Greetings, everyone. For some time, I've been studying the concept of price inertia (just do a quick Google search for the academic papers on the subject) and developed a EURUSD M15 trading strategy based on it. With a little help from my friends (c), I got it coded into a fully automated EA which I am now keen to test live in the form of a public diary on these boards. Below is a graph of the bot's backtest with a rather excessive lot of 0.2 on a 1K initial deposit for the period of 2010 until now. It's not over-optimized and was done with the base settings that stem from the very concept. I sure could get a much nicer profits curve by means of optimization but don't really see any point in it, as the market never repeats that much anyway. I do not intend to convince anyone of the strategy's long-term profitability, as that is one thing I am not quite sure about myself. My plan is to set up a 1K live ECN account on the first day of every month over the course of a whole year. Considering the excessive 0.2 starting lot size I will be going with, that should spare me from a need to wait out some months for an account's recovery from a large drawdown, like would be the case with a single trading account. Also, even if I happen to blow an account, I will still be able to keep going and end up net profitable (hopefully, highly net profitable) in a year's time. Needless to say, I will post a fully verified MyFXBook link every time a new 1K live account is set up (that is, on the first day of every month over the course of a year). I picked Tickmill as my broker for this experiment - this offshore shop offers tight spreads and makes it possible to instantly open additional accounts in their website's client area. Running up to 12 live accounts will eventually put some decent load on my VPS but I don't mind upgrading it at some point to a costlier package for the sake of completing the experiment. By the year's end, I will be proven either right or wrong about the profitability of my strategy. If the former is the case, I will end up having a publicly documented online track record, lots of fun, and lots of friends from here. As a token of my appreciation for staying with me over a year, I will then share my EA at no cost in this thread (which I wouldn't do just yet, as I don't want anyone to blow their accounts in an event I'm wrong with my trading strategy). I will post the first live account's verified MyFXBook link here exactly on April 1, 2016. Until then, feel free to ask me whatever you would like to know or just share whatever thoughts you might have.
If you are not willing to share your EA now, can you at least roughly tell what kind of strategy the EA is running on, without going into very detail. I have developed many EAs as this but they were all curve fitted. It did not work in forward testing. But that was me, it could be different for you.
As I briefly mentioned, the strategy is based on the concept of price inertia. Basically, it implies that once the price has reached a certain degree of acceleration (increased growth/fall over a fixed unit of time), it is reasonably expected to move for some more time in the acceleration direction before the momentum exhausts. I don't curve-fit for the very reason you mentioned - the market doesn't repeat and you can't expect the same (or even similar) results on a forward test. If the virgin concept survives on backtests without any optimization (like this one did), it makes sense to try it live but never optimize it beforehand. Some might disagree here with me (I know a lot of people believe in the parameters optimization) but that's not what my personal experience has been so far - or yours, for that matter.
As promised, here's the first 1K account's fully verified MyFXBook link: http://www.myfxbook.com/members/babyjake1961/01042016/1573036 The bot will be placing anwhere between two to six simultaneously opened trades, each with a starting lot of 0.1 (making the total starting lot 0.2). The first trades can be placed as soon as today (for some of you, tomorrow), since it's the US NFP Friday and substantial price movements might occur, should the NFP data surprise to either the upside or the downside.
A brief analysis of yesterday's trades. As the price gained sufficient acceleration in the NFP/ISM aftermath, a total of four trades opened with a 15-minute interval yet only the first two managed to reach the profit zone before the momentum ran out of steam and the price reversed. So we netted a small loss of 1.18% (almost a break-even, considering the risk level). What pleases me the most is yesterday's live trades fully matching those on a backtest for yesterday, which means the algorithm is functioning properly. As all the time in the world is ours, we shall continue when the next market opportunity presents. As a side note, never trade in a time of your worst financial desperation! Remember that markets are heavily manipulated and technical analysis no longer works. That's why the "gurus" have switched to bucket-shopping and b-booking while never trading themselves, not even during the "live sessions" when they are doing the "guru stuff" in a webinar mode to get a hold of your cash. So if you are cash-strapped and must push at high leverage under stress and time limit, it will always go against you to take you out for good.
It does but the best results are on M15. On smaller timeframes we get more trades with a lower profit factor (lots of short-living retracements on a volatile market that trigger stop-losses) and on higher timeframes we get much fewer trades while it's a rare occurrence that the price would continue accelerating at the same pace for hours.
Those of you who want a copy of the bot can get it here: http://www.elitetrader.com/et/index...ar-on-autopilot-call-for-beta-testers.298966/
Having evaluated the forum feedback, I am now canceling my offer of the bot's beta test, as the forum members somehow feel I would be "indirectly supplying advice through the use of a bot". At their insistence, the bot shall remain proprietary, private, and confidential at all times while the forum members will be fully left to their own means. I will, however, go ahead with my own year-long live test of the bot as planned and it will be extensively documented in this thread. I am additionally attaching the bot's full backtest report for the period of 01.01.2010-04.04.2016 which was produced at the request of a forum member.