Blog · Oct 27, 2026

Trading journal metrics: what to measure, how to calculate it, and when to trust it

Start with net P&L after fees and results in R; then expectancy, profit factor and max drawdown; then rule adherence. Read them per setup and only on enough trades.

Each of these numbers can be calculated correctly and still mislead you. A win rate can look healthy on a book that loses money, and a profit factor can look strong on twelve trades.

This guide covers the trading journal metrics worth tracking for futures, which double as your main trading performance metrics: the formula for each, the question it answers, and when it has enough trades behind it to deserve attention. Every formula is worked through on one illustrative 10-trade MNQ book, gross and net side by side. It applies the same way to live, sim and prop accounts.

Start here: the six numbers that matter first

If you track nothing else, track these six.

Outcome numbers need a sample. Process numbers like rule adherence mean something from the first trade.

Get the inputs right first

Four input problems cause most bad numbers in futures journals.

Net after fees, not gross

Every futures trade pays commissions, exchange and platform fees. Record them as one round-turn amount per contract and subtract them from every trade. Our example uses an illustrative $1.30 round turn per MNQ contract; check your own with the round-turn fee calculator or your statement.

It follows that a scratch is a loss. A trade that exits at its entry price makes $0 gross and loses the fees net. Count scratches as zero and your win rate, average loss and profit factor drift away from your statement.

R needs a planned stop written before entry

R is the result divided by the dollar amount you planned to risk. Planned risk is the stop distance in ticks, times the tick value, times the number of contracts. A 40-tick stop on 2 MNQ is 40 × $0.50 × 2 = $40. If you need the tick values for ES, NQ, MES and MNQ, they're $12.50, $5.00, $1.25 and $0.50 per tick, with four ticks per point (source: CME Group contract specifications).

The stop has to be written before the entry; a stop chosen after the exit fits whatever happened. If you size from the stop, work out the position size from your stop in ticks before the trade.

File each trade on the CME session day, and tie the day first

The CME equity index session for a trade date opens at 5:00 p.m. Chicago time the evening before. A trade at 7:00 p.m. Chicago on Monday belongs to Tuesday's trade date, and your broker statement agrees. File trades on the CME session day, not the calendar date, or your daily totals, daily drawdown and weekday breakdowns will all be off by one.

Then tie the day to your statement before you read any metric for it. If the journal says +$213.00 and the statement says +$209.00, something is missing: a fee, a partial fill, a trade on the wrong day. Every metric built on that day inherits the error.

One account per book

Keep live, sim, eval and funded accounts in separate books. Sim fills are cleaner than live fills, and mixing accounts produces averages that describe no account you actually trade. Prop traders should keep separate books for eval and funded accounts, because the rules and drawdown differ. Compare accounts side by side; don't merge them.

Outcome metrics

Outcome metrics describe results, and they need the largest samples.

Win rate, and why it means little alone

Win rate = winning trades ÷ all trades, counting net after fees. In our example book, 6 of 10 trades are net winners, so the win rate is 60%. The scratch counts as a loss, because it lost $2.60.

On its own, win rate says little about whether a book makes money. A book can win 80% and lose money, or win 35% and make money. It becomes useful only next to the average win and average loss.

Average win, average loss and payoff ratio

Average win = sum of winning trades ÷ number of winners. Average loss = sum of losing trades ÷ number of losers. Payoff ratio = average win ÷ average loss, using the size of the loss.

Payoff connects to win rate through one formula: break-even win rate = 1 ÷ (1 + payoff). With a payoff of 1, you need to win half your trades to break even:

From gross numbers these are break-even rates before fees, and the real bar is higher; from net numbers, fees are already inside. That's the honest answer to "win rate vs risk reward": neither is good or bad alone. Each tells you what the other has to be.

Expectancy in dollars and in R

Expectancy is the average result per trade. The simplest version is net P&L ÷ number of trades. The same number comes out of the longer formula: (win rate × average win) − (loss rate × average loss).

In R, expectancy is the average R per trade. It's the better version when your size changes, because a 1-contract trade and a 4-contract trade count equally.

