FAQ
Quick-answer page on Hedgicore, the Hedgicore Engine and how the platform works. For the full technical methodology, read The Hedgicore Engine Methodology Paper (v2.0, May 2026) on the methodology page.
Hedgicore is a stat-arb platform for crypto perpetual futures. The platform builds an adaptive spread between perps and lets you apply standard technical indicators (Z-score, MACD, Bollinger, Stochastic, Nadaraya-Watson) to trade it. Backtest as you tune. Pairs today, multi-leg portfolios on the roadmap. Proprietary Hedgicore indicators ship as a separate add-on.
Hedgicore is built by the team at Bonton AI. Bonton AI is the parent firm; Hedgicore is the retail and prosumer product. Team bios are on the /about page.
No. Hedgicore is an analytics platform. It does not execute trades, hold customer funds or provide financial advice. Users execute trades themselves on their chosen exchange based on the alerts and analytics the platform provides.
Three tiers for individual and team use (Free, Pro, Team) plus two add-ons coming later (the Hedgicore Indicators and the AI Strategy Builder). Plus two business paths (Hedgicore Business for custom limits and API access, Bonton AI for white-label and embedded). Pricing is published at launch and waitlist members get founding-member pricing locked.
Binance Futures at launch. Bybit and additional venues are on the roadmap.
The Hedgicore Engine is the modelling layer that produces the adaptive spread Hedgicore is built around. It handles the cointegration check, the regime drift in the hedge ratio and the funding awareness that crypto perpetuals require. The Engine is the part of Hedgicore that does not exist anywhere else. Internal estimators and parameter logic are not published. Design principles are in the methodology paper.
Cointegration (Engle and Granger 1987, Johansen 1991, Vidyamurthy 2004, Pole 2007, Krauss 2017), statistical arbitrage (Avellaneda and Lee 2010, Gatev, Goetzmann and Rouwenhorst 2006) and the foundational indicator constructs (Appel 1979 for MACD, Bollinger 2002 for Bollinger Bands, Lane 1950s for the Stochastic Oscillator, Nadaraya 1964 and Watson 1964 for kernel regression). Full reference list in the methodology paper.
Hedgicore ships the core standard indicators today: Z-score, MACD, Bollinger Bands, Stochastic Oscillator and Nadaraya-Watson. More indicator coverage is planned. The indicators work the same way they work on any charting platform; the difference is what they are computed on (the Hedgicore spread).
Five proprietary Hedgicore indicators calibrated specifically for the Hedgicore spread. Stretch is built on the Z-score, Flow on MACD, Envelope on Bollinger Bands, Pulse on the Stochastic Oscillator, Glide on Nadaraya-Watson. They ship as an add-on later this year.
The Hedgicore Engine that produces the adaptive spread is the part of the platform that does not exist anywhere else. The specific estimators, parameter logic and calibration methodology are not published. The indicators that run on top are standard technical indicators; they work the same way they work in every other tool.
The Engine's outputs are validated against the academic literature for cointegration and mean-reversion, against held-out historical data via walk-forward partitioning and against live forward-tested data on an internal test universe. Each Engine version is dated and documented.
Adaptive equilibrium for each pair (continuously updated), perpetual funding payments at the venue cadence, exchange fee schedule at the configured tier, realistic side-of-book fills (not mid-price), equal-USDT-notional position sizing per leg, take-profit and stop-loss evaluated bar-by-bar and walk-forward partitioning for parameter optimisation.
Tick-level intra-bar microstructure, liquidation cascades, listing and delisting events, order-book impact at large notionals and cross-venue arbitrage costs. The methodology paper §5 documents these omissions explicitly so users can size their expectations accordingly.
The backtest is built to be a defensible model of what a strategy would have done if traded live, not a perfect simulation. Real live performance will differ from backtest performance due to the modelled omissions above and to ordinary out-of-sample variance. Historical performance is not indicative of future results.
A single historical path is one draw from an unknown distribution (Lopez de Prado 2018). Reporting a single point estimate is misleading about the actual reliability of the strategy. The Hedgicore methodology paper roadmap §6 includes resampling and Monte Carlo diagnostics for the Pro tier so users can see a distribution of outcomes, not just one point.
Trading carries risk of loss. Hedgicore is an analytics platform that gives you better information; it does not eliminate market risk, leverage risk, liquidation risk, exchange counterparty risk or execution risk.
Yes. Hedgicore does not guarantee profitable strategies. The platform provides analytics and alerts; the trade decisions and the risk management are the user's.
Hedgicore is an analytics tool, not a broker or fund manager. It is not registered as a financial advisor. It does not execute trades or hold customer funds. Users are responsible for compliance with the regulations that apply to their own trading activity.
General-purpose charting tools display indicators applied to raw inputs and do not model the substrate problems specific to crypto perpetuals. Hedgicore is a purpose-built platform for pairs trading on crypto perpetuals. The indicators are calibrated against an adaptive equilibrium, not raw price. The backtest models funding and exchange fees explicitly. The Engine handles regime drift in the hedge ratio. For the details, see Z-score on a crypto pair is not what you think on the blog.
Read the methodology paper, then ask the same questions of any other platform you are evaluating. Documented methodology, named team, disclosed backtest scope, foundational citations, public indicator definitions and a prominent risk disclaimer are the criteria that matter. Hedgicore was built to meet each of them.
Verify against public criteria rather than taking anyone's word. The methodology paper, the named team, the disclosed backtest scope, the foundational citations, the public indicator definitions and the risk disclaimer on every page are all designed for verification. If any criterion is unclear, that is a documentation bug we want to know about.
Email and Discord channels are on the /about page. Technical questions about the methodology go to the technical team and get substantive responses.
The Hedgicore Engine Methodology Paper, v2.0, May 2026, available at hedgicore.com/methodology.
Three layers. (1) The blog at hedgicore.com/blog has five posts, from a plain-English intro to mid-level technical pieces on hedge ratios, backtest reading and half-life. (2) The methodology paper at hedgicore.com/methodology is the principle-level Engine description. (3) The research papers at hedgicore.com/research are the formal treatment with equations and full reference lists, both pairs-trading-implementation and statistical-arbitrage-technical-guide.
Want the full technical picture? The methodology page has the v2.0 paper. The blog expands each section into plain English.
