Trading practiceUpdated 3 min readBy the FXNOD team
How to review and improve your trading results
Review your trading once a week from your trade history and journal. Measure the win rate, average win, average loss, expected value and largest drawdown. Then split the trades by setup, time and whether you followed your plan, to find the group that loses. Change one thing, keep it for at least fifty trades, and measure again.
Step 1: get the facts
| Number | How to get it |
|---|---|
| Trades taken | Count them |
| Win rate | Wins divided by trades |
| Average win and average loss | Total won over wins; total lost over losses |
| Expected value per trade | Net result divided by trades |
| Largest drawdown | The biggest fall from a peak in the balance |
Step 2: split the trades
The total hides the answer. Sorting the same trades into groups reveals it. Most traders find one group that carries nearly all of the losses.
- By plan: trades that followed your rules against trades that did not.
- By setup: each type of entry on its own.
- By time: session, day of the week, first hour against the rest.
- By sequence: trades taken straight after a loss against all others.
- By market: each instrument separately.
Step 3: separate process from outcome
| Followed the plan | Broke the plan | |
|---|---|---|
| Won | Good trade | Bad trade that happened to pay: the most dangerous kind |
| Lost | Good trade with a normal loss | Bad trade |
Judge yourself on the columns, not the rows. A planned loss is the system working. An unplanned win teaches a habit that will cost more later.
Step 4: change one thing
Pick the largest leak
The group that lost the most, not the mistake that annoys you most.
Write one rule against it
For example: no trade within ten minutes of a loss.
Hold it for fifty trades
Fewer than that and you cannot tell the change from chance.
Measure the same numbers again
Keep the rule if the numbers improved, drop it if they did not.
Reviewing a bot
For automated trading, review runs. Record each run's settings, how it ended and its result, then compare by setting: does the bot do better on one market, one stake rule, one contract length? Change one setting between tests. Also review your own decisions: every time you stopped a bot early or restarted it with a different stake, write down why.
What not to do
- Do not review after every trade. One result is noise.
- Do not change several things at once.
- Do not judge a change by its first five trades.
- Do not skip the review after a good week. Good weeks hide bad habits.
Questions people ask
- How often should I review my trades?
- Weekly for the numbers and patterns, monthly for the larger trend. Daily review is useful only for checking whether you followed your rules.
- What is the most important trading metric?
- Expected value per trade, because it combines win rate and payoff. After that, maximum drawdown, because it decides whether you can keep going.
- How many trades before I change my strategy?
- At least fifty with the same rules, and more if the strategy wins very often or very rarely. Changing sooner means reacting to luck.
Test your rules before you fund them
Turn a rule into a bot in Bot Forge by answering questions, run it on your Deriv demo account, and read the result trade by trade.
Get started