Of everything we tested, six families of intraday Nasdaq strategies passed both tests: recent Nasdaq and the years 2006 to 2019, which had not been used for anything. They all follow price, anchor to the New York cash session and trade little. And they're all slow: the most studied one takes about 11 months to pass a 50K with 1 contract.
The bar
To say a strategy "holds up" we asked for four things at once: make money on 2019-2026 micro Nasdaq with fitting, validation and test periods, and with costs; make money on 2006-2019 Nasdaq, with the financial crisis and the sideways years inside; neighboring settings must also make money (a plateau, not a peak); and beat the chance control where it can be checked. We don't give the parameters: what's useful is understanding what they have in common and what they're really worth.
The ones that passed
| Family | What it does, in one sentence | t 2019-2026 | t 2006-2019 | Neighbors that win | Verdict |
|---|---|---|---|---|---|
| Noise band | intraday momentum: follow price when it breaks strongly out of its usual range | 3.34 | 2.95 | 100% | very robust |
| Opening drive | follow the direction of the session's first minutes | 2.46 | 2.68 | 95% | very robust |
| VWAP pullback | enter pullbacks to VWAP in the direction of the morning trend | 3.45 | 2.22 | 100% | robust |
| Morning breakout | buy the break of recent highs in the morning | 3.09 | 2.17 | 98% | robust |
| V reversal | buy the opening V-shaped reversal | 2.92 | 1.78 | 94% | acceptable |
| Cash-session close | the move around the close of the cash session | 3.58 | 1.66 | 96% | acceptable |
| 30-min ORB (improved) | opening range breakout | 2.09 | 0.07 | 43% | doesn't hold up |
| Opening ORB (improved) | opening range breakout | 2.47 | 0.27 | 66% | doesn't hold up |
t: how many times larger the average result is than its noise; below 2, it can be explained by chance. Neighbors: share of nearby settings that make money in both periods.
Where the noise band comes from. It starts from a published academic paper: Carlo Zarattini, Andrew Aziz and Andrea Barbon (2024), Beat the Market: An Effective Intraday Momentum Strategy for S&P500 ETF (SPY), Swiss Finance Institute Research Paper No. 24-97. The paper studies it on the S&P 500 ETF; we measured our own version on the Nasdaq, with our parameters and our tests. In our first tests, 96% of 45 settings made money out of sample: a broad edge that doesn't depend on fine-tuning.
What they have in common
- They follow price. No intraday strategy that trades against the move passed the bar.
- They anchor to the New York cash session: the open, the morning and the close.
- They trade little, so costs weigh less.
- Their filters are measured relative to price or volatility, not in fixed ticks. With the Nasdaq four times higher than in 2019, a fixed-tick filter ended up letting only 5% of bars through: the strategy had switched itself off.
- They only work on the Nasdaq (more in this article).
- They make little per trade: statistically real, economically modest.
The fine print: the VWAP pullback
The final version, with 1 MNQ contract and costs, made $19,050 in seven years, with a profit factor of 1.31 and 7/8 positive years. Sounds good. Now the important part:
- It makes about $17 per trade and wins 39% of the time. It has strung together up to 15 losses in a row.
- In a 50K account with 1 contract it blew up between 4% and 7% of accounts, but took a median of about 11 months to reach +$3,000.
- With 2 contracts it passed in about 4 months, but blew up between 18% and 26%.
A real, robust and slow strategy: the opposite of what's sold on YouTube. And even with the best ones, the worst drawdown you've seen is not the worst you'll see: on the noise band, the largest realized drawdown was about $18,000 per NQ contract, and the Monte Carlo 95th percentile about $46,000.
Coming from the literature doesn't make it good
The intraday market momentum of Gao, Han, Li and Zhou (2018) lost $10,444 on micro Nasdaq with costs; the opening gap fill made money in 4% of 216 variants, about what chance would give. Every published idea has to be tested again with your data and your costs.
The best of all doesn't fit in a prop firm account
The most robust idea we found isn't intraday: it's daily trend following, the classic rule of Moskowitz, Ooi and Pedersen (2012) applied without tuning anything. It made money in both halves of the data in 20/20 window and rebalance combinations, and its correlation with the intraday ones is -0.04. But it holds positions for weeks or months, and APEX and Lucid require you to be flat at the end of each session; besides, its drawdowns don't fit in a $2,000 cushion. The most solid edge needs a personal account, and prop firm accounts prohibit exactly what makes it work.
What it means for you
- Systematic intraday edges exist, but they're small: a robust strategy passes a 50K in months, not days. If someone promises you'll pass in a week with a system, they're selling you luck or size.
- Look for ideas that follow price, anchored to the cash session and with few trades.
- Always ask for the second test: other years of the same market.
- Size with the Monte Carlo, not with the worst drawdown you've already seen (how many contracts to trade).
How it was calculated
Micro Nasdaq 2019-2026 with a $0.62 commission per side and 1 tick of slippage, split into fitting, validation and test; Nasdaq 2006-2019 as a second test, used only once per setup; grids of neighboring settings and chance controls. We don't publish the parameters. Lab results, simulated and hypothetical.
Simulated and hypothetical results, with the limitations that implies. Not investment advice. Firm rules change: check them on each firm's website. This article contains no affiliate links.