Just following up on my weekly gap strategy - I've enjoyed the long (4 day) weekend away so catching up on my weekly screenng today for the remainder of the week ahead. The results of last weeks gap entry screen were as follows:
None of the positions are due to close until the end of this week (no stoplosses yet, thankfully). The vast majority of these gap setups never get filled, despite the fact that they have 4-5 weeks to fill (hit the entry level), which is ok imo, that's pretty much what I am looking for in an effective filter.
The screen for this week is below, showing what stocks I will be watching for this strategy this week. Slightly more setups identified than last week (27 vs 20) and because I was away I missed one which was triggered yesterday, so will have to buy on open tomorrow:
You will note that I track variances to the strategy - very important requirement for systems trading imo since variances they are probably the biggest risk to the strategy achieving results that are in line with the historical backtest.
Variances come in at least 2 types - systems variances and execution variances.
An example of a system variance is the daily gap up open entry. It occurs regularly & I have the statistics of how often it has happened so I have quantified what adverse impact it has had historically. Tracking it enables me to see whether or not its occurrence is in line with historical data or whether the frequency is starting to increase and hence potentially leading to system failure. Similarly with gap down stoplosses.
The results above show already 2 examples of execution variance (read execution error). Naturally the less of these there are the better, but it is important to track to see how often it occurs and how much it costs (something that is not possible to estimate in historical backtesting).
