Today seems to be the day for waxing cynical about the prospect of hedge fund regulation. John Carney at DealBreaker writes:

It's a pretty simple formula: regulate an industry and you instantly politicize it. Which is another way of saying that you monetize the industry for politicians...

All the other talk -— about “systemic risk” or pension funds or low-liquidity real estate millionaires -— is just the sound of a policy in search of a rationale. And that policy, of course, is the enrichment of politicians. That's always the policy.

Of course loyal readers (hi mom!) will know that I think "systemic risk" is very real, and that it will hit us all like a rocket-propelled two-by-four, soon. But I quite agree with the cynicism about policy and politicians. So what is to be done?

Here's a simple suggestion. Investment funds should not be permitted to be limited liability entities. As legal entities, they should be restricted to organization as ordinary partnerships.

This isn't really regulation at all — It simply amounts to the state declining to confer the privilege of limited liability to certain kinds of organizations. Limited liability is rife with moral hazard problems, but most people (including emphatically myself) would argue that its advantages far outweigh its disadvantages for nonfinancial businesses.

But limited liability, like copyright, is a legal oddity conferred for a specific purpose: to encourage entrepreneurs to start and invest in risky but productive ventures. The businesses that investment funds put their money in should certainly be limited liability ventures. But the risks taken by the investment funds themselves are speculative financial risks. When a fund invests without leverage in a corporation, the fund's own limited liability status is worthless. With or without limited liability, the fund can lose all of, but no more than, the value of its investment. But if a fund borrows from a bank to invest ten times its own money in that same corporation, the fund's limited liability status is a big deal. It lets the funds investors reap investment gains from much more money than they own, while risking no more than the same meagre amount as in the unleveraged case.

There's no reason the state should grant investment funds a special dispensation to encourage this kind of risk. Fund investors should be allowed to take speculative financial risks, sure. But fund investors should be responsible for all the money they lose if things go sour. They shouldn't be able to let the bankruptcy of some shell corporation or LLP shield them from the consequences of speculative financial foolhardiness.

It's one thing to socialize the risk faced by an entrepreneur starting a productive business. It's quite another thing to socialize the risk of taking on leverage to achieve speculative financial gains. Limited liability is a privilege, not a right, and an oddity from a libertarian perspective. It should not be extended to leveraged investment funds.

Steve Randy Waldman — Tuesday October 31, 2006 at 9:51pm permalink
Felix (mail) (www):
Just so I'm clear on this: Are you proposing that the funds' principals (ie the fund managers, who receive the fees) would be the partners, or are you proposing rather that all investors in the fund (everybody who gets the benefit of the leverage) would be the partners? In other words, are you proposing that an investor should, theoretically, be able to lose more than he invested? Or are you just saying that the personal assets of the principals should be at risk, and not just those assets they have invested in their own fund?
11.1.2006 10:50am
Steve Randy Waldman (mail) (www):
Felix, I'm proposing that all investors be partners with unlimited liability. Investors would remain protected by limited liability, in as much as the funds invest in limited liability instruments. If the fund is not leveraged and purchased stock, then fund participants can lose no more than they invest.

But for funds that choose to invest borrowed money, my contention is that the net positive externalities of financial speculation are not sufficient to justify a state grant of limited liability. Indeed, investors in leveraged funds could, in a rout, lose more than they have invested.

Why should wealthy, "accredited investors" who invest on margin via hedge funds receive more protection than a retail investor who does the same without an LLP between herself and the bank? Why should a fund investor who indirectly takes positions in derivatives markets be better protected than say, me, when I write an option or take a futures position as an individual?
11.1.2006 12:07pm
Felix (mail) (www):
OK -- One more question -- would the unlimited liability be joint and several? In other words, say a hedge fund has ten equal investors: nine individuals, and one big pension fund. Hedge Fund goes bust, with net liabilities of $1 billion, and the individuals can't come close to covering their share. Is the pension fund now liable for the full $1 billion in losses, or just its own $100 million share?
11.1.2006 12:51pm
Steve Randy Waldman (mail) (www):
Felix -- That's a very good question, and implicitly an excellent point.

