Well I used to exit at 3:40pm NY time….this did two things…One was keep the results closer to the backtest as in the backtest it will still take a trade at 3:59pm…The other was to allow me to check back and see over time if there was any benefit to be gained.

I did have a look after about 6 months and found it was really 50/50….no discernible difference to trades taken after 3 pm…some were up and some were down.

At the risk of applying logic, we are trading systematically after all and often what you think is sensible proves in testing not to be so, one would think that if a stock is falling at 3:30pm the chance of it going up before 4pm and making you some money is slight. However the stock market has all these dummy moves, players screwing around with the closes, pushing stocks one way and the other, and some of the winners after 3:30pm are probably due to that sort of thing.

So having said all that I am running at 3pm cut off now. I just figured that was as good a time as any.

Prolly just trying to squeeeeeze that extra trade or two Scott

This is an interesting topic, in today’s electronic/algo markets 30min can still make a difference. Even if after few months of trading/monitoring my MOC systems, my gut feeling is telling methat the bread and butter trades seem to be ones filled in open/first hours of session. Currently closing system around the last hour (manually so I “feel” the volatility of the market)

Another observation (a bit more tricky as it derails the MOC principle) from my backtest analysis, it seems to be profitable to keep the position overnight when the loss for the day is meaningful ( above the 6/8% range)
Anyone has been investigating this line of tought?
Cheers