Trading Pullbacks in Wall Street’s Best Stocks

There are two common approaches to trading: trading stocks that are breaking out (championed by IBD among others) and trading stocks that are pulling back (championed by L. Connors among others). As a professional statistician, I’ve been long interested in these two approaches as they apply to fundamentally sound stocks, especially those with remaining value in their price. My work over the years shows that this select group, also, is more profitably traded utilizing some sort of pullback strategy. Let me share an eye-opening example.

The following chart shows the performance of the two investment strategies over a 479 week period (9.2 years from 12/17/99 to 2/13/09) through both bullish and bearish times. The S&P 500’s performance is included as a benchmark (and the y-axis is a log scale). Each strategy begins with $10,000 and has a one-week hold period. Over this period, the S&P’s equity curve dropped to $5,587, while the first strategy’s equity curve rose to $2,096,347 and the second one dropped to $2,127. So what are these two strategies that differ so drastically in their performance? Surprisingly, both are trading strategies limited to fundamentally sound stocks.

Pullbacks chart

Both begin by requiring their stock membership be limited to those having a Zacks ranking of 1 or 2 (~850 stocks rated as either a strong buy or a buy). Both require their respective picks to be in the top 50 percent in terms of price performance of this group over the past 24 weeks and, from that group, the top 50 percent price performance over the past 12 weeks. The two then differ in their fourth requirement: the strong performing strategy picked the worst 10 performers over the past week (often those in strong pullback), while the weaker performing strategy picked the top 10 performers over the past week (breakout and strong momentum risers). These are both sets of fundamentally sound stocks that have performed well over the past three and six month periods then either pulled back for the last week or continued their strong run higher.

Those ten exhibiting the worst pullback each week outperformed the next week: taking their equity curve from $10,000 to $2,096,347. The second group, by contrast lost over 75 percent of the original $10,000. No doubt, if you had stayed in the trade longer this second group might have performed better, but I think this is clear evidence that entering on a pullback is far superior to entering on a breakout for the short-term trader.

Larry Connors’ group has written much about the usefulness of the two-period RSI for qualifying stocks and ETFs in pullback. Its formula simplifies to

RSI(2) = 100 * (Adv2)/{ (Adv2) + } where (Adv2) and (Dec2) are the average advance and declines in price change over the two periods expressed as absolute values. After three price data, if price advanced both times, RSI(2) = 100; if it declined both times, RSI(2) = 0; and if it advanced once and declined once, RSI(2) = the percentage advancement. After the fourth price is collected, the current and are averaged with the newly calculated ones and then fed back into the RSI(2) calculation. Thus, RSI(2), in calculating the percentage price advancement over the past two periods, is a measure of the extent of breakout (high percentages) or the extent of pullback (low percentages). Multiple, consecutive down or up trends in RSI(2) is a measure of the extent of the pullback or breakout.

In this study, one to four consecutive daily RSI(2) values falling below 25 were used as a trigger for a buy and, once the trade was entered, a RSI(2) value greater than 70 was used as an exit signal. All entries and exits were made at the close of the day.

Twenty stocks were chosen to demonstrate these strategies executed between January first and July seventeenth of this year. The 20 were fundamentally sound stocks with value remaining in their price, as determined by a series of fundamental screens and their, respective PEG ratios. Each had been a TSM pick over the past month. In the 80 sets of results shown in Chart I (20 stocks times four separate strategies), only once did a loss result and that was limited to -0.03 points.

ES chart

Requiring RSI(2) to end the day once below 25, buying the close, then closing the position at the close over the next few days when RSI(2) rose to 70 or beyond resulted in a 83.84% win rate for 229 trades with a total of 253.17 points earned (1.11 per trade). Requiring RSI(2) to end two consecutive days below 25, then entering and closing the position as above, resulted in a 87.83% win rate for 115 trades with a total of 132.16 points earned (1.15 per trade). The win rated climbed to 91.67% for the last set of data where four consecutive RSI(2) values below 25 were required for entry, and the number of trades fell off to 36, but a similar 0.96 points per trade continued to be earned.

While each of these four strategies provided very good trade entry and exit points for these quality stocks, I prefer the combination win rate to number of trades offered by the second strategy.

Finally, let’s consider how buying and selling pressure changes for a quality stock that still has value at its current price. Simply put, once identified, everyone wants to own it (that’s you and me as well as institutions) so buying pressure builds and drives price higher. Price is driven higher to a point where it becomes overbought (often over a short period of time): investors stop buying; traders sell their positions and short sellers even step in. Selling pressure is able to temporarily win out, and price falls. The pullback begins, but all the while, these three groups (institutions, investors and traders) look for a point to re-enter: at the support of a major moving average (20-, 50- or 200-day), at a prior price reversal (earlier valley or hill) or at a major Fibonacci level (38.2, 50 or 61.8%). Though my work has shown there’s something symmetrically pleasing about the Fib levels that causes people to react there, these areas of support at not magical, just self-fulfilling because smart money reacts there. Trading (or investing in) quality stocks in pullback presents a low-risk, high-profit opportunity for the individual as well as the institutions.

Click here to read the continuation to this article, as Richard Miller takes a look at thirty-two stocks to demonstrate another strategy aimed at trading pullbacks.

Richard Miller, Ph.D. – Statistics Professional, is the president of and .

Get the best ETF setups before Wall Street – click here.