Volatility indicators in a trader's toolkit — a practical survey

One trader asks for the best volatility indicator; another wonders why the same stop loss works one week and fails the next. These are two sides of the same coin. EUR/USD can sit in a forty-pip range one week and travel one hundred and fifty the next — any honest risk plan has to bend to that. This is a survey of the tools I actually use — ATR, Bollinger Bands, Keltner Channels, historical and implied volatility — and how to combine them into one read on the regime.

Why bother measuring volatility at all?

So that position size and stop loss stay consistent across regimes. A fifty-pip fixed stop is generous when ATR drops to thirty, and absurdly tight around an ECB meeting with ATR at ninety. Beyond that, volatility narrows the strategy set: low volatility rewards mean reversion and range trading; high volatility rewards breakouts and demands caution on leverage. The broader treatment is in market-regime trading.

One caveat to repeat: no volatility indicator forecasts the future. ATR and standard deviation measure the past. Bollinger Bands and Keltner Channels describe dispersion as of now. Implied volatility from currency options is a consensus on the size of the move over thirty days, not its direction. These tools describe the regime; the decision is still yours.

ATR — the workhorse of the toolkit

The Average True Range was introduced by J. Welles Wilder in 1978 in New Concepts in Technical Trading Systems. The idea: instead of taking only today's high-low, account for any gap between yesterday's close and today's open. A fourteen-period ATR is the smoothed version of that and tells you, in pips, how much an average session has covered over the last two weeks. The full breakdown lives in ATR as a measure of volatility.

In practice ATR handles three jobs. Stop distance: the standard multiplier for swing and position trading is 1.5–2 times current ATR. Position size: when ATR rises, the lot count shrinks so monetary risk stays the same fraction of the account. Trailing stop: roughly two ATRs lets a trend run without throwing you out on noise. A scalper shortens the period to five or seven; a swing trader on D1 may stretch it to twenty-one or twenty-eight. For the formula itself, see the basics of ATR and true range.

Bollinger Bands and Keltner Channels — two answers to the same question

Bollinger Bands were developed by John Bollinger in the early 1980s: a twenty-period simple moving average with side bands two standard deviations above and below it. Under a roughly normal distribution the envelope contains about ninety-five percent of observations — a move outside the bands is, by construction, an event that happens about once every twenty candles.

Keltner Channels are their younger cousin from a different mathematical family. Instead of the standard deviation of price they use a multiple of ATR (1.5 or 2); the channel reacts more slowly to single outsized candles and reflects persistent volatility more honestly. Many traders display both at once and watch for the case in which Bollinger Bands sit fully inside the Keltner Channel — the TTM Squeeze, a marker of extreme volatility compression. After such a squeeze the market almost always produces a sharp breakout, though direction is not guaranteed.

The three Bollinger configurations match three regimes. Narrow bands during a squeeze — wait for the breakout. Wide parallel bands with price walking one of them — look for trend continuation. Stable width with price around the middle — range, play mean reversion. Treating bands as hard support and resistance during a strong trend produces nothing but small losses.

Historical and implied volatility — two different answers

To compare instruments or to see where the current market sits in its own history, on-chart indicators are not enough. That is when you reach for historical volatility in its raw form: the standard deviation of logarithmic returns over a fixed window (usually twenty or thirty sessions), annualised by the square root of the sessions in a year. A single number — annualised HV for EUR/USD at 7.8 percent, say — tells you whether the market is calm, normal or hot, without looking at a chart.

On the other side stands implied volatility from currency options. It does not look at the past; it captures what option markets expect the size of the move to be over the coming thirty days. For US equities that index is the VIX — but the VIX is an equity index, not a currency one. The FX equivalents are CVIX (Deutsche Bank's DB Currency Volatility Index, from a basket of G7 options) and JPVIX (implied volatility of USD/JPY options from the Osaka Exchange). Those are the proper FX compass. VIX still works as a broad risk backdrop — above twenty-five, risk currencies (AUD, NZD) typically weaken and safe havens (JPY, CHF) strengthen; AUD/JPY-to-VIX correlation sits between minus 0.65 and minus 0.80. For the wider intermarket context see intermarket analysis on ForexMechanics.

"Bollinger Bands answer a question: are prices high or low on a relative basis? In the past volatility was thought to be static; today we know that volatility is a dynamic quantity, indeed very dynamic." — John Bollinger, Bollinger on Bollinger Bands, McGraw-Hill, 2001.

How to fold these tools into one regime read

Volatility indicators only become useful in combination, each with a single job. Start with the macro backdrop — a glance at CVIX or JPVIX (VIX for completeness). High implied volatility means the market is pricing larger moves, so I reduce size and avoid mean-reversion plays. Low implied volatility invites range trading, with the quiet warning that calm markets are often a prelude to a breakout.

Second step, the local regime: Bollinger Bands and a Keltner Channel on the same timeframe. Bands inside the channel — wait, a squeeze rarely lasts. Bands flared with price walking one of them — look for trend continuation. Stable width with price around the middle — play mean reversion with a tight stop beyond the opposite band. Third step, ATR: sets the stop at 1.5–2× the current reading and drives position size so monetary risk stays the same fraction of the account, typically one percent.

An illustrative example: EUR/USD on H4. CVIX moderate, ATR thirty-five pips, bands wide, price tags the upper band after an impulse. A 1.5-ATR stop is fifty-two pips; a target at reward-to-risk 1:1.5 puts take-profit seventy-eight pips above entry. On a €10,000 account with one-percent risk (€100) and a pip worth about €1.92, the position works out to roughly 0.19 of a standard lot. Illustrative, not backtested.

What to do tomorrow

Volatility theory only pays rent once you configure the tools on your own platform. The five steps below take an afternoon; from the next morning you trade with a discipline most retail traders never reach.

1. Add ATR(14) to your two main daily charts and study the last sixty sessions. Write the minimum, maximum and median values into the journal — those three numbers are your volatility calibration, and every future trade is opened only after comparing the current ATR against that range.
2. Overlay Bollinger Bands (20, 2) and a Keltner Channel (20, ATR 2) on the same chart. Scroll back and mark every instance where the bands sat fully inside the channel. Track what happened over the next five to ten sessions — that one exercise shows how often a squeeze really precedes a breakout on your instrument.
3. Replace the daily VIX check with CVIX or JPVIX. Pick the FX volatility index that fits your pairs and add it to the morning routine. Note in the journal where the current level sits relative to the last twelve months — that is your risk backdrop under the local chart picture.
4. Compute annualised historical volatility for EUR/USD by hand. In a spreadsheet pull thirty closes, calculate logarithmic returns (ln of today over yesterday), take the standard deviation and multiply by the square root of 252. One percentage figure, comparable across instruments — it teaches you to discuss volatility in numbers.
5. Write two risk rules and stick them above your monitor. First: every stop loss comes from the current ATR multiplied by 1.5 to 2, never from a fixed pip count. Second: position size is calculated from a one-percent risk budget and the stop distance, never from a pre-decided lot count. Those two rules separate disciplined traders from the rest.