Testing the strategy initially with 30 trades
First of all, you need to record 30 trades using a demo account.
If a trading strategy is not profitable over 30 trades, it is likely that modifications to the strategy are required or that the strategy is not worth using on a real live account. At this stage you may need to find an alternative strategy and test in the same manner.
Every time you change or modify your strategy, you must backtest it again with 30 trades.
The reason why you need to test a strategy with at least 30 trades is because many strategies regularly have losing streaks of several trades in a row. You need a sample size just large enough to incorporate several cycles of winning and losing streaks and to see if the sample is initially profitable.
*Every trading strategy should be tested with at least 30 trades. 30 trades provide an initial scope as to whether the strategy has the potential to work.
If, after the initial 30 trades, you find that the sample is not profitable, then you may want to modify the strategy or change it altogether. If the sample is profitable, then you can go onto the next stage of backtesting.
Carry on testing until you have 100 trades
Once you have found your initial sample size to be profitable over 30 trades, the next stage is to continue collecting trades until there are a minimum of 100. A larger sample size of 100 trades provides a clearer picture of the potential performance of the strategy. With a sample size of 100 trades you can find the following information:
*The next stage is to continue collecting trades until there are a minimum of 100. You can find what kind of drawdown you may have had, any losing streaks and the amount, the ratio of winners to losers.
• What kind of drawdown does the strategy seem to produce?
• How often are losing streaks likely to occur?
• What is the ratio of winners to losers
Having the answers to these questions will help you later when you are experiencing a performance downturn. If, for example, your initial sample showed that you had a drawdown of a certain percent, then if you experience a similar drawdown when trading live, you will understand that this is a normal part of the strategy.
Benchmarks help you determine how much of a down turn in performance you can accept
Once you have at least 100 trades, you can now see the extent of a drawdown period with your strategy and you can create benchmarks.
A benchmark is the maximum negative impact that you allow on your trading strategy. The following is a list of example benchmarks you can create with your backtesting:
• The maximum amount of losing trades in a row.
• The maximum percentage drawdown of your account.
• The maximum risk to reward you can use in the strategy.
Once you have your benchmarks, you can start to trade your strategy on a live account. You can use these benchmarks to monitor if your strategy continues to work in different market environments. For example, if the maximum drawdown in your initial 100 trades was 15% of your account, then you can use this as a benchmark while trading with live money.
As long as the drawdown does not exceed more than 15% of your account, then the strategy is within the normal limits – you can confidently continue trading until the strategy becomes profitable again. If a losing streak results in a drawdown of over 15%, then the benchmark from your testing has been broken – this highlights that there could be something wrong and the strategy may no longer be functioning as normal.
This is the point when you would stop trading on a real account and assess the strategy in the same manner as before – going back to a demo account and testing another 30 trades.
If the results are positive on these 30 trades, then the benchmark can be adjusted and live trading can continue. If the testing does not produce a profitable result, then it can be that for whatever reason, your strategy no longer works. At this point, you can go back to adjusting the strategy or testing a new one until you find a sample size of 30 trades that are profitable.
*After a benchmark is broken, take another 30 trades on your demo account to backtest to see if you are ready to take your strategy to live trading, or if you need to continue analysis.
You must continually monitor the performance of the strategy to see if a benchmark is broken, even if the strategy performs well, because the market environment can change at any time causing your strategy to become ineffective.
Summary
• to backtest a strategy, first you need to record the data from 30 trades on a demo account.
• 30 trades is enough of a sample size to see several cycles of winning and losing streaks.
• if the initial backtesting of 30 trades is profitable, then you can go on testing until you have 100 trades.
• after you have a minimum of 100 trades, then you can create benchmarks – a maximum negative impact that you allow on your account.
• when you have your benchmarks, you can then trade on a live account and monitor your trade until a benchmark is broken.
• if one of your benchmarks is broken, then you need to test the strategy again on a further 30 trades.
•if the further 30 trades is profitable, then your benchmark can be adjusted and you can continue live trading. If it is not, then you must modify or change your strategy and start the testing over again.
