A system for trading GBPUSD based on daily data was machine designed by Price Action Lab on June 5, 2012. This post is an update on the performance of this system.
The GBPUSD system was machine designed by Price Action Lab software on June 5, 2012 using an in-sample period from 01/04/1999 to 05/31/2010 and an out-of-sample period from 06/01/2010 to 06/01/2012. Click here for the development details. This system returned nearly 28% from 06/01/2012 to 12/31/2012, or about 44% on an annualized basis.
In another post we reported that the system return for 2013 was 37.50% based on an initial trading capital of $20,000 per standard lot. The GBPUSD system evaluated in this report is the exact same system developed more than two years ago.
Below are the original results of 17 patterns, 11 long and 6 short, that satisfied the performance criteria specified on the search workspace for the in-sample and for 150 pips profit target and stop-loss:
Each line in the results corresponds to a price pattern that satisfies the performance parameters specified by the user. Index and Index Date are used internally to classify patterns. Trade on is the entry point, in this case the Open of next bar. P is the success rate of the pattern, PF is the profit factor, Trades is the number of historical trades, CL is the maximum number of consecutive losers, Type is LONG for long patterns and SHORT for short patterns, Target is the profit target, Stop is the stop-loss and C indicates the type of the exit, in this case it is “pts”. Last Date and First Date are the last and first date in the historical data file
Updated performance results
Price Action Lab generated code for Amibroker and a system was implemented in that platform that combined the long and short patterns using the OR Boolean operator. Recall that the profit target and stop-loss are both 150 pips and the entry is the open of the day following a pattern formation. Commission and slippage of 2 pips per side were used but since this is a position trading system with a large profit target the effect of friction is minimal. Daily GBPUSD data were downloaded from Finam using the Amibroker downloader for the period 01/02/2014 to 09/30/2014. The entries, exits and equity curve are shown on the chart below:
|Avg. bars in trades||8.08|
* Based on a the performance simulation of 20,000 random coin toss systems
Based on 20,000 random long/short coin toss systems the realized return was significant at the 0% level. The distribution of returns is shown below:
Note that backtested systems that are significant at a level close to 0% are often curve-fitted but out-of-sample or forward results near the 0% level show high significance. As shown on the above screenshot from Price Action Lab analytics, the system performed better than 100% of random systems. For significance at the 5% level, a 6.06% would suffice, as also shown on the above results.
FAQ: How long will these patterns remain profitable? This is unknown. However, the pattern search with Price Action Lab can be repeated, usually every year, and the system can be updated in a forward fashion.
Disclosure: no relevant position at the time of this post.
Disclosure: no relevant positions.
Charting program: Amibroker