What is a worked example of Strategy Review?

What it means: Strategy Review in plain terms

A Strategy Review is a structured look back at how a trading plan performed, compared with the plan’s rules and assumptions. In an informational context, it focuses on understanding why outcomes did or did not match expectations, rather than predicting future results.

A worked example means you see a complete, numerical or step-by-step scenario where every input is stated and every calculation can be repeated. The goal is independent verification: a reader should be able to follow the same assumptions and reproduce the same intermediate numbers.

How a worked example works (mechanics)

A worked example typically uses the following mechanics:

1. Define the “strategy” mechanics you are reviewing (for example, position sizing rule, entry/exit rule, holding period assumption, and the decision workflow).
2. Separate stable mechanics from variable conditions. Stable mechanics are the rules that the strategy intends to follow. Variable conditions include market volatility, spreads/fees, slippage during entry/exit, and the environment in which orders were executed.
3. Set assumptions for every number you calculate. This includes costs, units, and whether you treat returns gross or net of costs.
4. Compute simple performance measures from recorded outcomes (for example, average return per trade, win rate, drawdown range) based only on the provided scenario data.
5. Test failure modes. A good review also notes at least one way the method can break (for example, the data does not represent real execution, or rules rely on hindsight).

Worked scenario example with explicit assumptions

Below is a fully specified example you can recalculate.

Strategy review setup (assumptions):

• You review 5 trades from historical logs.
• The strategy’s mechanics are simple:
  • Each trade risks a fixed amount, and exits after a fixed outcome is realized.
  • You record realized profit/loss per trade.
• Returns are measured in account currency.
• Costs are treated as a per-trade drag of $2 (spread + commission + typical execution cost). This $2 is assumed constant across all five trades.
• The raw, before-cost profit/loss for each trade (from the scenario record) is:
  1. +$15
  2. -$8
  3. +$7
  4. -$6
  5. +$10

Step 1: Convert raw P/L to net P/L Net P/L per trade = raw P/L − $2.

• Trade 1: $15 − $2 = +$13
• Trade 2: -$8 − $2 = -$10
• Trade 3: $7 − $2 = +$5
• Trade 4: -$6 − $2 = -$8
• Trade 5: $10 − $2 = +$8

Step 2: Compute summary metrics

• Total net P/L = 13 + (-10) + 5 + (-8) + 8 = +$8
• Win trades = 3 (trades with positive net P/L)
• Win rate = 3 / 5 = 60%
• Average net P/L per trade = total net P/L / 5 = 8 / 5 = +$1.60

Step 3: Separate mechanics from variable conditions In this scenario, the strategy mechanics (the plan for entering/exiting and the decision rules) are assumed unchanged across trades. What varies are outcomes and the environment captured in the net cost assumption ($2). If, in reality, spreads and slippage were larger during some trades, then the fixed $2 would be a poor approximation and the review’s conclusion would change.

Material limitations and risks (what can fail)

1. Assumption mismatch: If the assumed costs ($2 per trade) are not representative, net results can be overstated or understated. This is a measurement risk.
2. Limited sample size: Five trades provide an unstable estimate of any metric. Outcomes can look positive due to chance, even if the underlying mechanics do not reliably translate.
3. Execution realism: If historical records ignore slippage, partial fills, or order timing effects, the review may evaluate something closer to a backtest-like record than real execution.
4. Hindsight bias: A review can accidentally incorporate information that was not available at decision time. That turns a diagnostic process into a re-labelling exercise.

How to verify independently

To independently verify a worked example, recreate the calculations from the stated assumptions:

• Recompute net P/L from raw P/L and the given cost assumption.
• Recompute totals, win rate, and averages from the net P/L list.
• Check whether any change in assumptions (especially costs) materially alters the summary metrics.