Why PnL Is Not a Single Number

The Illusion of a Final Balance

Profit and loss is often presented as a definitive number, sitting neatly at the top of a dashboard. Traders glance at it, interpret it as truth, and move on. But this figure is a projection, not a verdict. It reflects a set of assumptions about valuation, timing, cost, and risk that vary depending on the system, platform, and even the moment of observation.

A portfolio’s reported PnL can shift dramatically without any trade being executed. Change the reference price. Switch from mid-price to bid. Revalue using the index instead of the mark. You’re holding the same assets, but the number moves.

Some platforms display multiple versions side by side: realized, unrealized, mark-to-market, projected, net of fees. Others collapse it into a single figure and hide the methodology. The difference isn't cosmetic. It changes how traders interpret risk and performance.

Realized vs Unrealized vs Allocated

Realized PnL refers to positions that have been closed. The system calculates entry and exit, deducts fees, and posts the result. But even here, definitions vary. Some systems treat partial closes differently, breaking a single position into fragments. Others pool all fills and apply weighted averages.

Unrealized PnL reflects positions that remain open. This is where ambiguity expands. The valuation source matters: last trade, mark price, index, or a liquidity-adjusted estimate. Each choice has downstream effects on margin, liquidation triggers, and internal risk models.

Allocated PnL arises in multi-account structures or pooled portfolios. When a desk runs a strategy across several clients or systems, it needs to allocate profit and loss — not always proportionally. Some flows are easier to attribute than others. Some hedges benefit one account more than another.

Valuation Mechanics Inside the Platform

Different exchanges apply different pricing references. A position on a perpetual futures market might use a mark price fed by an index. A spot position might reference last trade. Some platforms average over time or apply a dampening function to reduce volatility in PnL readouts.

This creates tension between visible PnL and true economic outcome. A trader may see a negative mark-to-market but still exit in profit. Or see positive numbers vanish at settlement. Systems enforce discipline based on calculated risk, not on how the chart looks.

Platforms with internal matching engines often use synthetic reference prices. When there's no external anchor — such as in low-liquidity altcoin markets — the exchange builds its own pricing logic. This introduces noise into unrealized PnL. That noise can cascade into user actions, especially when paired with automated strategy triggers.

Implications for Strategy and Reporting

When backtesting or simulating, PnL metrics tend to simplify. Execution costs are estimated, spreads ignored, slippage smoothed. But in live systems, these details are volatile. A model that appears profitable on paper might perform poorly in production, simply because its definition of PnL lacks granularity.

Auditing becomes harder when PnL lacks traceability. Traders may try to reconcile performance with historical fills, only to discover mismatches in how the platform grouped, netted, or timestamped executions. This complicates compliance, client reporting, and risk reviews.

Systems that treat PnL as a fluid set of records — not a static number — are better suited for automation. They allow downstream systems to revalue, recompute, and adjust logic without rewriting execution history.