RTATC2011 — Day 3 — Principal Idea
June 26, 2011 (Last updated on May 28, 2013) by Andriy Moraru
OK, two weeks since my last RTATC2011 journal entry and it’s time to share my progress and offer first interesting downloads. So far, I’ve considered two different trading strategy ideas for my ATC 2011 EA:
- Use Total Power Indicator to measure the bull/bear power and enter long position on bull prevalence, short — on bear.
- Pure price action system that I’ve been testing on demo for about 6 months already.
Both trading ideas involve ATR-based trailing stop. The TPI-based EA, would close previous position if a new signal (in the opposite direction) is received.
Total Power Indicator EA
You can freely download the TPI-based EA that I’ve used for backtesting (don’t forget to download the Total Power Indicator itself, otherwise the EA won’t work). The strategy is very simple. Each new bar we’d check the state of the Bear and Bull buffers and if Bull > Bear we’d go long, if Bear > Bull, we’d go short. Opposite signal would close previous position. ATR-based position sizing would be used for money management and the ATR-based trailing stop (without take-profit) would be used for a normal profit taking and loss prevention. The EA was tested on all 12 pairs offered in ATC on timeframes M15, M30 and H1 (I don’t see any point in using other timeframes in a championship that lasts less than 3 months). The period for testing was chosen since January 1, 2011. Unfortunately, only EUR/USD has shown positive results during back-testing. But even these results weren’t satisfying enough and looked quite random.
Price Action EA
My back-testing version of the price action EA can also be downloaded for your own tests and use. This expert advisor is extremely simple. Initial direction for position is selected randomly. An ATR-based trailing stop is attached to a position. If the stop-loss triggers, a new position in the opposite direction is opened. That’s it. Position sizing used in testing is reverse-Martingale — we linearly increase position size as our account balance grows. Backtest conditions were the same as with the TPI EA. EUR/GBP and EUR/AUD at H1 timeframe proved to be the only good combinations. That’s pretty strange, because I am forward-testing this EA since December 12, 2010 and it shows good results also for EUR/USD and GBP/CHF. Nevertheless, it looks like I’ll have to stick with EUR/GBP and EUR/AUD, because otherwise my EA won’t pass ATC backtest requirement.
I decide to go with the second EA idea (price action) for two reasons: I like its backtesting results more than TPI EA’s results and I have some positive forward-testing experience with this EA.
Finding the main idea for your EA is a necessary and important step in preparation for the ATC, but there are still more things to do:
- Implement some Z-Score optimization (more on it in the next RTATC2011 entry).
- Prepare the EA for trading 2 currency pairs.
- Optimize the EA without overoptimizing it (if I only knew how to do that…)
- Find optimal position sizing rules.
- Prepare all the trading routines to work in ATC conditions.
- Testing, testing, testing…
- Maybe a better idea for taking the initial position instead of random. After all, first trade can be very important in ATC.
P.S.: 2 months and 28 days left to register in ATC 2011. 630 participants registered so far.
If you have any questions regarding my proposed idea for the ATC or if you want to suggest another idea for an expert advisor for this year’s contest, please reply using the form below.