Nothing dramatic happened. No gap through a weekend, no stop blown apart by a news print, nothing you could point at afterwards and name. The account just leaked for eleven weeks, and one Tuesday you closed the platform and didn’t open it again.
Your volatility tools flagged none of it, because none of it was volatile. It was just long.
By the end of this you will be able to measure the half of a drawdown that most risk tools throw away: not how far the account fell, but how long it stayed down.
Key Findings
- What it measures: the ulcer index combines the depth of a fall below the running high with the length of time spent below it, in a single number.
- Why ATR misses it: average true range and standard deviation are direction-blind, so a strong up day and an equally sized down day contribute the same amount.
- Where it came from: Peter Martin and Byron McCann introduced the measure in their 1989 book The Investor's Guide to Fidelity Funds, to rank funds by how hard they were to hold.
- How to read it: there is no universal threshold; the value is only interpretable against another value from the same instrument, timeframe and lookback.
What does the ulcer index actually measure?
Distance below your own high water mark, weighted by how long you stay there. That is the whole idea. Every bar, the calculation asks one question: how far below the best level reached so far are we right now? A run of answers near zero produces a low reading. A run of answers deep in the red produces a high one.
Martin and McCann built it for fund investors who kept abandoning perfectly reasonable funds partway through a bad stretch. Their argument was that the standard risk numbers described the wrong experience. Investors don’t quit because a fund is jumpy. They quit because they have been underwater for months and the bottom keeps not arriving.
That argument transfers to a trading account almost unchanged.
Why doesn’t ATR tell you this?
Because ATR doesn’t know which way price went. It measures the size of a typical bar’s range, and a violent rally registers exactly the same as a violent collapse. Standard deviation has the same blind spot: an outsized winning week widens the spread of returns and gets scored as risk.
Maximum drawdown gets the direction right and the duration wrong. It reports one number, the worst peak-to-trough fall, and then stops. Two strategies can post an identical worst-case fall while one recovers in a fortnight and the other grinds sideways for half a year. Max drawdown calls those equal. Your patience does not.
| Metric | What it measures | A sharp rally counts as | Sees time spent underwater |
|---|---|---|---|
| Average true range | Typical bar range over a lookback | Risk, same as a selloff | No |
| Standard deviation | Spread of returns around their mean | Risk, same as a selloff | No |
| Maximum drawdown | The single worst peak-to-trough fall | Nothing | No |
| Ulcer index | Depth of falls below the running high, over time | Nothing | Yes |
How is it calculated, without the maths?
Three steps, and you can follow all of them in your head.
First, track the highest close reached so far within the lookback window. Second, for every bar, record how far below that high the current close sits, as a percentage. Third, square those percentages, average them, and take the square root.
The squaring step is the part that matters. Squaring makes a deep retracement count for far more than a shallow one, so a single ugly plunge dominates a month of minor dips. A long queue of small negatives still adds up though, which is exactly why a slow bleed registers here and nowhere else.
What counts as a high reading?
Nothing, in isolation. The value scales with the instrument, the timeframe and the lookback you picked, so a figure lifted from a monthly fund chart tells you nothing about your 15-minute account. Anyone quoting a fixed cutoff has imported it from a context that probably isn’t yours.
Use it as a comparison tool instead. Same data, same window, two candidates, lower wins.
| Where you point it | What the reading answers |
|---|---|
| Equity curve of a live account | How hard your own record has been to sit through |
| Equity curve of a backtest | Whether the losing runs are survivable before you fund it |
| Price chart of an instrument | How punishing that market has been to hold long |
| Two strategies side by side | Which one earns its return with less time below the high |
RelicusRoad Pro
Have you been trading for a while but have never made consistent profits or are you new to FOREX trading and want to get a head start? Try RelicusRoad and you'll never look back.
Get RelicusRoad ProCan you run it on price instead of an equity curve?
Yes, and it answers a different question there. On a price chart it describes how uncomfortable an instrument has been to hold from the long side, which is useful when you are choosing between markets rather than between systems.
Be honest about the asymmetry, though. The measure only counts distance below a running high, so it is a long-side metric by construction. Short sellers get no information from it unless they invert the series first. That is a real limitation, not a footnote, and it is the main reason I keep it on the equity curve rather than the chart.
Where does the ulcer index let you down?
It is backward-looking. It tells you what a stretch of history felt like; it forecasts nothing.
It is also sensitive to the lookback. Shorten the window and recent calm dominates; lengthen it and an old crisis keeps scoring against you long after conditions changed. Pick a length that matches your holding period and then leave it alone, because quietly re-tuning it until the number looks acceptable is self-deception with extra steps.
The bigger failure is upstream. Feed it an equity curve produced by signals that shifted position after the bar closed and the reading will be beautiful, because the drawdowns you actually lived through were edited out of the record. That is the same mechanism behind backtests that fake their own results , and no downstream metric can detect it. Verify the signals first.
How does this change what you put on the chart?
It changes what you demand from a tool before you trust its history.
If the arrow you traded can still move to a better location after the candle finishes, then the equity curve you are grading was never available to you. Every risk number computed from it, ulcer index included, is measuring a version of the account that didn’t exist. RelicusRoad Pro is built around that constraint: a signal prints, and it stays printed. What you review at the end of the month is the record you could have executed.
That won’t tighten a drawdown on its own. An indicator sharpens timing; it doesn’t fix position sizing, and no tool removes losing runs from trading. What a fixed signal does give you is an honest curve to measure, which is the precondition for every risk decision that follows, including how much size a system with a long recovery profile deserves. Pair it with proper risk control and the ulcer index stops being trivia and starts being a filter.
Frequently asked questions
What is a good ulcer index value? There isn’t a universal one, and any article quoting a fixed threshold is borrowing it from a context that may not be yours. The reading scales with the instrument, the timeframe and the lookback length you chose, so a number from a monthly fund chart means nothing next to a number from your 15-minute equity curve. Use it comparatively. Run it across two strategies on the same data with the same settings, and the lower reading is the one that spent less time hurting.
Is the ulcer index better than standard deviation for measuring risk? It’s better at measuring one specific thing: downside pain. Standard deviation treats an unusually good week as risk, which is why a strategy that occasionally rips higher can look dangerous on paper while being perfectly comfortable to hold. The ulcer index ignores gains entirely and only counts distance below your running high. That makes it sharper for judging whether you could sit through a losing stretch, and poorer if you genuinely need a symmetric measure of dispersion.
Does the ulcer index repaint? The calculation itself doesn’t. It reads closed data and produces a value that stays where it lands. The risk sits upstream: if you feed it an equity curve produced by entry signals that shifted after the candle finished, the curve you are measuring is not the curve you could have traded, and the reading will flatter you.
Can I use the ulcer index on MT4, MT5 or TradingView? Yes, in one form or another. It is not a default indicator on the MetaTrader platforms, so you are usually installing a community version or reading it out of a strategy report. TradingView carries community scripts for it. Confirm the lookback the author chose before comparing its output against anything else.
What is the Martin ratio? It is the return-to-pain version of the same idea, sometimes called the ulcer performance index. Instead of dividing excess return by volatility the way a Sharpe ratio does, it divides excess return by the ulcer index, so upside moves stop being penalised and only time underwater counts against the score.
Grade your own equity curve on time underwater rather than worst-case depth, and if you want signals that stay where they printed while you do it, start with RelicusRoad Pro .