Sunday, December 2, 2007

Expand your knowledge

I came across an interesting research which examined whether profit can be improved using moving averages crossover as a replacement for BAH (Buy and Hold). The research examined the QQQQ in the years between 1991 and 2001. At this point some of you may say of "course SM is better than BAH", which is also true according to the research, but the interesting part is by how much.

The research shows using MA crossover resulted in up to twice the profit for the same period.

This gave me an idea how I can try and improve the system. I will publish the results in the following week.



Anonymous said...

love to see your results.... :)

Guy said...

It's worth noting that this result is totally explained by avoiding the disastrous bear market from April 2000 onwards. As noted in the reference, the moving average cross-over system is worse than buy and hold in choppy periods. If you can buy and hold, but sell when a real bear market hits you would be better off than either system.

Ofir Amitai said...

Guy hi,

The article discusses the difference between BAH and MA and not its combination. Selling when a bear market hits as you suggest contradict the meaning of BAH. As to the performance in choppy period the article points that the MA outperformed BAH in all tests, though one thing you should have in order to really see it happens in your portfolio (during bear markets ) is mental strength and believe in your system.

Guy said...

Sorry. I wasn't clear enough.

I agree that MA beats BAH over the time period tested, and I am certainly no believer in BAH. What I was pointing out, is (a) the reason it does is that MA manages to avoid the bear market starting April 2000, and (b) that MA has the disadvantage of underperforming in choppy markets (going up and down a lot without a long-term trend). In choopy markets MA can get you whipsawed (you buy after prices have been going up for a while, which in a choppy market is just before they are about the go down. then you short after prices have been going down for a while, which in a choppy market is just before they are about to go up again).

Looking forward you have to make a judgment call which market conditions are more likely to prevail. If you fear a sustained bear market go for MA. If you fear a choppy market you should stay away from MA. If a mix of both, then *probably* a well chosen MA will outperform, but not nearly by as much as the comparison they made suggests.

Yet another way to put it, is that MA is a form of trend following. Trend following in general does well when there is a trend to follow, and does poorly in choppy markets. In fact, in choppy markets you would do well trading on a counter-trend system. It all depends on the market.

Guy said...

By the way, your stock ARTG is a great example for why BAH is dangerous. It IPOed in 1999 around 9, and a year later was up something like 1,500% (!) And yet, if bought and held until now you would be showing a return of -50%.

And yet, there are other examples when MA is dangerous.

The ARTG example is so extreme because its 2000 valuation was pure hot air.

Ofir Amitai said...

Obviously one does not build a system purely on a single indicator but combine it with other parameters. For example ARTG was picked using the system which has 10 more conditions before a stock is picked.

Having said that I thank you for your comments and would be happy to see more of them.