This repository holds the result files behind the backtest published at https://www.blitz-trading.com/backtest and the out-of-sample results that followed it, so that the SHA-256 values printed on the site can be checked by anyone.
It is a results archive, not a trading system. The engine source is not here.
See the bot behind these numbers. Blitz is a free DCA ladder bot that runs on your own exchange account. Read the full backtest page, follow a real Blitz bot day by day in live-bot-record, or plan your own ladder with the DCA Ladder Calculator. Backtest results are not a guarantee of future returns.
| Path | What | First release |
|---|---|---|
btc-futures-s14-40-n18-u75-20260824-public.report.json |
BTC historical backtest report, 2021-01-01 → 2026-08-24 | snapshot-20260824 |
oos-20260825-20260930/ |
First out-of-sample window for the BTC and ETH public presets, 2026-08-25 → 2026-09-30 | oos-20260825-20260930 |
Every tagged release of this repository is archived by Zenodo (CERN), which issues a DOI. The archived copy is held by Zenodo, separately from this repository and from blitz-trading.com. Within 30 days of publication, the record owner can delete the record, or start a correction of its files that keeps the same DOI and must be published within 45 days of the original publication. After that, files change only in exceptional cases (such as exposed personal data) through Zenodo support. The SHA-256 values printed here, on the website and in each GitHub release let anyone check that an archived file has not changed.
| DOI | |
|---|---|
| All versions (cite this) | https://doi.org/10.5281/zenodo.22809596 |
Historical snapshot (snapshot-20260824) |
https://doi.org/10.5281/zenodo.22809597 |
Out-of-sample window 2026-08-25 → 2026-09-30 (oos-20260825-20260930) |
https://doi.org/10.5281/zenodo.23166756 |
The JSON inside the Zenodo archive for snapshot-20260824 hashes to the same
SHA-256 as the file in this repository and the value printed on the website
(verified 2026-09-17). The four JSON files inside the archive for
oos-20260825-20260930 hash to the values in that folder's README and in the
GitHub release (verified 2026-10-05).
| File | Bytes | SHA-256 |
|---|---|---|
btc-futures-s14-40-n18-u75-20260824-public.report.json |
12,240 | 1c3306db60746d534e4466ff590cc4fa03938d90f951906a84f7d3a62c122a0a |
The hash above is the value printed on the website as Report JSON SHA256. The file is byte-identical to the one the site links to.
The website also prints two more hashes for the same snapshot — State SHA256 (engine end-state) and Excel SHA256 (a spreadsheet rendering). Those files are not included in this release.
| Public snapshot ID | btc-futures-s14-40-n18-u75-20260824-public |
| Instrument | BTCUSDT perpetual futures |
| Strategy key | s1-4 +40% · N18 · U75 · 5x · Cross · TP ROI 1.5% |
| Input data | Binance BTCUSDT aggTrades (trade-by-trade), 2021-01-01 → 2026-08-24 |
| Engine source commit | 8cff0b0500550cbc3d310eb9e63c17d7af74a714 |
| Report generated | 2026-08-25 |
The strategy parameters are the ones published on the site: an 18-step entry ladder (N18) with steps 1–4 covering +40% of the ladder's range, 75% wallet usage (U75), 5× leverage, cross margin, and take-profit at +1.5% margin ROI (a +0.30% price move at 5×).
# Linux / macOS
sha256sum btc-futures-s14-40-n18-u75-20260824-public.report.json
# Windows (PowerShell)
Get-FileHash .\btc-futures-s14-40-n18-u75-20260824-public.report.json -Algorithm SHA256
Compare the output with the hash in the table above and with the Report JSON SHA256 line on https://www.blitz-trading.com/backtest. If any of the three differ, the file has been altered.
These are the values the website reports. They are reproduced here only so a reader can confirm the file and the site agree; the file is the source.
| Field | Value |
|---|---|
| Starting wallet (USDT) | 133,333 |
final_wallet |
2,201,183.98 |
profit |
2,067,850.98 |
roi_pct |
+1,550.89 |
tp_count (take-profits filled) |
14,018 |
liq_count (liquidations) |
0 |
rebal_count (wallet-growth resets) |
29 |
max_step (deepest ladder step reached) |
18 of 18 |
total_fee (trading fees paid, USDT) |
629,345.95 |
total_volume (USDT) |
899,065,645 |
| Maximum account drawdown | −81.31% (published on the site; not a field in this file) |
Trading fees are deducted from the reported result. Funding and slippage are not separately published for this snapshot.
| Key | Meaning |
|---|---|
final_wallet, profit, roi_pct |
End wallet, net profit, and ROI over the full period, in USDT and percent |
tp_count, cycles_count |
Number of take-profit fills; equal to the number of completed cycles |
liq_count |
Liquidations during the run |
rebal_count |
Times the ladder was re-sized after wallet growth (the run reinvests via a +10% wallet-growth reset; the live bot does not reinvest automatically) |
max_step |
Deepest ladder step the run ever reached |
total_fee, total_volume |
Trading fees paid and notional traded, in USDT |
monthly, yearly |
The same counters broken down by calendar month and year |
open_at_end |
Position state at the end of the run |
- Not live returns. This is a historical simulation on recorded trades. Historical paths cannot reproduce future liquidity, fills, funding, or slippage. Trading involves risk of loss.
- Not out-of-sample. This file covers the backtest period through
2026-08-24 only. Out-of-sample results are in
oos-20260825-20260930/. - Not the engine. The code that produced this file is not published. What is published is the result, its provenance, and its hash.
- Not custom settings. The result covers exactly the preset above.
The site's full methodology, limitations, yearly drawdown, per-cycle MAE, capital utilization, ladder-depth distribution, and fee/funding variants are at https://www.blitz-trading.com/backtest.
The data in this repository is released under
Creative Commons Attribution 4.0 International (CC BY 4.0),
the same license declared in the site's Dataset structured data.
Blitz Technologies Ltd. · https://www.blitz-trading.com