https://quantorsaas.app/blog Long-form on signed builds, regime-aware trading, and self-custody. Engineering, not marketing. en Sun, 19 Jul 2026 18:00:00 +0000 https://quantorsaas.app/blog/what-a-signature-doesnt-prove https://quantorsaas.app/blog/what-a-signature-doesnt-prove Fri, 28 Aug 2026 09:00:00 +0000 A bad one-minute market-data print both tripped a stop and became the price the exit was booked at, putting a loss the market never delivered inside a performance record signed with Ed25519. The signature stayed valid throughout: it proves the published bytes were not altered, not that they were computed from sound inputs. Full numbers, the free reproducible check against public exchange bars that found it, why the signed artifact was not re-signed, and what the fix still does not cover. Methodology Signed performance https://quantorsaas.app/blog/we-preregistered-a-strategy-test-it-failed https://quantorsaas.app/blog/we-preregistered-a-strategy-test-it-failed Sun, 19 Jul 2026 18:00:00 +0000 We designed a mean-reversion strategy, committed the pass/fail rules to git before running the test, ran it exactly once on ~9 years of real Binance data — and it failed: 39-43% win rate versus the assumed 65%, 45-56% of trades ending at the disaster stop versus the pre-declared 20-25% kill line. Full numbers published as-is, plus a funding-rate carry study (a real premium, currently too thin to build for). Pre-registration, one-shot testing, and published failures — the discipline behind Quantor's signed performance records. Methodology Pre-registration https://quantorsaas.app/blog/we-ran-264-walk-forwards https://quantorsaas.app/blog/we-ran-264-walk-forwards Tue, 24 Jun 2026 15:00:00 +0000 We ran 264 out-of-sample walk-forward tests on our own crypto strategies — 6 strategies, 11 major pairs, 2 timeframes, 2 cost levels — on real public Binance data, after fees and slippage, benchmarked against buy-and-hold. Zero showed a credible return edge that beats holding; 108 reduced drawdown. Risk reduction, not alpha — and the result is reproducible with one command on free public data. Methodology Walk-forward https://quantorsaas.app/blog/lower-drawdown-vs-higher-returns https://quantorsaas.app/blog/lower-drawdown-vs-higher-returns Sun, 22 Jun 2026 15:00:00 +0000 Returns get the headlines; drawdown decides whether you survive. The asymmetric recovery math (a 50% loss needs a 100% gain to break even, 80% needs 400%), why path matters as much as average return, and what Quantor's own walk-forward honestly found across strategies and timeframes after costs: no durable return edge over buy-and-hold, but lower drawdown in 108 of 264 tested configurations (partly the mechanical effect of holding cash part of the time). Risk reduction, not alpha — and we don't promise returns. Risk Drawdown https://quantorsaas.app/blog/are-crypto-trading-bots-a-scam https://quantorsaas.app/blog/are-crypto-trading-bots-a-scam Sun, 22 Jun 2026 14:00:00 +0000 Some crypto trading bots are outright scams, most overpromise, and a few are honest tools. The four red flags (guaranteed yield, custody of your funds, unverifiable track records, black-box claims), whether bots actually beat buy-and-hold after costs (our own walk-forward: lower drawdown, not extra profit), and the checklist for an honest bot — applied to Quantor itself. Industry Crypto bot scam https://quantorsaas.app/blog/crypto-market-manipulation https://quantorsaas.app/blog/crypto-market-manipulation Sun, 22 Jun 2026 13:00:00 +0000 The six most common forms of crypto market manipulation — pump-and-dump, spoofing, stop-hunting, wash trading, insider/early information, and leverage liquidation cascades — the observable tell for each, and the only sane thing an automated system can do about them: detect a dirty tape (volume spikes, violent wicks, illiquid bars, wide spreads, lopsided order-book walls) and refuse to open a new position into it. Quantor trades spot only, no leverage, on a liquid-majors allowlist. Risk Market manipulation https://quantorsaas.app/blog/verify-trading-bot-track-record https://quantorsaas.app/blog/verify-trading-bot-track-record Sun, 22 Jun 2026 12:00:00 +0000 Almost every crypto trading bot advertises a track record; almost none can be verified. The four tests a real, auditable record must pass — cryptographically signed and dated, live rather than backtest, every trade