In the February 2011 CI Online Exclusive, I introduced Leslie Masonson’s Stock Market Dashboard from his book “Buy—Don’t Hold” (FT Press, 2010) (this article is available online at the Computerized Investing website). Masonson uses the Dashboard to gauge the market’s direction in order to decide whether it is time to buy, sell or sit on the sidelines. Research cited by Masonson provides a compelling argument for why it is useful to know where the market is headed: Fully 70% of an individual stock’s return is a result of the stock market’s overall trend. In other words, a rising tide raises all boats (and vice versa).
Masonson’s original Stock Market Dashboard used eight indicators that told him when to buy, hold, or sell. However, follow-up backtesting performed by Masonson proved how difficult it is to time the market. After publishing his book, Masonson discovered that the Dashboard would have failed to generate a sell signal during the market collapse of 2008–2009. If you develop a market timing system that would have failed to get you out of the market during the biggest market downturn since the Great Depression, it’s time to go back to the drawing board. And that is what Masonson did, revising the dashboard so that, according to him, it would have gotten you out of the market in 2008 while generating higher returns and smaller drawdowns. He released his findings on his website in March 2011 (www.buydonthold.com).
The original Stock Market Dashboard consisted of eight indicators:
In his Version 2 Stock Market Dashboard, Masonson uses only four of the original eight Dashboard indicators:
Masonson points out that three of the four indicators are based on the performance of the NASDAQ composite index. He has chosen to focus on this index because, he believes, the NASDAQ market tends to lead the overall market when changing directions.
While many investors track the 200-day moving average to gauge the trend in the market, Masonson uses a 100-day moving average. He does so because it generates quicker buy and sell signals, thereby producing better profits in both bull and bear markets.
When the NASDAQ composite index closes above its 100-day moving average, Masonson views this as a bullish signal, while index values below the 100-day moving average are bearish.
One pitfall of using a moving average as a trading signal is the potential for “whipsaws.” Masonson defines whipsaws as “buy and sell signals near a moving average.” When the market is moving within a trading range (moving sideways), moving averages tend to flatten out, increasing the likelihood that the index can alternate between closing above and below the line in short succession. Acting on these buy and sell signals can result in a string of unprofitable trades and higher trading costs. It is for this reason that Masonson does not rely solely on moving averages as a trading tool. In his opinion, using all four of his market dashboard indicators in conjunction will reduce the risk of whipsaw trades.
The moving average convergence-divergenceis a trend-following indicator developed by Gerald Appel. It is calculated by subtracting the 26-period exponential moving average from the 12-period EMA (periods can be days, weeks, months, etc.). Exponential moving averages place greater weighting on recent prices, making them more reactive to recent price moves relative to simple moving averages, which place equal weighting on each periodic price.
The indicator also consists of a “signal line,” which is a nine-period moving average of the MACD. The indicator generates buy and sell signals when the MACD line crosses the signal line—a crossover from below is a buy, and a crossover from above is a sell.
Since the MACD makes use of moving averages, it too is sometimes prone to generating false short-term signals. This again highlights the importance of using multiple indicators when determining when to buy and sell.
Beyond using crossovers between the signal line and MACD to identify possible buy and sell points, traders also look for divergences between the direction of the MACD and the underlying security or index. When you see the MACD making new highs, while the underlying security or index is falling, this points to a possible reversal in the index or security.
For an introductory discussion of the MACD, see this issue’s Technically Speaking column.
Investor sentiment is often used as a contrarian indicator—individual investors tend to be most bearish at market bottoms and most bullish at market tops.
One sentiment indicator that is widely followed by professional traders and the financial press is AAII’s Investor Sentiment Survey. Each week, AAII members can vote online as to where they think the market is going over the next six months—up (bullish), down (bearish), or sideways (neutral). The results are published each Thursday at the AAII website (www.aaii.com/sentimentsurvey).
The final piece of Masonson’s revised Stock Market Dashboard is the NASDAQ summation index. The NASI consists of two parts. The first is the difference between the number of advancing stocks and declining stocks in the NASDAQ composite index. The second is the five-day exponential moving average of the summation ( line. When the NASI line crosses above the EMA, this is considered a buy signal, while a crossover to the downside is a sell signal. As confirmation of the crossover buy and sell signals, Masonson adds an MACD of the NASDAQ composite index. An NASI buy signal is confirmed when the MACD line crosses above its signal line; likewise, an NASI sell signal is confirmed when the MACD crosses below its signal line.
