By Benjamin Roure L'Hermitte
A trading dashboard is the screen that gathers the few numbers about your trading worth monitoring at a glance, and it is judged on one thing: has any number on it ever made you change a behaviour? One that stacks impressive metrics leading to no decision never will. Here are the ones that change something, and the ones you can remove without losing anything.
- A useful metric is one that makes you act differently next week. The rest is decoration.
- Cumulative P&L is the least actionable number on the board: it says where you stand, never why.
- Breakdowns beat averages: by hour, by instrument, by size, that is where the leaks show up.
- A dashboard is read cold, outside market hours, or it turns into an emotional thermometer.
Most trading dashboards show the same blocks: equity curve, win rate, average win to average loss, streaks. The MetaTrader 5 help on the testing report, for instance, lists the share of winning trades, the average profit and average loss per trade, and the longest winning streak. Those numbers are accurate, easy to compute, and of little use on their own, because they describe a result without ever naming a cause. A 44% win rate tells you neither what worked nor what to stop. It does not even tell you whether you make money: at 44% you break even, before costs, when your average win is about 1.27 times your average loss (56 ÷ 44).
This article starts from use, not from a feature list: which number to look at, how often, and above all what decision it enables. It doubles as the grid for judging any tool before subscribing, ours included.
What a dashboard must answer, and what it should not display
The right question is not "how much did I make", it is "what do I repeat next week, and what do I stop". A trade dashboard built for the first question gives an exact number and no follow-up. Built for the second, it puts recurring situations and their cost in front of you. Stephen Few draws the same line: in his article 'Dashboard Confusion' (2004) he defines a dashboard as a visual display of the most important information needed to achieve one or more objectives, consolidated and arranged on a single screen so it can be monitored at a glance, and adds that such a screen tells you what is happening, but not why: diagnosis requires further detail.
In practice a metric earns its place if it can trigger a precise action. Results by hour of day qualify: if the first half hour costs you regularly, the decision that follows is obvious. A streak of consecutive wins does not: whatever its value, it tells you nothing to do, and mostly risks feeding confidence at the worst moment. Removing is not a detail: in Stephen Few's list of thirteen common pitfalls in dashboard design, every unnecessary piece of information results in wasted time, and when everything is visually prominent, nothing stands out.
| Keep | What it triggers | Remove | Why |
|---|---|---|---|
| Results by hour | Cut or shrink a time window | Cumulative P&L | Describes, does not explain |
| Results by instrument | Drop a market that does not suit you | Current streak | No action, confidence bias |
| Real risk per position | Fix your sizing | Total trade count | Volume is not performance |
| Plan versus execution gap | Re-apply a rule you dropped | Leaderboards | Compares incomparable contexts |
The P&L calendar, the block you read at a glance
A calendar colouring each day by its result is a block people come back to willingly, and understandably so: an anomaly is easier to spot in a month grid than in a sorted table. Stephen Few says it of dashboards in general: presented properly, pictures of data can be scanned and understood much faster than the same data presented as text. A systematically red Monday jumps out where an average would have diluted it. It is the heart of any PnL tracker, and the easiest block to build yourself in a spreadsheet.
Its value is not the monthly total, it is the shape: losses concentrated on a few days, or spread everywhere? One day wiping out a week of work? An immediate re-entry after a large loss, the signature of revenge trading? Those are questions a smoothed equity curve never asks.
The trap is looking at it during the session. A calendar consulted while the market is open risks becoming an emotional thermometer that pushes you to size to the mood of the day rather than to the plan. The tendency has been measured among professionals: in Coval and Shumway's study in the Journal of Finance (2005), Chicago Board of Trade proprietary traders who lose in the morning are about 16 percent more likely to take above-average risk in the afternoon than those with morning gains. That result is about those traders, not about a retail trader at a screen, but it shows that a loss on the day can weigh on the risk taken next. Read the calendar at the weekend, or in the evening, with no position open.
Breakdowns, where the leaks actually are
A global average always hides a sub-population dragging it down. The useful work is slicing: by entry hour, by instrument, by weekday, by position size, by holding time. The aim is to check whether one or two of those slices concentrate most of your losses: when they do, you have a decision. The MetaTrader 5 testing report draws part of that breakdown itself, with its diagrams of profits and losses by hours, weekdays and months.
The breakdown by size is one of the most revealing. If you discover, once you plot it, that your largest positions are also your worst, you are looking at sizing driven by confidence rather than by risk. That is the kind of finding no monthly total will ever surface.
