Rising and Falling Window — The Gap as Support and Resistance

Most candlestick patterns promise to predict the future — and most, as this series keeps showing, do it at coin-flip odds. The window is different. It's the only "pattern" that leaves something permanent on the chart beyond a forecast: a hole, a concrete price range where nobody traded with anybody. The Japanese called the gap a window and treated it not just as a continuation signal but, above all, as a level: a rising window is supposed to defend price from below, a falling window to block it from above. Bulkowski's measurement gives this tradition partial support: continuation works clearly above coin-flip odds (75% and 67%), but the support/resistance role is far more modest than the legend claims. Let's go through it.

What the Pattern Looks Like

The window is the simplest construction in the entire catalog — two candles and a no-overlap condition:

Rising Window:

• Trend before the pattern: up.
• Condition: yesterday's candle high (including its shadow) sits below today's candle low. An empty space is left between the candles — the window.

Falling Window:

• Trend before the pattern: down.
• Condition: yesterday's candle low sits above today's high.

That's it. No requirements on color, body size, or sequence. The gap comes from simple mechanics: between sessions (or in a thin market) new information or orders arrived that jumped price in one move — an earnings report, a dividend, a macro headline. The market "leapfrogged" a certain price range without trading in it for even a moment.

Japanese tradition adds two rules on top of this. First: the window supports the trend — as long as it stays open, the side that created it controls the market. Second: the interior of the window is a support zone (rising window) or resistance zone (falling window) — price returning to the window should be stopped there, and a full close of the window ("closing the window") means the signal has expired.

A crypto caveat that matters more than usual in this article: 24/7 spot markets don't have classic gaps — every BTC/USDT candle opens at the close of the previous one. You'll see gaps on CME bitcoin futures (the weekend break) and on session-based markets. On spot, the role of "leapfrogged prices" is played by violent full-bodied candles and by the daily/weekly opening gaps tracked under the NDOG/NWOG convention, covered in the article on NDOG and NWOG. For the full typology of gaps (common, breakaway, measuring, exhaustion), see the article on types of price gaps.

What the Numbers Say

Statistics from Bulkowski's tests (~4.7 million daily candles, US stocks — standard caveat: this is a different market and volatility regime than crypto, the numbers don't carry over 1:1):

Rising Window:

• Continuation: 75% of cases. For a catalog where most patterns hover around 50-60%, that's a very good result — and a rare case where theory and measurement agree.
• Frequency: 20/103. Windows happen often — you'll find them on charts without much trouble, though the higher the timeframe, the rarer they get.
• Performance rank: 42/103. An honest note from the measurement's author: this rank mostly measures the trend around the window, not the window itself — the gap is more a symptom of a strong trend than its cause.
• Support in the gap: 20% of cases. Only one time in five did price returning to the window form a local low inside it before closing the gap. This is the most sobering number in this article.
• Time to close: median 11 days, mean 79 days. Half of all windows close within two weeks; the tail of the distribution stretches out for months.

Falling Window:

• Continuation: 67% of cases. Weaker than the rising variant, but still clearly above coin-flip odds.
• Frequency: 23/103 — also common.
• Performance rank: 7/103. Surprisingly high — a downtrend with gaps can move very fast. Again, though: credit goes to the trend, not to any magic in the window.
• Resistance in the gap: 25% of cases. Three times out of four, price returning to the window passes right through it without forming a local high.
• Time to close: median 9 days, mean 55 days.

The conclusion from this table: the window is a good continuation pattern (75%/67% is concrete) and a mediocre level (support/resistance works in 20-25% of cases before the gap closes). Japanese tradition was right about the trend and overstated the wall.

How to Trade It (and How Not To)

How NOT to trade it: fading the gap automatically. The play "the gap will fill, so I'm going against it" ignores both key numbers: 75%/67% continuation and an average time-to-close measured in weeks. Even if the gap eventually closes, your stop can take you out long before that.

