A majority of investors today are stock pickers. Mark Twain once said, "Whenever you find yourself on the side of the majority, it is time to pause and reflect." Questioning what we believe is difficult at the best of times. Whether you spend seconds or hours reflecting, stock picking in all its forms, while sometimes questionable in effectiveness, will always be an engaging activity adopting never ending mass interest. But is "interesting" better for you financially? Everybody is trying to put forth effort to do better when it comes to investing. With limited time and resources, are you better off picking which stocks to buy or when to buy? To shed light on this question, we simulated stock picking vs. market timing to quantify the risk associated with each approach. We found that stocking picking produced more tail risk.
We ran simulations using real market data from 2002 thru 20111. To decouple ourselves from any biased selection methodology, we implemented random selection. To represent market timing, we implemented tactical asset allocation, which departs from stock picking. Tactical asset allocators feel that "when" is more important than "what" focusing on the broader market rather than individual stocks. Recognizing that a rising tide lifts all boats, they "manage in the time domain," when to be or not to be in the market, without needing to quote Shakespeare and actively allocate between asset classes. It is a union of active managing and the instant built-in diversification of index investing. Rather than buying individual stocks, they take a position in the broad market thru many available broad market ETFs like, S&P500 index ETF (SPY), S&P500 equal weight ETF (RSP), S&P100 index ETF (OEF) or index future contracts like, E-mini S&P500, E-mini Nasdaq100. The approach can be described as index market timing.
To simulate index market timing, at the beginning of each year, we had a random generator choose (with replacement) an investment amount of either 100% stock or 60% stock (& 40% cash2). The stock portion was an investment in the equal-weighted performance of S&P100 stocks. Of course, the 100% represents an aggressive position and the 60% represents a more conservative position. After running the simulation 500 times, the ending portfolio annualized returns of the 10-year period are shown below.
(click to enlarge)
To simulate stock picking with the equivalent stock exposure and portfolio turnover as the first simulation, we ran an 80-20 stock-cash portfolio with 20% turnover. This allows for an apples-to-apples comparison. We had a random number generator choose 20 stocks among the S&P 100 at the beginning of each year, for 80% of a portfolio. The remaining 20% was invested in cash2. At the end of each year, new stocks were randomly selected to achieve 20% portfolio turnover. After running the simulation 500 times, the ending portfolio annualized returns of the 10-year period are shown below.
The results showed an equal mean and standard deviation. Risk, often defined as the uncertainty of returns, is commonly measured by calculating the standard deviation of returns. The lower the standard deviation, the lower the risk. At first, it seems neither approach yields any difference, especially with the mean and standard deviation being equal. But stock picking resulted in a comparatively higher kurtosis, which just means more of the deviation is due to a few extreme differences from the mean. A fisherman's net catches all types of fish, but fish that are too small like minnows will pass through the net and too large like whale sharks will break the net. Facing increasing chances of minnows or whales sharks would be financially unfavorable to a fisherman.
For a brief explanation on kurtosis, let's look first at a normal distribution (following chart), where returns cluster around the mean with 99.73% within and .27% beyond 3 standard deviations of the mean. So a tail outcome has only .27% chance of occurring.
When kurtosis increases while the standard deviation stays the same, more of the deviation is due to fatter tails in the distribution. This increases the chances of a tail event.
Probabilities that are extremely low are a special case and are sometimes overlooked. Harvard's Public Health Center for Risk Analysis calculated the probability of a fatal car accident to be .000149. Some take numerous precautions to reduce accident risk, yet some ignore this risk altogether.
In finance, fatter tails are regarded as undesirable because of the additional risk involved. Though infrequent, more results occur far away from the mean. A sharp positive return would be welcomed. But a sharp negative return could cause great damage to a portfolio. No one wants to increase their chances of a car accident. Further, a sharp loss has lasting adverse effects. A sharp gain/loss followed by a sharp loss/gain of the same percent does not get you back to breakeven in nominal terms.
Of course the inverse is also true as lower kurtosis reduces the chance of an extreme event. For the same amount of stock exposure and portfolio rotation, index market timing's kurtosis of -.40 was lower than stock picking's +.16. After repeating the simulation additional times, and then simulating with market data from other periods, we consistently found that index market timing produced lower kurtosis. The idiosyncratic risk of individual stocks is the predominant source of a fatter tail in this comparison study. Many pundits have proclaimed how foolish it is to time the market. We have empirical evidence that active managing in the time domain allows one to reduce the chances of an extreme outcome. The need to calculate tail risk is an important lesson learned from the recent market turbulence. Additionally, more institutional investors have adopted a more dynamic approach to asset allocation3.
The greatest stories told everyday are the fortunes of a stock. The discovery or re-discovery of new and old wonders is creatively foretold in a variety of daily narratives, making possible thought provoking discussion and interesting financial news. Economist Burton Malkiel stated that:
a blindfolded monkey throwing darts at a newspaper's financial pages could select a portfolio that would do just as well as one carefully selected by experts.
The Wall Street Journal ran an expert vs. random dart throwing simulation for 14 years, but declared no clear winner. Our study did not seek to find a winner, but to measure the risk sources between approaches (stock picking & index market timing) regardless of the investment competency of a manager. Overlooked by many, a key to investment success is keeping losses within a comfortable zone. Identifying risk sources and careful strategy selection is important to help reduce sharp negative effects on a portfolio. So investment approaches that minimize tail risk should be considered as a part of your portfolio.
1 using S&P100 constituents as of 12/31/2011. Did not account for constituent changes over 10 yr period for both simulations.
2 cash paid zero interest.
3 Fabozzi, Focardi, & Jonas (2010). Investment Management after the Global Financial Crisis.
Disclaimer: Article is for educational purposes only and does not constitute financial advice.