Re: A question on how to compare buying signals

[Bob Fulks <bfulks@xxxxxxxxxxxx>]

At 10:31 AM +0200 9/6/01, Bengtsson, Mats wrote:

>Is there anyone who has a good way of looking at buy signals more isolated,
>to see which of them that deilvers good results? Any tips on how to do this
>and what to look for when comparing (like for strategies, I compare kelly
>value and Sharpe ratio).

One technique is to build a trading system using your buy signals
followed by exits some fixed number of bars in the future:

    if BarsSinceEntry >= 3 then Exitlong at market;

Then calculate the Sharpe Ratio of the results. If the Sharpe Ratio
is higher than that of a similar system with random entries, your
entries are better than random.

You can also look at the percent winning trades of the system. With
random entries and exits X bars later, you would expect about 50%
winners. If your entries are better than that, you would expect
greater than 50% winners.

Think about what makes a good entry. After a good entry, the price
moves in your direction AND moves far enough to cover your fixed
trading costs (commissions and slippage). A good counter-trend entry
might initially move in the wrong direction for a few bars then turn
to move in the correct direction so you need to allow some time
before deciding if it is good entry.

Bob Fulks