FT Alphaville had an article recently where they featured an academic paper Complexity, Innovation and the Regulation of Modern Financial Markets by Dan Awrey. It’s not a mathematically complex paper, though it deals with complex financial instruments and it is quite relevant to our present troubles.
Let me start by quoting the beginning of the abstract:
The intellectual origins of the global financial crisis (GFC) can be traced back to blind spots emanating from within conventional financial theory. These blind spots are distorted reflections of the perfect market assumptions underpinning the canonical theories of financial economics: modern portfolio theory; the Modigliani and Miller capital structure irrelevancy principle; the capital asset pricing model and, perhaps most importantly, the efficient market hypothesis. In the decades leading up to the GFC, these assumptions were transformed from empirically (con)testable propositions into the central articles of faith of the ideology of modern finance: the foundations of a widely held belief in the self-correcting nature of markets and their consequent optimality as mechanisms for the allocation of society’s resources. This ideology, in turn, exerted a profound influence on how we regulate financial markets and institutions.
I have consistently been a critic of modern portfolio theory, the Modigliani and Miller capital structure irrelevancy principle, the capital asset pricing model and, the efficient market hypothesis. I have also criticized the idea that incomplete markets are a problem, and that derivatives are needed to complete markets. Trying to make liquid markets out of assets that are naturally illiquid is a fool’s bargain. I have also argued that derivatives should be regulated as insurance contracts, and subject to the doctrine of insurable interest.
Why don’t academic finance theories work? Quoting again from the paper:
These theories share a common and highly stylized view of financial markets, one characterized by, inter alia, perfect information, the absence of transaction costs and rational market participants. Yet in reality financial markets – and market participants – rarely (if ever) strictly conform to these assumptions. Information is costly and unevenly distributed; transaction costs are pervasive and often determinative, and market participants frequently exhibit cognitive biases and bounded rationality.
In short, men are not hyper-rational calculators, like the Vulcans of Star Trek. Markets for goods and services might be efficient, but not those for assets, where values are not as easily estimated.
Asset markets are frequently reflexive. As an intelligent former boss once said to me, “When does a company look its best? Immediately after receiving a loan. That’s why we wait a few years before shorting a bad company that has just received a significant loan.”
The paper encourages study to understand how markets work in practice, rather than how they work ideally to neoclassical economists. I heartily agree; I think that the more one studies the structure of asset markets, the more they will understand that there are many approaches to the market, and the popularity of strategies depends a lot on past success.
Nevertheless, taking a broad look across the financial system, it is possible to identify at least six – in many respects intertwined and overlapping – sources of complexity: technology, opacity, interconnectedness, fragmentation, regulation and reflexivity.
Technology – things move faster, but can people keep up with it. More data can be gathered by connected players. It is one reason that I am a low turnover investor. There are too many playing the short duration game in the market.
Opacity – few can truly understand the economics of most securitizations, or whether the subordination levels are right or not. Investing in “dark pools” is rarely wise. And as for the credit guarantors that I have criticized, they could not understand the risks they were taking, because they assumed the credit boom was normal and perpetual.
Interconnectedness – What could be more interconnected than the financial guarantors? Or the banks who lent to one another, directly and indirectly?
Fragmentation – Securitization makes it tough for owner to understand what is happening several links down the chain. Guess what? It’s a lot easier to have your own mortgage loan department, and watch over your own loans.
Regulation – Financial regulation is fragmented, and often co-opted by those regulated. Because the US government does not get this, market players arbitrage regulators. Another aspect of it was the moral hazard engendered by the Fed and other regulators, giving the impression that there would be rescues available in any real crisis.
Reflexivity – I have talked about this above.
I would add a seventh source of complexity – leverage. People underestimate the effects of leverage on managements in managing assets. When the possibility of bankruptcy arrives, the effects are discontinuous, violent. This is especially true when debts are layered, where A owes B, who owes C, who owes D, on thin equity bases. A’s failure to pay has ripple effects, leading to a cascading failure.
An eighth source of complexity is use of short-term debt to finance long term assets. An early precursor to the crisis was the failure of off-balance sheet subsidiaries that were financed with short-term loans. Similarly, firms that relied on repo funding to finance their inventory of assets had a rough time of it as repo haircuts rose rapidly, and in tandem.
A ninth source of complexity was similar: margin requirements in derivative agreements, where a credit downgrade could lead to a call on capital at the worst possible moment. This happened with AIG.
A tenth source of complexity which enabled much of the above nine is one that many writers don’t want to talk about, because it exposes many of “victims” to be enablers: yield-seeking. Why be a lender to complex investment vehicles? You get more yield. None of them have blown up. They are highly rated. Why not go for the higher yield? As I have said before, it takes failure to mature an asset class. All new asset classes look pristine; nothing starts with failure. Typically the best deals get done first – best quality, best incremental yields. Then competition drives down quality and yield spreads, but raises quantity.
Without the yield-seeking, many complex financial instruments would never get issued. Someone had to buy the “safe” tranches of CDOs, CDO-squareds, ABS CDOs, etc., in order to sell the deals. Many European banks bought them because they didn’t have to put much capital against them for risk purposes, and the added yield helped them meet earnings targets, at a cost of greater illiquidity.
An eleventh source of complexity was the failure of accounting to properly account for these new instruments with volatile fair values. Fair Value accounting, using market values and their approximations was a step in the right direction, and bitterly opposed by those who were financing illiquid, opaque, financial instruments, while their funding was short-dated with too little equity.
More will come in part 2, soon. Still without power — a hard providence, be we are surviving it.