Value investors in the stock market spend a great deal of time studying the words and actions of Warren Buffett. He has delivered amazing returns in the stock market for decades and created great wealth for himself and others. Buffett applies the principles outlined in Graham and Dodd’s “Security Analysis” better than anyone ever has. After studying under Benjamin Graham, Buffett obviously improved the ideas in the book and earned more than any other investor in history. Students and practitioners spend hours analyzing every word that Buffett utters, but traders have tended to believe they couldn’t learn from a master of the buy-and-hold philosophy. Nothing could be further from the truth.
Buffett likes to say, only half-jokingly, that his favored holding period in a stock is forever. Occasionally he does seem to sell, with Moody’s offering an example of a stock that he has recently been selling. When he originally bought the stock, it looked like a good fit for Berkshire Hathaway. Moody’s was a financial services company that had a secure business model as a government sanctioned ratings agency. In hindsight, the ratings agencies weren’t that smart, and investors in them including Buffett have suffered losses. Over time, Buffett has reduced his position in the stock, and seems like he will eventually sell all of his shares. The reasons he had for originally buying the stock no longer apply, and rather than argue with the market, Buffett is cutting his losses.
Even though Buffett hopes that he’ll never sell, he does seem to have a plan for selling. It looks like he sells a stock when the business changes, as is the case with Moody’s. He’ll also sell to take advantage of what he considers to be a better investment opportunity. A few years ago, when he bought the Burlington Northern railroad, he decreased several positions to raise cash. Subsequent filings with the Securities and Exchange Commissions show that he bought back some of the stocks he sold as the mountain of uninvested cash he holds increased.
Traders look at things in a different time frame than Buffett does. But there are valuable lessons that short-term traders can learn from the master of long-term investing. The first lesson is that knowing when to sell is important. Traders should have the sell levels defined before they even enter the trade.
Minimizing trading costs is also important. Buffett buys and sells some of his positions over the course of several months. He’s a very large investor and a simple market order from him would move the price significantly against him. That’s a trading cost he can avoid by buying in smaller pieces and using patience to acquire a position over time. A trader buying or selling a single emini contract on the S&P 500 won’t necessarily move the market, but should still pay attention to trading costs. Commissions add up on trades, and these costs can have a very large impact on long-term performance.
To understand the impact of trading costs, we will look at test results from a simple 10-period, long-only moving average system. All tests will be run on the SPDR S&P 500 ETF which trades under the symbol SPY. Unless otherwise noted, trading costs of $10 per transaction will be subtracted from each trade.
Looking at a 10-minute timeframe, the results would have been disastrous. From 2006 into August 2011, the system would have lost almost $90,000 over five years on more than 4,400 trades. Only one in six trades, on average, would be a winner. Results improve quite a bit if we can ignore trading costs. Without commissions, the system breaks even.
Moving up to hourly bars, the results improve some, losses only total $13,000 in five years and one on four trades is a winner. Daily bars decrease the loss to only $10,000. With weekly bars, the system breaks even after trading costs. Monthly bars offer an annual profit of about $1,000 a year and 57% of trades are winners. On average, you only trade once a year with monthly bars instead of almost 900 times a year on the smallest time frame. On any time frame, you would be in the market about 55% of the time and in cash the rest of the time.
These results are interesting. First, we notice that Buffett does seem to be onto something. Trading activity isn’t necessarily a good thing. Brokers do make money every time you buy or sell and overcoming these costs is a significant hurdle to success.
A moving average system doesn’t do well in the shortest time frame for two reasons. First, the costs are high with the system needing trading profits of nearly 100% a year just to break even after commissions. Second, this idea isn’t appropriate for an intraday traders. The moving average tries to capture trends, and in the short term, which includes a 10-minute chart, stocks are mean reverting. Therefore, the results really aren’t unexpected, even before considering trading costs. Systems need to begin with a logical rationale, and there was none in this case. These results were intended to show that trading costs can be significant.
It did pretty well in the long-term. Monthly performance helps identify a possible way to boost trading profits. The 10-month moving average is a well-known strategy for long-term investors and is extensively detailed in “The Ivy Portfolio” by Mebane Faber and Eric Richardson, a book written in 2009. The general ideas and results of their tests are also available in a paper that can be downloaded for free at the Social Science Research Network web site. Faber’s “A Quantitative Approach to Tactical Allocation” is at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=962461. Searching for technical analysis topics at that web site will honestly provide hours of entertainment for geeks, and many profitable ideas to traders.
The greatest benefit of using the 10-month moving average is in avoiding bear markets. It’s a timing tool to avoid the worst part of the steepest stock market declines and has been in cash for most of the bear markets since the the S&P 500 index was first introduced. In Faber’s paper, he shows that from 1973 to 2008, this simple timing model beat the S&P 500 by a percentage point a year and reduced drawdown by half. The worst drawdown with the 10-month moving average was only 23.3%. Similar results are shown for international stocks, government bonds, commodities, and real estate.
We started by pointing out that Buffett thinks long term and traders can benefit from this mind set as well. Small traders will need to use leverage in their accounts to make returns they can live off. Futures trading is usually done in small accounts holding cash for the margin balance. Instead of holding cash with a very low return, an ETF that is above its 10-month moving average could increase the returns of a small account. Real estate investment trust ETFs have very nice risk/return profiles and might be a good alternative to a stock fund.
Buffett wants to hold forever. And if the ETF stays in a bull market, this strategy will allow you to hold and profit from that trend forever. But Buffett does have a selling plan, and the discipline of the system rules provides a selling plan for traders. This long-term strategy, cross margining an ETF with futures, also offers a way to increase potential returns without dramatically increasing risk.
The bulk of the trading account gains would still come from a short-term strategy. As I’ve written previously, short-term can mean daily strategies in the futures market. Combining the 10-month moving average system with a 20-day breakout system on a diversified basket of commodities can provide a steady equity curve and profits that make it possible to trade for a living.
The 20-day breakout strategy minimizes trading activity while delivering decent returns with lower than average risk. Managing trading costs is as important as managing risk, especially in small accounts.
Small accounts will not immediately see Buffett-sized gains, but Buffett took decades to become a billionaire investor. He’s worth studying, even for those aspiring to become day traders.
By Michael J. Carr, CMT