner debt have puts to the taxpayer similarly adds nothing to the comparison. The comparison rests on the structure of the put. The put structure is aggregate in the case of the bank. It is subdivided in the case of the Geithner plan. This means the collective outcome is different in the two cases, depending on the constituent outcomes. It is the effect of the subdivision of the put that creates all of the extra risk for the taxpayer in the Geithner plan.”

    Also left at:

    http://brontecapital.blogspot.com/2009/04/

    little-bit-of-careful-thinking-and-why.html