SPY's Donchian Ceiling Fell in October. Price Never Did.

Sometime in mid-October, on an hourly SPY chart, the amber upper boundary of a Donchian Channel held flat above the candles — then stepped lower. The candles hadn't dropped by the same amount. So why did the line move?

The answer, laid out by the analyst Investrade in a TradingView chart note on the State Street SPDR S&P 500 ETF (BATS: SPY), is mechanical: the step-down reflects a change in the highest high remaining inside the indicator's lookback window, not a fresh decline in price.

What does a 20-bar Donchian Channel actually measure?

The chart runs Donchian Channels with length 20 and offset 0. The upper boundary marks the highest high across the latest 20 chart bars. The lower boundary marks the lowest low across the same window. The midpoint is hidden so the two extremes stay visible.

One nuance Investrade flags: twenty hourly bars are not twenty trading days. The calculation follows the bars that exist under the chart's session settings — count what's on the screen, not calendar days.

Why does the upper line step down while price sits below it?

The mechanics come down to a rolling window. As each new bar enters, the oldest bar leaves. The upper line stays flat while the same highest high remains in the window — unless a newer bar exceeds it. Once that extreme expires out of the 20-bar lookback, a lower high can become the highest of the remaining bars, and the line drops.

That's exactly what happened in mid-October. "That move is a change in the highest high remaining in the lookback, rather than a candle falling by the same amount," Investrade wrote. Follow the flat section into the middle of the month and the downward steps that follow, and the picture is clear: the rolling reference is catching up with the price history it now contains.

In plain words: the ceiling didn't break. It simply aged out of old data.

What about the lower boundary?

The same logic runs in both directions, and the lower line deserves separate reading. A fresh lookback low can push the lower boundary down. An old low expiring out of the window can move it up — even without any equivalent upward jump in price.

The practical consequence: channel width responds to two forces at once. New extremes entering the window widen or shift it; old extremes leaving it do the same in the opposite direction. Reading only the candles misses half the equation.

Does offset 0 change anything?

Yes, and it matters for timing. With offset 0, the current, still-forming bar is included in the calculation. A new lookback high or low can therefore move the relevant boundary on that very bar, in real time.

That immediacy cuts both ways. "A touch alone is not an instruction to buy or sell," Investrade cautions. The boundary responds to data, not to intent.

How can traders verify this themselves?

Investrade offers a concrete exercise. Choose a completed bar where the upper line steps down. Count back through its 20-bar window and compare it with the preceding window. Two questions do the work: which high left the window, and which high replaced it as the new maximum?

Checking that single swap makes the step-down easier to interpret — and harder to mistake for a price signal that was never there.

The analyst's note is educational material; the securities shown are illustrations, not recommendations, and past performance doesn't guarantee future results.

The takeaway for SPY watchers: before treating a Donchian boundary move as a trend call, check whether the window simply rotated.