Sunday, January 21, 2007

3 Month Low - Using Weekly Data

Wow! I read the stockfetcher manual (what a concept - a man reading instructions). Anyway I found out how to convert the filter to work with weekly values. Here is the final result.

Show stocks where close is between 15 and 35
and average volume(90) > 500000
and weekly close 1 week ago reached a new 3 month low 1 week ago
and weekly close 1 week ago < weekly open 1 week ago
and weekly close > weekly open
and weekly low > weekly low 1 week ago

The first 4 lines set up the dip and the next two set up the buy conditions. In this regard I want to ensure that the “dip week” is a down week and the buy week is an up week and that the lows are going up.

So the filter goes from 3 lines to 6 lines but what is the result?

I tested it against three time periods.

06/30/2006 – 11/01/2006 there were 74 trades, a 66% win rate, Reward/Risk of 2.52 and a 129.34 ROI.
09/01/2006 – 12/22/2006 there were 51 trades, a 71% win rate, Reward/Risk of 3.66 and 124.57 ROI.
09/29/2006 – 01/19/2007 there were 52 trades, a 69% win rate, reward,risk of 3.84 and 120.57 ROI.

These are excellent results. The fewer trades in the two later periods just show that there have been fewer dips to be bought in recent months. That’s a function of the market going up in a relentless manner.

Stockfetcher computes reward risk by: compute the product of the number of winners and average gain of the winners. Next, compute the product of the number of losers and average loss of the losers. The reward/risk is computed by dividing the winner product by the losing product. (copy right –

The higher the reward/risk factor the better the filter.

So that’s the saga of the 3-month low filter – I will be using this one for identifying dips from here on out. The most current one is LPNT – do your own DD.

1 comment:


Theirs a lot of data to digest.