rather than a flattering window, and re-checkable by you — a red-flag checklist for unverifiable claims, and the exact Ed25519 mechanism Quantor uses to publish results you can check with openssl. Plus the honest caveat: verifiable is not the same as profitable. Transparency Verifiable performance https://quantorsaas.app/blog/transparency-caught-our-own-bots https://quantorsaas.app/blog/transparency-caught-our-own-bots Mon, 09 Jun 2026 18:00:00 +0000 Two of our four signed-performance bots were running on the 1-minute timeframe, where round-trip fees and minute-bar noise dominate the signal. Our own transparency machinery — signed monthly snapshots plus a per-decision journal — surfaced it. The fix (M1 to H1), the honest data, and why we published the messy transition month with a dated changelog instead of restarting the track on a convenient date. Transparency Signed performance https://quantorsaas.app/blog/paper-mode-trading-guide https://quantorsaas.app/blog/paper-mode-trading-guide Fri, 29 May 2026 10:00:00 +0000 What PAPER simulates faithfully (strategy code, market data, regime classifier, 14 risk gates, decision tape, audit chain), what it deliberately approximates (slippage as fixed haircut, instantaneous fills, no partial fills, no exchange downtime, no funding-rate impact), how to read a PAPER P&L honestly (trade count, max drawdown, profit factor, Sharpe, edge vs buy-and-hold), and the seven-item PAPER-to-LIVE checklist. Trading Strategy validation https://quantorsaas.app/blog/why-i-built-quantor https://quantorsaas.app/blog/why-i-built-quantor Tue, 19 May 2026 14:00:00 +0000 A personal post from the founder. The $4,200 SOL FOMO trade at 2 AM, the pattern (strategy good, operator broken), what I tried first that didn't work (discipline, scheduled trade windows, third-party bots, Pine Script), and what Quantor actually is in two sentences. Plus the disclaimers — Quantor doesn't make money for you, PAPER is not LIVE, crypto can lose 80%. Founder story Trading psychology https://quantorsaas.app/blog/no-yield-promise https://quantorsaas.app/blog/no-yield-promise Sun, 11 May 2026 20:30:00 +0000 Most crypto bot landing pages lead with a yield number. The compounding math makes those claims physically implausible; the regulatory landscape makes them illegal where it applies; the historical pattern of how the yield-promising bots end shows the failure mode. Walks the math, the regulation, the three terminal states, and a five-question buying checklist for any bot — including ours. Industry Trading bot scam https://quantorsaas.app/blog/14-risk-gates https://quantorsaas.app/blog/14-risk-gates Sun, 11 May 2026 20:00:00 +0000 Before a single real order leaves the api, fourteen independent checks fire — symbol allowlist, regime classifier, kill switch, beta allowlist, notional sanity, subscription, bot count, plan permissions, Telegram link, strategy compatibility, exchange-key verification, and the final risk evaluator. Walks each gate with the exact reason code the api emits. Risk Live trading safety https://quantorsaas.app/blog/self-custody-by-construction https://quantorsaas.app/blog/self-custody-by-construction Sun, 11 May 2026 19:30:00 +0000 Quantor never touches your funds, by construction. The exchange API key has trade-only scope, the database stores only an AES-GCM blob, the master key lives in a separately-administered Secret Manager, and the exchange itself refuses withdrawal. Even if the entire company disappeared tomorrow, your money sits in your exchange account untouched. Security Self-custody https://quantorsaas.app/blog/regime-detector https://quantorsaas.app/blog/regime-detector Sun, 11 May 2026 19:00:00 +0000 A 50-line classifier — realized daily volatility plus drawdown from peak — decides when Quantor refuses to start LIVE bots. Walks through the algorithm, the four thresholds, the historical windows (2020-03-12, Luna, FTX, 2024 yen unwind) where it would have fired, and the upside we accept giving up in exchange. Risk Market regime https://quantorsaas.app/blog/signed-builds-ed25519 https://quantorsaas.app/blog/signed-builds-ed25519 Sun, 11 May 2026 15:30:00 +0000 Most crypto-bot SaaS ask you to trust them. Quantor signs every production deploy with an offline Ed25519 key and publishes the public key. Walks through the exact mechanism, threat model, and three verification recipes. Security Ed25519