Saturday, November 5, 2011

Methinks Thou Dost Protest Too Much

Any sufficiently advanced technology is indistinguishable from magic.

— Arthur C. Clarke

Attentive Readers will realize that I have used my durable and insightful epigraph before, specifically in a post which defended my industry against accusations of malfeasance arising from the common tendency of merchants in any economy reliant upon buying and selling to conceal the true costs and profits embedded in their activities. It was my contention then and is now that no law, human or otherwise, compels a vendor to offer buyers of its wares the “best price”—whatever that may be—or, indeed, prevents it from doing what profit-maximizing enterprises are commonly presumed to do: maximize profits. As long as said vendor is not selling faulty merchandise to inappropriate customers in a fraudulent manner, we should not expect to know nor require it to reveal all of its secrets.

This, however, is not that post.

For those of you with half a brain will (or should) realize that my idyllic little précis of laissez-faire capitalism skips lightly over two critical assumptions: that 1) all this happy buying and selling take place in reasonably competitive markets, where other vendors compete to offer the same good or reasonable substitutes therefor, and 2) the manufacture and sale of these goods does not impose intolerably noxious externalities on the society in which they are sold. The first of these can be seen as simply a special case of the latter, in which the externality which society should naturally seek to limit is economic rent-seeking in all its forms: monopoly, oligopoly, producer or factor cartels, preferential government regulation, etc. Of course, this tends to assume that the economy should be servant to society, rather than vice versa, which belief seems unhappily out of fashion nowadays.1 Go ahead, call me a dreamer.2

A cynic might say that politics is nothing more than a neverending argument over the size and distribution of economic rents in society. But let us set that question aside for now. Instead, I would like to focus on other kinds of externalities: those corrosive and destructive injuries to society which are generated as ineluctable byproducts of the activity of certain unsavory economic actors, like arms dealers, child pornographers, and television reality show producers.

And investment banks.

* * *
First, some history.

One of the principal functions of investment banks is the distribution of economic risk in society, from those who wish to sell it (and its associated productive return) to those who wish to buy. In the past, investment banks generally worked pretty well as conduits for risk, passing it from natural seller to natural buyer pretty effectively while skimming a small percentage off the top as recompense for their services. On the wholesale securities side of the house, they acted as large, temporary warehouses, buying and selling securities and derivatives on behalf of clients and maintaining minimal stocks in inventory to satisfy unforseen demand. It was a model which required little equity capital to support it, so investment banks levered up with short-term financing of their short-term assets and earned a nice return on the shareholder or partner equity they employed. Operating with so little equity entailed substantial risk, as a simple mistake or unexpected market shock could send the entire house of cards tumbling down. But because they tended to deal in liquid, easily marketed instruments, failed investment banks could be liquidated with relatively little disruption to their counterparties or the financial markets. Of course, the shareholders or equity partners got wiped out, but that was understood as part of the game. Live by the sword, die by the sword.

But then came the Great Moderation, and the industry changed. Investment banks merged and converted into universal banks, with commercial lending, mortgage businesses, and retail depositors, and they began swelling like mutant ticks on a hemophiliac dog. They began to warehouse more and more securities and derivatives to accommodate increased trading volumes on the market-making side. They began to warehouse more and more financial instruments for their own proprietary trading efforts. And they began to manufacture securities and derivatives, like mortgage-backed securities, credit default swaps, and other “structured products,” to meet investors’ insatiable demand for adequate returns in a seemingly riskless world. But as their balance sheets ballooned, these banks stuck with the tried and true risk management philosophy they had developed over decades as pure investment banks: mark your assets to market in real time, get out of losing positions early, and never hold risky assets in inventory without hedging them. Unfortunately, this is a strategy which depends at its core on operating in liquid, transparent markets, where prices are well known, trading volumes are robust, and hedging instruments are effective and liquid themselves. It also depends on a key principle which every trader knows: it doesn’t matter whether the markets are liquid or not if your position has become so large that you effectively are the market.

