Every number on this site comes out of three JSON files, and here they are. Recompute our claims, grade the trades under your own rules, or plot the whole thing yourself. If our arithmetic is wrong we would rather you found it.
Check our headline claims without taking our word for them.
counts, window, both conventions, per-channel breakdown
Recompute the record yourself, or re-grade it under your own rules.
pair, side, leverage, entry, exit, stop, R, %, $ on $100, outcome, channel, date
See what is tracked and how stale each row is.
what we cover and when it was last re-measured
curl -s https://rebatemax.trading/data/v1/scoreboard.json | jq '.conventions'
{
"scaleout_r": 0.4145,
"holdexit_r": -0.0732,
"pct_loss_scaleout": 21.9,
"pct_loss_hold": 54.4
}import requests
r = requests.get("https://rebatemax.trading/data/v1/receipts.json").json()
losses = [t for t in r["receipts"] if (t.get("pnl_pct") or 0) < 0]
print(len(r["receipts"]), "graded,", len(losses), "closed at a loss")
# recompute our own headline from the raw rows
print(sum(t["usd_100"] for t in r["receipts"] if t.get("usd_100")))Free, no key. Static JSON over CDN — hammer it if you like.
Updated nightly, alongside the scoreboard and receipts pages. Every file carries its own generated_at.
Attribution appreciated — link rebatemax.trading if you publish anything built on it.
Not investment advice, and not a forecast. It is a historical record of third-party signals, measured — not endorsed.
Want a paper-trading API to point a bot at instead? That needs accounts and keys, so it is not built yet — our paper terminal and backtester run in your browser with no account at all. Tell us if you would use a programmatic version and it moves up the list.