One guardrail goes with every slice: sample size. A bucket holding six trades proves nothing, and an honest dashboard shows the observation count next to each line. The confidence interval calculation presented in the NIST statistics handbook, from the US standards institute, gives the measure: two winning trades out of six, or 33%, are compatible at 95% with a true win rate anywhere between about 10% and 70%; twelve out of forty, or 30%, with a range of 18% to 45%. Without it you mistake a coincidence for a pattern, and cut a profitable time window on the strength of three bad sessions.
Measure the risk you actually took, not the one you planned
You know what risk you intended to take. Do you know what risk you took? Planned risk has a unit: in the Van Tharp Institute's trading concepts, R is the initial risk of a trade, what you lose if your planned exit is hit. The gap comes from positions added along the way, stops moved, and sizes chosen on instinct after a loss. A useful dashboard shows effective risk per position, as a percentage of capital at the time of the trade.
That measurement matters mostly through its spread. A trader who believes they risk 1% everywhere, and whose positions actually range from 0.4% to 3.5%, a spread of nearly one to nine (3.5 ÷ 0.4 = 8.75) where a single loss at 3.5% costs as much as three and a half losses at the planned risk, does not have a strategy problem, they have a sizing problem, and that is a far quicker fix than changing method.
Two measures round out this block, because they describe the path of a trade rather than its outcome: MAE and MFE, the furthest price travelled against you and in your favour while the position was open. The MetaTrader 5 help describes them as the maximum potential loss and the maximum potential profit that occurred during the lifetime of a position.
Tracking the gap between plan and execution
The least spectacular block, and probably the most useful: the share of trades that followed your own rules. It assumes the rules are written, which is already a harsh filter, but it turns an impression into a number. "I follow my plan most of the time" becomes a proportion, measured over dozens of trades.
This split separates two problems that are easy to confuse. If your compliant trades are profitable and your off-plan trades lose, your method is fine and your execution is what costs you. If both populations lose equally, the method itself is the problem. Those two diagnoses call for opposite fixes, and without the split you work blind.
The prop firm dashboard: what a funded account must watch on top
On a funded account two numbers outrank everything else, because breaching them can cost you the account: the maximum daily loss and the trailing drawdown. One example: according to Topstep's help page on its Maximum Loss Limit, the account is liquidated immediately if the balance hits that limit at any point during the trading day, unrealized P&L included. A prop firm tracker worth its name shows them as distance remaining, not as value reached: what matters is what is left before the breach, not the ground covered.
The specific difficulty is that the limit moves, under rules specific to each firm. Topstep's trails the end-of-day balance, never moves down and locks once it reaches the starting balance. At FTMO, according to FTMO's trading objectives page, the daily loss limit is recalculated every day at midnight Central European time from the account balance, and it is equity, open positions included, that must not drop below it. These rules change: read your own firm's at the source. A tracker using a generic formula misleads you exactly when being wrong is fatal. That is the first thing to check in any tool claiming to follow a funded account.
Spreadsheet or dedicated tool: what each one really does
A spreadsheet does everything above, provided you give it time and accept manual entry. Its real advantage is freedom: you decide every column, and your data stays with you in a format everyone can read. It is a perfectly defensible choice, especially at the start, and a trade journal spreadsheet template is plenty to find the first leaks.
Its limit arrives with volume and consistency. Manual entry takes daily discipline, and an incomplete journal can be worth less than no journal, because it makes you draw conclusions from a biased sample, the trades you felt like recording. A dedicated tool earns its keep through automatic import, not through its charts.
| Criterion | Spreadsheet | Dedicated tool |
|---|---|---|
| Data entry | Manual, must be kept up | Import from your broker |
| Layout freedom | Total | Whatever the vendor planned |
| Cross-breakdowns | Build them yourself | Immediate |
| Effort to sustain over time | Entering every trade | Checking the import |
How to judge a trading dashboard before subscribing
Three questions are enough. First: does the tool read the broker you actually use? Without that everything else is theoretical, and you fall back into the manual entry you were trying to escape. Second: does it show the number of observations behind each statistic? A tool that concludes without saying from how many trades is selling confidence, not measurement.
The third is about the exit: can you get your data back in an open format? Your history belongs to you, it is the one asset you genuinely build month after month, and a tool with no export makes you captive. All three answers are available before you pay, and they sort the candidates faster than any feature comparison.
The Tradoshi dashboard
Tradoshi shows the P&L calendar, breakdowns by hour, instrument and size, real risk per position, and funded-account tracking with distance remaining. Manual backtests are journaled alongside live trades rather than in a separate tool, so the same board serves both. Trades arrive through a broker connection or a statement import, precisely so the board does not depend on data entry that is hard to sustain over time.
What we deliberately do not do: no leaderboard between users, and no statistic displayed without its sample size. The first compares incomparable situations, the second lends the appearance of proof to a handful of trades. A dashboard that stays quiet when it does not know is more useful than one that always has something to say.