Trading Behaviour Analysis Deep Dive

When you click into the Behaviour card on the results dashboard, a full modal opens with four panels arranged in a 2x2 grid, an interpretation block, and a collapsible statistics row at the bottom. This is the most psychologically oriented view in the platform — it answers the question: what would it actually feel like to trade this strategy?

The four panels are: Win/Loss Asymmetry (top left), Streak Behaviour (top right), Trade Outcome Sequencing (bottom left), and Behaviour Profile (bottom right). Below the grid, Observed Behaviour Interpretation provides a two-column summary of characteristics and implications, followed by a collapsible Additional Statistics row.

At the top, a subtitle states: "Psychological and structural patterns derived from N trades." The word psychological is deliberately chosen. The Performance Metrics modal tells you what the numbers are. This modal tells you what those numbers feel like when you are the one holding the position.

The Win/Loss Asymmetry panel shows four horizontal bars: Avg Win, Avg Loss, Best Win, and Worst Loss. Each bar is scaled proportionally against the largest value, and each carries a signed currency amount.

How to read it

The visual comparison is immediate. If the green bars (wins) are materially longer than the red bars (losses), the strategy depends on asymmetry — winning trades are bigger than losing trades. If the bars are roughly equal, the strategy is balanced. If the red bars dominate, the strategy needs a high win rate to survive.

Avg Win is the mean profit of all winning trades. Avg Loss is the mean absolute loss of all losing trades. These two values produce the payoff ratio (Avg Win / |Avg Loss|), but seeing them as bars reveals the relationship more intuitively than a single number.

Best Win and Worst Loss are the extremes. A Best Win that is many times larger than the Avg Win suggests concentrated returns — the strategy's profit depends on catching one or two exceptional moves. A Worst Loss that dwarfs the Avg Loss signals tail risk — most losses are manageable, but occasionally the strategy takes a hit that is far larger than typical.

The caption

Below the bars, a generated caption describes the asymmetry pattern:

• "Wins are materially larger than losses — payoff-driven behaviour" — appears when the average win exceeds the average loss by more than 50%. The strategy depends on letting winners run, not on being right often.
• "Losses exceed wins — requires high accuracy" — appears when the average loss exceeds the average win by more than 20%. Each loss is larger than each win, so the strategy must win frequently enough to compensate.
• "Balanced win/loss magnitude relationship" — the default when neither side dominates.

What to look for

Compare the ratio of Best Win to Avg Win. If the best single trade is, say, 3x the average winner, a non-trivial share of the total profit came from one trade. That trade might not repeat. Now look at Worst Loss relative to Avg Loss — if it is 3x larger, there is a tail event in the data that the strategy did not protect against. Both patterns matter for setting realistic expectations about live performance.

The Streak Behaviour panel answers the hardest question in trading psychology: how bad will the bad stretches get?

Max Loss Streak and Max Win Streak

These are displayed as prominent numbers — the maximum consecutive losses (in red) and maximum consecutive wins (in green). In the example shown, the max loss streak is 9 and the max win streak is 3. This immediately tells you: the strategy's worst losing run was three times longer than its best winning run.

A max loss streak of 9 means that at some point during the backtest, the strategy lost nine trades in a row. In a live environment, that means nine consecutive entries where you risk capital and lose. Each loss compounds — not just financially, but psychologically. By trade 5 or 6, most traders begin questioning whether the strategy is broken. By trade 9, many have abandoned it entirely. The platform shows this number explicitly because it is the single best predictor of whether a trader will actually stick with a strategy.

Loss Streak Distribution

Below the headline numbers, a histogram shows the frequency of loss streaks by length. In the example, there were 8 streaks of length 1 (a single loss before a win), 5 streaks of length 2, 3 streaks of length 3, and so on — with one extreme streak reaching 9.

The distribution shape matters. If most streaks are short (length 1-2) with a single long outlier, the strategy is generally well-behaved but has one structural vulnerability — possibly tied to a specific market regime. If the distribution is flat (plenty of 3s, 4s, 5s), losing runs of moderate length are a normal part of the strategy's operation, not an anomaly.

Bars at length 5 or above are highlighted in a stronger red, a visual cue that these streaks are psychologically significant.

The caption: psychological preparation

If the max loss streak is 5 or above, the caption reads: "Extended losing streaks require psychological preparation." This is not generic advice — it is a specific finding from the data. The strategy has demonstrated that it will, under historical conditions, produce losing runs of this length. Any deployment plan needs to account for this reality.

What the streaks mean for position sizing

Max loss streak directly constrains position sizing. If you size each trade at 2% of capital and experience a 9-trade losing streak, you would draw down roughly 16-18% (compounding slightly reduces the impact). If you size at 5%, that same streak costs you approximately 37%. The streak data tells you what loss sequence to plan for, and position sizing is where you respond to that information.

The Trade Outcome Sequencing panel plots cumulative P/L against trade index — a line chart where each point represents the running total after each trade closes.

Why trade index, not time?

