My Home About Me

Understanding The Efficiency Portfolio Testing
By Van K. Tharp, Ph.D.

Next week we’ll be doing another update to the IITM efficiency portfolio, but before I do I’d like to explain what I’m doing and why I’m doing it. A few of our readers seem to think I’ve completely lost it with the portfolio testing. However, if you are one of them, then you might want to examine your own beliefs about the markets and what you believe it takes to develop a system, and most importantly why an educational process like this would press your buttons. For example, one reader writes:

“...the latest attempt with the efficient portfolio trading 'strategy' has bothered me for some time. Is this a joke? A test? An attempt at showing that if one discards all the advice and methodology that IITM offers one will fail?? Trading using a random non-proven strategy and modifying it as one goes along seems completely contrary to how I interpret what IITM recommends and teaches in general about system design and about trading processes and business plans…What does back testing tell you? You are not just grabbing stuff from thin air and pushing it into subscribers mailboxes, are you? Why would anybody want to see your "indicator" if you have no idea what it means? If you have not tested it at all? “

“Come on! I have had some hope that the stuff provided by IITM has been authentic and credible and the last year or so I have spent much time evaluating concepts and building my systems/methodology based much on ideas you have provided. This has been the way I have seen as offering the most potential so far. I have got some promising results but am not finished testing my system and business plan yet….what is mentioned above though makes me actually doubt that what you say has any relevance at all. I really hope that is not the case. There are too much non-tested irrelevant comments out there anyway. I would not like to put IITM on that list.“

So first of all, thanks for your email; it gives me a basis for this article.

Now let me mention a few keys to trading success.

First, you can only trade your beliefs about the market. You cannot trade the “markets” as such. Consequently, anything you do must go through your many filters. If the efficiency portfolio testing is pressing your buttons, it’s probably a good thing. You can learn something about yourself, so I’d suggest you take a look inside and see what is going on. What is annoying you about it and why?

We’ve develop a very extensive process for helping people do this and the testing of the efficiency portfolio is actually part of the process. I’m not giving any trading recommendations…I’m simply illustrating some very early testing processes that I recommend you go through. And there is probably no better scrutiny for testing an idea than doing it in front of thousands of readers. Those processes (and I’m shortening it for this article) include:

Deciding who you are and laying out your objectives. I did that in the early articles.

Describing your beliefs about the market is the next step. I’m a strong believer in trading great trends and I believe that such trends will continue. And by the way, I’ve done extensive testing with efficiency. I described a methodology very similar to this for over a year in Market Mastery during 2001-2002. My point in doing so was to show that you could trade positive efficiency stocks even during a bear market. As an illustration, look at Autozone and DeLux Checking during 2001. The testing was profitable until we reached the bottom of the bear market when there were no positive efficiency stocks. In addition, I’ve been trading efficient stocks for a long time with positive results and some of my best results have been shorting negative efficiency stocks. I’ve also taken one time period, selected a set of positive efficiency stocks and traded it in real time with a 25% trailing stop. In about six months I made about 19% with that portfolio. I was also trading negative efficiency stocks real time and doing well with them that the same time. Thus, I have extensive experience in real trading with this concept already and I’m convinced it works. What I’m not sure about is if being fully invested according to the efficiency of the market works.

The next step is to work out how you want to actually trade. Then trade it or paper trade it to see if it works for you. My purpose in doing the efficiency portfolio is to illustrate an example of this process. And I’m trying out some new things such as selecting the number of positive and negative stocks according to the efficiency of the market. I BELIEVE that this process that I’m following is much better than back testing because you are actually doing it and can see what happens rather than having the computer do things and spit out results that are always replete with errors that you do not know about. I’ve only changed the rules when it was clear that the testing was not meeting my objectives because it was generating far too many trades. If you want to keep the trades longer, widen the stop and that’s what I did. It’s still generating more trades than I really want, but I’m going to stick to these rules in the testing for a while. However, this process allows you to get a real feel for how your idea works in the markets and what could happen. And you can even gather preliminary R-multiples for it. This is the step that I’m illustrating in Tharp’s Thoughts on a monthly basis.

The next step, once you are satisfied with the result of the preliminary testing is to really trade the concept with small position sizing (perhaps for a year). This is where you collect real R-multiples.

And if you like the results, then you can trade the system with full position sizing.

