Correlation between trade entry time & profitability?

Back in December I ran some analysis over a batch of 40-50 trades I had completed in preceding weeks.  Click here to go that blog post.   Looking at that data I came to the conclusion that I was making profits in the first half of the day (the London session) and then giving the profits by trading in the afternoon (the New York crossover session).

I ran the same analysis for the 40 trades I completed last week – last week I traded both in the mornings and in the afternoon from Monday to Thursday.  Interestingly this time data the data looked quite different.

Last week the pattern of profitable mornings and non-profitable afternoons did not exist. If anything it looked as if the pattern was now reversed.  Hail to those who always go on about not making conclusions from sample sizes.  Thought this was worth sharing – nothing significant I would say at this point, but regardless an intriguing observation.

The graph below shows the cumulative performance for last week.  How was the graph compiled? The 40 trades were ordered by entry time. The earliest trade, entered around 7:30am was assigned the first spot in the data series.  The second earliest trade was assigned the second spot, and the cumulative profit/loss at that point comprised the profit/loss of those two trades.  Then the third trade was taken into account and so on.  The net result of the 40 trades was a profit of 1R – hence that is where the line finishes:

cum-performance-by-entry-time

 

For comparison, here are the graphs for the analysis completed in December for another similar-sized batch of trades:

 

time-of-entry

 

Bring on next week, and a raft of central bank interest rate decisions.  And will the Dow hold above 20,000?

This entry was posted in Reversals Strategy, Trend-Following Strategy. Bookmark the permalink.

4 Responses to Correlation between trade entry time & profitability?

  1. Statisticians might say you can draw conclusions from a 30 observation sample.

    But in trading, market conditions change, (low vol/high vol, bull, bear, consolidation) and your sample is only valid in similar market conditions to those when the sample was taken.

    You really need at least a 300 trade sample, and the trades in that sample need to come from many different types of market conditions. Not just from say a single six month period.

    Looking back over the last 10 years, i would want to include trades trades from 2008 and 2009 and well as trades from the last 7 years.

    Personally I back test intraday every trade going back to 2002. I have thousands of sample trades for systems that trade often and hundreds of sample trades for systems that trade only a few times per month.

    The sample covers, bear markets, bull markets, very slow markets and insanely volatile markets.

    This gives me a good idea of what to expect as market conditions change.

    Liked by 1 person

  2. I got my ass kicked uisng 19 years of data about options expiry- it means squat. Real time trading has shown me time and again that markets are pretty much random,you can read patterns if you like,some people do and they win for a short time,but over the long term- I’ve been trading badly for 16 years, the market is just a squiggly line. Enjoy the journey and all power to you.
    PS I have constructed a system with amazing results on DAX -it took about 10 minutes,no optimisation,and I’m tracking it in real time

    Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s