Expectancy is what past trades in this sample earned on average, not a forecast. A positive expectancy on 15 trades is a reason to keep collecting data, not to size up. The expectancy calculator can do the arithmetic; the reading is up to you.

Profit factor, gross vs net

Profit factor = sum of winning trades ÷ sum of losing trades, with losses taken as a positive number. A profit factor above 1 means the wins added up to more than the losses over the sample.

Calculate it net of fees. Gross profit factor is always higher, because fees shrink every win and enlarge every loss. In our example, gross is 2.98 and net is 2.62: same fills, different answer. Your statement agrees with the net one.

There's no universal "good" profit factor. It depends on sample size, fees, the contract and how you trade.

Profit factor vs expectancy vs win rate: when they disagree

These three can tell different stories about one book.

• High win rate, low or negative expectancy: many small wins, a few large losses. Look at your R distribution for losses well beyond −1R.
• Low win rate, positive expectancy: few wins, but large ones. Normal for trend-following styles. Expect longer losing streaks.
• Profit factor and expectancy point the same way, but one is barely above break-even: fees may decide it. Compare gross and net.
• Strong numbers on a small sample: none of the three can be trusted yet. See "When to trust a metric" below.

If you kept only one, expectancy in R is closest to the real question, but you need win rate and payoff to understand why it is what it is.

Risk and path metrics

Outcome metrics ignore order. Two books with the same total can take very different routes.

Max drawdown in dollars and in R

Max drawdown is the largest drop from a running peak to a later low. Calculate it on the running total of closed-trade net P&L: keep a running peak, subtract each new running total from that peak, and take the largest difference.

Report it in dollars (what it cost) and in R (how many full planned losses it was). In our example the max drawdown is −$43.60, or −1.09R, from a peak of $167.00 to $123.40.

Max drawdown only grows with time. A short book shows a small one simply because there hasn't been time for a bad stretch.

Largest losing day vs your daily loss limit

Group trades by session day, find the worst day and compare it with your daily loss limit. If the worst day is bigger, the limit isn't working, and that's a process problem. Count days that hit the limit and days that went past it.

Prop accounts: distance to the limits

On a prop eval or funded account, the metric that matters most is one you check before you trade: how many dollars stand between your current balance and the trailing drawdown floor, and between today's net and the daily loss limit. Write both down at the start of each session. Read how trailing drawdown works for your firm's version, and check the firm's current rules, which change.

Losing streaks

Losing streaks happen even on a book with positive expectancy; at a 50% win rate, runs of several losses come up by chance over many trades. Track your longest streak and decide ahead of time what you'll do during one. A streak doesn't make the next trade more likely to win, so don't size up to win it back.

R distribution

List every trade's R. With stops placed before entry, losses should cluster near −1R, a little past it because of fees and slippage. In our example, the stop-outs are −1.065R (fees only) and −1.09R (fees plus a 1-tick slip).

Losses well beyond −1R usually point to slippage on a fast move, a stop moved away from price, or a position larger than planned. The last two are mistakes to tag.

Process metrics (judge these on every trade)

Process metrics measure whether you did what you planned. They don't need a large sample.

Rule adherence rate, and the net and R of off-plan trades

Rule adherence = trades that followed the plan ÷ all trades. Then split the book in two: on-plan trades and off-plan trades. Report net and average R for each.

If off-plan trades lose money, cutting them helps. If they make money, that's more dangerous, because it rewards breaking the plan. Either way, the plan can't be measured until most trades follow it.

Cost of each mistake tag

For each mistake tag, such as chase, moved stop away, size above max, traded after daily stop, averaged down or revenge entry, add up the net P&L of the trades carrying it. The tag with the largest cost is next month's priority. To pair each tag with the state you were in before entry, see the trading psychology journal.

Plan written before entry (% of trades)

The share of trades with setup, stop and planned risk written before entry. This makes R honest; if it's low, every R number is partly reconstructed.

Fees as a share of gross

Fee share = total fees ÷ gross P&L. In our example, $26.00 of fees on $239.00 gross is 10.9%. Watch it if you scalp: as targets shrink, the same fee takes a larger share of each win.

Futures-specific metrics

Futures add a few metrics that change how you read the rest.

