Moving averages: long on talk – short on action

Anthony Trongone, Ph.D., CFP, CTA, dr.trongone@gmail.com

Although the technical literature is replete with articles discussing the efficacy of moving averages to forecast future stock prices, after analyzing the results of a 3, 5, 8, 13, 21, 50 and 200 day simple moving average, they were extremely disappointing. In looking back at 643 trading days (January 2, 2002 – July 21, 2006, for a daytrader taking a long position in the cues (QQQQ), there was not a single success story – across all seven timeframes.

So why do we continue to subscribe to trading moving averages?

A moving average is the best known smoothing technique. Although the timeframe can be flexible, the more days in your analysis the less emphasis that is given to the more recent performance. Despite the timeframe, however, a simple moving average distributes the same weighting for each score.

Its predictive function is to follow emerging patterns of success or failure, but since each score gets the same weighting; it is always below the actual price when they are rising, but above the actual price when they are falling. When rising prices begin falling or falling prices begin rising a moving average is slow to respond to these sudden changes; therefore, it is a “lagging” indicator.

Traditional 200 day simple moving average

In using the traditional moving average approach found in books on technical analysis, you take a long position when price (blue line) rises above the 200 day SMA (red line); but offset your position when the price descends below the red line. One drawback comes when the cues are trading away from the 200 day SMA (1). Since the system restricts you from buying the cues until its price crosses over the smoothing line, you are missing a large portion of the upswing.

In shorting the cues, you sell when the price falls below its 200 day SMA; but offset your position when it breaks above this lagging indicator. Unfortunately, in applying this system, you miss the opportunity to short the cues at its peak (2). When you finally place your short position in April -05, the cues fall (3), but recovers back to the red line (4), such an occurrence completely erases your earlier profits.

This article, however, uses a different decision making process, which resembles the procedure found in business books on quantitative analysis. Instead of using a crossover technique, it allows the individual to enter into a position at the opening of trading (9:30 ET); whereas, offsetting the position at the close of trading (16:00 ET)

Cues (symbol: qqqq)

With average daily trading volume of 97 million shares, the cues (QQQQ) are a good representation of investor psychology. They represent the intraday movements of a price index (NASDAQ-100); however, they trade as a single security, allowing investors to participate in the collective performance of a portfolio of 100 companies. With such diversification, a large percentage decline in a particular company has a less dramatic impact on the cues. There are, however, more compelling incentives for trading the cues, such as tighter spreads, a longer trading session, and short-sellers can take a short position on a downtick. And, since it has an almost perfect correlation with the mini nasdaq 100 futures, trading can take place throughout the night.

According to the New York Times (July 21, 2006, p. C9), the Nasdaq composite index receives 41% of its value from technology companies.

In this example, we will calculate a 3 day simple moving average (3 DAY SMA) by taking the average of the opening price (9:30 a.m.). Although many people calculate the closing price; thereby, generating a signal at the close of trading, in this study we restrict our trading to the regular trading session.

BUY SIGNAL – when the opening price is > 3 DAY SMA

SELL SIGNAL – when the opening price is < 3 DAY SMA

07/18/06 RESULTS:

OPENING PRICE = 36.220

3 DAY SMA = (36.220 + 35.955 + 36.310)/3 = 36.162

SINCE 36.220 > 36.162 = BUY SIGNAL (OPENING PRICE > 3 DAY SMA)

BUY AT OPEN (36.22)

SELL AT CLOSE (36.15)

P/L = $7 LOSS ON A 100 SHARE LONG POSITION IN THE CUES

07/17/06 RESULTS:

OPENING PRICE = 35.955

3 DAY SMA = (35.955 + 36.310 + 36.660)/3 = 36.308

SINCE 35.955 < 36.308 = SELL SIGNAL (OPENING PRICE < 3 DAY SMA)

SELL AT OPEN (35.955)

BUY AT CLOSE (36.030

P/L = $7.50 LOSS ON A 100 SHARE SHORT POSITION IN THE CUES

Cues performance

Since any system appears flawless if the instrument you are studying has been moving in the direction of your trading advice, it is important to monitor its performance. On January 2, 2004 the cues began the year trading at $36.79, the closing price on July 21, 2006 was $35.70; therefore, the price change was negligible.

Buy signals

The table indicates the results of the 3, 5, 8, 13, 50 as well as the 200 day simple moving average. Across all seven timeframes the results put you in a losing proposition. In fact, three of the timeframes (8, 13, 200 SMA) were unable to produce a .500 percentage; whereas, the best winning percentage was 52% (155 – 143). The four lowest SMA’s had an overall loss surpassing $10; whereas, the 3 day SMA’s loss was $12.15.

The average loss of the lower SMA’s were similar: losing approximately $3.70 on a 100 share long position in the cues during the regular trading session (9:30 – 16:00 ET). The best performance was an average loss of $1.16 for the 50 day SMA. From interpreting these results it appears as if we can improve our success by incorporating more trading days, which puts less emphasis on recent trading activity.

Short position

This table showing short positions demonstrates the success of using a simple moving average when a short position was taken during the previous 643 trading days. In a short position, a winning trade occurs when the price of the cues declines during the regular trading session. Although each timeframe was profitable, the results were unimpressive.

According to the statistics, a longer timeframe produces better results. The average return was more profitable as well as generating a better winning percentage. The 200 day SMA gave us the best performance for shorting 100 shares of the cues. In comparison to a long position, there were significantly less shorting opportunities in the 200 day SMA category.

Conclusion

In looking at the previous 643 trading days, a simple moving average across different timeframes was unable to improve our predictive ability. But despite substandard results, you can still construct a system which will enhance your trading.

Since each security has its own best formula for success, it is important to perform a statistical analysis to produce the most profitable solution. And, it is necessary to occasionally run an analysis monitor the effectiveness of your parameters, because the success is often fleeting.

Although it appears as if moving averages are ineffective, in my next article, we will begin to examine some of the ways in which they can be applied to strengthen your forecasting acumen.

Anthony Trongone, Ph.D., CFP, CTA

Dr. Trongone is the Director of Executive MBA programs in China + Taiwan for Centenary College.

Please contact him by writing: dr.trongone@gmail.com

For further discussion join our forum and share your ideas!