While the media’s master narrative is generally correct, an analysis of recent coverage reveals that it suffers from several problems. A major shortcoming is that the coverage tends to conflate all quantitative hedge funds and the investment strategies used by those funds and leaves the impression that it is the quantitative approach itself that is the problem. This becomes particularly apparent when one considers the discussion of “market-neutral” funds and strategies.
Recent coverage reports huge losses at market-neutral hedge funds and cites “non-quants” like ClearBridge Advisors’ Hersh Cohen in New York. Cohen denigrated market-neutral approaches in the WSJ and noted that “[he now realizes] that what sounded impressive was not much more than a thing we saw played out in 1987 and 1998.”
Or, one can consider the comments of oft-cited “Black Swan” pusher and former quant Nassim Nicholas Taleb, who tells the Washington Post that current-day quants are "very smart in front of a textbook but not smart enough to understand very elementary things in reality." In fact, “most” of the legions of PhDs toiling away on Wall Street, says Taleb, are merely “idiot savants brought to industrial proportion."
In an effort to explore the recent media criticism of market-neutral investing, we contacted ValuEngine View newsletter portfolio manager Eric Stokes to discuss his thoughts on the topic. Stokes, the author of “Market Neutral Investing” practices a hybrid form of the market neutral approach whereby he uses his own picks for the long side - which are based on his analysis of technicals as well as ValuEngine’s quantitative models, and ProShares S&P ETF (SH) for his shorts.
Stokes used the ValuEngine View portfolio results to illustrate the protection afforded by his preferred strategy. The portfolio, which is “long only” and rebalanced monthly, recently took a hit of 15.5% as the market gave up most of this year’s gains. However, if one had hedged the portfolio by investing an equal amount in Pro-Shares SH or any other similar ETF, the decline would have only been 3.2 %. That modest loss, notes Stokes “would have been far better than the S&P’s loss of 9%.”
Stokes cautioned investors to “remember not to throw the baby out with the bathwater.” He noted that “the problems of the supposedly market neutral large hedge funds were related to their reckless utilization of leverage to garner large returns as well as their inability to pick good shorts. With equity buyout firms buying up"dogs" based on attractive cash flow”- Stokes likes to use the Yellow Pages as a prime example of this phenomenon - ”it is tough to make sure that shorts really do"lose."”
The better coverage of the hedge fund meltdown notes that many of the big losers utilized strategies similar to their competitors along with huge multiples of leverage, and then, when faced with margin calls, liquidated their positions all at once. Thomson Financial claims that "many of the highly sophisticated quant funds employed similar investment approaches and held similar core holdings. This resulted in the funds selling similar long stocks and covering similar short positions." This factor, say the astute journalists, resulted in stock prices for long picks decreasing in value while the value of short picks increased.
But this role reversal between the long and short side did not effect all market- neutral investors. Consider the experience of ValuEngine client Forrest Warthman. Warthman uses ValuEngine Institutional software to trade and track several portfolios based on our quantitative models. Warthman’s tracking data for his VE Standard 20 Market-Neutral Strategy shows that the portfolio had a positive return of 2.82% for the period July18-August 10, 2007 while the S&P lost 9.02% for that timeframe. Warthman’s data for his VE Forecast 22 Market-Neutral Strategy demonstrates similar results with gains of 1.29%. For Warthman, the VE models’ short picks performed exactly as needed.
So, if one cannot condemn the market-neutral strategy or the use of quantitative models, what is to blame for the recent “unpleasantness?” One might start with an old standby, greed. It seems clear that many hedge-fund managers, mesmerized by the magic of leverage on their portfolios, resorted to ever increasing lines of credit in an effort to boost returns and personal gains. When one gets 20% off the top and is managing billions of dollars, it is easy to forget that high multiples of leverage will kill you if things go wrong like they did on August 9th, 2007.
One might also consider that many hedge funds were not using market-neutral strategies at all, but were actually engaged in various forms of statistical arbitrage whereby price differentials in securities, bonds, credit instruments, etc. were “found” by the firms’ quant models and then exploited. This is also an issue frequently glossed over in the ongoing orgy of quant bashing. These sorts of quant funds were not buying and selling stocks based on “old-fashioned” notions like value, they were exploiting loopholes in the system by making thousands or even millions of trades over brief timeframes. These sorts of quant funds also resorted to huge multiples of leverage, watched their loopholes disappear, and then--worst of all, liquidated in a panic as their credit dried up when the subprime crisis spread through the system.
What is truly disturbing about all of this quant bashing are the claims that the models were ill-equipped to predict the future or incapable of reacting to the crisis. That is nonsense. For the past several months, ValuEngine’s quantitative models indicated that ALL tracked sectors of the stock universe were “overvalued.” The last time that this occurred was six weeks prior to the May, 2006 market correction. Regular readers of ValuEngine’s bulletins were constantly reminded of this fact. Of course, timing is always the issue, and when any method of analysis solves this question the markets will cease to exist!
Here are a few things to remember the next time you read an article claiming that “quant has failed” or that “market-neutral approaches do not work:”
1. Not all market- neutral investors are losing money, many did not lose that much relative to recent gains, and many have already reversed recent losses.
2. Not everyone bought Exxon (XOM) and not all quant models provide the same picks.
3. Smart quant-based portfolio managers use human beings to check out model picks via fundamental and technical analysis as well as common sense.
4. Leverage always seems like “money for nothing” but when things go wrong, it will kill you.
5. Despite technological advances, investing still requires a commitment to value, steady growth over time, and the power of compound interest rather than get-rich-quick schemes. A bad month--or even a year, means little over the course of a thirty-year portfolio. Chasing the markets’ ups and downs is counterproductive;
6. “Garbage in, garbage out.” Quant models are only as good as their programmers and their data.
7. You will never see a computer model on CNBC red-faced and blubbering for the Fed Chairman to “open the window!”