Are these results good enough to trade?

I have been investigating forex trading for almost a year now and have at last devised a strategy that seems to consistently produce a profit during back-testing so that only one of the last 8 years would have seen a loss. It is a simple trending strategy based on a combination of technical indicators, and as such suffers from the well-recognised problems of needing trends to make any decent return as otherwise the general market ups and downs produce a gradual loss. I don’t regard this as the final strategy, but merely a stop along the way which may give some further insights.

The question I have is whether the current results are good enough to start trading in the hope of providing some income while I learn a bit more (always good to get paid to learn and helps with the motivation).

For 1 hourly time-frame eurusd over the last 8 years the strategy averages ~100 trades / year producing ~7 pips profit / trade with a maximum draw-down of ~700 pips and a ~40% hit rate.

It is currently in demo and if continues to be successful I intend to start on £1/pip, possibly working up towards £10/pip at which point I should get ~£4000/year on average assuming a 3pip dealing loss / trade provided the market stays within the parameters of the last 8 years which seem to have covered some turbulent times. My fundamental outlook is that trends will continue to arise in the short to near term as the tectonic plates of economic adjustments due to the on-going credit problems continue to move.

The only thing I can’t simulate properly in back-testing is the stop loss, which I can only do at the time-frame level and from that 75 pips seems a good value for maximum pips. At the often-quoted 1% of total at risk this would mean that I would need to fund £7500 for every £1/pip that I wanted to trade.

As a comparative newbie I would welcome any thoughts / observations on my plans, thanks.

There are testing environments where you can simulate stop losses. Prodigio from ThinkOrSwim comes to mind, but I'm sure there are many others.

RandomOne said:

I have been investigating forex trading for almost a year now and have at last devised a strategy that seems to consistently produce a profit during back-testing so that only one of the last 8 years would have seen a loss. It is a simple trending strategy based on a combination of technical indicators, and as such suffers from the well-recognised problems of needing trends to make any decent return as otherwise the general market ups and downs produce a gradual loss. I don’t regard this as the final strategy, but merely a stop along the way which may give some further insights.

The question I have is whether the current results are good enough to start trading in the hope of providing some income while I learn a bit more (always good to get paid to learn and helps with the motivation).

For 1 hourly time-frame eurusd over the last 8 years the strategy averages ~100 trades / year producing ~7 pips profit / trade with a maximum draw-down of ~700 pips and a ~40% hit rate.

It is currently in demo and if continues to be successful I intend to start on £1/pip, possibly working up towards £10/pip at which point I should get ~£4000/year on average assuming a 3pip dealing loss / trade provided the market stays within the parameters of the last 8 years which seem to have covered some turbulent times. My fundamental outlook is that trends will continue to arise in the short to near term as the tectonic plates of economic adjustments due to the on-going credit problems continue to move.

The only thing I can’t simulate properly in back-testing is the stop loss, which I can only do at the time-frame level and from that 75 pips seems a good value for maximum pips. At the often-quoted 1% of total at risk this would mean that I would need to fund £7500 for every £1/pip that I wanted to trade.

As a comparative newbie I would welcome any thoughts / observations on my plans, thanks.

Click to expand...

Are you saying your average win is 7 pips and your stoploss is approximately 75 pips? Thats not clear as you also say that the hit rate is 40% which couldnt be possible with a 1:10ish RR ratio.

Assuming you are taking an average of 7 pips on each winning trade my guess is that slippage and widening spreads will kill you. You wont see this demo trading but when you go live you will see it, especially with spread betting.

Random, which software are you using? Even fairly basic software now allows for programming of stops, and I would recommend a factor of ATR, e.g. 1 ATR away from entry.

Also, reporting results in pips is not very useful, for many reasons (e.g. pip size is relative and can change, as equity grows you're risking less if you keep the same amount per pip and so on). Can you give us your CAR (compound annual return) in %, along with max drawdown in % for the entire period?

Howard, thanks for the info, I will check that out.

Pboyles, the average profit per trade is 7 pips including the loss making ones. The profitable trades make more than the losses on average so that overall there is a profit. I have tried to cut down the losses, but every time it reduces the winning trades by a factor more than the profit/loss ratio so the overall effect is negative. Basically it seems you need to take risk to be in early otherwise waiting until you are more certain increases the hit rate, but reduces the remaining profit. The slippage is another issue, I was hoping for as little as 3 pips/trade on average as this is not a strategy specifically looking to trade at volatile times, but rather trades just after the hour whenever I get the signal (hopefully SMS triggered when I look into it). However, there is no substitute for experience, hence my post.

Meanreversion, I am using ProRealTime for historic reasons really, although I see MT4 is quite popular. I think the problem in back-test is that it downloads all the statistics for each time-frame, e.g. hourly, while a stop loss in reality would be triggered on the tick level. I may be wrong however.

Thanks for tip on ATR, I will see if I can program this in via a suitable indicator rather than the constant value used at the moment and whether this will improve the results.

My results are just as they come out of the software for a flat £1/pip stake. From what I can see compounding would not be helpful in this case as the hit rate is too low and I have observed that at such levels you tend to over-stake after a successful deal only to therefore loose a larger amount on the next deal which is more likely to be a loss. Similarly, after a loss, the stake is reduced and then you gain less on the next successful deal. Compounding seems to work overall when you have runs of losses or successes, which this current strategy does not produce enough of. I did simulate compounding, but the results were worse at the end of the year.

As a result, I am not sure I can give a CAR or MAXDD in % (feel free to correct me if I am wrong). In raw pips the results were ~5600 pips over 8 years if that helps.

For me it's good enough... keep it up man =)