When a rose is not a rose: TAF is not “just” the discount window

Yves Smith at Naked Capitalism and Accrued Interest have both taken very measured views of the Fed’s new Term Auction Facility, which gets its big start on Monday. As Smith puts it, TAF is the “discount window with no stigma and another pricing mechanism”. He also notes that the Fed’s novel device looks just like the same old auctions the ECB has been using as a monetary policy tool for years, and observes pessimistically that “the gap between non-dollar Libor and the ECB’s target rate is higher than the spread between Libor and Fed funds. Not a good sign.” (Bloomberg reports that Friday’s markets shared Smith’s skepticism.) Accrued Interest, responding to my ravings, suggests that TAF is no more a “bail-out” than any other lending at the discount window.

Yves Smith and Accrued Interest are two of the finest financial bloggers out there. If you are dallying here without having read every word they’ve written, your priorities are out of whack. Nevertheless, both are missing a forest of difference among trees of similarity. To call TAF the discount window without the stigma is like calling a person a corpse that is not dead.

In theory, TAF and the discount window are quite similar. In practice, it never mattered what the discount window was, because it was so little used. TAF is the discount window remastered. It is designed to be used, in quantities heretofore inconceivable. James Hamiltion, in an excellent discussion of the facility, shows a graph of historical direct lending by the Fed. Thanks to TAF, the graph through December 2007 will show a spike so tall you’d need a screen three times as tall as a typical monitor just to accommodate. Prior to TAF, the Fed had no effective tool for getting funds to banks in trouble, since loans at the Federal Funds rate are only available to banks that other banks trust, and borrowing at the discount window is expensive, financially and reputationally. With TAF, the Fed can lend directly to the sketchiest banks, and on very generous terms, without spooking depositors. As the ECB’s experience suggests, this funding mechanism can made quite routine. Some commentators view TAF as a temporary means of addressing end-of-year liquidity issues, but I suspect that TAF will be an active part of the Fed’s monetary policy arsenal for the foreseeable future.

If TAF is so similar to the ECB’s facility, why should it matter? After all, the ECB hasn’t succeeded at calming European interbank markets, yet. But, in the words of the maestro (Eric Clapton), “It’s in the way that you use it.” In order to reduce interbank spreads, central banks have to (1) auction money in quantities sufficient to ensure that the “market price” is quite close to the interbank rate policymakers view as healthy; (2) accept a wide variety of bank assets as collateral, and place high valuations on those assets relative to the market bid; and (3) credibly signal that it is willing and able to continue these policies in perpetuity, or until the crisis is abated, whichever comes first. Point (1) immediately drops the cost of borrowing by troubled banks to “ordinary” levels. However, this alone will not immediately reduce spreads on loans are still made between banks. Private banks will still fear counterparty defaults even if the central bank is blasé. Interbank spreads might even widen, as lenders view failure to anticipate and secure adequate funding at auction as a sign of trouble. Point (2), over time, will remedy this. Easy lending against a broad panoply of assets helps ensure that banks have the capacity to bid for all the funding they need. At the same time (as Yves Smith was quick to point out), the price at which assets are accepted as collateral suggests a floor for accounting valuations, helping to assure counterparties that borrowers won’t be thrown into crisis by a sudden write-down. Again, the effect here won’t be immediate — as long as collateral values and market bids are divergent, there will be valuation uncertainty. But, that’s where point (3) comes into play. There’s an arbitrage between exchange value and long-maturity collateral value. If the traders view collateral values as stable, eventually market bids will rise to approach the central-bank set asset prices.

This medicine will take time to work. Central banks will have to establish credible expectations of continued support. Don’t expect “forward looking markets” to reprice spreads until central bankers establish a track record. The ECB is not auctioning enough money yet, but it will. I suspect that, despite Europe’s institutional head start, it will be Bernanke’s radical Fed that shows the way.