My original suggestion of organizing investment funds as ordinary partnerships implies joint and several liability, so that in your example, the pension fund would be on the hook for obligations that investors with shallower pockets do not meet.

Pro rata liability might be worth considering. Joint and several liability increases the risk of faced by fund investors relative to direct investors, thus discriminating in favor of retail investors rather than creating a "level playing field". But pro rata liability is also problematic. Despite a lot of historical and theoretical interest, no common legal form currently apportions liability in this way. From a practical standpoint, it's easy to see why. It would be very hard for lenders and counterparties to judge the creditworthiness of a fund with many investors responsible only for their small piece.

None of the alternatives are ideal. But given existing organization forms, I'd argue that joint and several liability is clearly superior to full limited-liability for investment funds. Investors can mitigate their own risks completely by choosing funds that restrict their investment to stakes in limited liability entities, and leveraging themselves individually as much or as little as they please. Investors who choose to join funds that take very risky bets ought to be responsible for those bets, and if that means policing the creditworthiness of their coinvestors, that seems like a good outcome to me. Investors like pension funds, which are deep-pocketed fiduciaries for risk intolerant beneficiaries, probably would have to steer clear of funds that are highly leveraged or that enter into risky derivatives contracts. Again, this seems to me a feature rather than a bug.
11.1.2006 8:30pm
Douglas Knight:
Limited liability doesn't socialize risk. The risk is assumed by the lender.

In your 10 June piece you talked about limited liability and principal-agent problems. Yes, limited liability and leverage exacerbate the principal-agent conflict between investors and managers. But banks do understand this. They see hedge funds failing all the time and keep lending to them.
Systematic risk is another matter. Banks are probably making poor decisions about lending to hedge funds (though, maybe they're rational, too, once you take their limited liability into account). A really bad collapse involving the FDIC obviously would socialize risk. But the bankruptcies hurt society mainly because the businesses were good for society, not because society insured the risk. Your position seems to be that these businesses are not productive, so they should be treated as second-class businesses. I think they are productive, of liquidity, though I'm not entirely sure.

Also, I'd like to comment on the libertarian perspective. Yes, limited liability is odd from the typical libertarian point of view of. But that's because most libertarians are rabidly dogmatic about contracts. Why should the power of the state be used to enforce contracts? It's pretty arbitrary where you draw the line. In fact, if you really believe in unlimited freedom of contract, then you should be in favor of limited liability—-it's just a detail in the contract with the lender. (Again, limited liability doesn't socialize risk, but rather transfers it to a knowing party.)
11.1.2006 11:23pm
Steve Randy Waldman (mail) (www):
Douglas -- Good points all. "Limited liability socializes risk," is a shorthand, not entirely accurate for reasons you suggest, but not as inaccurate as you claim either. Limited liability entities are supposed to prominently advertise that fact (e.g. "Inc.", "LLP", "SRL", "Ltd."), the idea being that all counterparties know what they are getting into. So you can claim that anyone with exposure to a limited liability entity has explicitly contracted to take on their risk.

This analysis is theoretically correct, but fails a real world smell test:

1) It is perfectly possible for unlimited liability entities to arrange for nonrecourse credit arrangements backed only by business assets. In an idealized world of rational actors without transaction and information costs, the default arrangement wouldn't matter. In the real world, the popularity of LL entities argues that which arrangement "normally" prevails matters a great deal.

2) Banks are not the only creditors to LL entities. Most employees, customers, and suppliers, are direct creditors as well. When a company goes bankrupt abruptly, the costs can be spread quite widely, often among parties who could not have been expected to do a thorough analysis of the company's financial situation and business risk.

3) Even among diligent creditors, there are hard informational problems in evaluating firm solvency, meaning that creditors must assume a form of model risk in dealing with many LL entities. Think Barings -- creditworthy, not fraudulent (as an organizations), great books, broke in an instant thanks to one employee. In a world where the defaults were set up differently, lendors might seriously discriminate against LL entities. In the real world, that would mean discriminating against the vast majority of business customers of any scale.

