Wednesday, June 18, 2008

How dangerous are OTC markets?

(Naked Capitalism) This week's Institutional Risk Analytics has an alarming title: "Is Risk Management Even Possible in an OTC Marketplace?" By all indications, the article points to a strong "no".

As much as I am a harsh critic of so-called financial innovation, the headline goes further than the case the article makes. OTC markets covers a large territory. The Treasury, corporate, and agency bond markets are OTC. Because those instruments are offered publicly (ie, the size and terms of each issue are known) and the valuation process for these securities is pretty straightforward, the potential for trading in them to cause a large scale problem is no greater than in other simple products (remember, it's possible to create disasters via simple operator error, such as believing as Walter Wriston did, that countries don't go bankrupt).

Even in derivatives-land, some products are sufficiently straightforward, such as simple ("plain vanilla") interest rate and currency swaps, as to not pose undue risks if properly managed. But the problem is that financial firms have come to have high return requirements, and these simple products aren't very lucrative (indeed, some are marginally profitable but the dealers stay in the market because most large players prefer institutions that make markets in the full range of financial products).

A general pattern in the last 15+ years is that large financial firms have take on more and more risk in order to keep their returns high and their profits growing. Some of that is via leverage, some of that is via launching or expanding the market for new products (think CDS and CDOs), some of it is by taking risks they shunned before (think of leveraged loans).

I've only excerpted some bits of this article, but in case you decide to read it in its entirety (which I do recommend), I wanted to address some quotes that give a misleading impression. One is at the start of the article:
OTC derivatives had been legally permitted for the first time in 1993 by a regulatory exemption that Wendy Gramm had adopted as virtually her last act as CFTC chair.

That's not correct. OTC equity options and FX options had been around well before 1990. Swiss Bank formed a joint venture, which was a precursor to a full buyout, of O'Connor & Associates, a Chicago-based derivatives trading firm. The reason for the deal was O'Connor's leadership in OTC equity and FX options. Um, those are derivatives, and regulated by the CBOE, which is a self-regulatory organization under the supervision of the SEC. Bankers Trust was another big player in those markets back then. I hate to sound pedantic, but it makes me crazy when people make definitive statements that are just plain wrong.

Putting this quibble aside, the article has a very good section on how the FDIC is gearing up for a big increase in bank failures. The rest of the article focuses on the problems posed by OTC markets. Some excerpts:
One veteran federal regulator, who provided comments on our upcoming article for the Journal of Structured Finance, put to rest some of our fears that Washington is clueless about the nature of the OTC problem....:
Regulator: You suggest in your article that the issues raised by OTC markets have not been discussed within the regulatory community, but in fact it has and is being discussed intensively. Whether or not the OTC market needs to be officially "regulated" seems debatable, especially since regulators do not have the legal authority to enforce such a change.

The IRA: So the US and other nations must simply accept the fact that OTC markets are here to stay and that these inefficient markets will periodically destroy a large financial institution? That seems like a recipe for disaster, financially and politically.

Regulator: Part of the problem is cultural. Today we think that all markets are at a minimum weak form efficient. The Efficient Market Hypothesis is taught as gospel. The underlying assumptions of modern economic thought -- ready prices, informational symmetry, and rational expectations -- are all suspect and have been for a long time. We tend to view economics and modeling as a science governed by laws similar to the laws of nature. We believe that markets can be confined to probability spaces we understand and can reasonably estimate. This is not true.

The IRA: So you agree with our view that risk management of OTC markets is essentially impossible? Or does your statement apply to all financial markets?

Regulator: The reality is that economics is a social science and the attempts to make it "hard" are always going to run aground on the reality that human actions don't follow the "simple" laws of nature; they learn, adapt, herd, swarm, fall prey to trends, forget, remember, forget again - and in a semi-rational and sometimes irrational manner. The probability spaces are impacted by things traditional theory freely jettisons in order to make the model tractable. Therefore, risk models, upon which so much of the OTC market rests, are simply a way to communicate and express views of value -- usually rather naive and simplistic views -- and miss huge chunks of the real underlying "human action" risks.