Some people cannot trade anything unless it is somehow tested by a computer and the computer spits out positive results. However, this does not fit my personal beliefs system for all of the following reasons:

I don’t know of any good software for testing. NONE. All of them have errors built into them and you also have the possibility of data errors. Thus, 1) I would not trust the results of a computer back test and 2) I would not trust the results if the back test rejected the system not knowing if it was or wasn’t due to computer errors. It could cause me to totally reject a great method. However, I have back tested it to the extent that the software I’m using will allow (see comment five below) and it tests out very well.

I want to know the expectancy and the R-multiples generated by the system. However, I don’t know of any software that will give me that the way I want it.

The key to success and meeting your objectives is position sizing. Most of the testing software is designed to optimize indicators, and does not even consider position sizing which is the key factor in trading success.

What if the concept you are testing is very discretionary? Most good traders can come up with great ideas for trading, but they are very difficult to test because you have to figure out how to get the computer to think the same way you do. My efficiency concept is very discretionary and this is only way I know how to test it. I’m simply looking for a market that is going up in a nice straight line. The computer won’t give me that. However, it will give me a selection of 100+ stocks that I can look at and choose from. How can you test that with a computer? You cannot. The chart below illustrates one of the top efficiency stocks defined by my algorithm as of November 3rd. I would never trade anything that looks like that. So how could I test a portfolio that rejects those? Read my interview at the end of Trade Your Way to Financial Fre-edom to see what I say about software and back testing in general.

I currently use AIQ to screen for my stocks in the portfolio I’m testing. AIQ has a feature in which I can buy all the stocks with an efficiency greater than some value that we’ll call X and sell it after a fixed holding period. That concept actually tests out very well, vastly outperforming the S&P 500 during the same holding period. AIQ claims it's one of the better methods they’ve seen. But I would never trade most of the stocks that it is testing such as the one in the chart above. And the only way I know to reject those stocks is for me to do it by personally looking at each of them.

When you are testing the stock market over time, you’ll find that a) many stocks disappear over time and b) many new tickers also appear. In addition, stocks split and have dividends, all of which make the backtesting more complex or less accurate and often impossible to do.

I also want to say that people always equate the indicator or the setup with the system. And this is a total fallacy. I call it efficiency testing, but efficiency really refers to the straight line stocks that I like to trade. The concept I believe in is that what goes up will continue to go up. Furthermore, I believe (but with less certainty) that the smoother the line, the more it is likely to continue. I could use any trend following indicator to find stocks that have the “look” I want. I just happen to use efficiency.

I’d also like to mention that I do trade efficiency stocks in my personal portfolio. However, my criteria are much more rigid than the one I’m using in this portfolio. I might only find one or two that I really like each quarter.

However, efficiency does have the interesting quality by which I can measure the efficiency of the entire market and match my portfolio (percentage of long and shorts) to what the market is showing. In the portfolio testing (and again this is just to illustrate the testing process to all of you), I’m looking at the efficiency of the overall market…which might be 60% positive and 40% negative. Thus, I have to find the top six positive efficiency stocks and the bottom four negative efficiency stocks for the portfolio to be fully invested. Every week/month my portfolio could change depending on the efficiency of the market. How could one possibly back test that concept? I have no idea. However, I can look at it over time, like I’m doing in the newsletter and determine how useful it is.

My impression is that most people are finding the efficiency testing very useful. However, if it clashes with your beliefs, then examine why. And if you are still happy with your beliefs, then great…you can skip that article each month. However, if enough of you think it’s not useful, please let me know and I’ll stop doing it. Or, if you are getting value from the process, then we’ll keep going.

About Van Tharp: Trading coach, and author Dr. Van K. Tharp, is widely recognized for his best-selling book Trade Your Way to Financial Fre-edom and his outstanding Peak Performance Home Study program - a highly regarded classic that is suitable for all levels of traders and investors.

You can learn more about Van Tharp by clicking the Efficiency Portfolio Testing header link above.

Have a good weekend.

options   oil   technicals   stock   investor   charts   Short selling opportunities   Technical chart patterns   support/resistance levels   ETFs (exchange traded funds)   Deron Wagner   Basic Materials sector   IRS denies trader tax status   trader tax status   mark to market accounting   Patience in the markets   Dow Monthly Performance   Dow Performance since 1900   Amateur Investors   Nasdaq Composite chart pattern