P&L in ticks per contract vs dollars

Dollar P&L depends on size; ticks per contract don't. Divide P&L by tick value and contracts to compare 1-lot and 5-lot days. Our example book made 239 ticks per contract gross. Fees of $1.30 per contract are 2.6 MNQ ticks, so net is 213 ticks per contract.

Results by contract

Report ES, MES, NQ and MNQ separately. One ES tick is worth ten MES ticks, and fees are a much larger share of a micro trade. Don't merge micros and minis into one dollar average; if you must combine them, do it in R.

Session buckets

Split results by time of day using Chicago time, for example: overnight (Globex), the RTH open (8:30 a.m. to about 10:00 a.m.), midday, and the close (into 3:00 p.m.). Use your own boundaries, but keep them fixed so weeks can be compared.

Roll week and news days

Tag trades taken in roll week, when volume moves to the next contract, and on major scheduled data days. If they're worse than the rest, add a rule about trading them.

Hold time

Bucket trades by hold time, for example under 2 minutes, 2 to 10, 10 to 30 and longer. Short holds pay the same fee as long ones.

MAE and MFE (optional)

Maximum adverse excursion (MAE) is how far a trade went against you before closing; maximum favorable excursion (MFE) is how far it went your way. They can hint at stops that are too tight or exits that are too early. They need price data during each trade, which a broker export usually lacks. A journal without them is still complete.

Worked example: a 10-trade MNQ book

All figures below are illustrative. The setup: 10 MNQ trades, 2 contracts each, every trade with a 40-tick planned stop. Planned risk per trade is 40 × $0.50 × 2 = $40.00. The illustrative round-turn fee is $1.30 per contract, so $2.60 per trade. R is net ÷ $40.00.

Now every metric, gross and net side by side:

To check expectancy by the long formula: 0.6 × $57.40 − 0.4 × $32.85 = $34.44 − $13.14 = $21.30. In R: the six wins average +1.435R and the four losses average −0.821R, so 0.6 × 1.435 − 0.4 × 0.82125 = 0.861 − 0.3285 = +0.5325R. The net payoff of 1.75 implies a break-even win rate of 1 ÷ 2.75 = 36.4%; the book won 60%.

Fees turned a flat trade into a loss, cut profit factor from 2.98 to 2.62 and took 10.9% of gross. The net average loss is smaller than the gross one only because the $2.60 scratch joined the losers.

Most important: 10 trades is far below the roughly 30 per setup this article treats as a minimum, so these numbers prove nothing about this trader or setup. They show arithmetic, not an edge.

Breakdowns: by setup, time of day, weekday, side

Once the overall numbers are right, the useful questions are about parts of the book: which setup carries it, which session drags, whether shorts work as well as longs.

Why splitting shrinks samples

A 120-trade month split by four setups gives about 30 trades each. Split by four sessions too and you have about 7 per bucket; by side as well, 3 or 4. Break down by one thing at a time, starting with setup.

The many-slices trap

Slice a book enough ways and one slice will look excellent by chance. Check 20 combinations and one or two can look like an edge even if every trade were a coin flip. Decide which breakdowns you'll read before you look, and treat a great slice as a question to test on the next trades, not an answer.

When to trust a metric

About 30 trades per bucket, as a rule of thumb

A common working rule is to wait for about 30 or more trades in a bucket, such as one setup or one session, before you change size or drop a setup because of its numbers. It's a habit, not a law: thirty trades can still mislead you.

Wide vs narrow ranges

A win rate from a sample is an estimate. A confidence interval writes that uncertainty as a range: wide with few trades, narrower with many. Using a standard 95% interval, 6 wins in 10 trades fits a true win rate from about 31% to 83%. 60 wins in 100 trades fits about 50% to 69%. Same win rate, very different certainty. A percentage without its trade count is half a number.

Before you have enough trades, judge process metrics only

Until a setup has enough trades, judge process instead: rule adherence, plans written before entry, mistake-tag counts and statement tie-outs.

