Editor's note: This is part two of a three-part series. Click here to read part one
and click here to read part three
Last week, I discussed the logic behind quantitative allocation of capital to trading strategies. The ideas were based on work by physicist John Kelly at Bell Labs in the 1950s, which was extended and popularized by mathematician, blackjack card counter, and hedge fund innovator Ed Thorp. These ideas came to Wall Street via a loose community called “rocket scientists” in the early 1980s.
Rocket scientists were quantitative students who dropped out in the 1970s. It seemed to us at the time that almost anyone who knew any math was using it either to destroy all life on earth, support economic systems that were inflicting worldwide misery, or to do nonsense work. Our objection to the last two groups crystallized in the claim that these people wouldn’t bet a nickel on the results of their models. They would dash off op-ed columns advocating massive economic interventions with less real thought than they would put into buying a car for themselves.
In response, we quants turned to gambling. We learned to beat casino games and play world-class poker; we reinvented sports betting. We took mathematical theories to Las Vegas to find out what really worked when your own money was at stake, rather than what got applause in the faculty lounge or impressed bureaucrats. A lot of elegant theories got tossed aside, and we refined a robust body of knowledge. We then took that to Wall Street where we ran quantitative trading strategies, again with our own money at stake.
After five or so years of Wall Street success by quants, qualitative traders started to notice. On many trading desks, the quants were enlisted to do what we now call “risk management.” They analyzed each trader's pattern of profits and losses, factored in valuation, liquidity, and funding risks, and came up with an optimal capital allocation. In the process, the quants usually managed to improve the trader’s performance. Even the most experienced traders seldom knew which were their most profitable trades, which should be made bigger, and which were their least profitable or even money-losing trades. Simple statistical analysis led to dramatic improvements in trading results.
This brings us up to about 1987, which is when a new set of players arrived. Banking disasters in the 1970s and early 1980s convinced central bankers that banks needed minimum capital standards. Until that time, banks had reserve requirements (a fraction of liabilities that had to be matched by liquid assets), but usually not binding capital requirements (a fraction of assets that had to be matched by subordinated liabilities or equity). Bank examiners looked at a bank’s capital, but generally applied judgment rather than fixed rules.
The first major set of capital standards was adopted in 1987 and has become known as Basel I. Even at inception, it was known that Basel I was not sophisticated enough to capture the risks in modern banking. Work immediately began on Basel II, which was an attempt to apply quantitative methods to estimate minimum capital requirements not just for traditional bank businesses, like lending, but for complicated trading and derivatives businesses. Regulators and economists enthusiastically adopted the mathematical techniques invented by quants to determine optimal capital levels for traders and trading systems.
If you weren’t playing close attention, you might have missed the error that killed the financial system. Quants and economists were using the same word -- “capital” -- for two entirely different things. Economists were talking about a cushion of equity that could be used to pay losses if things went wrong, sparing depositors and other creditors, not to mention governments. Quants were talking about the denominator to use when converting daily profits and losses into percentage returns. Quant capital was designed to let profitable risk-taking businesses grow exponentially while avoiding blow-ups, not to make investors whole if the business failed.
Economists and regulators thought the problem was that banks took so much risk relative to their equity, so they might not be able to pay off losing bets. Quants thought the problem was individual risk takers acting in ways that guaranteed blowing up. In the example from last week (see Whodunit? Part I: Rocket Scientists on Wall Street
) in which the trader needs five times the capital, economists thought the bank should issue more equity capital to protect against excessive losses from the trader. This misses the point entirely.
What has to change is the trader’s behavior. The reason he blows up is not that he doesn’t have enough capital to pay his losses. The reason he blows up is that he increases his bets too much after a win, and decreases them too much after a loss. With 10% gains and losses, if he has a run of eight wins in a row (something that is expected to happen more than once per year), he will double his capital, and thus double the size of his bets. With five times the capital and therefore only 2% gains and losses, the same eight wins means only 17% larger positions. It is the accumulated losses from volatility drag that guarantee eventual ruin, not lack of capital.
Again, this is just a simplified example. Real qualitative traders do not usually increase or decrease bets directly in proportion to capital (real quants, on the other hand, usually do). But all traders who wish to harness the power of exponential growth have some mechanism for adjusting trade sizes in response to results. A properly tuned mechanism leads to long-term success; an improperly tuned mechanism leads to ruin.
Why would economists and regulators adopt the methods developed by rocket scientists? Few economists took the ideas of John Kelly or Ed Thorp seriously. Paul Samuelson, who wrote the most influential economics textbook in history and won a Nobel Prize, even wrote an article criticizing Kelly entirely in words of one syllable
in order to get through to simple-minded people who took Kelly seriously. Moreover, there was plenty of academic theory on how to set capital standards.
I was there at the time, and I can tell you: Economists and regulators liked rocket scientist methods because they had been accepted on the trading floor, and also in the executive suite. Few took the time to understand them. They did not seek out quantitative traders and line risk managers; they mostly relied on presentations to top executives. The major academic forum for our work was the triennial International Conference on Gambling and Risk Taking run by Bill Eadington at the University of Nevada at Reno. This was attended by a few maverick economists, and no financial regulators.
Next week, in the final installment, we’ll see what happened when economists and regulators took quantitative methods developed to optimize behavior of individual risk takers, and applied them to the entirely different problem of capital structure for financial institutions. Hint: it wasn’t pretty.
No positions in stocks mentioned.
The information on this website solely reflects the analysis of or opinion about the performance of securities and financial markets by the writers whose articles appear on the site. The views expressed by the writers are not necessarily the views of Minyanville Media, Inc. or members of its management. Nothing contained on the website is intended to constitute a recommendation or advice addressed to an individual investor or category of investors to purchase, sell or hold any security, or to take any action with respect to the prospective movement of the securities markets or to solicit the purchase or sale of any security. Any investment decisions must be made by the reader either individually or in consultation with his or her investment professional. Minyanville writers and staff may trade or hold positions in securities that are discussed in articles appearing on the website. Writers of articles are required to disclose whether they have a position in any stock or fund discussed in an article, but are not permitted to disclose the size or direction of the position. Nothing on this website is intended to solicit business of any kind for a writer's business or fund. Minyanville management and staff as well as contributing writers will not respond to emails or other communications requesting investment advice.