Scalping is the strategy where backtesting matters most and where the average tool lies hardest — because the thing that decides a scalper’s P&L is fees, and most backtests quietly ignore them. At high frequency, commission and spread don’t shave your edge; they are the game. "Best scalping backtesting software" really means: does it model costs honestly, does it have low-timeframe data, and can you take enough reps to find out the truth.
We have a hard number on this. When we ran a 1-minute momentum-scalping system on BTC over 12 months, it placed 15,653 trades and paid $10,850 in fees — more than the entire $10,000 account, which it wiped out (-100%). Disclosure: Secuora is ours; that finding came from our own engine, and we publish it because a scalper needs to internalize it before risking a cent.
What actually matters
- Honest cost modeling — commission per side (and spread/slippage where it applies); a scalping backtest that ignores fees is fiction.
- Low-timeframe data and replay — 1m candles (and Secuora’s derived 30-second view) so you can rehearse the actual cadence you trade.
- High rep count, fast — scalping edges are thin; you need a large sample to separate edge from noise.
- Per-trade review — see how often a "winning" gross trade turned into a net loss after costs.
- A free tier to stress-test the math before paying.
The honest shortlist
Replay down to 1m with a derived 30s view, simulated orders, and a deterministic engine that models commission; honest gap — no tick-by-tick or Level 2 depth.
Deep tick data, market-replay, and per-contract commission modeling — but desktop-bound with a steeper learning curve (free with a funded account).
Second-level replay data, broker auto-import, and analytics suited to high trade counts — but no permanent free plan (as of June 2026).
Free, and modeling commission per trade in a sheet teaches the lesson viscerally — but no replay, no execution feel, and unusable at scalping volume.
Where Secuora fits
Secuora’s scalping value is the combination of low-timeframe replay (1-minute, plus a derived 30-second view) and a deterministic engine that applies commission to every fill — so when you rehearse a scalp, the net number you see already has fees in it. You trade candle by candle with simulated market/limit/stop orders, SL/TP and partial closes, and each rehearsed trade pushes to the journal in one click.
The AI backtester at /backtest/ai is useful here too: describe a scalping rule in plain English ("on the 1m, enter on an opening-range break with a 1.5R target") and it compiles to a deterministic engine that runs it with costs on — the AI writes the rules, never the numbers, so it can’t flatter a thin edge into a fake one.
Honest gap: Secuora is not a tick/Level-2 platform. It models commission on bar data; it does not replay the order book tick by tick or model queue position. For tick-precise futures scalping with depth-of-market, NinjaTrader is the better tool. Secuora is for testing whether a scalping idea survives fees at all — which, in our research, is where most of them die.
The fee math that ends most scalping systems
Here is the finding stated plainly: our 1-minute momentum scalper on BTC took 15,653 trades in a year and paid $10,850 in commission on a $10,000 account — it lost the entire account, and fees alone exceeded the starting balance. The same system on ETH paid roughly $9,530 in fees and also went to zero. This isn’t a quirk of one system; it’s the structural problem with scalping — your gross edge has to clear a cost wall that scales with frequency, and most edges don’t. we backtested 60+ strategies across 6 markets — over 80,000 simulated trades, fees on — and published every result (none were profitable after costs). The honest takeaway: before you scalp live, rehearse with fees switched on and watch how many gross winners become net losers. That single habit saves accounts.
Frequently asked questions
Why did your 1-minute scalping backtest lose so badly?
Fees. Over 12 months the 1-minute momentum scalper placed 15,653 trades on BTC and paid $10,850 in commission on a $10,000 account — more than the whole account — and finished down 100%. At scalping frequency the cost of trading dwarfs a thin gross edge. That’s the central lesson of scalping backtesting: model the fees, or the result is fiction.
What is the best backtesting software for scalping?
For tick-level futures scalping with depth-of-market, NinjaTrader is the deep specialist. For a free, fee-aware replay where you can rehearse 1-minute (and a derived 30-second view) scalps with commission applied to every fill, Secuora fits — with the honest caveat that it models commission on bar data, not tick-by-tick order-book depth.
Can I backtest a scalping strategy for free?
Yes — Secuora’s free plan includes bar-by-bar replay (including the 1-minute timeframe on crypto and the FX majors over a short window) with a deterministic engine that applies commission. There’s also a no-sign-up demo at /backtest/demo. The point is to see whether your scalp survives fees before you fund an account.