Tested by BacktestingMax

VWAP cross strategy backtest

Does the VWAP cross strategy make money in day trading?

Mostly no. VWAP is the average price of the day so far, weighted by volume. The VWAP cross strategy buys when price closes above VWAP and sells when price closes below it. We ran 21 backtests from 2020 to 2025, after trading costs: 3 stock indices, 3 forex pairs and gold, each on 5-, 15- and 30-minute charts. Only 3 of the 21 made money: the DAX on 15- and 30-minute charts, and the Nasdaq 100 on 15-minute charts. Of those 3, only the two DAX tests made money in both 2020–2023 and 2024–2025. Every forex, gold and 5-minute test lost money. In 12 of the 21 tests, the trades made money before costs, but the trading costs were bigger.

21 backtests 2020–2025 Trading costs included Tested 12 October 2026

Made money

3

of 21 tests, using the basic rules

Made money in both periods

2

2020–2023 and again 2024–2025

Best backtest

+215%

DAX, 15-minute chart: account growth 2020–2025

Only the first 2 hours after the open

3

tests made money in both periods, compared with 2 using the basic rules

Made money in both periods

• DAX, 15-minute
• DAX, 30-minute

Lost money on every market

• 5-minute charts

The results

Results on every market

Each box shows one backtest of the VWAP cross using the basic rules, from 2020 to 2025, after trading costs: the % result, and a small chart of the account balance. The solid line is the account balance and the dashed line is the starting balance. Green results made money; red results lost money. Under each result: whether it made money in both 2020–2023 and 2024–2025, or passed all four of our checks (see "How a test passes" further down).

The account balance

The best test and the middle test

What we found

What the 21 backtests show

1. On forex and gold, the VWAP cross lost money on all 12 tests: EUR/USD, GBP/USD, USD/JPY and gold, each on 5-, 15- and 30-minute charts. In 11 of the 12 tests, the account was down more than 90% from its starting balance at some point. In 6 of the 12, the trades lost money even before trading costs.

2. On 5-minute charts, the VWAP cross lost money on all 7 markets. A 5-minute chart has more candles, so more candles close on the other side of VWAP. That means more trades, and every trade pays trading costs. On the S&P 500, the 5-minute chart made about 5 or 6 trades a day, compared with about 2 on the 30-minute chart. From 2020 to 2025, the S&P 500's 5-minute trades made $17,613 before costs and paid $26,679 in costs, so the $10,000 account lost 90.7%.

3. On the stock indices, the trades made money before costs in all 9 tests. But in 6 of the 9, the trading costs were bigger than that profit, including all three S&P 500 tests (see the table below). The DAX was the only market where the VWAP cross made money in both 2020–2023 and 2024–2025: on the 15-minute chart (+215% from 2020 to 2025, the best backtest) and on the 30-minute chart (+108%). Even in the best test, the biggest fall from the highest balance was 56%. No test passed all four of our checks.

4. The Nasdaq 100 made money from 2020 to 2023 on all three charts, then lost money on all three in 2024–2025. So its results did not last into the newer years. On the 15-minute chart, the $10,000 account grew 106% by the end of 2023, to about $20,600. In 2024–2025 the account fell 40%, to about $12,300. So over the whole test, the account was still up 23%.

5. Most trades lost money. In all 21 tests, only 14% to 34% of the trades made money (this share is the win rate). Price often crosses VWAP and then crosses back within an hour, and a trade opened on that kind of cross usually ends with a small loss. On the stock indices, most of the profit came from the trades that stayed open for 4 hours or more, on days when price stayed on one side of VWAP.

Trading costs

On the stock indices, trading costs were bigger than the profit in 6 of 9 tests

Each row is one backtest of the VWAP cross using the basic rules, from 2020 to 2025, on a $10,000 account. 'Made before costs' is what the trades made before paying trading costs. Trading costs here are the spread: the gap between the buy price and the sell price.

Every trade risked 1% of the account at that time, so the dollar amounts grow and shrink with the account.

Time of day

On the DAX, the profit came from the first 2 hours

Each bar adds up the results of all the DAX trades opened in that hour, Frankfurt time. The results are in R: 1R is the amount risked on one trade. The bars include all three DAX charts using the basic rules, and all years from 2020 to 2025. Green bars made money in total; red bars lost money.

• 09:00–10:003,518 trades +243.3R
• 10:00–11:003,470 trades +215.3R
• 11:00–12:002,453 trades −11.3R
• 12:00–13:001,919 trades −60.1R
• 13:00–14:001,776 trades −76.4R
• 14:00–15:001,778 trades −25.3R
• 15:00–16:001,873 trades +16.3R
• 16:00–17:002,117 trades −90.1R
• 17:00–17:30901 trades −38.7R

On the S&P 500 and the Nasdaq 100, the pattern was different. Their trades opened in the first 30 minutes after the 09:30 New York open lost 103.8R in total. Their trades opened from 13:00 to 15:00 New York time made 191.1R in total. This fits the result of the first-2-hours change further down: the first-2-hours rule helped on the DAX, but did worse than the basic rules on 4 of the 6 S&P 500 and Nasdaq 100 tests.

The strategy

What the VWAP cross is

VWAP stands for volume-weighted average price. VWAP is the average price since the session opened. A session is the part of the day the strategy trades, for example 09:30 to 16:00 New York time on the S&P 500 (see the first rule). Candles with more volume (more buying and selling) count more in this average. VWAP starts again at every session open, so it draws a new line each day. Many day traders see price above VWAP as a sign that buyers are stronger, and price below VWAP as a sign that sellers are stronger. The VWAP cross strategy buys when a candle closes above VWAP after the candle before it closed below VWAP. It sells (goes short) in the opposite case.

VWAP

Each candle's typical price, (high + low + close) ÷ 3, weighted by the candle's volume

Volume

The volume in our price data. Forex has no central exchange, and our stock index and gold prices are CFD prices (contracts that follow the market's price). So this is not the volume of the whole market, and our VWAP can differ from the VWAP on an exchange chart.

Stop-loss

1 ATR (the average candle size over the last 14 candles) beyond VWAP, set when the trade opens

Take-profit

None. A close on the other side of VWAP, or the end of the session, closes the trade.

Risk per trade

1% of the account

The rules we tested

1. Each market has a session, and VWAP starts again at each session open. On the S&P 500 and the Nasdaq 100, the session is the New York stock market's hours: 09:30 to 16:00 New York time. On the DAX, the session is 09:00 to 17:30 Frankfurt time. Forex and gold trade 24 hours a day, so for them the session runs from 08:00 London time (the London open) to 17:00 New York time (the end of the forex day).
2. Buy when a candle closes above VWAP and the candle before it closed below VWAP. The trade opens at the next candle's open.
3. Sell (go short) when a candle closes below VWAP and the candle before it closed above VWAP. The trade opens at the next candle's open.
4. Set the stop-loss when you place the trade: 1 ATR below VWAP for a buy, 1 ATR above VWAP for a sell. ATR is an indicator of how far price moves in an average candle, measured over the last 14 candles of the same chart. The stop-loss does not move after that.
5. When a candle closes on the other side of VWAP, close the trade at the next candle's open. If that next candle also closes on the other side of VWAP, open a trade in the new direction at the following candle's open. This also applies when the stop-loss is hit during a candle that closes on the other side of VWAP. After any other stop-loss, wait for the next cross.
6. Close any open trade at the end of the session. Don't open a new trade on the session's last candle. On the stock indices, don't trade on that exchange's holidays. On the S&P 500 and the Nasdaq 100, also skip the half days when New York closes at 13:00.
7. Size every trade so that, if the stop-loss is hit, the trade loses 1% of the account.

The method

How we tested it

The test

Markets

S&P 500, Nasdaq 100, DAX, Gold, EUR/USD, GBP/USD and USD/JPY

Charts

5-minute, 15-minute and 30-minute candles

Dates

January 2020 to December 2025

Account

$10,000, risking 1% of the account on each trade

Trading costs

Included. Every trade pays that market's normal trading cost: the spread, which is the gap between the buy price and the sell price.

How a test passes

We call a test passed only if it meets all four checks. A strategy can make money in one backtest and still fail one of these checks.

• At least 100 trades. With fewer trades, a good result could be luck.
• It made money from 2020 to 2023 and again from 2024 to 2025. This shows whether the strategy kept working in the newer years, not only in the older ones.
• A profit factor (the money won divided by the money lost) of at least 1.2 from 2024 to 2025, after trading costs.
• In 95% of 1,000 random shuffles of the trades, the account must not fall more than 30% from its highest point.

Improvements

Can the VWAP cross be improved?

We tried four changes that traders often make to the VWAP cross: trading only in the first 2 hours of the session, a 200 EMA trend filter, taking only crosses made on high volume, and a fixed 2:1 take-profit. The table below explains each one. We ran each change on the same 21 tests. We picked the best change using 2020–2023 only: the change that made money in the most tests, and in a tie, the one with the better middle test. Then we checked it on 2024–2025, the years we did not use to pick it. If you pick a change and judge it on the same years, it almost always looks better than it really is.

One change helped, but only on the DAX: trading only in the first 2 hours of the session. In 2020–2023, the years we used to pick, none of the four changes made money in more tests than the basic rules, which made money in 6 of 21. The best of the four was trading only the first 2 hours: 5 of 21, the same as the 200 EMA filter, but with a smaller loss in the middle test (−53.8%, compared with −77.0%). In 2024–2025, the years we kept for the check, the first-2-hours rule made money in 7 of 21 tests, compared with 3 using the basic rules. The first-2-hours rule made money on all three DAX charts in both periods. But from 2020 to 2025, the first-2-hours rule did worse than the basic rules on 4 of the 6 S&P 500 and Nasdaq 100 tests, and lost money on every forex and gold test. The middle test for 2024–2025 lost a little less with the basic rules (−15.5%, compared with −22.9%), mostly because by 2024, 4 forex and gold accounts using the basic rules had almost no money left to trade. The fixed 2:1 take-profit did worst: none of its 21 tests made money in 2024–2025. It closes trades that would have stayed open for hours: with it, the trades in the 9 stock index tests lost 1,120.3R in total, compared with a gain of 131.8R using the basic rules.

Questions

Questions traders ask

class VwapCrossIntraday(Strategy):
    """The intraday VWAP cross, as day traders learn it, plus the changes they try.
    VWAP (volume-weighted average price) is the average price of the session so far,
    weighted by volume: the sum of (high + low + close) / 3 x volume over the session's
    candles, divided by the sum of their volume. It starts again at every session open.
    Sessions, each on its own city's clock (daylight saving included):
      US indices (S&P 500, Nasdaq 100, Dow, Russell 2000): 09:30-16:00 New York, the stock
      market's hours. No trading on days the New York Stock Exchange is closed or closes
      early at 13:00.
      DAX: 09:00-17:30 Frankfurt. No trading on days the Frankfurt exchange (Xetra) is closed.
      Forex and gold trade around the clock, so they use 08:00 London (the London open)
      to 17:00 New York (the end of the forex day). So does any market not in the table.
    Standard (all switches off): buy when a candle closes above VWAP and the candle before
    closed below it; sell short on the opposite cross. A close on the other side of VWAP
    closes the trade, and the next candle opens the trade in the new direction if it also
    closes on that side. After a stop-loss, wait for the next cross. The stop sits
    1 x ATR(14) beyond VWAP; size so that the stop costs 1% of the account. No new trade
    on the session's last candle; a trade still open then closes at the session close.
    The changes, one at a time:
      morning = 1   only open trades in the first 2 hours of the session; open trades
                    still close on the usual rules
      trend   = 1   only buy above the 200 EMA, only sell below it
      volume  = 1   only take a cross made by a candle with more volume than the
                    average of the last 20 candles
      exit_rr = 2   also take profit at 2 times the risk
    """
    params = {"atr_mult": Param(type=float, default=1.0, min=0.25, max=5.0),
              "morning": Param(type=int, default=0, min=0, max=1),
              "trend": Param(type=int, default=0, min=0, max=1),
              "volume": Param(type=int, default=0, min=0, max=1),
              "exit_rr": Param(type=float, default=0.0, min=0.0, max=10.0)}
    lookback = 20
    # Session hours: (clock of the open, open, clock of the close, close, exchange calendar),
    # with the times in minutes after midnight.
    hours = {"USA500IDXUSD": ("America/New_York", 570, "America/New_York", 960, "nyse"),
             "USATECHIDXUSD": ("America/New_York", 570, "America/New_York", 960, "nyse"),
             "USA30IDXUSD": ("America/New_York", 570, "America/New_York", 960, "nyse"),
             "USSC2000IDXUSD": ("America/New_York", 570, "America/New_York", 960, "nyse"),
             "DEUIDXEUR": ("Europe/Berlin", 540, "Europe/Berlin", 1050, "xetra")}
    around_the_clock = ("Europe/London", 480, "America/New_York", 1020, "")
    # Days with no trading, 2020-2025: the exchange is closed, or (New York) closes at 13:00.
    closed_days = {
        "nyse": "2020-01-01 2020-01-20 2020-02-17 2020-04-10 2020-05-25 2020-07-03 2020-09-07 2020-11-26 "
                "2020-11-27 2020-12-24 2020-12-25 "
                "2021-01-01 2021-01-18 2021-02-15 2021-04-02 2021-05-31 2021-07-05 2021-09-06 2021-11-25 "
                "2021-11-26 2021-12-24 "
                "2022-01-17 2022-02-21 2022-04-15 2022-05-30 2022-06-20 2022-07-04 2022-09-05 2022-11-24 "
                "2022-11-25 2022-12-26 "
                "2023-01-02 2023-01-16 2023-02-20 2023-04-07 2023-05-29 2023-06-19 2023-07-03 2023-07-04 "
                "2023-09-04 2023-11-23 2023-11-24 2023-12-25 "
                "2024-01-01 2024-01-15 2024-02-19 2024-03-29 2024-05-27 2024-06-19 2024-07-03 2024-07-04 "
                "2024-09-02 2024-11-28 2024-11-29 2024-12-24 2024-12-25 "
                "2025-01-01 2025-01-09 2025-01-20 2025-02-17 2025-04-18 2025-05-26 2025-06-19 2025-07-03 "
                "2025-07-04 2025-09-01 2025-11-27 2025-11-28 2025-12-24 2025-12-25",
        "xetra": "2020-01-01 2020-04-10 2020-04-13 2020-05-01 2020-12-24 2020-12-25 2020-12-31 "
                 "2021-01-01 2021-04-02 2021-04-05 2021-12-24 2021-12-31 "
                 "2022-04-15 2022-04-18 2022-12-26 "
                 "2023-04-07 2023-04-10 2023-05-01 2023-12-25 2023-12-26 "
                 "2024-01-01 2024-03-29 2024-04-01 2024-05-01 2024-12-24 2024-12-25 2024-12-26 2024-12-31 "
                 "2025-01-01 2025-04-18 2025-04-21 2025-05-01 2025-12-24 2025-12-25 2025-12-26 2025-12-31"}
    def init(self, ctx):
        tz_open, self.open_at, tz_close, self.close_at, calendar = self.hours.get(ctx.symbol, self.around_the_clock)
        self.closed = set(self.closed_days.get(calendar, "").split())
        self.open_clock = ctx.clock(tz_open)
        self.close_clock = self.open_clock if tz_close == tz_open else ctx.clock(tz_close)
        self.atr = ctx.atr(period=14)
        self.ema = ctx.ema(source=ctx.close, period=200)
        self.avg_volume = ctx.rolling_mean(source=ctx.volume, period=20)
        self.day = None      # the date of the session being traded
        self.pv = 0.0        # this session: sum of typical price x volume
        self.vol = 0.0       # this session: sum of volume
        self.side = 0        # +1 / -1: the last candle closed above / below VWAP (0: none yet)
        self.armed = False   # True from a cross until a trade is opened in its direction
        self.loud = False    # the last cross was made on more volume than the 20-candle average
    def on_bar(self, ctx):
        # VWAP counts the session's candles only, and starts again at each session open.
        in_session = (self.open_clock.minute_of_day >= self.open_at
                      and self.close_clock.minute_of_day < self.close_at
                      and self.open_clock.date not in self.closed)
        if in_session and self.open_clock.date != self.day:
            self.day = self.open_clock.date
            self.pv, self.vol, self.armed = 0.0, 0.0, False
        if in_session:
            self.pv += ctx.hlc3[0] * ctx.volume[0]
            self.vol += ctx.volume[0]
        vwap = self.pv / self.vol if in_session and self.vol > 0 else None
        close = ctx.close[0]
        side = 0 if vwap is None else (1 if close > vwap else -1)
        prev_side, self.side = self.side, side
        avg_volume = self.avg_volume.value
        if prev_side != 0 and side != 0 and side != prev_side:
            self.armed = True  # a cross: a new trade may open in its direction
            self.loud = avg_volume is not None and ctx.volume[0] > avg_volume
        if not in_session:
            if ctx.position is not None:
                ctx.close_position()  # only if the session's last candle was missing from the data
            return
        if vwap is None:
            return
        last_candle = self.close_clock.minute_of_day + ctx.timeframe_minutes >= self.close_at
        pos = ctx.position
        if pos is not None:
            if last_candle or (pos.is_long and side < 0) or (pos.is_short and side > 0):
                ctx.close_position()  # the session close, or a close on the other side of VWAP
            return
        if last_candle:
            return
        a = self.atr.value
        if a is None:
            return
        if not ctx.gate("fresh_cross", self.armed):
            return
        minutes_open = self.open_clock.minute_of_day + ctx.timeframe_minutes - self.open_at  # at this candle's close
        if not ctx.gate("first_2_hours", ctx.params.morning == 0 or minutes_open <= 120):
            return
        ema = self.ema.value
        trend_ok = ctx.params.trend == 0 or (ema is not None and (close > ema if side > 0 else close < ema))
        if not ctx.gate("ema_trend", trend_ok):
            return
        if not ctx.gate("cross_volume", ctx.params.volume == 0 or self.loud):
            return
        rr = ctx.params.exit_rr
        if side > 0:
            stop = vwap - ctx.params.atr_mult * a
            tp = close + rr * (close - stop) if rr > 0 else None
            ctx.buy(size=ctx.size_by_risk_pct(0.01, stop=stop, entry=close), sl=stop, tp=tp, tag="vwap_long")
        else:
            stop = vwap + ctx.params.atr_mult * a
            tp = close - rr * (stop - close) if rr > 0 else None
            ctx.sell(size=ctx.size_by_risk_pct(0.01, stop=stop, entry=close), sl=stop, tp=tp, tag="vwap_short")
        self.armed = False

Your turn

Test the VWAP cross yourself

Change the settings, the market or the stop-loss and see the result in seconds. Ask the BacktestingMax AI: Backtest a VWAP cross on the DAX and the S&P 500 15-minute charts, trading only the first 2 hours of the session.

Free account. No credit card needed.

Backtest on historical data with trading costs, run by BacktestingMax. Past results do not guarantee future results, and nothing here is financial advice. Updated 12 October 2026.