ATR Indicator: NVDA's Median ATR Is 3.2 Times SPY's

ATR, or Average True Range, measures how far a price typically travels in a day. We express it as a share of the closing price, which we call ATR%. On 26 October 2026, the 14-day ATR% of SPY (the S&P 500 ETF) was 0.95%. For NVDA (Nvidia) it was 2.83%. For GLD (the gold ETF) it was 1.55%.

Over 5,504 trading days, the median ATR% was 3.2 times as large for NVDA as for SPY. We also tested ATR as a stop-loss distance on all three. Prices here count dividends.

The Average True Range is a measure of volatility. Volatility is how far a price moves from day to day. ATR counts the largest of three distances: the day's low to its high, or either end of the day's range to the previous close.

ATR has the same units as the price. An ATR of 5 on a stock at 100 means the stock typically travels about 5 in a day. We divide ATR by the closing price to get ATR%, so that a low-priced stock and a high-priced fund compare on one scale.

A day's true range is the largest of three numbers:

Take a day with a high of 105, a low of 100 and a previous close of 98. The three numbers are 5, 7 and 2, with each gap counted as a positive distance. The true range is 7, because the price gapped up from 98 and the day's range alone would miss that.

The 14-day ATR is built from these true ranges. The first value is the mean of the first 14 true ranges. After that, each day's ATR is 13 times the previous day's ATR plus the day's true range, with the total divided by 14. Each new day counts for one fourteenth, so a large day fades out slowly.

The gap term matters often. From 9 December 2004 to 26 October 2026, a gap widened the true range on 34.5% of SPY's trading days, 31.2% of NVDA's and 52.9% of GLD's.

The window runs from 9 December 2004, the first date GLD has a 14-day ATR, to 26 October 2026. That gives 5,504 trading days for each name.

Chart 1 shows the median ATR% over the window and the reading on 26 October 2026.

The median ATR% was 1.09% for SPY, 3.48% for NVDA and 1.20% for GLD. NVDA's median ATR% was 3.2 times SPY's.

On 26 October 2026, SPY's reading of 0.95% sat below the readings on 63.7% of days in the window. NVDA's 2.83% sat below 73.3% of days, while GLD's 1.55% sat below only 24.4%. Each reading means something only against the same name's own history.

Chart 2 shows SPY's ATR% from 19 February 1993, its first reading, to 26 October 2026. That is 8,479 trading days.

The highest reading in SPY's history since 1993 was 7.92% on 27 October 2008. The highest in 2020 was 6.57% on 23 March. The highest in 2025 was 3.87% on 10 April.

The lowest reading in the same history was 0.37% on 18 October 2017. SPY's median ATR% for the year 2017 was 0.60%, against 2.06% for 2008.

ATR averages 14 days, so one large day lifts it and then fades over the following weeks.

A stop loss is an order that sells an asset once its price falls to a set level. An ATR stop sets that level a multiple of the ATR below the entry price. A 2 ATR stop on a stock bought at 100, with an ATR of 3, sits at 94.

Across the entries described below, the median distance of a 2 ATR stop below the entry close was 2.18% for SPY, 6.98% for NVDA and 2.40% for GLD.

We tested this on every trading day from 9 December 2004 that had 20 more trading days in the file. That gives 5,484 entries for each name.

The stop level was the entry close minus a multiple of that day's ATR. A stop counted as hit if any of the next 20 closes fell to the level or below.

We compared this with a fixed stop 5% below the entry close. The only change between the two rules is how the distance is set. Chart 3 shows the result.

The fixed 5% stop was hit in 15.8% of SPY windows, 45.6% of NVDA windows and 16.6% of GLD windows. A 5% fall is large for SPY and small for NVDA.

The 2 ATR stop was hit in 37.8% of SPY windows, 35.0% of NVDA windows and 41.2% of GLD windows. The 2 ATR hit rates differed by about 6 points across the three names, against about 30 points for the fixed 5% stop.

Share of 20-trading-day windows with a close at or below the stop. The median distance of the stop below the entry close is in brackets. 5,484 entries per name from 9 December 2004. Source: YX Insights price data

A wider multiple was hit less often. The 1 ATR stop was hit in at least 56.5% of windows on every name, while the 3 ATR stop was hit in 20.9% to 26.1%.

The entries overlap, so the 5,484 windows are not independent. The test uses daily closes, charges no costs and never re-enters after a hit. A real stop order can trigger during the day and fill below its level.

We reran the test at 10 and 60 trading days. We also reran it with the day's low as the trigger instead of the close. That makes six settings, counting the main one.

In all six, the gap between the highest and lowest hit rate across the three names was smaller for every ATR multiple than for the fixed 5% stop.

The widest ATR gap was 7.2 points, for the 2 ATR stop at 60 days. The narrowest fixed gap was 25.7 points, at 10 days.

ATR measures the size of moves, not their direction. The reading is the same whichever way the price moves.

Our explainer on standard deviation in investing covers a second measure of volatility, built from closing prices only. Our guide to Bollinger Bands covers a band built on that measure, while what a drawdown is covers falls from a high.

For the wider idea, see our explainer on what volatility is.

ATR is the average size of a day's price range, including gaps. In our data the median was 3.2 times as large for NVDA as for SPY. Setting a stop distance from ATR evened out how often the stop was hit across the three names, while a fixed 5% stop did not.

The Academy is free to read, here on the web. The Long Trend, our free weekly letter, comes by email.

This explainer is part of the YX Insights Academy. Each one takes a single idea and checks it against real data.

The same approach runs through everything else we publish:

Good places to start on the website:

A good ATR value depends on the asset, because ATR scales with its price level and volatility. In our data the median 14-day ATR was 1.09% of the close for SPY, 3.48% for NVDA and 1.20% for GLD. A reading is high or low only against the same asset's own history, such as the 0.37% to 7.92% range SPY showed.

An ATR stop sits below the entry price by a chosen multiple of the ATR on the entry day. A 2 ATR stop sat a median 2.18% below the entry close for SPY, 6.98% for NVDA and 2.40% for GLD. Across the three, it was hit within 20 trading days in 35.0% to 41.2% of windows.

A 14-day ATR is a smoothed average of the last 14 days' true ranges. A true range is the largest of three distances: high to low, high to the previous close, or low to the previous close. The first value is the mean of 14 true ranges, then each new day counts for one fourteenth.

No. ATR averages the size of each day's price range, including gaps, in price units. Standard deviation measures how widely daily returns spread around their mean, using closing prices only. Both describe volatility, but they use different inputs. Our explainer on standard deviation in investing covers the second.

DISCLAIMER: This article is intended for educational purposes only. Any information or analysis in this note does not constitute an offer to sell or a solicitation of an offer to buy any securities. Nothing in this note is intended to be investment advice, nor should it be relied upon to make investment decisions. Any opinions, analyses, or probabilities expressed in this note are those of the author as of the note's date of publication and are subject to change without notice.