Bitcoin Prior-Day Push + Delta Divergence
Crypto like bitcoin trades every day, all day, so a session here is the Eastern calendar day, midnight to midnight. Over the last 20 sessions it took 6 trades, won 17%, net -4.01 R (hypothetical).
HypotheticalHypothetical results on past sessions: no fees, no slippage, fills at bar prices. Order flow is approximate: estimated from where each 1-minute bar closed in its range.
Rules
- BTC, 5-minute bars. Longs and shorts. Entries every day, 8 AM–4 PM ET.
- Mark yesterday's high, low and midpoint, midnight to midnight Eastern.
- Trigger: price trades below the prior-day low (above the prior-day high for shorts), then a 5-minute close back above the prior-day low (back below the prior-day high for shorts), within 6 bars. Enter on that bar's close.
- Take it only if a new 12-bar low printed on higher cumulative delta (sellers fading) within the last 6 bars (a new high on lower delta for shorts); otherwise skip it.
- Stop 25 dollars beyond the furthest price of the setup.
- A stop usually goes where the trade's idea is invalidated. If price takes out the furthest price of the sweep, the rejection has usually failed.
- Target the prior-day midpoint.
- At most 2 trades a day, one at a time. Flat by 8 PM ET.
How it works
Crypto like bitcoin trades every day, all day, so a session here is the Eastern calendar day, midnight to midnight. Yesterday's high and low are two prices most traders watch. This setup waits for a 5-minute bar to push past one of them and then close back inside. That bar is the trigger. Only then does it read order flow, who was buying and who was selling. Delta is the volume bought minus the volume sold. Session delta adds that up since midnight. Here it is estimated from 1-minute bars: each bar's volume is split by where the bar closed in its range, so it's a rough read, not a tape.
A divergence is a new price extreme that session delta doesn't follow. A new high on lower delta means the push had less buying behind it. That's a sign of fading pressure. It isn't a reversal signal on its own, which is why the setup also needs the close back inside. The entry is that close. A stop usually goes where the trade's idea is invalidated. Here it sits 25 dollars beyond the furthest price of the push. If price takes out the push, the rejection has usually failed. The target is the prior-day midpoint, the price between yesterday's high and low. The setup takes at most 2 trades a day. A sweep is a push past yesterday's high or low that doesn't hold. A close back inside is a reclaim. R is the distance from entry to stop, so a stop-out costs about 1R.
Here's a saved example from Wednesday, November 18, on BTC. The prior-day high was 65,383.29 and the low was 63,256.75. Their midpoint was 64,320.02. At 9:40 AM, a 5-minute bar pushed below the prior-day low, down to 63,139.63, and closed back above it at 63,305.65. That was a long trigger. It was skipped. The new low at 63,139.63 took out the earlier 63,520.91, with session delta of −150 against −105 at the earlier low. Delta didn't weaken, so the push still had force. The bar closing at 9:45 AM pushed below the prior-day low, down to 63,001.00. At 9:50 AM, a 5-minute bar closed back above it at 63,316.99. That was another long trigger. Every check passed. The new low at 63,001.00 took out the earlier 63,139.63, with session delta of −99 against −150 at the earlier low. Delta was weaker there, so sellers were fading. The long went on at 63,316.99. The stop went 25 dollars beyond the sweep low, at 62,976.00. That's 340.99 dollars of risk. If price takes out the sweep low, the rejection has usually failed. The target was the prior-day midpoint, 64,320.02. BTC hit the stop at 62,976.00 in the 10:47 AM minute. The trade closed there for −1.00R. That was the day's one trade. The bar closing at 12:00 PM pushed below the prior-day low, down to 62,342.04. At 12:30 PM, a 5-minute bar closed back above it at 63,275.52. That was another long trigger. It was skipped. The push bottomed at 62,342.04, above the earlier low of 62,202.64. So there was no divergence to read.
In your own replays, mark yesterday's high and low first. When price pushes through one, compare session delta at the new extreme with its value at the last extreme. If delta didn't follow the price, watch for the close back inside. Note how often the push kept going anyway. Divergence only says pressure is fading, and a fade can stall or fail. Treat these hypothetical results as a small sample, not a verdict.
Session by session
| Date | Trades | Result | Skipped |
|---|---|---|---|
| Oct 30, 2026 | 0 | No trade | 3 |
| Oct 31, 2026 | 0 | No trade | 0 |
| Nov 1, 2026 | 0 | No trade | 1 |
| Nov 2, 2026 | 1 | -1.00 R | 1 |
| Nov 3, 2026 | 0 | No trade | 0 |
| Nov 4, 2026 | 0 | No trade | 0 |
| Nov 5, 2026 | 1 | -1.00 R | 1 |
| Nov 6, 2026 | 1 | +1.20 R | 0 |
| Nov 7, 2026 | 0 | No trade | 0 |
| Nov 8, 2026 | 0 | No trade | 0 |
| Nov 9, 2026 | 0 | No trade | 0 |
| Nov 10, 2026 | 0 | No trade | 3 |
| Nov 11, 2026 | 0 | No trade | 0 |
| Nov 12, 2026 | 0 | No trade | 0 |
| Nov 13, 2026 | 1 | -1.00 R | 1 |
| Nov 14, 2026 | 1 | -1.21 R | 0 |
| Nov 15, 2026 | 0 | No trade | 0 |
| Nov 16, 2026 | 0 | No trade | 0 |
| Nov 17, 2026 | 0 | No trade | 5 |
| Nov 18, 2026 | 1 | -1.00 R | 2 |
Saved replay
Nov 18, 2026: the most recent session with both a trade and a skipped signal.
Bitcoin Prior-Day Push + Delta Divergence questions
How did Bitcoin Prior-Day Push + Delta Divergence do?
Over the last 20 sessions it took 6 trades, won 17%, net -4.01 R (hypothetical).
What is the Bitcoin Prior-Day Push + Delta Divergence drawdown?
The deepest losing run was 4 R, from Oct 30 to Nov 18.
Do the Bitcoin Prior-Day Push + Delta Divergence results include fees?
Hypothetical results on past sessions: no fees, no slippage, fills at bar prices.
What triggers Bitcoin Prior-Day Push + Delta Divergence?
BTC, 5-minute bars. Longs and shorts. Entries every day, 8 AM–4 PM ET.