Market Exposure Timeline Deep Dive

The Market Exposure Timeline modal opens from the Exposure card on the results dashboard. It answers a question that return figures never answer on their own: how much of the test period did the strategy actually spend in the market?

Two strategies can produce identical returns over the same period while having completely different exposure profiles. One might be in the market 90% of the time, riding every swing. The other might be flat for months, entering only when a very specific condition is met, and still match the same equity curve. The risk characteristics of these two strategies are fundamentally different — and return alone cannot distinguish them.

The modal is divided into four panels arranged in a 2x2 grid. The top row shows when the strategy was exposed (timeline and density). The bottom row shows how long each position lasted and what the overall exposure structure implies.

The top-left panel displays a horizontal bar representing the entire backtest duration. Green segments mark periods where the strategy held an open position. Grey gaps represent time spent flat — no capital at risk, no exposure to market moves.

Reading the Bar

Dense green clustering in one part of the bar with long grey stretches elsewhere is the first warning sign of regime dependency. A strategy that only traded during 2021's bull run, for example, will show a concentrated block of green with empty space on either side. That pattern means the strategy's entire track record comes from a single market environment.

Evenly spaced green segments across the bar suggest the strategy activates consistently regardless of broader market conditions. This is more encouraging for forward deployment because it implies the entry logic doesn't depend on a specific price environment to trigger.

Summary Metrics

Below the bar, three numbers quantify the visual pattern:

• Participation — the percentage of total test duration spent holding a position. Calculated as the sum of all trade durations divided by the total backtest window. A 9% participation rate means the strategy was flat 91% of the time.
• Median hold — the middle value when all trade durations are sorted. Median is used rather than mean because a single multi-week outlier trade would distort an average among otherwise short-duration trades.
• Longest gap — the maximum stretch of time between one trade closing and the next opening. A 5.4-month gap means the strategy's entry conditions were not met for over five months. This isn't necessarily bad, but it has implications for patience, opportunity cost, and whether the strategy could realistically be followed in live trading.

The top-right panel shows a rolling percentage of time the strategy spent in market, plotted across the test period. The x-axis divides the backtest into twenty equal windows (each representing 5% of the total duration), and the y-axis shows how much of each window was occupied by an open position.

How Density Is Computed

For each of the twenty windows, the system calculates how much trade time overlaps with that slice of the backtest. If a trade spans multiple windows, only the overlapping portion counts toward each window's density. The result is a percentage from 0% (completely flat) to 100% (continuously in market) for each point.

Reading the Curve

A flat line near 0% across the entire chart confirms what the timeline bar already showed: the strategy is almost never in the market. A curve that oscillates between 0% and high values reveals clustered participation — the strategy alternates between bursts of activity and total inactivity.

A consistently elevated line (say, 40-60% across most windows) indicates steady engagement. This is typical of trend-following systems that maintain a position whenever the trend filter is active.

Interpretive Caption

Below the chart, a single sentence summarises the density pattern. "Exposure is clustered rather than evenly distributed" appears when some windows show zero density while others exceed 50%. "Exposure is relatively consistent" appears when all windows fall between 20-80%. Otherwise, the default "Exposure varies across the test period" applies — a neutral description that avoids over-interpreting mixed signals.

The bottom-left panel groups every trade into one of four duration buckets and displays the count in each. The buckets are fixed:

• <1d — intraday trades, opened and closed within 24 hours
• 1-7d — short-term swing trades lasting one to seven days
• 1-4w — medium-term positions held for one to four weeks
• >1mo — long-duration positions exceeding one month

What the Distribution Reveals

A strategy with all trades in the 1-7d bucket behaves very differently from one split across 1-7d and 1-4w. The former is a consistent short-term system. The latter might be a short-term system that occasionally catches a larger move and holds through it — or a system without a clear time-based exit logic.

Duration distribution also connects to the risk profile. Longer-duration trades are exposed to more overnight and weekend risk, more news events, and more regime shifts within a single position. Short-duration trades avoid these risks but face higher transaction costs relative to their profit potential.

The screenshot shows 32 trades in the 1-7d bucket and 26 in 1-4w, with zero intraday trades and only 4 exceeding one month. This is a multi-day positioning strategy that occasionally extends into multi-week holds — matching the "Multi-day positioning" style classification shown in the exposure profile below.

