Search This Blog

Tuesday, March 3, 2009

"An Errror Laden Machine"

John Plender in the FT presents a nice summary of "what went wrong" including many of the points that have been highlighted here several times. I did find his description of the financial decision making process as "an error laden machine" and its tendency towards disaster myopia (which is pretty much the same as underestimating the "black swans")

....The message of all this misery is summed up by Michael Lewitt of Harch Capital Management, a fund manager who was quick to identify the risks in the credit bubble. "Virtually every strategy institutional investors followed, or were advised to follow by their consultants or funds of funds", he says, "turned out to be a complete disaster". Even if that verdict errs on the sweeping side, it is clear that mainstream investment strategies failed to deliver. Why - and what needs to change to prevent a repetition?

A good diagnostic starting point is the phenomenon that academics call "disaster myopia" - the tendency to underestimate the probability of disastrous outcomes, especially for lowfrequency events last experienced in the distant past. The risk of falling victim to this syndrome was particularly acute in the recent period of unusual economic stability known as the "great moderation". Investors were confronted by falling yields against a background of declining volatility in markets. Many concluded that a new era of low risk and high returns had dawned. Their response was to search for yield in riskier areas of the market and then try to enhance returns through leverage, or borrowings.

and this myopia created a lack of awareness of risk:

The sedative was exacerbated in the bubble, according to a recent paper by Andrew Haldane, director for financial stability at the Bank of England, by badly flawed risk models. "With hindsight, the stress-tests required by the authorities over the past few years were too heavily influenced by behaviour during the golden decade" of 1998-2007, he says. So many risk management models were pre-programmed to induce disaster myopia. The input into the models was based on highly unusual macroeconomic circumstances that differed materially from longer-term historical experience. Risk was thus mispriced on a dramatic scale because of model-enhanced myopia

And then there is the issue of compensation. Plender writes:

Among hedge funds, disaster myopia is more cynically entrenched by a poor alignment of interests between managers and their investors. Hedge fund fee structures rarely allow investors to claw back fees if years of profits are wiped out by a single year's giant loss. Research by Harry Kat, professor of risk management at the Cass Business School in London, confirms just what this would lead one to suspect. Many hedge fund managers take on "tail" risks in derivatives markets, which produce a positive return most of the time as compensation for a very rare negative return. In effect, the funds have been writing catastrophe insurance.Then the catastrophe happened. Arbitrage strategies that took market liquidity for granted also foundered

(In fact the situation was even more insidious: traders within financial institutions, as well as hedge funds were all incentivized to take on the aforementined tail risk. And since the trades created positive returns "most" of the time there was tremendous incentive to leverage these positions. ....And when things blew up and took down the institutions the losses were so great that the financial system was endangered and the companies were too big too fail. AIG took enormous leveraged bets on tail risks and ironically based on its high credit rating based on its "stable" core insurance business. Was able to leverage those bets with cheap credit.)

The error laden machine was alive and well in portfolio aqnagement as well

Equally unfortunate has been a botched approach to portfolio diversification. This powerful tool allows investors to achieve higher rewards for a given degree of risk, or the same reward for a lower level of risk. Yet in alternative asset categories it has failed to do that, despite the use of sophisticated mathematical modelling of correlations between asset classes. Hedge funds, private equities and commodities have underperformed in unison.

John Kay, a fellow Financial Times columnist, points out in The Long And The Short Of It , a new book on investment, that the endowments of Harvard and Yale did well in hedge funds and private equity in the 1990s. But asset classifications can change their meaning. As the sector grew, hedge funds became less a bet on an individual's skills, more a conventional run-of-the-mill fund.

At the same time, private equity firms, bloated on credit, turned into a highly borrowed play on the stock market. Returns became increasingly correlated with other investments. The endowments, along with other investors who accepted consultants' conventional wisdom on alternative assets, have suffered in consequence. Prof Kay's message is that diversification is a matter of judgment, not statistics, and that a model will tell you only what you have already told the model. It can never replace an understanding of market psychology and the factors that make for successful business.

A more fundamental point is simply that diversification cannot work well in a credit bubble because virtually all asset categories are driven up by leverage. Then when the bubble bursts, deleveraging affects asset categories indiscriminately. Equally fundamental is that fund managers tend to move in herds because that reduces the risk of their losing client mandates. Minimising business risk takes priority over the interests of beneficiaries.

(This dependence on leverage imo is fundamental to the difficulties of "alternative asset classes" during the deleveraging. Private equity, hedge funds ,real estate were not generating superior returns through manager skill (alpha) they were simply leveraging themselves to generate higher return. There was no true alpha (superior risk adjusted return)...just more return delivered with more risk. When the credit was no longer easily available and everyone had to liquidate at the same Warren Buffett says "when the tide runs out you lear who is swimming naked".)

It is hard to argue with the articles conclusions:

So despite the complexity of today's markets, the lessons in all this are oddly homespun. Mathematical models should not be relied on without a proper understanding of the economic conditions and behaviour that fed them. It is foolish to put blind faith in credit rating agencies. Do not invest in what you cannot understand. Shun arbitrage strategies that assume permanent access to liquidity. Avoid investment vehicles that inflict swingeing charges in exchange for what in most cases will amount to market performance or worse. Treat leverage with due care. Recognise that the conventional wisdom of the consulting fraternity is not conducive to contrarian behaviour, one of the keys to successful investing. Above all, beware what Charles Mackay, the 19th-century historian, called the madness of crowds.

No comments: