Tuesday 14 August 2007

S&P 500 Bullish: Revised COTs Composite Numbers

Sheesh, what a nightmarish close for the bulls today. Perhaps we should have gone sailing after all, like that guy wrote in Forbes?

Perhaps not. I've just posted some revised data for my U.S. Composite Equity Indicator that suggests the Commitments of Traders are far more bullish than I had thought.

I was trying to see if my composite indicator gives profitable signals for the S&P 500 when I realized I made a mistake in how I originally calculated the indicator. The indicator is based on a composite of my setups for four U.S. equity indexes: the S&P 500, Dow Jones Industrial Average, Russell 2000 and NASDAQ 100. My apologies for the error.

When I checked the revised numbers against the S&P 500, I found they produced an interesting new trading setup for this index. Check out the signals and revised data here:

S&P 500/COTs Composite Equity Indicator

My deepest apologies for giving incorrect information for this indicator in past entries on this blog.

I had previously written that the indicator has lately been neutral or mildly bearish. In fact, it flipped to bullish for the S&P 500 on March 30 and has been generally bullish ever since. It even just gave a renewed bullish signal in last Friday's report. This trade is, incidentally, still in the money despite the ongoing market bloodbath.

The setup is based on trading the S&P 500 when the composite indicator is 0.8 standard deviations or more from its 156-week moving average. In other words, we get a buy when it hits 0.8 or more.

Click the "Trading Setups" link above to see how this new setup fared. It's actually my most statistically robust indicator for any equity index in terms of market-beating profitability, but there's a caveat: The data is somewhat limited. Unlike most of the COTs data, which goes back to 1995, the composite indicator's data starts just in 1997. That's because the COTs started to include data for the Dow Jones industrials only then. The setup's first signal came in 2000.

But wait a second, you say. Isn't the COTs data for the S&P 500 giving me a bearish signal? Yes, indeedy, it is. So which one is right? I compared the results of the two setups starting from the same time in 2000. The results were virtually identical in terms of profitability. The composite indicator had a somewhat higher past drawdown, but the S&P 500 COTs setup had slightly inferior results of statistical robustness.

In the end, I think I'll stick with my existing S&P 500 setup for the time being. (It's on a fairly profitable bearish signal at the moment.) The reason: It's based on a longer range of data that gives me more representative market conditions. But I will also add the new setup to my weekly updates. It may provide interesting insights about the direction of the broader market. I'll also compare the composite indicator to other indexes to see if I can find other tradable setups.

No comments: