VWAP — the volume-weighted average price, anchored to the session open — is the line institutional desks watch all day, which is why retail strategies cluster around it. The trend reading is the simplest: hold long while price trades above VWAP, hold short while it trades below, flipping on each cross. The thesis is that VWAP separates the day’s buyers from its sellers, so being on the right side of it keeps you with the dominant flow.
We tested that cross-trend reading mechanically: enter on the cross of the day-anchored VWAP, go flat at session end, with a 1.5×ATR(14) stop and a 1.5R target, on 5-minute candles from 12 months of real Binance 1-minute data on BTC and ETH — both directions, 1% risk, 0.05% commission per side, the same deterministic engine behind Secuora’s AI backtester. The headline is the harshest in this batch: the rule fired thousands of times per symbol, not a single month finished green, and both equity curves were driven all the way to zero. As the page below explains, that is a story about anchoring and frequency at least as much as about VWAP itself. Full numbers inside.
Verified Result
No edge: net negative after costs across 2 markets.
| Market | TF | Trades | Win | PF | Max DD | Net |
|---|---|---|---|---|---|---|
| ETH | 5m | 2,752 | 39.8% | 0.65 | 100.0% | -100.0% |
| BTC | 5m | 2,902 | 38.2% | 0.34 | 100.0% | -100.0% |
How the SVS 20 breaks down ▾
12 months of real 1-minute data, fees on (0.05%/side), $10k start, 1% risk. How the score works →
The exact rules we tested
- Anchor VWAP to the session open and recompute it each day on 5-minute candles.
- Enter long on a cross up through VWAP; enter short on a cross down through VWAP.
- Cross semantics — the signal fires once on the cross, not on every bar price stays on one side.
- Stop 1.5×ATR(14) from entry; target 1.5R; go flat at session end.
- No session filter beyond the daily anchor — crypto trades 24/7, so every cross is taken, both directions.
- Risk 1% of equity per trade; 0.05% commission per side; 10× max notional leverage.
Results
Binance spot 1-minute klines (data-api.binance.vision), aggregated per strategy timeframe · starting balance $10,000 · risk 1%/trade · Commission 0.05% per side; no spread/slippage modeled (BTC/ETH spot spreads are sub-basis-point); position size capped at 10× notional leverage. Generated 2026-06-12 by the Secuora ai-strategy deterministic runner (same engine as the in-app AI backtester).
| Month | Trades | Win rate | Net P&L |
|---|---|---|---|
| 2025-06 | 235 | 41% | −$7,410 |
| 2025-07 | 267 | 39% | −$2,125 |
| 2025-08 | 236 | 38% | −$371 |
| 2025-09 | 208 | 33% | −$73 |
| 2025-10 | 272 | 39% | −$17 |
| 2025-11 | 249 | 39% | −$4 |
| 2025-12 | 263 | 37% | −$1 |
| 2026-01 | 263 | 34% | −$0 |
| 2026-02 | 185 | 41% | −$0 |
| 2026-03 | 235 | 42% | −$0 |
| 2026-04 | 228 | 38% | −$0 |
| 2026-05 | 261 | 39% | −$0 |
| Month | Trades | Win rate | Net P&L |
|---|---|---|---|
| 2025-06 | 207 | 43% | −$4,411 |
| 2025-07 | 230 | 47% | −$1,880 |
| 2025-08 | 229 | 38% | −$2,048 |
| 2025-09 | 249 | 37% | −$1,223 |
| 2025-10 | 240 | 43% | −$205 |
| 2025-11 | 227 | 43% | −$92 |
| 2025-12 | 247 | 37% | −$101 |
| 2026-01 | 243 | 31% | −$33 |
| 2026-02 | 198 | 42% | −$3 |
| 2026-03 | 228 | 39% | −$3 |
| 2026-04 | 221 | 34% | −$1 |
| 2026-05 | 233 | 45% | −$0 |
Assumptions (how loose terms were pinned down)
- Enter on the cross of the day-anchored VWAP; flat at session end; 1.5x ATR stop; 1.5R
The anchor problem and the frequency problem, compounding
Two flaws stacked to produce the worst result in this batch. The first is the anchor. VWAP was designed for a market with a real session open — equities reset every morning, and the line means something because everyone re-anchors at the same bell. Crypto never closes, so anchoring VWAP to an arbitrary daily boundary draws a line through a market that did not reset, and the "cross" it generates is often just price oscillating around a number with no structural significance. The second flaw is frequency: a bare cross-trend rule on 5-minute candles flips constantly, firing thousands of times a year per symbol. Multiply that trade count by 0.05% per side and the fee line alone is enormous — on both symbols the commission paid was larger than the entire starting balance.
Put those together and the outcome was not a slow bleed but a controlled demolition: zero green months on either symbol and both equity curves to zero. The win rate around 38–40% is almost a distraction — even a respectable hit rate cannot survive a 1.5R target whipsawing on a line that does not mark real supply and demand, with a fee charged on every flip. The lesson generalizes past VWAP: a frequently-firing rule with no genuine edge is not neutral, it is a fee subscription, and the more often it trades the faster it pays out. If VWAP has a use on 24/7 markets it is almost certainly as a mean-reversion reference or a bias filter, not as a high-frequency cross engine — and the unfiltered baseline here is what any improvement has to beat.
How to backtest the VWAP trend on Secuora
VWAP is a built-in primitive of the AI backtester, so the cross-trend rule reproduces in seconds — and the first thing to read is the fee line.
- Open /backtest/ai and describe it in plain English: "go long on a cross up through the day-anchored VWAP, short on a cross down, 1.5×ATR stop, 1.5R target, flat at session end, 5-minute candles." It compiles to the same engine that produced this page.
- Run it with costs on and read total fees against net P&L before anything else — that single comparison is the whole story of this page.
- Change one variable per run: trade VWAP as a mean-reversion fade instead of a trend cross, use it only as a directional filter for a separate setup, or move to a higher timeframe so the line flips far less often.
- Sign up free and use the replay terminal — add VWAP as one of your two free indicators and replay a chopping day bar by bar to see how many crosses were noise around the line.
- Journal the variants worth keeping, with confluences and screenshots attached, so the version you trade live is the version that survived the fee line in testing.
Methodology, in one paragraph
Data: Binance spot 1-minute klines (data-api.binance.vision), aggregated per strategy timeframe, June 1, 2025 – June 1, 2026 (12 months). Execution: Secuora’s deterministic strategy runner (the same engine behind the in-app AI backtester) — single position at a time, entries at the close of the signal candle, commission 0.05% per side; no spread/slippage modeled (btc/eth spot spreads are sub-basis-point); position size capped at 10× notional leverage, starting balance $10,000, 1% risk per trade. Swings are confirmed fractals with no look-ahead. These are mechanical results: no discretion, every signal taken. Past performance does not predict future results; this is research, not financial advice.
Frequently asked questions
What is the VWAP trend strategy?
VWAP is the volume-weighted average price, usually anchored to the session open. The trend reading holds you long while price is above VWAP and short while it is below, flipping on each cross, on the idea that VWAP separates the session’s buyers from its sellers and you want to be on the dominant side.
What win rate does the VWAP cross have?
Around 38–40% on 5-minute BTC and ETH in our 12-month test — but the win rate is beside the point here. The rule fired thousands of times per symbol, not one month finished green, and both equity curves went to zero because the commission on that trade count exceeded the entire starting balance. The exact figures are in the results table above.
Does VWAP work on crypto?
As a high-frequency cross-trend engine on 24/7 crypto, it failed badly in our test — both accounts went to zero, driven mostly by fees on an enormous trade count and by a daily anchor that has no real meaning in a market that never resets. VWAP may still be useful on crypto as a mean-reversion reference or a bias filter; it was the constant flipping, not the line itself, that did the damage.
How do I backtest a VWAP strategy myself?
Two ways on Secuora: describe the rule in plain English at /backtest/ai — VWAP is a built-in primitive that compiles to the same deterministic engine used here — or sign up free, add VWAP in the replay terminal, and trade it bar by bar with simulated orders. Either way, check the fee line against net P&L first.