Thursday, May 15, 2008

Looking for Mr. Goodbar - Part 4

Continuing with yesterday's theme and the emergence of XLE as the top performing ETF of our test sample, let's now break down the XLE into its top 10 component stocks and examine returns relative to the ETF itself. An initial assumption would be that the individual components perform better than the ETF since they have greater volatility than an index. We should also expect the individual stocks to entail greater risk. Whether any enhanced performance is worth a risk premium is the focus of this backtest. For the sake of comparison I included both 5 and 10 minute bars for the test, again to see whether volatility in the component stocks skews performance relative to bar size.
Our results are signicantly different on 5 and 10 minute bars. On the 5 minute, XLE essentially breaks even for our 500 bar backtest and places 9th out of 11 in terms of short term gains. Only the top 3 stocks, SLB, CVX and OXY show relatively attractive returns. For 5 minute bars our best strategy is to find the highest correlation performers and trade them in lieu of the XLE.
For the 10 minute bars, the situation changes. Here XLE places 5th out of 11 for our 500 bar test. And, while SLB and CVX hold top positions, APA and COP join their ranks, closely followed by the XLE.
These results support our initial conclusions from yesterday.
Bar size does clearly effect ROI and win/loss ratios. Different stocks are subject to different accumulation/distribution forces that one can only guess at. An important point with this type of analysis is that we don't have to try and figure out why this change in sentiment is occurring, we just have to know when it is happening.
This information will provide a useful risk management tool to trade not necessarily the top relative strength stocks, but the top technical alignment stocks (and ETFs).
Above are the IBD top 20 stocks, backtested on 5 and 10 minute bars with our RSI (3,90,15) long only, no stop model. I include this test as another example of how to select a technically reliable trading vehicle (stock or ETF). Note that in both 5 and 10 minute bars, 14 of the 20 stocks make money over the 500 bars test. Of the 6 stocks that fall below the zero line, only 2 are the same in the 2 time frames, CLR and SOM. That's not to say that these are weak stocks, the point to emphasize is that it's tougher to make money trading them using this system. Of course, you could find another system that would have completely different results. . and that's a valid approach. The goal is to find close technical alignment of the stocks (or ETFS) that you trade in order to maximize your returns and minimize your risk. Another approach is to monitor scans like the IBD20, which is a dynamic list of high momentum, high relative strength stocks, then cull the technical aberrants and trade the top 5 or 10 as a basket, depending on your capital, and risk management plan.
This is technical approach to trading that looks at "relative strength" in a different way. As presented over the last few days, this approach has been examined as a daytrading system using 5 or 10 minute bars (other time frames may also prove viable, but have not been evaluated). Looking at the charts is barely necessary if auto-traded. The need to monitor the NYAD, TICK and VIX, the number of stocks making new high, new lows, etc, etc. becomes unnecessary. Adding a short side component would increase the numbers of daily trades and, hypothetically, total ROI. I'll be looking at further refinements of this approach in upcoming posts as it provides the direction for my current trading evolution.

No comments: