Relative Strength Trading Explained
Karabi - Sxoinias 1999
http://billusko.tripod.com/kallithearmwrestlingclub/id4.html
October 11, 2005 Note: While FastBreak limits you to 10 stocks and/or funds, the Pro Version has essentially no limit. Using it, I have developed systems holding as many as 30 things.
Definition of Relative Strength Trading
This is the definition of relative strength trading,
(1) Select a list of K funds or stocks.
(2) Select the number (N) of things you want to hold at a time.
(3) Select the cut-off point (M), where K > M > N.
(4) Select the method of calculating relative strength.
Rank the stocks in the list, where the lower number is the higher rank. This means that the funds are ranked 1 through K.
Buy the top N funds.
When a fund is ranked below M, sell it and buy the top-ranked fund that you do not already own. It is possible to have a different method of ranking for the buy and sell decision. So you would sell when the sell rank falls below M, and buy based on the buy ranking list. You might also, in order to minimize the frequency of trading, insist on a minimum number of days to hold a thing.
Ranking Method
There are (at least) the following ways to measure relative strength:
(1) Ranking of a collection of stocks or mutual funds - which of them have gone up the most over the last N days.
(2) Slope: fit a curve to the stock chart and take the derivative at the end point. The curve could be linear, quadratic, exponential, or something else. Your choice.
(3) Most Anchored Momentum (MAM): compute a short exponential moving average and a longer simple moving average. Take the ratio at the end point (present time). This was developed by Rudy Stephanel in an article in Technical Analysis of Stocks and Commodities.
(4) Exponential moving averages: a short one and a long one - take the ratio short to long at the end point.
(5) Accutrack: this is a proprietary indicator developed by Fasttrack. It compares two funds based on their recent returns. It can be used to give you a visual comparison of a fund to a money market account, for example. If Accutrack is positive, then the fund is a candidate to buy. Using Fasttrack, you can compare a set of funds, two by two, taking the top fund every time to compare it with the next fund on the list. In this way you can rank a list of funds.
The first method (1) is the weakest of all of the methods, since it does not take into account the direction of the thing you are trading. A stock that has moved from 50 to 100 and then back to 75 over the last 30 days, for example, will probably have a good ranking - after all it is still up 50%, and that will probably be better than most stocks. But a look at the chart might indicate that the stock is on a downward trajectory. Methods (2), (3), (4) and (5) attempt to measure the current direction of the stock, and so are likely to be more reliable.
There is an excellent presentation on MAM by Rudy Stefenel that also looks at other methods of ranking, and explains the rationale behind MAM.
FastBreak supports all of those ranking methods. So far as I can see, exponential curve fitting, MAM, and Accutrack are the most powerful. However, MAM seems to be the best choice, particularly for stocks.
Note that FastBreak also allows you to have stops and buying "pre-conditions", called "flow control". This means that the number of combinations of parameters in a system is essentially unlimited. The "Pro" version of FastBreak has a genetic search algorithm which will help you to find the best parameters. See the FastBreak site for a more detailed description.
Predictive?
This method does not try to predict a market trend before it happens. Sometimes people analyze of the market and they say that because of certain fundamental factors - e. g. inflation, world economic problems - a certain thing will be a good investment, and will move up in the future. The relative strength method, on the other hand, only looks at a current trend and hopes that the trend will continue. It is like betting on the horse that is ahead, holding your bet so long as he is in the running, and then if he falls back, betting on the horse that has taken his place at the front. Of course, you cannot exactly do that at the racetrack.
Tools and Cost (Cost figures as of Jan 21, 2004)
There are three basic computer tools:
Fasttrack, a database of mutual funds and stocks - $336/year for mutual funds, $588/year for both funds and stocks.
FastBreak, a tool to develop relative strength trading systems - $499 for the simple version, and $1900 for the "Professional" version. One time charge.
Trade, a programming language which can be used to calculate indicators, perform market timing, and also develop relative strength trading systems - free.
How much time does this take? Often people say that they simply do not have the time to trade stocks, or anything else. This is utter nonsense, and some of the people that have said this are living experts on re-runs of Lost in Space, or have memorized all the lyrics of the Red Hot Chili Peppers, so they do not get any sympathy from me.
So long as there are no net problems (and there seldom are), here is the schedule:
Download the Fasttrack database – 2 minutes at most (after 7:30 Eastern Time).
Run any Trade programs you may have developed for signals, etc. – 5 minutes should be plenty.
Run your FastBreak systems to see if any trades are called for – 5 minutes at most.
If trades are called for, log onto your broker the next day, and input the trades – the amount of time depends on the number of trades.
If you are on travel, as I used to be when I was working as a Systems Engineer in the standards area, pick a hotel with high speed Internet. Marriott Courtyards are offering that now, as are some others. Transfer the stuff to your laptop (or always use a laptop - some are extremely powerful now), run the systems and trade from the hotel.
How long does it take to develop a system? If you look at my page on suggestions on how to trade stocks, you can see some guidelines, and you can develop a decent system in less than an hour, depending on how long it takes to download the lists of stocks. If you have the Pro version of FastBreak, the computer will do the work for you in several hours, depending on its speed.
When you are trading stocks, you do need to check the news before trading. There may be something in the news which would indicate that it does not make sense to buy a stock no matter what its rank. A good example is that if someone offers to buy out a company, the stock will take a sudden jump and perhaps appear at the top of the list. However, after that it is likely to flat-line, particularly if the offer is all cash and people believe that it will likely be accepted. Of course, you might also read that some of the officers have left on an extended vacation to Brazil.
If you really do not have much time, or cannot trade stocks during the day, then just use TimingCube or some other good timing system and you only have to trade 2-3 times per year.
If you really do not have the time or inclination, then Fundbuster is where you should go. Buy rather than build. Harry Larson has been extremely successful in developing systems, and the subscription price is definitely right. You can join his site for a nominal fee, and he will inform you whenever any of his systems make a trade. Check it out, and look at his links as well. They are very informative. At this point, I recommend trading one or more of his hedging programs, since it is becoming impossible to trade funds actively.
If you want somebody to manage your money, my friend Dave Serbin has an excellent record and is ready to offer that service. Take a look at his Yahoo Group, and see the model portfolios and how they have done. The group is free, of course, although you have to sign up with Yahoo.
Results???
Of course, good results cannot be guaranteed. If you optimize a system over a period of several years when the market has done very well, and you get a backtested result of, say, 50% per year, you cannot expect to get 50% every year. What you gain through the development of a system is a way of making decisions that has worked in the past. The next year you might get 60%, or you might even lose money. It all depends on the market conditions. After all, if the best fund in your family only goes up 5%, you can hardly expect to do much better than that. The key is to make the best use of the opportunities that the market gives you.
There are a lot of people who do not believe in backtesting. They say it means nothing. But what is the alternative? A backtested system is chosen because it has worked in the past. Should you choose something that has not worked in the past?
[ Top ] [Introduction] [The Dogs] [End of Funds?] [Couch Doggie Portfolio] [Pick Good Funds] [Stock Trading Systems] [ Trading Systems ] [Simple and Profitable] [Journal] [ Timing ] [Selection ] [ Common Sense ] [ Statistical Noise ] [ Mechanical Hibernation ] [Best Ever ] [ LTBH Results ] [ Beat Market ] [ Technical Voodoo? ] [ Charts Do Not Lie ] [ Bad Advice ] [ Analysts ] [ Gamblers, etc. ] [ Relax ] [Links ]