The costs of complexity II

The Economist underlines what I think is going to be a growing debate about financial regulation:  The push for simple rules:

But it did not do what a regulator charged with supervising exceedingly complicated institutions should: focus on the simplest rules possible where there is the least room for interpretation. Banks need to obey high capital standards, backed by robust leverage ratios that are not easily gamed.

"Boring banking vs fancy finance"

Krugman's aphorism describes more than a sense of nostalgia for the good old bankers of our parents. Boring banking had regulatory and "moral" implication advantages that fancy finance, for all the (alleged) efficiency it brings to financial markets, seems to lack. Matthew Yeglesias takes the latest LIBOR scandal to illustrate the point and concludes: 

The lesson of Libor is that regulators need to recognize that bankers have cast aside the clubby values of yore, and they need to respond in kind. Banks will try to abide by the letter of the law, but where loopholes exist, they’ll be ruthlessly exploited—through dishonest means if necessary—and the financial cops need to have a fundamentally suspicious attitude toward the regulated entities. Time and again, when tighter regulation of trading is proposed, the concern is raised that stringency will push activity to foreign centers. In the short run, that’s almost certainly true. Banks will want to move to wherever they’re most likely to be able to get away with more shady dealings. But an economic development strategy based on turning your country into an appealing location for dishonest banking is just going to get you a financial system that’s rotten with dishonesty. It’s time to stop being surprised and start realizing that these are the inevitable fruits of a regulatory system that’s weak by design.

High Frequency Trading and Market Quality

 A new report by Credit Suisse looks at the data and finds that HFT has not made markets worse off. Their chosen measures of market quality is volatility, spread, the messaging traffic and the risks of flash crash events. While one must applaud the effort of looking at a set of objective data, the report says nothing about allocational efficiency. Yes ! HFT certainly helps to bring down volatility, but it then means that HFT, like most financial innovation these past decades, helps allocating risk - and not capital.

Does finance has a natural size limit ?

The murky world of hedge fund suggests that there is some truth in it. From Bloomberg's interview with Richard Maraviglia, a small hedge fund manager who outperformed better-known rivals managing billion: 

“If you want strong outperformance, a smaller fund has a better chance of producing that,” said Maraviglia, who plans to close his hedge fund to new money after receiving commitments from investors for the $250 million. “How big an allocation can a big fund make to a great trading idea for it to make much of a difference in their total portfolio?”

 This points to a dilemma between the size and efficient captital allocation. Would the same reasoning be valid at the whole financial sector level ?

How to regulate the financial sector

 

She has a point, being surprised at the level of public attention JP Morgan received because it lost some of its own money (close to 7 billion in risky investments). But the JP Morgan was upheld as the very last bank that did not make these "risky investment". The surprise of the public is also legitimate, even though it is none if its business.

Her first point, however, on the regulation of CDS, has to be commented. The central argument she is making is that CDS bring valuable information to a lot of market player. That does not say anything about the quality of this information. The spread of Credit Default Obligation (CDO)  is commonly attributed to the development of a mathematical formula called the "gaussian copula", by David X. Li. His formula was quickly adopted by Wall Street, because it gave a convenient approximation of the complex reality behind CDOs. The central problem that lies beneath this issue is to know what kind of information financial market are giving participants. Too much and too complex information can be conducive to a loss of information. That is also the price of complexity. Prices are supposed to be the magic signal making coordination possible. Do prices of all traded financial assets traded reflect truly economic fundamental today ?

More on the Gaussian copula here, and here, and here. 

Financialisation watch

The Boston review adds the historical background to the now more and more accepted argument that the political weight Wall Street is getting from the phenomenon of financialization changes the political economy landscape of the U.S: 

American politics has always been dominated by wealth, but only rarely have wealthy people from a single industry dominated politics as much as those from the financial industry dominate both parties today. I can only think of two examples: the Southern planters before the Civil War and the railroad tycoons in the 1860s and 1870s, before the rise of wealthy manufacturers in the late 19th century. The recent financialization of politics was made possible by the bubble economy. Over time, even without reform, the inflated financial share of the economy is likely to shrink, with influence shifting partly to economic elites in other sectors, including perhaps the energy sector and a partly-revived manufacturing sector. As a rule, ordinary Americans have been better off when rival economic elites are forced to compete for their votes than when a single industry supplies the donors and much of the personnel of both parties.

"The costs of complexity: just because it's difficult to quantify, it does not mean its zero !"

Complex environments (the western financial systems, arguably) can be very costly. That's the reason why  big is not necessarily better in complexity economics: there is a “clear trade off between the benefit of scale and the coordination costs and constraints created by complexity” (Beinhocker: 151). These are costs inherent to any system and applying complexity economics theories to financial systems requires a level of abstract thinking I am not willing to do on a sunday afternoon. That would be a "too big to fail" metaphor, but at the level of entire financial system. Regulatory costs are easier costs to reflect upon.

It is necessary to point out right away here that deregulation does not mean less rules, but the contrary. The British financial revolution of the 1980s illustrate that very well. The Financial Services Act (FSA) passed after the Big Bang gave birth to a myriad of regulatory agencies. He also noted a change in “regulatory culture”: “The FSA may have revolutionized life in the City even more than the Big Bang, for in matters of regulation it replaced the informal with the formal, the flexible with the rigid, and the personal with the legalistic” (Vogel:93). This shows as well that financial regulation also is embedded in broader social and cultural structures that evolve over time.

But as regulation are getting more complex, it becomes - irony -  a source of risk in itself too. I believe the regulatory problems that Barclay's LIBOR fraud (for the latest development of the saga, go there) are actually symptomatic of a problem of discipline within the financial sector that the multiplication of rules did not help to ciment. Yes ! FT Philipp Augar is right "Banking supervisors would be well advised to leave as little as possible to management discretion and to go for bold, simple rules that are easy to understand and possible to enforce". The "too big to regulate" has to be considered too.