Many commentators view TAF as an odd, desperate attempt by the Fed to do something, or to seem to do something, in a desperate situation. That is not my view. I think the Fed is acting quite deliberately here, that it is working out of a playbook that the Chairman developed and described years ago. I am optimistic that the Fed’s approach, if pursued tenaciously, can succeed in undoing the widespread perception of risk and instability in the banking system.

Of course, “optimistic” is not quite the right word here, because I oppose the whole enterprise. All the talk about moral hazard and burning houses misses the point. The financial sector has underwritten gargantuan misallocations of real resources in the United States, and profited handsomely from doing so. The cost of these errors is not just a matter of homes foreclosed, or a period of commercial turbulence and unemployment. We face the prospect of a serious decline in living standards, increased danger of conflict between large powers (which hopefully, but not certainly, will be confined to the economic sphere), and the possibility of social instability of a kind we’ve not witnessed for decades. There are actors in this drama who “deserve” to be punished, under a nineteenth century view of moral hazard. But that is not why we should endure rather than resist a painful restructuring of the financial sector. If we could let bygones be bygones and move on, that’d be great. But the reason for institutional accountability in any setting is not to shame and punish, but to create the conditions under which future actors will not misbehave. Contra Nouriel Roubini, the real economy cannot be “bailed out”. It must be built and repaired via the wise application of present scarce goods and services, rather than salved with promises of future goods and services. We require a financial sector capable of aggregating widely dispersed information into wise choices regarding the use of present resources. We absolutely do not have such a financial sector right now. Which is why the very expensive financial sector that we do have should be let go.

Existing players, including the Fed, won’t see it this way. The Fed will lend to support collateral values at levels higher than conventional valuations would merit. They will work with financial institutions to roll over loans as necessary, even against decaying collateral, rebundling assets to maintain some veneer of plausibility when underlying cash flows develop poorly over time. The proud titans of Wall Street can and probably will be bailed out. The rest of us are not so lucky.

Update History:
  • 17-Dec-2007, 3:42 p.m. EST: Cleaned up / reworded third paragraph quite a bit.

10 Responses to “When a rose is not a rose: TAF is not “just” the discount window”

  1. Bravo! I am a big fan of Yves Smith and Nouriel Roubini but think they missed the boat here. As far as I am concerned, the details don’t matter. What’s going on is something akin to the old federal farm support program with crop “loans”. The ONLY thing of siginificance in the Fed’s action is the collateral it will now accept for loans and what value it will place on the collateral. Inflation here we come! A fine critique!

  2. Ravi writes:

    I’m not sure I follow this. You’re pointing at the collateral the Fed will accept and the valuation the Fed will put on it as a price floor on the price of that collateral. But every time the collateral question has come up with the TAF, I’ve heard that it will take the same collateral as the discount window. So, if collateral policies (and the willingness to keep taking “bad” collateral) are the issue, how can the TAF be qualitatively different from the discount window when it accepts the same range of collateral at the same prices?

  3. IA — Thanks!

    Ravi — You’re right to point to the similarity. Conceptually, there ought to be no difference, and I think that Bernanke originally intended to use the discount window to do the job that is now falling to TAF. (After all, he said he would use the discount window.) My claim is that, when Bernanke cut the discount premium in August and persuaded several big banks to borrow, he’d hoped to resuscitate the traditional discount window, so he could use it to support collateral values. Banks in trouble perceived the costs of discount borrowing — the 50 bp spread, but more importantly the bad press and loss of counterparty confidence — to be not worth the risk. I had expected Bernanke to simply cut the discount spread further (at least against certain collateral), until it was such a good deal that many banks would use it. But, lending more favorably against poor collateral than against good makes the bail-out too transparent, and eliminating the discount premium across the board might have led to borrowing in indeterminate , large quantities, requiring the Fed to engage in constant, massive open-market operations to prevent inflationary reserve growth.

    TAF is a clever way to square this circle. It’s a way of getting people to actually use the discount window, while controlling the quantity of borrowing. If the discount window were usable by banks in trouble, or the discount premium could be dropped without upsetting other aspects of monetary policy, there’d be no need for TAF. But, in practice, TAF is likely to “work”, where the discount window did not.