Furthermore, when both the NASI and MACD are showing positive divergence with the NASDAQ composite—reaching new highs while the index is declining—Masonson considers this a very strong buy signal. Likewise, negative divergence with both indicators is considered a strong sell signal.
Just like with the original dashboard, Masonson assigns a +1, 0 or –1 value to each of the Version 2 Stock Market Dashboard indicators.
Figure 1 is a daily one-year chart from StockCharts.com for the NASDAQ composite index ($COMPQ) with a 100-day simple moving average. A buy signal (+1) is generated when the index rises above the 100-day moving average. A sell signal (–1) is generated when the index crosses below the moving average.
For the week ending February 17, 2012, the $COMPQ was well above its 100-day moving average, which it had crossed above in late December 2011. As a result, we assign a buy rating (+1) to this Dashboard indicator.
Figure 2 is a daily chart from StockCharts.com of the NASDAQ composite index ($COMPQ) along with the 50-day moving average and MACD. A buy signal (+1) is generated when the MACD line crosses above the signal line; a sell signal (–1) is generated when the MACD line crosses below the signal line. If no new signal has been generated in the last six months, a neutral signal (0) is generated.
For the week ending February 17, 2012, the MACD line (black line in lower chart) was slightly above the signal line (red line in lower chart)—60.229 versus 60.224—having crossed above it toward the end of December 2011. However, given the fact that the difference is so slight, we rate this a neutral signal (0).
Figure 3 is the weekly reading for the AAII Investor Sentiment Survey from AAII.com as of February 15, 2012. For Masonson, a buy signal (+1) is generated when the percentage bullish reading drops to 25% and then rises the next week; a sell signal (–1) is generated when the percentage bullish reading rises to 50% and then drops the following week; if no new signal has been generated in the last six months, a neutral signal (0) is generated.
As of February 15, 2012, 42.7% of AAII members voting were bullish on the direction of the market over the next six months, which was 8.9 percentage points lower than the previous week. Since the bullish percentage is above 25% but below 50%, a neutral signal (0) is generated.
Figure 4 is a StockCharts.com chart plotting the NASDAQ summation index ($NASI) and its five-day EMA in the upper pane and the MACD of the NASDAQ summation index in the lower pane.
A buy signal (+1) is generated when the $NASI rises above its five-day EMA and is confirmed by the MACD crossing above its signal line on or near the same day. A sell signal (–1) is generated when the index falls below the five-day EMA and the MACD has a confirming sell signal on or near the same day.
As of February 17, 2012, the $NASI was just above its five-day EMA (632.64 versus 632.40). However, its MACD line had crossed below the signal line earlier that week. Therefore, neither a buy nor a sell signal is being generated by this indicator, and we assign it a 0 value.
With the original Dashboard, Masonson established a singular composite indicator to help him decide when to buy and sell. He assigned a +1, 0 or –1 value to each of the eight Dashboard indicators on a weekly basis, meaning the composite could have ranged from –8 (most bearish) to +8 (most bullish). Composite scores of +3 or higher were considered buy signals, when you should consider having the majority of your money in the market. When the composite value dipped to –3 or lower, Masonson felt it was time to sell all equities and hold 100% cash until the score rose back to +3. Readings between –2 and +2 were considered “indeterminate,” resulting in no action at all until the market indicated a pronounced up- or downtrend.
For Version 2, Masonson no longer adds the individual indicator values together to arrive at a composite net signal. Instead, if three of the four indicators have +1 values, the Version 2 Dashboard issues a “buy” signal. If three of the four are –1, this is a “sell” signal. Otherwise, no action is taken.
As of February 17, 2012, only one of the Dashboard indicators showed a buy signal (NASDAQ composite index 100-day moving average crossover), while three were neutral (MACD on NASDAQ composite, AAII Investor Sentiment Survey bullish percentage and NASDAQ summation index moving average crossover with MACD confirmation). Therefore, no action is indicated at this time.
To quote from the conclusion of my original Stock Market Dashboard article, “no market timing system is 100% perfect,” and Masonson’s backtesting of his original Dashboard proved this point. While some systems and market timing “experts” claim to be able to pinpoint market tops and bottoms over a few market cycles, it is virtually impossible to consistently predict them over the long term. This is because each market cycle is unique. As a result, systems should be monitored and, if need be, adjusted over time. By revising the Stock Market Dashboard, Masonson hopes to generate better-timed signals that will ultimately generate additional profits.