RESOURCES / BACKTESTING FOUNDATION
Portfolio Backtest Metrics Explained
No single metric is a portfolio conclusion. Return must be read together with its path and consistency.
Questions about portfolio metrics.
Why should return never be read as a standalone portfolio conclusion?
Ending value and annualized return describe where the portfolio finished, but not what an investor had to endure to get there. Two portfolios can share a return while differing sharply in volatility, maximum drawdown, recovery time, and consistency across start dates. PortfolioBacktest places these measures together so a higher result can be read alongside its risk exchange. The practical question is whether the additional return was efficient enough, understandable enough, and holdable enough for the objective—not whether one number is larger in isolation.
What is the difference between maximum drawdown and volatility?
Maximum drawdown is the largest peak-to-trough decline before a new high; it describes a particularly painful historical episode. Volatility summarizes the dispersion of periodic returns across the whole window, so it captures frequent movement even when losses recover quickly. A portfolio can have lower volatility but a severe drawdown, or a larger day-to-day range without a long-lasting loss. The report also shows recovery from the worst drawdown and the distance below the prior high, which turns an abstract metric into a more realistic account of the path.
How should I use relative performance and alpha language?
Relative performance compares the portfolio with a matched SPY reference over the same dates. The gray SPY-matched line is the zero line: above it means the portfolio outperformed, below it means it underperformed. The relative result states the ending difference, while the curve shows when that difference opened, narrowed, or reversed. ‘Alpha’ in this context should be read as historical excess performance versus the chosen reference, not as proof of skill or a forecast. The risk table explains the cost of that excess return.
What does recovery time tell me that drawdown alone does not?
Drawdown tells you the depth of a loss; recovery time tells you how long the investor remained below the previous peak. A shallow loss that lasts years can be harder to plan around than a deeper loss that recovers quickly. PortfolioBacktest reports the peak, trough, recovery date when available, and whether the portfolio or reference has recovered. Reading these dates beside the environment replay helps identify whether the damage was a short shock, a prolonged regime, or a sequence of smaller setbacks.
How does result consistency change the meaning of a metric?
Consistency asks whether the direction and approximate size of a finding survive reasonable changes in start date, market regime, or nearby parameter choices. A portfolio can have an impressive all-period return but lose its advantage in most shorter windows. Conversely, a lower-return allocation may repeatedly reduce drawdown even when its ending-value rank changes. Treat consistency as a confidence qualifier, not as a guarantee. It tells you how conditional the historical statement is and which follow-up test is worth running before making a decision.