Traders are usually focused on maximizing profits and moving averages (MAs) are one of the tools some traders use to meet that goal. An MA is applied to smooth price data and help to identify the trend. One of the earliest references to this strategy can be found in the classic technical analysis text book, Technical Analysis of Stock Trends by Robert Edwards and John Magee. In the first edition of their book, in 1948, they wrote:
And, it was back in 1941 that we delightedly made the discovery (though many others had made it before) that by averaging the data for a stated number of days…one could derive a sort of Automated Trendline which would definitely interpret the changes of trend…It seemed almost too good to be true. As a matter of fact, it was too good to be true.
Edwards and Magee mused if this worked they move to an island and trade from the beach. A little testing forced them to abandon that dream. While MAs are useful, they are not the Holy Grail of investing. Before looking at the best way to use an MA, let’s review what an MA is.
An average is “a single value (as a mean, mode, or median) that summarizes or represents the general significance of a set of unequal values.” The value is calculated by adding the data points we’re working with and dividing that summed value by the number of items in the sample. In the stock market, prices change over time and the average can be taken over any consecutive period of price changes. This is a moving average, where the sample moves forward in time and the divisor in the calculation remains fixed.
The moving average is designed to smooth the short-term trends and filter out volatility, helping traders spot the longer-term trend. The direction of the trend is defined by the position of the last data point used in the calculation relative to the MA. If the most recent price is above the MA the trend is up. A downtrend occurs when the most recent price is below the MA. The chart below shows the 50-week MA would have allowed traders to hold positions throughout the long uptrend.
The next chart shows traders would have less clarity with a shorter-term MA.
These two examples highlight one of the challenges associated with MAs – which time period should we use? There is also the challenge of determining how to calculate the MA.
Many traders have attempted to refine this simple tool by adding varying degrees of complexity. In addition to the simple moving average, analysts can apply exponential, weighted, or adaptive averages. All, to some degree, overweight or underweight part of the data as opposed to the simple moving average which equally weights all data.
The exponential moving average (EMA) overweights the most recent price data, which means it tries to stay closer to the price action. This attempts to address one of the most noted flaws of MAs which is that they significantly lag behind the price in trending markets. Lags lead to delayed trade signals that mean missing big moves that develop after a trend reverses. The EMA is designed to minimize the lag between the current price level and the moving average, theoretically making it possible to take quicker action on trend reversals. The formula to calculate an exponential moving average is:
EMA = (Weight * Close) + ((1-Weight) * EMAy)
where Weight is the smoothing constant selected by the analyst
Close is the closing price of the security being studied
and EMAy is the value of the EMA from yesterday
The weight is calculated as 2/(N+1), where N is approximately equal to the value of the time period one would use in a simple moving average calculation. A common weighting value is 0.181, which is close to a 20-period simple moving average; or 0.10, which is approximately a 10-period moving average.
Other types of weighted moving averages are developed using this same principle. They either overweight or underweight selected data points. A front-weighted MA gives more importance to the more recent data. A back-weighted MA underweights the most recent data.
The Encyclopedia of Technical Market Indicators includes test results of hundreds of technical indicators and concluded more complex MAs have “considerable intellectual appeal” but “tests fail to show any real practical advantage.” Testing generally shows more complex calculations offer better results sometimes but at other times the simplest calculation works best. Overall, the results obtained with the simple MA tend to be similar to the variants over the long-term.
Most of the refinements in calculating MAs are designed to reduce the lag, or the distance between the average and the current level of prices. Unfortunately, none of this math solves another shortcoming of MAs which is that following their trading signals will inevitably lead to a large number of losing trades. In New Concepts in Technical Trading Systems, Welles Wilder estimated that all markets show a significant and tradable price trend approximately a quarter of the time. Most markets spend about 75% of the time confined within relatively narrow ranges, when MA buy and sell signals will be repeatedly generated as prices rapidly move above and below the moving average. No matter how it is calculated, any MA will be prone to this problem, known as whipsaw trades, most of the time.
Despite the large number of losing trades, some MA strategies have been shown to work well. The 10-month MA is a popular strategy that would have helped traders avoid the bear market in 2000 and in 2008. As global stock markets declined by 50% or more, investors following this strategy would have been comfortably out of the market.
For the ten years ending in December 2009, this strategy showed an annualized gain of 14.71% while the S&P 500 was virtually unchanged. Books and articles were written explaining how well this strategy worked but those authors ignored some important facts.
Over those same ten years, the strategy showed a loss when applied to the NASDAQ 100, NASDAQ Composite index, the Dow Jones Industrial Average or the Russell 3000 Index. When stocks struggled through a bear market in the 1960s and 1970s, a buy and hold investment in the S&P 500 would have performed better than the 10-month MA.
The conclusion for the 10-month MA is the same we find for any other MA — works well sometimes in some markets.
The most apparent advantage of the 10-month MA is that it avoids extended bear market trends and can help investors avoid large losses.
For long-term investors, avoiding losses is as important as capturing gains. As an example, the S&P 500 increased by more than 80% in the year after the March 2009 low. After that gain, the buy-and-hold investor would still have an account balance more than 20% below the high seen in the summer of 2007. Avoiding the loss would have helped them reach a new high in their account value much sooner. This is because losses and gains are asymmetric, with bigger percentage gains being required to make up for losses.
MAs can help avoid losses by alerting investors to move to cash during extended market declines. They can also provide signals for getting back into the market. But they shouldn’t be used as an “all or none” indicator. MAs work best when combined with other indicators as part of a complete trading strategy. For example, MA could be combined with a stochastics indicator. Stochastic sell signals might be ignored when price are above a long-term MA since they are likely indicating a short-term pullback. But a stochastic sell signal while prices are below the MA could be an indicator to increase cash since this is likely to be a significant pullback or even a bear market.
Traders are likely to discover the value of MAs is in decreasing risk rather than maximizing gains. A break of a long-term MA is a warning to be cautious. That simple message could save investors thousands of dollars when the next bear market comes.