Friday, November 12, 2010

Why The Equity Curve Is Importnat In Evaluating Studies

While I don't always show it I do always look at the equity curve when evaluating studies to include in my analysis.  Last night while conducting my research I came across a great example of why this is important.

Inside days have generally suggested a bearish edge when the market is below the 200ma and no edge much better than upside drift when above the 200ma. I found it unusual that the inside day came with an unfilled gap down so I tested the possible effects under these circumstances.

At first glance the numbers seemed to suggest a downside edge. A closer look showed the numbers to be misleading. Here are the results in table format.

Base on this the next 1-3 days would seem to have a bearish inclination. But here is a picture of the equity curve.

As you can see it has been a long time since this setup has produced compelling odds.   Researchers should always take a look at the equity curve when considering whether to incorporate results into their analysis.

Another blogger who often makes this point is Michael Stokes of MarketSci.  He did it again in his recent Thanksgiving returns post yesterday.

1 comment:

tamas said...

Hi Rob,

I think this is a very smart post, because it exposes a typical and hidden mistake.