First of all, you need to record 30 trades using a demo account.
If a trading strategy is not profitable over 30 trades, it is likely that modifications to the strategy are required or that the strategy is not worth using on a real live account. At this stage you may need to find an alternative strategy and test in the same manner.
Every time you change or modify your strategy, you must backtest it again with 30 trades.
The reason why you need to test a strategy with at least 30 trades is because many strategies regularly have losing streaks of several trades in a row. You need a sample size just large enough to incorporate several cycles of winning and losing streaks and to see if the sample is initially profitable.
*Every trading strategy should be tested with at least 30 trades. 30 trades provide an initial scope as to whether the strategy has the potential to work.
If, after the initial 30 trades, you find that the sample is not profitable, then you may want to modify the strategy or change it altogether. If the sample is profitable, then you can go onto the next stage of backtesting.
Carry on testing until you have 100 trades
Once you have found your initial sample size to be profitable over 30 trades, the next stage is to continue collecting trades until there are a minimum of 100. A larger sample size of 100 trades provides a clearer picture of the potential performance of the strategy. With a sample size of 100 trades you can find the following information:
*The next stage is to continue collecting trades until there are a minimum of 100. You can find what kind of drawdown you may have had, any losing streaks and the amount, the ratio of winners to losers.
• What kind of drawdown does the strategy seem to produce?
• How often are losing streaks likely to occur?
• What is the ratio of winners to losers
Having the answers to these questions will help you later when you are experiencing a performance downturn. If, for example, your initial sample showed that you had a drawdown of a certain percent, then if you experience a similar drawdown when trading live, you will understand that this is a normal part of the strategy.
Benchmarks help you determine how much of a down turn in performance you can accept
Once you have at least 100 trades, you can now see the extent of a drawdown period with your strategy and you can create benchmarks.
A benchmark is the maximum negative impact that you allow on your trading strategy. The following is a list of example benchmarks you can create with your backtesting:
• The maximum amount of losing trades in a row.
• The maximum percentage drawdown of your account.
• The maximum risk to reward you can use in the strategy.
Once you have your benchmarks, you can start to trade your strategy on a live account. You can use these benchmarks to monitor if your strategy continues to work in different market environments. For example, if the maximum drawdown in your initial 100 trades was 15% of your account, then you can use this as a benchmark while trading with live money.
As long as the drawdown does not exceed more than 15% of your account, then the strategy is within the normal limits – you can confidently continue trading until the strategy becomes profitable again. If a losing streak results in a drawdown of over 15%, then the benchmark from your testing has been broken – this highlights that there could be something wrong and the strategy may no longer be functioning as normal.
This is the point when you would stop trading on a real account and assess the strategy in the same manner as before – going back to a demo account and testing another 30 trades.
If the results are positive on these 30 trades, then the benchmark can be adjusted and live trading can continue. If the testing does not produce a profitable result, then it can be that for whatever reason, your strategy no longer works. At this point, you can go back to adjusting the strategy or testing a new one until you find a sample size of 30 trades that are profitable.
*After a benchmark is broken, take another 30 trades on your demo account to backtest to see if you are ready to take your strategy to live trading, or if you need to continue analysis.
You must continually monitor the performance of the strategy to see if a benchmark is broken, even if the strategy performs well, because the market environment can change at any time causing your strategy to become ineffective.
Summary
• to backtest a strategy, first you need to record the data from 30 trades on a demo account.
• 30 trades is enough of a sample size to see several cycles of winning and losing streaks.
• if the initial backtesting of 30 trades is profitable, then you can go on testing until you have 100 trades.
• after you have a minimum of 100 trades, then you can create benchmarks – a maximum negative impact that you allow on your account.
• when you have your benchmarks, you can then trade on a live account and monitor your trade until a benchmark is broken.
• if one of your benchmarks is broken, then you need to test the strategy again on a further 30 trades.
•if the further 30 trades is profitable, then your benchmark can be adjusted and you can continue live trading. If it is not, then you must modify or change your strategy and start the testing over again.