4) Finally, as you note, the lending decisionmakers may have different interests than those of depositors, governments, and taxpayers, all of whom are implicitly guaranteeing many loans and pension funds. Any LL entity whose stakeholders will in practice be bailed out by a government clearly has its risks socialized. All banks, most pension funds, Freddie, Fannie, arguably LTCM and Chrysler, we're talking large numbers of large entities. And there is a perverse incentive here, since the larger the risks taken by an organization, the more painful the costs to other stakeholders, the more likely taxpayers will end up footing the bill even without an explicit guarantee. (See my agency costs piece for more on this.)

For all these reasons, I think that in practice it is more accurate to say that "limited liability socializes risk" than to assert that it does not. But the phrase is a bumper sticker answer to a complicated question.


You write, "Your position seems to be that these businesses are not productive, so they should be treated as second-class businesses. I think they are productive, of liquidity, though I'm not entirely sure."

My position is that limited liability does matter, and that a default it should only be granted to organizations whose risks the citizenry wishes to encourage. All business and risks have costs and benefits. Indeed, hedge funds are major liquidity providers, and liquidity is important to financial markets. But they take very large risks, and are better able to shift the costs of those risks to other parties than other kinds of enterprises. It is a judgement call, but in my judgement the benefits of "hedge fund entrepreneurship" to the general public do not sufficiently outweigh the external risks their activities create to merit a grant of limited liability.

If I didn't think systemic risk and bank agency cost issues were serious, I'd probably come to a different judgement.
11.2.2006 6:33am
Douglas Knight:
After reading your essay on multiple levels of agency costs, I can't complain about much. But I think it's easier to limit socialized risk by addressing the parts that are explicitly socialized or too big to fail (eg, don't let pensions invest in hedge funds) than trying to deal on the other end.

Your essay begins by saying that the decision of which companies should be limited liability isn't regulation. It's probably less vulnerable to regulatory capture than more complicated schemes, but it still is, as you admit, "a judgement call."
11.12.2006 12:05am
Aaron Krowne (mail) (www):
A very interesting idea.

To provide negative feedback for the use of large quantities of leverage, it may be sufficient to limit liability to the managerial partners of the fund, as opposed to extending it to the investors.

Yes, it would be nice to also de-incentivize "investing irresponsibly" in irresponsible hedge funds, but investors don't directly make decisions about how much risk and leverage to take on. Thus, it may not be necessary (or fair) to imperil them with over-unity levels of liability -- especially when they may be acting on falsified prospectus and reporting information.

And arguments about the depositors having deep pockets may not find a sympathetic ear with those that have noted institutional and pension fund participation.

Another idea for how to solve the "excess leverage"/insolvency problem through regulation, without really adding any more regulation, would be to simply bring back reserve requirements and not exempt broad money from their reach.
11.12.2006 12:06am
Steve Randy Waldman (mail) (www):
Douglas, I've no strong argument with anything you say. I'm a regulation-skeptic broadly (though I know I often don't come off that way), and worry that competitive managers will always find a way to take chances that benefit them and harm their under-represented fiduciaries. (See the CPDO stuff.)

The no-limited-liability suggestion is a way of squaring that circle, of making a change that dampens some of the agency costs without unduly burdening willing risk-takers' freedom to invest, or encouraging rent-seeking by politicians. It ain't perfect (politicians could still score on line-drawing questions), but seemed like a promisingly minimalist approach.
11.16.2006 5:51pm
Steve Randy Waldman (mail) (www):
Aaron -- The problem with only putting managers at risk is that they'd be small players taking on the liability created by very large pools of money, very highly leveraged. The position of hedge fund creditors would not be substantially improved, as managers would be bankrupt instantaneously if their funds went down. The approach would severely disincentive established managers from taking risk. But we'd end up with a world where established, wealthy managers run conservative, long-only funds, while young-whippersnappers with nothing to lose but a leased Porsche would risk personal bankruptcy in order to let institutional investors risk other peoples' money.

I'm not sure that the money genie can be put back into the bottle, at least not until it dies a traumatic natural death. But it'd certainly be worth a shot.
11.16.2006 6:03pm
Note: You don't have to register to comment! Just use the pull-down menu to comment as "Guest".

Post as: [Register] [Log In]

Remember info?