Risk vs Loss: Volatility, Downside Risk, and Drawdown
Learn how risk and loss differ, and how volatility, downside risk, and drawdown each measure a different kind of danger.
Category: Start Here
Content updated: . The tail-risk explanation now distinguishes negative return thresholds from loss magnitudes. This is not independent model validation.
In plain English
"Risk" and "loss" sound like the same thing, but they are not. Loss is money you actually gave up. Risk is the chance that things move against you before you find out. Picture a road trip: "risk" is how bumpy and unpredictable the road feels along the way, while a "loss" is only the moment you end up lower than where you started. This article explains three common ways to measure that bumpiness and those dips.
What it measures
Three different numbers capture three different worries. Volatility (measured as the standard deviation of returns) tells you how much returns bounce around an average, counting both up moves and down moves. Downside risk looks only at the bad moves and ignores the good ones. Drawdown measures the drop from a recent high point (peak) down to a later low point (trough), so it answers "how deep was the worst dip?" - Volatility: a stock returns +10%, -10%, +10%, -10%. It bounces a lot, so volatility is high, even though nothing crashed. - Downside risk: from those four moves, only the two -10% drops count. - Drawdown: if a price climbs to 100 then falls to 70, the drawdown is (100 - 70) / 100 = 30%. - Key point: a calm, slow grind downward can have LOW volatility but a LARGE drawdown.
How SurgeFlow uses it
These ideas underpin the risk metrics you see across the site, especially on the Backtesting tab. SurgeFlow reports volatility to show how jumpy a strategy's returns are, and maximum drawdown (often called "max drawdown") to show the worst peak-to-trough fall over the test period. It also reports tail-loss measures: Value at Risk (VaR), a tail threshold with a stated sign convention, and Expected Shortfall, the average return or loss beyond that threshold. A return threshold of -3% corresponds to a 3% loss magnitude. These historical estimates do not guarantee future loss frequencies. All of these are descriptive summaries of past results, not forecasts.
How to read it on the site
Open the Backtesting tab (/backtest/benchmark) and look at the performance and risk figures next to a strategy or benchmark. Read volatility as "how shaky the ride was," max drawdown as "the deepest hole it fell into," and VaR or Expected Shortfall as "how bad a bad day tends to look." A higher return next to a small drawdown is a different story than the same return next to a huge drawdown. - High volatility, small drawdown: choppy but never fell far. - Low volatility, large drawdown: smooth-looking but slid into a deep, lasting hole. - Compare two strategies on the SAME period, or the comparison is not fair.
What it cannot tell you
These numbers describe what already happened; they do not predict what comes next, and past performance is not a forecast. Volatility treats a +10% surprise and a -10% surprise as equally "risky," which is why downside-only measures exist. VaR and Expected Shortfall depend on assumptions and on the history you feed them, so a calm past can badly understate a future shock that the data never saw. None of these are buy or sell signals, and SurgeFlow does not give investment advice. - A low drawdown so far does not mean a bigger one cannot happen tomorrow. - A statistic looking precise does not make it accurate; it inherits the limits of its data. - Two assets with the same volatility can lose money in very different ways.