  4. Ravi writes:

    So, if I’m understanding you correctly, the problem is that the discount window accepts too wide a range of collateral? And that if banks were willing to use the discount window it would have the same problems as TAF?

  5. Ken Houghton writes:

    If the Discount Window returned to its pre-2001 actually being a DISCOUNT to FedFunds (the way it was when BofA, e.g., used it in the 1990s), then the reputational impairment might be worth it.

    But the Discount Window–now named in the manner of Clear Skies and Operation Iraqi Freedom–isn’t a discount, it’s a premium, and still has the attendant reputational risk (and transparency–hit the discount window and your competitors in the market know it, which impacts the prices you are offered (think the limping zebra in the herd, with the lions around it).

    TAF has (supposedly) no transparency, and will allow the same mispricings (Sep 2006 levels) at a lower cost. It’s a win-win-win for anyone holding sh*te to maturity (or, in reality, unable to sell it and therefore claiming its “for investment”).

    It’s a bailout, pure and simple, and will increase the money supply and inflationary pressures even as asset values drop.

  6. jp writes:

    “That is not my view. I think the Fed is acting quite deliberately here, that it is working out of a playbook that the Chairman developed and described years ago.”

    Another interesting link underscoring your point is here:


    Scroll down to the bottom to see the discussion concerning a proposed “auction credit facility”. That was back in ’02.

    “TAF is a clever way to square this circle. It’s a way of getting people to actually use the discount window, while controlling the quantity of borrowing.”

    Don’t forget that it also controls the time for which the loan will stay outstanding. The discount window was typically for overnight loans.

    I agree with you – the TAF will probably be made a permanent fixture at the Fed.

  7. Ravi — “problem” is a matter of perspective, but yeah. What’s interesting about either TAF or the discount window is that collateral values are much higher than market bids. (It’s not the variety per se, the Fed could take mud as collateral without much affecting asset prices, as long as its collateral valuation for the mud was comparable to the market bid for mud.) TAF really is very much like the discount window, and any form of lending would be similarly bad (for those who care about accountability and the importance of uncorrupted asset prices) or good (for those who view stability of the system, with the system effectively defined by large status quo players, as the most important thing). ECB’s recent “auction” in which all “bids” above a certain (low) rate were accepted was really precisely analogous to discount window lending at a low rate. The fact that there were takers (where as there are not takers at the Fed’s discount window) is what matters. If TAF breaks the taboo, and banks acquire a willingness to borrow at the discount window in quantity at some positive spread to Federal Funds (and maintain balances by letting loans rolling over), TAF will have done its job.

    Ken — I didn’t realize that was the case prior to 2001. That’s something I should know about. But your suggestion makes sense, the discount window could have a costly stigma attached, if there were a clear and significant benefit to bearing it. If the discount rate were dropped below Federal Funds, the Fed would have to work to ensure the maintenance of stigma, rather than fight it to avoid lending to all comers. A more transparent way of doing what the Fed is trying to do now would be to drop its lending rate to below FF, but only for actors publicly deemed to be in deep trouble and in need of a bailout, actors whose market equity values would have crumbled and whose existing management close to sacked prior to an infusion of manufactured money. Unfortunately, the current approach is to bail out but hide the bailout, pretending that bankruptcies of mortgage brokers will somehow discipline managers at banks whose hides are being discreetly pulled from the flames.

    jp — Thanks for the link. that is an interesting bit of history. It would be nice to know who proposed a “contingency” auction credit facility, and why.

  8. Steve Gelmis writes:

    Hi Steve, can you please, please consider changing your Blog’s RSS feed setting so that it passes the full text of each post through to the feed reader for those of us who read you that way? It’s really an annoying speed bump to have to click through to the website to read each post.

    Thanks in advance.

  9. Mark Thoma writes:

    I use this feed, and it comes through as full text:


  10. Mark — Thanks for that response… I’ve been away for the holidays.

    Steve G. — Hopefully the full feed is working for you. I am on your side here: I really, really prefer full RSS feeds to tease-and-click operations.