Most equity curves plot portfolio value over time. This chart deliberately uses trade index instead. The difference matters: a time-based chart compresses periods when the strategy is not trading (flat equity) and stretches periods of high activity. The trade-index chart normalises for activity — each trade gets equal horizontal space, so you see the actual sequence of outcomes without time distortion.

What the shape reveals

A smooth upward slope means the strategy generates consistent profits trade by trade. A staircase pattern (flat stretches interrupted by upward jumps) means the profit is concentrated in a few trades — the strategy sits through many neutral or losing trades before catching a winner that jumps the equity higher. A volatile zigzag indicates that winners and losers alternate frequently without clear momentum.

Look for flat periods. If the line sits at roughly the same level for 15-20 trades, the strategy was not generating any edge during that stretch. This might correspond to a specific market regime where the strategy's conditions were not met, or it might indicate a structural weakness where the wins and losses roughly cancel out.

Look for equity jumps. If the line leaps upward at one or two points, those trades drove a disproportionate share of the total return. Cross-reference this with the Win/Loss Asymmetry panel: if the Best Win is much larger than the Avg Win, those jumps are likely the best trades. The question is whether those specific market moves will repeat.

The dashed zero reference line

A dashed horizontal line marks zero. Any time the cumulative P/L dips below zero, the strategy was underwater — had you started trading from that point, you would have been in a losing position. How quickly the line recovers above zero and stays there tells you about the strategy's resilience during its weakest phase.

Caption

The caption reads: "Cumulative P/L by trade index — reveals flat periods and equity jumps." It is a reminder to read the chart for its structural character, not just its endpoint.

The Behaviour Profile panel classifies the strategy across five dimensions, each computed from the backtest metrics. Together they form a behavioural fingerprint — a compact description of what kind of strategy this is and what it demands from the trader.

Edge Type

Edge Type describes how the strategy makes money. The classification depends on win rate and payoff ratio:

• Payoff-driven — Win rate below 40% but payoff ratio above 2.0. The strategy loses often but wins big. Classic trend-following signature.
• Accuracy-driven — Win rate above 55% but payoff ratio below 1.5. The strategy wins often but wins small. Classic mean-reversion signature.
• Hybrid — Win rate above 50% and payoff ratio above 1.5. Both accuracy and asymmetry contribute.
• Balanced — Default when no extreme pattern dominates.

Win Dependence

How reliant the strategy is on win frequency:

• Low frequency — Win rate below 45%. The strategy does not depend on winning often.
• High frequency — Win rate above 55%. The strategy needs to win often to stay profitable.
• Moderate — Win rate between 45% and 55%.

Streak Sensitivity

How vulnerable the strategy is to losing runs. This is the only colour-coded field in the panel:

• High (red) — Max loss streak of 7 or more. The strategy has demonstrated extended losing sequences that will test any trader's conviction.
• Moderate (amber) — Max loss streak between 4 and 6. Losing runs are notable but manageable.
• Low (green) — Max loss streak below 4. The strategy recovers quickly from losses.

Streak Sensitivity is colour-coded because it has the most direct impact on whether a trader will actually stick with the strategy. A "High" streak sensitivity is not necessarily disqualifying — many profitable strategies have long losing runs — but the trader must be prepared for it.

Holding Tolerance

Based on trade frequency (trades per month):

• High patience required — Fewer than 3 trades per month. Long stretches between trades; the trader must remain disciplined during idle periods.
• Active monitoring — More than 15 trades per month. Frequent entries require ongoing attention.
• Standard — Between 3 and 15 trades per month.

Recovery Style

How the strategy recovers from drawdowns, derived from payoff ratio:

• Lumpy, event-driven — Payoff ratio above 2.5. Recovery comes in bursts from a few large winners.
• Stepped — Payoff ratio between 1.5 and 2.5. Recovery happens in noticeable steps.
• Gradual — Payoff ratio below 1.5. Recovery is slow and steady, compounding small wins.

Reading the profile as a whole

The five fields together tell a coherent story. The example in the screenshot — Balanced edge, Low frequency, High streak sensitivity, High patience required, Gradual recovery — describes a strategy that requires significant patience: it trades infrequently, experiences long losing runs, and recovers slowly. This is a capital-efficient but psychologically demanding approach. A trader who needs frequent positive feedback will not survive it regardless of the metrics.

Below the 2x2 grid, the Observed Behaviour Interpretation block provides a two-column summary: Characteristics (what the data shows) and Implications (what it means for the trader).

