February 27th, 2009, 9:55 pm
QuoteOriginally posted by: MartinghoulSorry to bring this back to the original subject ...Forget the fractional reserve thing, that was just a possibility. The idea is that these commercial banks should be a lot more conservative and a lot more regulated. I just found this really interesting speech by Volcker, who seems to agree with me. He also touches on a couple of other sensitive issues.Volcker Speech in CanadaInteresting speech.I don't have as negative a view of "financial engineering" as Volcker does. (I also find it especially sad and distressing that Volcker and grandson couldn't hold a session of mutual education because I suspect that both parties could have learned much from the exchange) Yes, there are major problems in applying "engineering" to human systems. The mathematical and scientific methods that underpin an engineering discipline can (will!) fail in self-modifying, anticipatory human systems such as markets, financial systems, and economies. But the human element in markets and economies does NOT entirely invalidate an engineering approach, it just means that the notion of mathematical precision must be tempered by how incentives and profits can distort the math and the data that underpins the discipline. Realizing that rational equations can and do lose against irrational animal spirits should temper the hubris of those that think math, physics, science, or engineering can work on economies. In other words, I'd not throw the baby out with the bath water even if the baby is especially immature and has taken a rather large and smelly dump in the world's collective economic diaper. Instead, I'd use the financial engineering tools but throw in generous safety factors and heuristics that reflect the simple fact that we know a great deal less than we think we know about the markets. Volcker also misses a key issue when he decries the supposed opacity of financial engineering. It's not that the math created opacity, it's that it created the illusion of transparency to those that understood the math. Quant finance converted a known unknown (i.e., the risks of mortgage pools) into what practitioners thought was a known known (i.e., the output of the equations for pricing mortgage-backed securities). In theory, the math actually increased the transparency in the system by attempting to model correlation from proxy data (e.g., commonly-reported prices rather than rarely observed credit events). Prior to developing that math, the risk was a known unknown and if I know that I don't know the potential pay-offs of an instrument, then I can make prudent choices. But if I think that the outcomes are transparent due to either a pricing formula or due to a hedging formula, then I can (and will) use much more leverage. Yet the math also created what became for most managers, investors, and regulators, a catastrophic unknown unknown (i.e., the instability of correlation, especially in the context of the limited datasets available). (We'll not discuss how our favorite yellow-suited crusader failed to convince the flocks of global villagers of the danger of the default-breathing correlation dragon!) The point is that the sequence of events was: known unknown (risk) -> data -> math -> seeming transparency -> confidence -> leverage -> more data confirming the original data (i.e., housing only went up) -> more confidence -> more leverage -> -> -> BOOM! I call this the curse of transparency because what-you-see-is-NOT-what-you-get.Another area that I more strongly disagree with Volcker is when he recommends having a smaller number of larger, regulated banks. That seems like an "put your eggs in one basket and then watch that basket" strategy. It's not surprising that a central banker would gravitate to a centralized solution (if you spend you life as a hammer .....). But what I find interesting is that it's the biggest banks that have created the most trouble for the financial system. Maybe I'm wrong and maybe all the shoes have not yet dropped, but it seems that bank failures to date have been concentrated among the very large and the very small. In contrast, mid-sized, super-regional banks have weathered the storm better than their larger or smaller competitors. That small banks fail in a downturn is no surprise -- they are simply less diversified by the very nature of the smaller numbers of accounts on both sides of their balance sheets and by their local nature. That large banks fail more than mid-sized banks is worth some thought and suggests more caution before advocating Volcker's one-watched-basket-of-behemoths strategy.The more insidious danger of regulation is that it actually replicates the same category of strategy that caused this problem in the first place (i.e., the strategy of creating illusory transparency). Regulation attempts to convert of known unknowns (e.g., risks of bankruptcies in the financial system) into known knowns (e.g., adequate reserves, Basel N standards, certified risk management processes, etc.). That's laudable goal, I'm sure. Like the elegant math behind all the toxic assets du jour, this regulation will be well-intentioned, well-thought-out, and 100% likely to create new well-hidden unknown unknowns. The danger isn't in the unintended consequences per se, but the the scale of them.My concern is in preventing single-point vulnerabilities in the global financial system. I see such vulnerabilities in the excessive sizes of institutions, excess centralization of regulation, and the excessive use of a single formula (model monocultures) in any market. The first two are relatively easy to prevent. The last is a far trickier beast that probably calls for more of what Paul advocates -- that people think for themselves.