Trading Journal for Gold Traders: What to Track

A trading journal for gold has to record the things that actually change an XAUUSD result: which session you were in, how wide the spread was, which level you used, whether a scheduled release was nearby, and whether you sized the trade from a stop or from a feeling. A generic forex log that stores symbol, buy/sell and profit will tell you that you made or lost money. It will not tell you whether your gold process is worth repeating.

If you only remember three questions after a session, use the shorter habit in the journal that protects your account. This page is the field list behind that habit. None of the example numbers below are a track record. They are blanks you fill with your own trades.

Why gold needs a gold-specific journal

Gold is one symbol with several personalities. The same "breakout" tag means a quiet Asia fakeout at one hour and a New York data spike at another. A journal that cannot filter by session will average those into one win rate and then give you advice that fits neither hour.

Contract math is the second reason. A one-dollar move in the gold price is not a pip in the forex sense, and two brokers can disagree on whether 0.01 or 0.10 is a pip. If you journal "pips" without writing the definition, your R multiples will not match your dollar P&L. Write dollars at risk and R, and keep pip value as a note in the header of the journal, not as a column you reinvent every row.

The third reason is behavior. Gold's speed makes revenge trades expensive. A journal that only celebrates P&L will hide a green day that was three impulsive adds after a loss. You want a column that can show a profitable day with a failing process. That column is rule-following, not profit.

You do not need fifty fields. You need fields you will actually complete on a losing trade, because losing trades are the ones you are tempted to leave blank. If a field is empty on more than a handful of rows, delete it or make it a checkbox. An abandoned column is not a process.

Core fields (session, setup, R, spread)

These are the columns that should exist before you add commentary.

Session. Use a fixed set: Asia, London, London–New York overlap, late New York, other. Define the clock once, in UTC or in your own time zone, and do not translate it differently each week. The point of the tag is to compare like with like. If you trade the overlap, you should be able to see that bucket alone. The map of those hours is in the best time to trade gold.

Setup. One name from a list you wrote down: Asia range breakout, level rejection, order block, news stand-aside, or whatever you actually trade. Free typing produces six names for one idea. A dropdown of five names produces a sample you can trust.

Direction and prices. Side, entry, stop, and target as you planned them, plus the prices you actually got. The gap between plan and fill is the spread and the slip. Store both. Editing the stop in the journal after you moved it on the platform deletes the lesson.

R. Risk unit, not pips. One R is the dollar amount you accepted if the planned stop was hit. A winner that made twice that amount is +2R even if the dollar figure looks large because the account grew. R lets a 0.01-lot trade sit next to a 0.10-lot trade without the bigger position dominating the review. If you do not know the dollar risk before you click, the position size comes before the journal row, not after.

Spread. Write the spread at the moment of entry in price terms, for example 0.25 if the quote was 0.25 wide. Gold spreads are not stable. A strategy that looks fine in a 0.20 spread and dies in a 0.80 spread is a session problem, and you will only see it if the column exists. You do not need a tick-by-tick average. One number at entry is enough.

Result. Planned R, actual R, and whether the trade was a win, loss, or scratch under your definition of scratch. Decide the scratch band in advance, such as between -0.2R and +0.2R, so a tiny winner is not counted as proof of an edge.

Example of one row, invented to show the shape: London overlap, setup "daily zone rejection," short, planned risk $80, spread 0.30, actual result -1R, rule followed yes. That single row is more useful than a screenshot with no labels. It is not evidence of anything until you have many like it.

Context fields (levels, news)

Core fields say what you did. Context fields say what the chart was doing.

Level. Name the support or resistance you used: prior day high, Asia low, daily zone, four-hour shelf. If you cannot name it, you did not have a location. "Price looked extended" is not a level.

Higher-timeframe bias. Daily up, daily down, or daily range. One word. This stops a short taken against a daily trend from hiding inside a bucket of shorts that were with the trend.

News. Not a paragraph. A tag: none, scheduled high-impact within 30 minutes, unscheduled headline. You are not writing a macro essay. You are marking whether the trade was a technical idea or a reaction to information. What moves gold is the background for which events deserve the tag. CPI, payrolls, and the Fed are the usual ones. A random opinion piece is not.

Session range note. Optional checkbox: "Asia range already wide" or "overlap just opened." This is how an Asia breakout journal stays honest. If you only log the breakouts you liked, the skip rule never gets a sample.

Attach a chart image if you will look at it. Mark the level, the entry, and the stop before you upload, not a clean chart that could be any trade. If you will not open the image during the weekly review, do not spend the minute saving it.

Psychology fields

Keep this section smaller than the numbers. A mood diary that replaces the R column feels deep and teaches nothing you can count.

Use a short emotion tag from a fixed list: calm, hesitant, rushed, revenge, overconfident. Add a second checkbox: "I wanted to make back the last loss." That box is the revenge flag. Count how often it is ticked and what those rows returned. The cost of the pattern is the sum of R on ticked rows, not a story about discipline.

Add one process score from 0 to 2:

• 0 means you broke a written rule (moved the stop, oversized, traded outside the session).
• 1 means the setup was allowed but the trigger was sloppy.
• 2 means you followed entry, size, and stop.

Average that score by week. It will move before your P&L does. A week with a positive P&L and a falling process score is a warning, not a celebration. The inverse, a red week with a high process score, is the week you do not "fix" by changing the strategy.

Do not ask yourself to write a paragraph on every trade. Ask for one sentence on the trades scored 0: what rule broke, in words a stranger could check. "I was anxious" is not checkable. "I added a second position after the stop on the first was hit" is checkable.

Review cadence

The journal fails when the review is a monthly archaeology project. Use three passes, and keep each one short enough that a losing day does not convince you to skip it.

Same day, five minutes. Fill the row before you leave the desk. Score the process. If you hit a loss limit, write "stopped" so tomorrow's you does not have to infer it from a flat equity line.

Once a week, thirty minutes. Filter by setup and by session. You are looking for a bucket with enough trades to be annoying, not for a conclusion from four samples. Note one thing to repeat and one thing to stop. If a setup is negative and the process score on those rows is high, the setup is the suspect. If the process score is low, you have not tested the setup yet. You have tested your willingness to follow it.

Once a month. Only now look at overall expectancy in R. Expectancy is average R per trade, wins and losses together. A high win rate with losers that are many R larger than winners can still be a losing journal. The monthly pass is also when you delete a field nobody filled in.

Do not change a rule in the middle of a week because one trade hurt. Write the proposed change, and start it on a date you can filter. Otherwise the journal mixes two strategies and blames the market.

Tools

A spreadsheet is a valid gold journal if the columns above exist and you fill them. The failure mode is friction: exporting fills by hand, forgetting the spread, and skipping the losing row because the sheet is on another laptop. An app is worth it when it removes that friction without deleting the fields that matter for gold.

TradYX is built only for XAUUSD, so the journal sits next to the session, the levels, and the replay rather than next to a watchlist of unrelated markets. The Trading Journal is the place to log the next gold trade with the fields on this page: session, setup, R, and the note on whether you followed the stop. MT4 and MT5 history can land in that journal so the prices are not retyped. You still write the setup name and the process score. No import knows whether you were revenge trading.

If you already keep the three-question habit after each session, do not replace it with this longer form. Attach this form to the trades, and keep the three questions for the day. The day score tells you whether you should have been trading. The trade rows tell you which idea, in which session, paid the bills.

Start with the core fields for twenty trades before you add a new column. A short journal you complete beats a perfect schema you abandon on the third loss.