Thursday, April 11, 1996
Before you sit down hoping to read a tale of my three body-building brothers, of whom I have none (sorry guys, muscles weren't our forte), let me tell you that what I'm really going to discuss actually has something to do with stocks. Many of you have followed the ongoing research into my Investing For Growth model that we've published as a primer (available in FoolMart -- get 'em while you can). In fact, many of you have been working on improvements to the model as well. I want to talk about one variation today that I'm excited about -- one based on Relative Strength.
If you're an IFG follower, you know the chief complaint with the approach is that it tends to hold on to a winning stock much too long after it ceases to rack up gains. In other words, IFG hasn't learned yet when the party's over. This is obvious in two particular cases in the 16-year history, both of which stripped hundreds of thousands of dollars off of the model's paper profits. I'm referring to US Surgical in 1992 and Micron Technology in 1995/6.
Both of these stocks were multi-baggers for IFG at one point. But in both cases, the model didn't kick the stocks out until long after they had given back 50%-70% of their massive gains. Value Line's ranking system is slow to downgrade a stock until it starts posting disappointing earnings, but in some cases, the stock's momentum is long gone before this actually occurs.
To compress several months of thinking into a paragraph, I've decided to try a different momentum factor in the IFG model. Keep in mind that the original IFG model is based on both fundamental factors and momentum factors (the Value Line Timeliness ranking over the next 6-12 months and the Industry ranking). Value Line's fundamental record has been proven over 30 years, so that isn't the problem, but the momentum element could stand a little fine tuning. That's where the Relative Strength ranking (found in Investor's Business Daily each day) comes in. By replacing the Industry Ranking screen with IBD's Relative Strength screen, we have a much more sensitive momentum element to add to Value Line's fundamental screen.
So how does it work? It's still as simple as ever. Take the 20 or so High Growth stocks from Value Line which have #1 rankings for Timeliness and then rank them by Relative Strength. Buy the ten with the highest rankings as long as the RS ranking is at least 80. (Buy fewer stocks if you like or if you're using this approach as only a portion of your overall portfolio). Then choose a holding period. I'm following both a quarterly and a monthly holding period in my database, but choose the holding period most comfortable to you. It's possible that a monthly turnover generates higher returns (it has so far in 1996), but it also means you trade more often, meaning higher commission costs and more spreads to overcome. At the end of your holding period, re-evaluate and sell the ones no longer on the top list and replace them with the new stocks, spreading your cash evenly across the new positions. (Don't readjust your holdings you keep; let those winners run as long as they stay on your list.)
Other investors are trying various sell disciplines, such as selling a stock whenever its Relative Strength ranking drops below 80 or 70, and this may work well, although the drawback to such an approach is that you must pay attention to the rankings on a daily basis. By adopting a comfortable and automatic holding period, you're still free to enjoy life without worrying about your portfolio daily. And with a reasonable holding period of 1 to 3 months, you're not as likely to have total stock collapses which stretch out for months before you get rid of a loser.
I'm in the process of gathering data to back-test this approach for as many of the 16 years as possible. Right now, we've only got the data completed for last quarter but should have 1995 data available soon. I'll continue to track the original IFG approach in our online IFG Area, but look for frequent updates to the Relative Strength variation in the IFG folder on the message boards. As always, as soon as I have any data, I'll post the results for all to see. Fool on!
Transmitted: 4/12/96