Scenario 1 — the window as a trend filter. The simplest use consistent with the measurement: a fresh rising window = the market is in continuation mode, look for longs, don't try to catch a top. A fresh falling window = the mirror image on the supply side (performance rank 7/103 is a reminder that declines with gaps can move fast). The window isn't an entry signal by itself — it's information about which side to stand on.

Scenario 2 — a return to the window as a decision point, not a guarantee. Price returns to a rising window: that's a sensible place to watch for a reaction (a bounce = the trend defending its structure), but the 20% figure forbids treating the gap like concrete. You trade the reaction if it shows up — a rejection candle, demand taking back the initiative — not the mere fact of touching the gap. The stop logically sits on the other side of the window: a full close invalidates the structure.

Scenario 3 — identify the type of gap before you trust it. Not all windows are equal: a breakaway gap (out of consolidation, on volume) has different consequences than an exhaustion gap (at the end of a long move) or an ordinary gap closed after two sessions. The window's definition alone doesn't distinguish between them — context does. We cover the typology and how to tell them apart in types of price gaps.

Stop loss and target. For trading continuation after a rising window: stop below the lower edge of the window (its closing is an exit signal in its own right), target at the nearest resistance. Mirror this for a falling window. Keep size moderate — 75% continuation sounds good, but one window in four fails right away.

Myth vs Measurement

Myth: "Every gap must get closed." Measurement: the median time to close is 9-11 days, but the mean is 55-79 days — the distribution has a long tail of gaps that stay open for months. "Always" in this slogan really means "eventually, possibly long after your position against the trend has already blown up."

Myth: "The window is strong support/resistance." Measurement: support worked in 20% of cases, resistance in 25%. The window is a place where a reaction sometimes happens — not a place where a reaction is guaranteed. Without confirmation from a reaction candle, it's just a rectangle on the chart.

Myth: "The gap gives you an edge because it shows strength." Measurement: 75%/67% continuation is a real, catalog-rare piece of concrete evidence — but the performance rank mostly measures the trend around the gap. The window confirms the strength of the trend; it doesn't create it. Buying "because there's a gap" without looking at the trend structure confuses symptom with cause.

Myth: "I trade windows on BTC just like on stocks." Definition: without a break in trading, there's no classic window. On 24/7 spot, work with the equivalents — CME opening gaps and the NDOG/NWOG convention — keeping in mind that the statistics in this article don't cover them.

Example Scenario

A stock after a lengthy consolidation: a quarterly report beats forecasts, the open gaps up, the day's low doesn't touch yesterday's high — a rising window that, given the context, looks like a breakaway gap. The measurement says: 75% odds of continuation, uncertain support in the gap (20%), half of these windows close within about two weeks. The plan: you don't chase the first candle. You wait for a pullback toward the upper edge of the window and a reaction from demand — a rejection candle, the initiative changing hands. The reaction shows up: go long, stop below the lower edge of the window, target at the nearest resistance, take partial profit. Instead of reacting, price closes the window and settles below it: no position, no debate — the structure has expired, and you've saved yourself a ride on the word "always."

Quick checklist:

• Is there a real hole between the candles' shadows (not just between the bodies)?
• Does the window align with the trend it appeared in?
• Do you know what type of gap this is (breakaway / measuring / exhaustion / common)?
• Are you trading continuation or a reaction in the window — not fading it "because it'll fill"?
• Are you waiting for a reaction candle when price returns to the window (keeping the 20-25% figure in mind)?
• Is your stop beyond the opposite edge of the window — with a close meaning exit?
• Is this a market with real gaps (stocks, CME), not 24/7 spot?

The window is one of the rare patterns that comes out of the measurement with a shield on: 75% and 67% continuation are numbers most of the catalog would envy. But even here the legend added more than was ever measured — the "window as a wall" turned into support that works once in five cases, and "every gap gets filled" turned into a median that hides a long tail. Trade what's been counted: trend behind the window — yes; concrete inside the window — only after confirmation.

FAQ

Rafał — Krypto Bez Ściemy

Trader and founder of Strefa Tradingu. He’s been breaking crypto down on YouTube for years — no hype, no signals, with a focus on market structure and risk management.

kryptobezsciemy.plInstagramYouTubeKlub Zero