In addition, investment banks began to take on more and more counterparty risk as they waded deeper and deeper into such activities as leveraged lending, prime brokerage (lending and clearing for hedge fund clients), and derivatives and other structured products. And this was not the simple counterparty trading risk of old, where your primary worry was whether the party you traded with would deliver a security. It was counterparty credit risk, incurred as part of a trade in which your ultimate profit depended on your counterparty’s ability to satisfy its financial obligations, like repaying a loan, delivering an unencumbered security, or paying off a derivative. And let’s face it: investment banks have historically been lousy at credit analysis. Oh, sure, they’re fine when it’s short-term, secured lending, like a margin loan collateralized by liquid, easily-marketable securities with transparent market values. But lending money (or, what is the same thing, contracting for delivery of future economic value under certain circumstances) to counterparties subject to multiple financial risks and multiple financial obligations over a longer period of time? Not so much. And this is a big problem, because it seems that investment banks as a group have become their own biggest credit counterparties in many markets, particularly derivatives.

* * *

The problem is neatly illustrated by a recent Bloomberg article on the European sovereign credit default swap market:

Five banks—JPMorgan, Morgan Stanley, Goldman Sachs, Bank of America Corp. (BAC) and Citigroup Inc. (C)—write 97 percent of all credit-default swaps in the U.S., according to the Office of the Comptroller of the Currency. The five firms had total net exposure of $45 billion to the debt of Greece, Portugal, Ireland, Spain and Italy, according to disclosures the companies made at the end of the third quarter. Spokesmen for the five banks declined to comment for this story.

While the lenders say in their public disclosures they have so-called master netting agreements with counterparties on the CDS they buy and sell, they don’t identify those counterparties. About 74 percent of CDS trading takes place among 20 dealer- banks worldwide, including the five U.S. lenders, according to data from Depository Trust & Clearing Corp., which runs a central registry for over-the-counter derivatives.

Gross exposures are many multiples higher, of course, but the banks like to advertise their net exposures instead. The problem is that net exposures are not the clean, unassuming things a layperson might think they are. Take the following scenario: Bank A sells a $100 million credit default swap on Underlying Company or Country X to Hedge Fund 1. Then, in order to hedge itself, it buys an identical $100 million CDS on X from Bank B. Bank A has completely eliminated its exposure to X and can sail off into the sunset, happily counting the money it made in spread between the two transactions, right? Wrong. Bank A has not eliminated its risk exposure at all, it has merely introduced a credit risk exposure to Bank B, which is now on the hook to pay off the CDS if X craters. But what if B craters? Bank A is still on the hook, and now it is completely naked short a $100 million CDS. Now Bank A could try to protect itself against Bank B’s default by buying a CDS on Bank B from Bank C or Hedge Fund 2, but I think you must begin to see that that merely introduces a credit exposure to Bank C or Fund 2. Of course in real life all these counterparties try to ameliorate this exposure by requiring frequently refreshed margin collateral on these trades, with the objective that any party’s true risk exposure at any point in time is simply the difference between the value of the collateral held (usually cash) and the net cost to replace the instrument in question.

The challenge to global financial stability posed by investment banks conducting these activities is threefold, in my humble opinion. First, the daisy chain of trades illustrated above clearly demonstrates that investment banks never completely eliminate the residual risk involved in buying and selling investment contracts like CDSs and other derivatives. There will always be some risk attendant on any transaction which has not been completely immunized (like, e.g., Bank A buying an offsetting CDS from Hedge Fund 1, which would have the effect of cancelling the original trade), whether this is direct credit exposure to your counterparty or basis risk introduced by trying to hedge counterparty credit risk indirectly, like via short-selling its stock. Each such trade adds residual risk to the bank’s balance sheet and, given the tremendous aggregate volume of gross derivative trades investment banks do, these residual risks can accumulate to a very large and scary extent.

Second, because most big banks have overall margin agreements (Credit Support Annexes) in place with each other that aggregate offsetting daily margin requirements across all trades outstanding between the firms, the collateral protection mechanism itself can trigger contagion both within and across tightly linked firms. A bank or large hedge fund faced with a substantial margin call in one market or security might liquidate positions in other, more liquid securities in order to meet its obligations. If substantial enough, this can cascade through the markets and the trading books of interlinked investment banks, causing broader market sell-offs and further associated margin calls. This sensitivity is exacerbated by the highly leveraged financial profiles of most major financial market participants, especially the large trading banks and derivatives dealers.

