Backtesting can be a useful tool in trading as it allows you to evaluate how a trading strategy would have performed in the past. By analyzing historical data, you can assess the effectiveness of your strategy, identify areas for improvement, and refine your approach.
However, it’s important to keep in mind that backtesting is not a guarantee of future performance. Markets can be volatile and subject to unpredictable events, so it’s possible that a strategy that performed well in the past may not be as successful in the future.
Additionally, it’s important to use high-quality data and be mindful of any biases that may be present in your analysis. For example, if you only test your strategy on a specific time period or set of market conditions, your results may not be representative of how it would perform in other scenarios.
In summary, while backtesting can be a valuable tool in trading, it should be used as part of a comprehensive approach that also takes into account current market conditions, risk management, and other factors.
Hmmm. Personally, I usually spend time backtesting strategies on demo before I try them out on live trades. I think backtesting is helpful, but just don’t spend too much of your time backtesting, and don’t expect that all of the results from your backtesting would exactly be the same once you go live.
It’s great to see such a wide range of responses to this question and interesting to see so many different ways everyone is using back testing, or not as the case may be.
Here’s a good example of how back-testing can highlight the risk (which is otherwise invisible) of a strategy failing -
In the clip, the presenter tests a simple mechanical strategy and finds it only works well on certain forex pairs, and only for certain periods across a 14-year back-testing horizon.
I have no idea why the same strategy works for years, and then doesn’t work for years. At 09:59 he has to plead with honesty, “Don’t ask me why these strategies were working really well and then just stop, that’s just what happens…”
Finding what isn’t working is maybe just as valuable as finding what is.
I am fairly new to trading but have been following stock prices and markets through passive investing prior. From the books I have studied, prices represent human emotions/feelings/predictions, and they are notoriously unpredictable. Hence the reason strategies work for only short stretches until market sentiment changes, and price action then follows a different set of rules.
I have been experimenting with trading algorithms on MT5 which has a super back test and optimization model. It further has a forward testing option on historical data which aims to minimize curve fitting of algorithms due to the unpredictability of markets. It has shown on every run, that settings on a backward optimization does not work on forward testing. Mostly I have seen that one set of trading parameters has excellent results on the back test but fails miserably on a forward test, and conversely, back test results that were poor can be the most profitable on a forward test. There lies a set of parameters in-between that could work.
This is my initial observations, still far more testing/learning to do. My initial thoughts are that back testing for 3 to 6 months seems more reliable than back testing for years, and that the trading parameters need to be optimized or at least checked if still valid every month or so. This comes back to the market sentiment statement I opened with.
Hi, can somebody give an email of a strategy that they actually use for trading and especially one which has been used in the back testing process?
I still don’t see how back testing is relevant / can be used effectively as i would have thought the two main variables i.e. current price and current time can’t be utilized historically
This is the most extreme example I can think of. Fundamentally speaking, COVID was in the news for a while before the price collapsed. The technical analysis provided great confirmation of which way the price would go to. Of course you still need to have your proper trading strategy in place with stop losses and position sizing.
Personally I would love to have been in the markets at that point.
Thanks, could you further elaborate? I am trying out different approaches similar to what you describe but would appreciate guidance, especially on market behaviour.
as you know, the market can move in three directions, up, down ( which mean trending ) and sideways, volatility can be “normal”, low and high. So you can build strategies separately for long and short for catch trend with stop order and mean reversal type strategies for sideways with limit order. Same thing for volatility “normal” and low volatility can be good for stop order, limit for high volatility. Unfortunately, it is necessary to do a research, because there are differences in behavior between instruments and time frames. In case of relevant topics High School - School of Pipsology - BabyPips.com little touch this topic. Regards Greg
In my own opinion, I think one does not have a strategy until it has been successfully backtested. I have benefited from it and it is the only reason I still trade today. Backtesting a strategy is as important as the strategy itself.
Personally, I usually spend time backtesting strategies on demo before I try them out on live trades. 
I think backtesting is helpful, but just don’t spend too much of your time backtesting, and don’t expect that all of the results from your backtesting would exactly be the same once you go live. 
