Beating the Quants at Their Own Game

As an unequivocal believer in the virtues of the fundamental stock analysis influenced by the teachings of Benjamin Graham and Warren Buffett, it is a bit odd for me to comment on technical analysis. Although I do believe that technical analysis can work under the right strategy executed by the right people with the right emotional temperament (my friends over at Tudor Investment Corp being the most obvious example), I have always been a fundamental investor in search of determining the fabled “intrinsic value”. The underlying principals of technical analysis seem too blunt in my opinion, everything in this world has some level of inherent value based on the utility it provides and a properly trained investor willing to do the necessary analysis should be able to determine this value within a range and trade accordingly.

Regardless of my theoretical beliefs, many market participants use technical analysis to drive their investment decisions. These collective actions result in real tangible changes in asset values and as such need to be understood even by fundamental investors. A fundamental investor need not agree with the reason why a stock is moving but it behooves him or her to understand why a stock is moving. Just as a hitter in baseball does need to know how to throw a curve ball but he does need to be able to recognize a curve ball and anticipate the ball’s movement.

Many quantitative investment strategies are based on some sort of trend following - very simply put they buy an uptrend and sell a downtrend. The “trend” is typically determined by a moving average. For the sake of this analysis I will be dealing with simple moving averages as opposed to exponential moving averages due to the fact that I was not prepared to spend my whole weekend creating the necessary calculations in Excel. The general take-away of this exercise will be the same for either. A simple moving average is the average price of the last “X” number of days. Popular moving averages used in technical analysis are the 50 day, 100 day, and 200 day. A technical investor wants a signal that is clear enough that they will not miss a big portion of the price move but also slow enough that they will not get whipsawed around from trading “noise”.

Given that moving averages are lagging indicators and given that we know 97+% of the inputs, I can tell you within a reasonable margin what the 200 day moving average for GOOG or GLD is going to be next week. It is here that I think the fundamental investor can exploit the technical investor's adherence to the hard rules of their trading systems.

With the dramatic sell-off at the end of 2008 further in the rear-view mirror, those low prices are going to be the primary inputs into the moving averages going forward. In other words, the high prices of August and September will be rolling off and the low prices will be setting the new moving averages going forward. This is will result in declining moving averages across the board which will likely result in a wave of buying from the investors utilizing technical analysis as more and more stocks trade above their respective moving averages. This is one contention that I have with technical analysis trading systems - even though the outlook for a certain asset has not changed or improved, a technical investor will be forced to buy that asset (assuming he follows his signals) simply because of the “roll-off” of old prices.

The following charts demonstrate this phenomenon for the S&P 500 (SPY). For the sake of this example I assume that the index value going forward does not change from Friday’s close (1/9/2008).

According to my analysis and assumptions, the S&P 500 50-day moving average will trough sometime next week (~January 23).

The S&P 500 100-day moving average will likely trough in mid-March.

The S&P 500 200-day moving average will likely trough in late June.

Obviously the biggest (and most impossible) assumption is that prices will not change going forward, regardless the moral of the story remains the same, the steep decline of this past Fall will result in a steep decline in moving averages of which we are currently approaching. This “roll-over” of prices will likely result in a wave of buying from quant investors utilizing moving average technical analysis.

A fundamental investor can exploit this likely source of buying in two ways. If one is bullish on the market they would want to buy now ahead of the pending decline in moving averages. Conversely if one is bearish they could wait for the moving averages to decline and use the quant buying to exit positions on strength.

Disclosure: None.

Comments

[Ferdinand E...:]

The smartest trader I ever met said that he did not believe in technical analysis, but he made a point of looking at it because it might have something to say that he did not want to miss.

Jan 13 08:08 AM

[kevinm:]

Excellent article.

Jan 13 08:41 AM

[Jason Rines...:]

I agree, well written. I am looking at April to pick up some equities for 5 year buy and hold. I think the 200 day MA will be effected upward by fiscal stimulus.

Jan 13 10:38 AM

[OldLimey:]

"This is one contention that I have with technical analysis trading systems - even though the outlook for a certain asset has not changed or improved, a technical investor will be forced to buy that asset (assuming he follows his signals) simply because of the “roll-off” of old prices."

Surely one reason that a moving average gets crossed is the slowing pace of price decline (or rise) in the underlying. Fundamental analysis may be one of the factors influencing that change of pace. Anyway, interesting article - thanks.

Jan 13 11:03 AM

[AndrewBaker:]

Technical analysis is used by many fundamental analysts for entry and exit points on a stock they like/stopped liking. Technical analysts will rarely if ever trade with a view of a five year holding period at the outset: they are short term traders. With that in mind, you don't care what's moving the market: only that it is moving. Essentially we are looking at two different trading styles. If I were trading as a chartist, and the price moved against me,say, 7.5%, I would sell (buy if it jumped similarly). I don't care what caused the move. A fundamentalist would probably shrug and think that there was plenty of time for the move to be reversed (if downwards) and brag about their great stock picking (if upward). There is room for both, and because there are both, the markets are more efficient as a result.

Jan 13 02:22 PM

[MILESCFA:]

I spent 25 years on the institutional buy-side and my last position was with a real "kick-the-tires" fundamentalist. However, the first thing he did when I mentioned a new name to him was he looked at a price chart. So despite your - AND HIS - skeptisism about technical analysis, people do it either consciously or unconsciously.

Since you're in the hedge fund world, I guess you must be right that "quantitative models" are trend following.... IN THE HEDGE FUND WORLD. But a trend following system would be more apty called a technical analysis model.

I was around when "quant models" were first being developed in the research depts of brokerage firms in early 1980s. MER's Rich Bernstein was one of the first, but Shearson-Lehman's Bob Jones was the best. Bob was hired by GS's money mgmt arm to run funds with his model, proof even then the Lehman sucked! (BTW: GS had one of the best money mgmt arms in the business back then and I think they still do).

The quant models developed were based mostly on FUNDAMENTAL data such as valuation and growth (especially EPS) measures. They SHOULD HAVE had a technical measure, but the skepticism about technical analysis was even more widespread back then.

Jan 13 03:49 PM

[secmaven:]

Sell at the resistance level, buy at the support level. Without a chart to look at at with a volume metric on the bottom the investor is absolutely clueless as to where these levels might be. We may not be able to identify these points with precision but to ignore their importance and fly blind is just plain dumb.

Jan 13 04:05 PM

[afortini@ao...:]

I eat quants for lunch.

Jan 13 06:32 PM

[Gisi:]

In my opinion, technical analysis is only useful in REINFORCING the right TIME to buy or sell a stock. I would never rely solely on techs in carrying out my trades simply because I don't think historical price movements are any indicator of how the stock is going to move a month, or even a week from now. I strongly believe that fundamentals drive a business, whether those are the company's customer base, global presence, management team, or whatnot. It's the stuff we can't control that typically causes huge price fluctuations... On the other hand, behavioral trading techniques, as you mentioned in the article, are most easily handled through tech analysis (though not necessarily all the time). And behavior is pretty predictable.

Great article!

Jan 13 11:06 PM