The IRA: Ludwig von Mises, the author of Human Action, would be pleased to hear your comments. So, again, you agree with our view that most alleged "risk management" systems deployed on Wall Street are really just tools used to do deals and have no real capacity to measure let alone limit risk?

Regulator: The risk and pricing models are often created in order to convince traders and end-user investors that all these financial transactions -- behind which are human behaviors and cash flows -- are "manageable" and can be properly understood with the right analytics, data and "secret sauce." This witchcraft and sorcery can turn a toad into a prince, a rotten apple into a juicy melon. But it's all spurious precision. It's all vaporware. The models are used to create liquidity, which spurs volume, which garners big commissions and large EPS for dealers. The senior managers know that the models and assumptions upon which the higher spread product is based for things like OTC complex instruments are garbage, but you need a "basis" upon which to talk and compare so you can drive business. This charade works 90% of the time when things are calm, but as Hynman Minsky wrote in 1980 "stability is ultimately destabilizing." When risk regimes change, those who don't "know" these truisms find themselves naked.

We then spoke to Bob Feinberg, our favorite observer of financial services policy on Capitol Hill, about the state of the banking industry and congressional efforts to address the subprime financial meltdown....
Feinberg: At the last CMRE event that Elizabeth Currier allowed you and I to attend, which was in 2004, Larry Kudlow called for monetary reflation in order to make the economy look healthy so George W. Bush could be re-elected. It's happening again this year on behalf of all incumbents. The late Bob Weintraub called this the presidential cycle. It doesn't always work. In 1992, Greenspan didn't accommodate "41" (aka George H.W. Bush), and for a time there was doubt as to Greenspan's reappointment because 41 blamed Greenspan for his defeat. Fighting deflation becomes the Fed's excuse for pre-election monetary expansion, even if there is no deflation in sight. The way Greenspan put it in 2004 was that deflation was a low-probability, high-value event that must be staved off, just to be safe.

The IRA: But is it even possible to reflate the US economy when the financial industry is in such a terrible state?

Feinberg: I stand by my previous view about the model being broken, but I don't think people realize that this can't be fixed, that industries have life cycles, and the banking industry is about a half century past its best-if-used-by date. Greenspan once said to the Senate Banking Committee that some people would say that the only thing banks do is live off the yield curve. Whenever that model isn't available, they have to resort to extremely risky gambits to try to earn the returns Wall Street demands. It's a 19th century business model that got an extension thanks to the ability to arbitrage securities, but once they actually try to create new products, these instruments must be risky and opaque. So they've ended up as GSEs, CDOs and CDEs (Capital-Destroying Entities). Another acronym is the Disaster-Prone Organization (DPO), an expression coined by Prof. Anthony F.C. Sutton in his book, St. Clair, in which he laid out the reasons for the failure of the coal industry, which was just as powerful in its day as the banks are now.

The IRA: So you see the US banking industry going the way of king coal? Obviously the neither the Fed nor the Congress is willing to admit that a big constituency like the banking industry is moribund.

Feinberg: The bottom line is that the system is broken and can't be fixed. The reason the banks keep coming up with opaque products is that they're trying to de-commoditize a business that is mature and adds little or no value to the economy. In fact, over time and on net, the banking system destroys value. I wonder what song and dance the geniuses at Treasury are going to come up with next to justify buying worthless CDOs in the name of "reliquefying the market." I'm fascinated by the application of Gresham's Law to this situation; that bad collateral is manifestly driving out the good. I even read that banks are making bad loans in order to create some bad collateral, because the Fed will buy it.

The IRA: I take it that you do not expect the Congress to propose significant reforms of market structure?

Feinberg: After LTCM, the President's Working Group did a report and found that the financial system was just fine. After Amaranth, I think, some congressional committee(s), certainly Senate Banking, asked the PWG to go back and take a look at what they said in 1998, and PWG came back and said it had nothing to add. They asserted that the opaque market for derivatives could best be monitored by the banking regulators. I think this assertion needs to be rethought given what's happened with Bear and LEH..

No comments: