Sneaky Pivot Strategy Backtest: Previous Day High and Low Reversal on Dow Futures

00The verdict in three lines

• Does the previous day high and low strategy work? On Dow futures it made money in every period: +$89 a trade after costs targeting the other end of yesterday's range, +$52 targeting the first bar's far end, and both beat random days at the 99.8th percentile.
• So why buried? In 2020 to 2023 the edge was too small next to the noise: a t-statistic of 1.32 to 1.58 against the 2.5 we require. Three of the four versions passed every other gate.
• Buried, cause No proven edge: not disproved, not proven. It is a candidate for a test on new data.

01What is the sneaky pivot strategy?

The source is a YouTube video, “How Trading Like an Idiot Makes Me $10,000/Month (15 Minutes a Day)” (The Rumers, about 2.6 million views), whose description adds that the results “are not typical”. The rules:

• One 15-minute chart, four lines: yesterday's high and low (the range high and range low), and the next earlier high above and low below them (the swing high and swing low).
• Buy only at the two lower lines, sell only at the two upper lines.
• Three candles: a strong first 15-minute candle that reaches one of the lines; a confirming “sneaky” candle (green after a drop to the lows); the entry when price crosses that candle, usually around 45 minutes after the open.
• Stop beyond the lows (or highs) that were tested; target “the top of the candle or somewhere back against the bigger seller”.

The video demonstrates it on YM in a replay and trades two stocks live. In both live trades price did not quite reach the line; the presenter calls that “close enough”, so we test a strict and a near version.

02Yesterday's high and low: what the data says

• The first 15-minute bar reached yesterday's low on 26.5% of 1,709 days and yesterday's high on 33.8%.
• The video's “phenomenon” is half true. When the range low broke after the first 15 minutes, price reached the swing low later that day on 53.8% of 771 days, and came all the way back to the range high on only 8.0%. When the range high broke, the swing high followed on 50.7% of 854 days. Not “almost every single time”.

03How we made the rules mechanical

• Lines: yesterday's regular-session high and low; the swing high is the session high of the most recent earlier day above yesterday's high (the swing low alike).
• Setup: the first bar (09:30-09:45 New York) reaches yesterday's low (strict) or comes within 10% of the 20-day average daily range of it (near) for a buy; the same at yesterday's high for a sell. A first bar reaching both: no trade.
• Sneaky candle: the first green bar (for a buy) among bars 2 and 3. Entry: one tick above its high, by the end of bar 4 (10:30). Stop: one tick below the lowest low since the open; a new low before the entry cancels the setup.
• Targets: T1 = the first bar's high (for a buy); T2 = yesterday's high. Otherwise out at the session close. One trade a day; sells are mirrored.

04How we tested it

YM (E-mini Dow futures, $5 a point), NinjaTrader 1-minute history, back-adjusted, January 2020 to 25 September 2026: 1,709 trading days. Costs: $15.76 a round trip per contract (commission plus one tick of slippage on each fill). Periods: 2020 to 2023, 2024 to March 2025, and a holdout from April 2025 already used (on YM by our trendline test, and the same months on ES and NQ). First-period bar t ≥ 2.5. Control: every trade moved to another day, entered at the same minute with the same stop and target distances, 500 times. A first version of the control entered at a price that was often the top of the minute and made random days look worse than they are; we found and fixed it before reporting any result. Engine check: the levels, both tolerances, setups both ways, the sneaky candle on bar 2 and bar 3, a cancelled setup, the targets and the session-close exit, and an independent re-implementation that matched every one of the 1,788 trades (snp_selftest_result.txt).

05Results: does the sneaky pivot work?

Cumulative net dollars per YM contract by month. * Holdout already used before this test.

How the trades ended

The T1 versions take a small target (a median 51 points) against a larger stop (107 points) and win about two trades in three, usually within a quarter of an hour. The T2 versions aim at the other end of yesterday's range, reach it about one time in five; those targets and the trades still open at the session close paid for the stopped ones. Both made money on longs and on shorts.

06Sneaky pivot or random day?

Moved to random days, the same trades lost about their cost; the real ones were better than 99% of the random sets in every version. Entering after a rejection at yesterday's high or low, rather than at the same time on any day, is what made the difference.

07By period and by year

* The holdout months were already used: on YM by our trendline test, and on ES and NQ by earlier tests.

Every version was positive in all three periods. There were losing years (SP-STRICT-T1: 2021, 2023, 2026; SP-STRICT-T2: 2023, 2024; SP-NEAR-T1: 2021, 2023; SP-NEAR-T2: 2023, 2024), which is why the first period's t-statistic stayed below the bar.

08The gates

The data-gap check removed 1, 4, 1 and 4 trades on days touching a known hole in the data; no verdict changed. No proven edge: positive, but not strong enough to pass the development gate.

Forward test (added 23 Oct 2026): SP-STRICT-T2 is now being tested forward, with its rules, start date (26 October 2026) and evaluation frozen in advance: a paper series computed by this engine on new YM data each month. Protocol and journal.

Replication on gold (added 23 Oct 2026): we ran the same rule on 22 years of gold (XAUUSD) minute data, with the day and the first candle fixed for gold (the 08:20 New York COMEX open) before any result. It did not hold up there: on gold the four versions ended between -$0.12 and -$0.34 an ounce a trade after costs, with two of six gates at best (the sneaky pivot on gold).

TWhat this shows, and what to take into your own tests

What the results show

• A reversal off the previous day's high or low, confirmed by the next 15-minute bar, made money on Dow futures in every period we tested: +$89 a trade after costs aiming back at the other end of yesterday's range, +$52 taking the first bar's far end.
• It beat the same trades moved to random days by a wide margin (99.8th percentile), stayed positive without its 10 best trades, and worked on longs and shorts.
• What it lacked was strength in the first four years: a development t-statistic of 1.32 to 1.58 against the 2.5 we require, because the edge per trade is small next to the noise.

What remains unknown

• Whether the edge holds on new data: the holdout here had already been used, on YM by one earlier test and on ES and NQ for the same months.
• The presenter's "close enough" eye and his optional rule for a very large first candle; we tested a strict and a near reading of the first.
• The stocks he trades live; we tested the Dow future he demonstrates.

Why it was still buried

Our first gate on strength asks for a development t-statistic of at least 2.5, because rules like this have been tried many times on index futures. None of the versions reached it (best 1.58). Everything else, the validation period, the holdout, the random control and the top-10 test, was passed by three of the four versions. "Buried" here means "not proven", not "disproved".

What you can apply to your own tests

• Mark yesterday's high and low before the open; this rule only trades there.
• Let the first two 15-minute bars show you a rejection before you enter, and put the stop beyond the low (or high) they tested.
• Treat a promising backtest as a reason to forward-test, not to size up.

A note on the numbers: drawdown figures are measured on closed trades unless a table says it includes open losses. Open losses inside a trade are not counted in a closed-trade drawdown, so real account risk is higher, especially for systems with no stop. Backtests are not live trading.

FAQCommon questions

Does the previous day high and low strategy work?

In our Dow futures backtest it made money in every period: +$89 a trade targeting the other end of yesterday's range and +$52 targeting the first candle's far end, after $15.76 costs. It passed five of our six gates; only the strength test on 2020-2023 failed.

What is the sneaky pivot strategy?

A 15-minute day-trading rule from a YouTube trader: mark yesterday's high and low, wait for the first candle to reach one of them, then enter when price crosses a confirming candle, with the stop beyond the tested low or high and the target back across the range.

Does price go to the next level after breaking the previous day low?

About half the time in our data: when the previous day's low broke after the first 15 minutes, the next lower level (the swing low) was reached the same day on 54% of days, and on highs 51%. Not almost every time, as the video says.

Why was a profitable strategy buried?

Because profit alone is not proof. Its edge per trade is small next to the noise: a t-statistic of 1.58 on 2020-2023 against our bar of 2.5 for rules tried many times. "Buried" here means not proven; new data is the next test.

How was the sneaky pivot tested?

On YM (Dow futures) 1-minute data from 2020 to 2026, built into 15-minute bars of the regular session, one contract, with rules written and dated before the data was run, costs included on every trade, results split into development, validation and holdout periods, a random-entry control as the benchmark, and every variant reported. Read the protocol.

09Check our arithmetic

Code provenance: our own code, written from the specification.

10Data notes

Source. NinjaTrader YM 1-minute history (YM DEC26, merge back adjusted), exported and checked on 22 October 2026; holes listed in data_holes_ym.json and never filled. A day needs its 09:31 bar and at least 300 one-minute bars. Fills: stop orders at their level, or a minute's open if it gapped through; one tick of slippage per fill is in the cost.

11Corrections

23 Oct 2026, addition: a paragraph in section 08 linking the forward test of SP-STRICT-T2. No result changed. 23 Oct 2026, addition: a paragraph in section 08 and a related link pointing to the replication of this rule on gold (its first sentence made more exact the same day). No result changed.

+More tests and guides

• Gold previous day high and low
• Trendline trading strategy
• Opening range breakout strategy
• Nr7 trading strategy
• Free prop firm challenge calculator
• What all the backtests found so far
• The backtesting protocol
• How to backtest a trading strategy without fooling yourself
• Backtesting glossary