Below the duration chart, the Exposure Profile section provides four categorical assessments. Each applies a set of fixed thresholds to the calculated exposure data.

Participation Level

Label	Threshold
Low	Below 25% time in market
Moderate	25% to 49%
High	50% to 74%
Sustained	75% or above

The example shows Low (9%), confirming the strategy is flat for the vast majority of the backtest.

Style

Based on median trade duration:

Label	Median Duration
Short-term positioning	Less than 1 day
Multi-day positioning	1 to 7 days
Swing positioning	1 to 4 weeks
Long-term positioning	Over 1 month

With a 3.0-day median hold, this strategy falls into Multi-day positioning.

Gap Structure

Based on the longest gap between consecutive trades:

Label	Longest Gap
Continuous engagement	No gaps, or longest gap under 1 day
Brief idle periods	Longest gap under 1 week
Extended idle periods observed	Longest gap under 1 month
Prolonged market absence observed	Longest gap 1 month or more

A 5.4-month longest gap triggers "Prolonged market absence observed" — a clear signal that the strategy can go dormant for extended stretches.

Operational Load

Based on trades per month:

Label	Trades / Month
Very low trade frequency	Fewer than 2
Low trade frequency	2 to 4
Moderate trade frequency	5 to 14
High trade frequency	15 or more

These four classifications work together. A strategy with High participation, Short-term positioning, Continuous engagement, and High frequency is a scalping-type system. One with Low participation, Multi-day positioning, Prolonged absence, and Very low frequency is a selective position trader. Both are valid approaches — but the risk management, capital allocation, and psychological demands are entirely different.

The bottom-right panel synthesises the exposure data into two narrative blocks: Exposure Pattern and Implications. These use the participation percentage to select from three interpretation templates.

Exposure Pattern

For strategies with over 70% participation: "Market participation is high, with capital frequently committed. The strategy maintains near-continuous market presence." This means you are almost always exposed — drawdowns will correlate closely with market declines.

For 40-70% participation: "Market participation is intermittent but sustained once positions are established. Exposure occurs in blocks rather than continuously." A pattern of engagement and disengagement — the strategy has an opinion about when to be in the market.

Below 40%: "Market participation is selective, with extended periods of inactivity. Capital is committed infrequently." The strategy waits. It may spend months flat. When it enters, the conditions had to be specific.

Implications

The implications section translates the pattern into forward-looking expectations:

• High exposure — "Returns will correlate closely with overall market movements. Drawdowns may align with broad market declines." The strategy doesn't protect during downturns because it doesn't exit.
• Medium exposure — "Strategy behavior aligns with position-holding rather than continuous engagement. Periods of inactivity are expected." Performance will come in bursts, with dead periods in between.
• Low exposure — "Strategy is highly selective. Performance depends on timing of specific entry conditions rather than market direction." Returns are decoupled from the market's overall trajectory — which is both the strength and the vulnerability of the approach.

Exposure data becomes most powerful when read alongside other dashboard cards.

Exposure + Performance Metrics

The Performance Metrics card shows total return, but return alone doesn't account for how long capital was at risk. A strategy returning 40% while in the market 9% of the time is dramatically more capital-efficient than one returning 40% while exposed 80% of the time. The exposure data lets you normalise returns by actual time deployed.

Exposure + Drawdown

Low-exposure strategies tend to have shorter drawdown durations because they spend most of their time flat. But when drawdowns do occur, they can feel disproportionately large because the strategy's infrequent trades carry more weight in the equity curve. Cross-referencing exposure with the drawdown envelope reveals whether drawdowns happen during active periods or bridge the gaps between trades.

Exposure + Regimes

The Regimes card shows which market conditions (trending, ranging, volatile) the strategy traded in. If the exposure timeline shows clustering in one part of the backtest, the regimes card will reveal whether that cluster coincides with a specific regime. A strategy that only activates during trending markets has regime-dependent exposure — which is fine as long as you understand and expect it.

Exposure + Behaviour

The Behaviour card's Trades/Month metric should align with the Exposure card's Load classification. If the behaviour card shows 0.6 trades per month and the exposure card classifies load as "Very low trade frequency," the diagnostics are consistent. Inconsistencies between cards would indicate a calculation discrepancy worth investigating.