How characteristics are generated

The platform evaluates three aspects of the trade data and produces one or two sentences for each:

Win/loss magnitude:

• If the payoff ratio exceeds 2.0: "Returns accrue through infrequent large winners"
• If the payoff ratio is below 1.0: "Relies on high win rate to compensate for smaller gains"
• Otherwise: "Balanced win/loss magnitude relationship"

Streak patterns:

• If the max loss streak is 5 or above: "Prolonged losing streaks are part of normal operation"

Trade frequency:

• If trades per month is below 3: "Short-term performance is noisy and unreliable"

How implications are generated

Each characteristic is paired with an implication:

• High payoff ratio → "Requires discipline through losing streaks"
• Low payoff ratio → "Edge erodes quickly if win rate declines"
• Balanced → "Moderate tolerance for win rate variation"
• Long losing streaks → "Best suited to patient capital"
• Low frequency → "Evaluate over months, not weeks"

Why this matters

The interpretation block bridges the gap between quantitative metrics and trading reality. A Sharpe ratio of 1.5 tells you the risk-adjusted return was healthy. The behaviour interpretation tells you that achieving that Sharpe required sitting through a 9-trade losing streak while the strategy traded only twice a month — which means the losing streak lasted over four months of real time. That context changes everything about how you plan to trade the strategy.

The interpretation is deliberately conservative and honest. It never says "this strategy is well-suited for you." It describes the demands the strategy places on the trader and lets you decide whether those demands match your temperament and capital.

At the bottom, a collapsible row displays four supporting values: Win Count, Loss Count, Median Loss Streak, and Trades/Month.

Win Count and Loss Count

The raw numbers behind the win rate. In the example: 27 wins and 35 losses, giving the 44% win rate seen in the Performance Metrics modal. Seeing the raw counts is useful for mental arithmetic: 35 losses in 62 trades means more than half the trades are losers. If you cannot tolerate losing more often than winning, this strategy is not for you — the numbers make that explicit.

Median Loss Streak

The typical (median) length of a losing run. This is distinct from the maximum loss streak shown above. In the example, the median is 1, meaning most losing "streaks" are just single isolated losses. The max streak of 9 is an outlier, not the norm. This distinction is important: if the median were 4, losing runs of that length would be routine, not exceptional.

Trades/Month

The average number of trades per calendar month across the backtest period, calculated as total trades divided by the time span in months. A value of 0.6, for example, means roughly one trade every 50 days. This feeds directly into the Holding Tolerance classification in the Behaviour Profile — anything below 3 triggers "High patience required". Low trades per month means long idle periods between entries, which sounds easy but is psychologically difficult for most traders. The urge to intervene during weeks of inactivity is one of the most common reasons strategies are abandoned prematurely.

The Behaviour modal is structured as a progression from facts to feelings to implications. The top row presents raw data (what happened). The bottom row classifies patterns (what kind of strategy this is). The interpretation block translates everything into trader-facing language (what this means for you).

A practical reading order

1. Glance at Streak Behaviour first. The max loss streak is the single most important number in this modal. If it is above 7, every other panel needs to be read through the lens of "can I survive this?"
2. Check the Win/Loss Asymmetry bars. Are wins materially larger than losses, or roughly equal? This tells you whether the strategy depends on accuracy or asymmetry — which shapes how you interpret everything else.
3. Read the Behaviour Profile. The five classifications form a compact summary. Pay special attention to Streak Sensitivity colour: green is comfortable, amber is notable, red requires planning.
4. Scan the Trade Outcome Sequencing chart. Is the equity smooth or lumpy? Are there flat periods? Do one or two jumps drive most of the profit? If so, cross-check against the Win/Loss Asymmetry panel.
5. Read the Interpretation block last. It synthesises everything above into plain language. The implications column tells you exactly what the data demands of you as a trader.

How behaviour connects to other cards

The Behaviour modal does not exist in isolation. It connects to every other card on the dashboard:

• Performance Metrics: The metrics tell you what was achieved. Behaviour tells you what it cost psychologically. A high Sharpe with "High" streak sensitivity means the returns were efficient but the path was painful.
• Trade Structure: If behaviour shows a payoff-driven edge with concentrated returns, the Trade Structure card will confirm whether the top 20% of trades drove the majority of profit.
• Drawdown Analysis: The max loss streak in Behaviour maps directly to the worst drawdown periods in the Drawdown card. A 9-trade loss streak will appear as a sustained drawdown in the underwater curve.
• Regimes: If the Behaviour Profile shows "High" streak sensitivity, the Regimes card can reveal whether that long losing streak occurred during a specific market regime — which tells you whether it is a structural vulnerability or a one-time event.

Common traps

Ignoring streaks because the total return is good: A 77% total return with a 9-trade losing streak means the strategy will, at some point, feel like it is completely broken. If you cannot hold through that, the total return is irrelevant.

Overweighting edge type labels: "Balanced" does not mean good, and "Payoff-driven" does not mean bad. They are descriptions of mechanism, not quality judgments. A payoff-driven strategy with a Sharpe of 2.0 is excellent. A balanced strategy with a Sharpe of 0.3 is not.

Assuming low streak sensitivity means low risk: A max loss streak of 2 is psychologically comfortable, but if each loss is large (check the Win/Loss Asymmetry panel), those two losses might represent a significant drawdown. Streak length and loss magnitude must be read together.

The Behaviour modal is the bridge between statistics and reality. The metrics tell you what happened on paper. Behaviour tells you what it would actually be like to trade it.