Preston, I think you are really on to something with your recent back testing. On one hand, your data indicates that the RS column direction is much more significant than the RS mode (RS buy or RS sell). On the other hand, the RS study mentioned in Tom's book indicates that stocks on an RS Buy significantly outperform stocks on an RS Sell signal.
The big difference is that your study is taking place with the DJIA in the 7000 to 9000 or so range. The Moody Parker study took place with the DJIA in the 2000- 3000 range. As the DJIA climbs, RS mode becomes less significant.
I looked at the Value Line stocks that we screened for VLPNF20 that didn't make the cut because the RS was on a sell signal. In all cases, RS was in a column of X's, the stock was trading above its bullish support line and the last P & F signal given was a buy signal:
ADC Telecom (ADCT) - Looking at the chart, a move up to 39 would be very powerful. In a flat market, this stock will not give an RS buy signal until it hits 50!!!!
Alcatel (ALA)- Another telephone stock on the move. Looks like a good buy on a pull back to 42. Everything else being equal, RS won't give a buy signal until 58!
Air Touch (ATI) - And yet another telephone stock on the move. Would've been a good buy at 53 or 54. It won't give an RS buy signal until it hits 62.
On the flip side, check out AMAT. AMAT is on an RS buy. The current RS is 3.45. The RS needs to fall to 2.0 to give an RS sell signal.
Even with the likely AMAT excitement Monday, it will end the day an RS buy unless it goes down to 18. It could lose 1/3 of it's value Monday and still be on an RS buy!!!!
Bottom line - Until DWA and Chartcraft either shift to the S&P500 or come up with a different multiplier, We should probably ignore RS Buys and RS Sells and instead concentrate instead on whether or not the RS is in a column of X's or O's. The effect of the rising DJIA has, of course, also slowed down the response that shifts RS columns. A three box reversal at RS = 6.0 is a lot slower than a three box reversal when RS = 18.0. If the DJIA keeps rising and we don't change our method of calculating RS, it won't be long before we say that "RS column shifts" last from 2 to 2 1/2 years on average and that "RS buy signals" last from 5 to 10 years on average.
Henceforth, we will ignore the RS buy/sell mode for selecting VLPNF20 stocks and will sell a holding whose RS shifts to a column of O's.
Sorry for being so longwinded my friend. Document your back testing carefully. You are definitely on to something.
later
duke60 |
