Have a bit more energy today to quickly report on the week’s trading. By lunchtime yesterday I was so tired that i had to take the afternoon off. Aiming to be at the office at 6am every morning took its toll this week. One of the days I was at my desk at 4.15am even – insane!
Let me tell you about it…..
This week the trading results were negative – in fact they would have been more dire were it not for a +4R trend-following trade on the DAX in Wednesday’s session. Last week was the 3rd week of live trading two trading strategies simultaneously. One more week to go and then will stop trading over the xmas break, and resume again at the beginning of 2017.
I reminded myself that the main objective of this 4-week trading block is not to maximise profit but to see if the mean-reversion strategy needs further refinement or tweaks. And secondly to see how good I am at actually trading it, and what mistakes I make. So it is to test the strategy and it is to test the trader (me!). From that perspective the trading block thus far has been very constructive as I am indeed finding a few tweaks and adjustments.
The main changes relate less to the trading strategy itself but more to the processes revolving around how to lay the strategy down on a daily basis – how many markets to monitor, how to monitor most effectively, how many hours to trade, what checklists to use?
In that sense I feel that I have been able to quickly make some good refinements so that I am placed well to trade live in the hedge fund office at the start of next year. I want to hit the ground running as much as possible.
Here are a couple of interesting statistics I found about the 40-50 trades I executed over the past three weeks. The first relates to a potential correlation between the time of entry (as to time of day) and results:
‘UTB123’ is the name of the mean-reversion strategy – thanks Paul Wallace! ‘K5’ is the name of the trend-following strategy. The graphs show that for UTB123, on the whole trades entered after 11am (London time) don’t actually add to the performance. On average they produce negative performance. This is true for K5 also, but to a lesser extent. Is it a case of me becoming more tired and not being able to concentrate as well? Does the strategy not work as well during the afternoon? Are the market conditions in the New York crossover sesssion less suitable than the London Open session?
Applying the 80/20 rule, one would simply say “Stop putting on trades after 11am!”.
With all these potential insights, one must remember that all this data is based on a sample size of only 50 trades – surely not statistically significant!
The other two interesting observations related to breakdown of UTB123 performance by markets, and breakdown by execution timeframe (i.e. the timeframe used for identifying the setup):
- By Market – Results on DAX, and other indices (5.5R profit, 14 trades) are strong whereas results on forex pairs (6.9R loss, 24 trades) are weak – incidentally the same correlation is true for trend-following – Indices (5.5R profit, 3 trades) whereas forex/commodities (-1.2 loss, 13 trades)
- By Execution Timeframe – 15M timeframe 9.7R loss (14), 5M timeframe (20) +8.6R profit, 1H & 1M (4) 0.0R
Based on these preliminary findings I have decided, for this coming week, to reduce the position size when placing trades on forex and commodity instruments, as well as as if I am executing on the basis of the 15M timeframe.
Ok, that’s all folks – have a nice weekend!
And remember – trading is hardest way to earn an easy living!