Metrics that mislead

• Overall win rate. Without payoff, it can't tell you whether the book makes money.
• Best trade. One trade is an anecdote. It says nothing about expectancy and tempts you to chase a repeat.
• Streak counts on their own. Winning streaks feel like skill and losing streaks feel like failure; both are normal in any sample.
• Gross P&L. It leaves out the costs your statement includes, and it turns scratches into free trades.
• Percent return on a tiny account. One good trade on a small balance produces a dramatic percentage that says nothing about repeatability.
• Any metric on a day that doesn't tie to the statement. If the day's journal net doesn't match the statement, every number built on it is wrong by an unknown amount.

A weekly and monthly metrics routine

Keep it short enough to do in a bad month. For the full weekly session, step by step, see the weekly trading review.

Where a journal like Tiedbook fits

A spreadsheet can compute every metric in this article. If you keep one, the futures trading journal template gives you the columns. The work a spreadsheet doesn't do for you is everything before the formulas: getting fills in, applying fees, filing each trade on its session day and matching each day to the statement.

Tiedbook does that part. It imports your platform's export and shows fee-adjusted net and win rate in dollars, R or ticks from the same fills. R uses one dollar risk you set in Settings, so if your risk changes from trade to trade, keep the planned risk in your note. The Edges view splits results by time of day, weekday, setup tag and hold time. Reports stay quiet until 30 trades: any bucket below that is greyed out, and each bucket shows a confidence interval for its win rate. Eval, funded and live accounts stay separate. The mistake rules count revenge entries, trades after your daily stop, size above your max, widened stops and averaging down. Trades file on the CME session day the calendar uses, and each day can be checked against your imported statement.

Metrics worksheet

Fill this in monthly. The last column matters as much as the number.

Frequently asked questions

What metrics should a trading journal track?

Start with six: net P&L after fees, results in R, expectancy, profit factor, max drawdown and rule adherence. Add average win, average loss and payoff to explain expectancy, and mistake-tag costs. For futures, add ticks per contract and results by contract and session.

Profit factor vs expectancy: which matters more?

Profit factor compares total wins with total losses; expectancy says what an average trade earns or costs. Expectancy in R is usually the more useful single number because it compares trades of different sizes. Read both net of fees, on enough trades.

What win rate do I need for a given risk-reward?

Use break-even win rate = 1 ÷ (1 + payoff), where payoff is average win divided by average loss. With a payoff of 1 you need 50% to break even; with 2, about 33.3%; with 3, 25%. Computed from gross numbers, these rates ignore fees, so the real bar is higher.

What is a good profit factor?

There's no universal number. It depends on sample size, fees, the contract and style, and a high figure on a small sample can be luck. Calculate it net, note the trade count, and compare it with your own history.

How do you calculate max drawdown?

Keep a running total of closed-trade net P&L and the highest value reached so far. At each trade, subtract the current total from that peak. The largest difference is the max drawdown; divide by planned risk per trade to express it in R.

What is an R-multiple?

An R-multiple is a trade's result divided by the dollar risk you planned before entering. Losing exactly the planned amount is −1R; making twice the planned risk is +2R. Planned risk is stop distance in ticks times tick value times contracts. It only works if the stop was written before the entry.

Should metrics be gross or net of fees?

Net. Your statement is net, and every metric should match it. Gross numbers flatter win rate, profit factor and expectancy. Keep gross only to see how much fees take.

How many trades before metrics mean anything?

A common rule of thumb is about 30 or more trades per bucket, such as one setup, before you change size or drop a setup. That's a working habit, not a guarantee. Until then, judge process metrics like rule adherence and plans written before entry, which mean something from the first trade.

What is MAE/MFE, and do I need it?

MAE is how far a trade moved against you before closing; MFE is how far it moved your way. They need price data during the trade, which a broker export usually lacks, so they're optional.

Which stats are worth tracking by hand?

Keep it to net after fees, R per trade, plan followed yes or no, and a mistake tag. From those four you can compute expectancy, profit factor, drawdown, rule adherence and mistake costs weekly.

Trading futures involves substantial risk of loss and isn't suitable for everyone. The examples in this article are illustrations, not trade recommendations.

Related

• Read what to write in each journal entry for the fields and questions that feed these metrics.
• Run your win rate and averages through the expectancy calculator.
• Plan a single trade's stop and target with the risk-reward calculator.