Third, the ineluctably bilateral nature of many of these structured products and derivatives means that, no matter how careful and conservative any one investment bank is in structuring and managing its risk profile, nobody can be assured they are not transacting with another AIG Financial Products or, less dramatically, that systemically dangerous net exposures are not accumulating in disturbing quarters. The chief reasons that AIGFP’s collapse exacerbated the financial crisis were because it did not post collateral (due to its AAA credit rating), it transacted in difficult-to-value, illiquid markets, and it accumulated huge net exposure to mortgage-backed securities. And yet investment banks and others gleefully piled into counterparty credit exposure with AIGFP (the “dumb money”) until it cried uncle. Wall Street piled into copycat trades and lending relationships with Long-Term Capital Management, too, in a 1998 dress rehearsal for 2008’s systemic collapse. The very nature of secretive, cutthroat competition in my industry means that none of us want to share information that might reveal the existence of unsafe concentrations of credit risk in the system.3 How else can one explain why French-Belgian bank Dexia was able to write so many interest rate swaps that it required a government margin call bailout to the tune of $22 billion? Last month.

* * *

The practice of counterparty risk management on Wall Street has improved mightily since the Panic of 2008. Given that disaster, it damn well better have. But given the nature of massively connected, highly leveraged investment banks acting as conduits and collectors of the risk of the financial system, and their historical blindness to risks like counterparty exposure and risk concentration which were the very risks which nearly killed them (and us), I am loathe to take them entirely at their word that everything is hunky-dory now. Short of requiring all derivatives and structured products to be cleared through global exchanges (with associated net position limits and centralized margin posting) and sharply limiting overall financial leverage at trading banks, I do not see a failsafe solution to this conundrum. Investment banks are bred in the bone to be highly competitive and take substantial risks. Their competitive risk taking added materially to the accumulation of dangerous stresses and vulnerabilities preceding the crisis, and there is no reason to believe it will not do so again.

I would be delighted to be proved wrong about this by those who know much more about the plumbing of the financial system than I do.4 What is to prevent the occurrence of another AIG Financial Products? How can existing system controls prevent or dampen the cascade of credit failures through the system? Are potential leverage-induced death spirals limited to markets with illiquid, opaquely valued securities? If so, what prevents them from spilling over via contagion into other markets? What is to prevent a major securities or derivatives market meltdown from forcing another massive government bailout?

And if you are brave, knowledgeable, and/or foolish enough to try to answer these questions, please keep in mind the admonition of another very clever man whom few now trust:

There are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns—the ones we don't know we don't know.

A wise man learns to plan for all three.

Related reading:
Committee on the Global Financial System, The role of margin requirements and haircuts in procyclicality (BIS CGFS Papers No. 36, March 2010)
Selling More CDS on Europe Debt Raises Risk for U.S. Banks (Bloomberg, November 1, 2011)

An early response:
Brandon Adams, Response for @Epicurean Deal (November 5, 2011)

1 An economy is simply the set of organizing principles and rules which a society establishes to allocate and employ resources for the benefit of its members. How these rules are established and maintained is politics. To assert otherwise, or claim as some do that society and politics have no proper claim on the organization or maintenance of economic activity (e.g., via regulation or taxation), is the height of folly or disingenuousness.
2 Those among you who cannot comprehend this concept and who would prefer to call me much less flattering names than “dreamer” are welcome to stock up on canned peaches and armor-piercing ammunition and join your fellow nutcases in Galt’s Gulch. The rest of us will come annihilate you when we can spare a moment. (Or just let you starve to death.)
3 For example, my best and most comprehensive source to-date for understanding the intricacies of the issues under discussion would not allow me to share them directly, in part because (s)he believed some of the generic information (s)he provided could provide a competitive advantage. And you people think I’m secretive.
4 All reasonable, informed, and specific responses are heartily welcome. I may publish or link to the most informative and interesting of these here. Please direct your responses to the email address found on this site, or notify me on Twitter (@EpicureanDeal) or by email if you have published it on another site. Please indicate if you would prefer no attribution.

© 2011 The Epicurean Dealmaker. All rights reserved.