Hypothesis-First
Every strategy begins with a falsifiable thesis. No curve-fitting. No data-mining without a prior economic rationale.
PILLAR 02
Where insight meets
capital discipline
The lab exists to hold intelligence to the strictest standard available: capital reality. Every strategy that leaves it carries a walk-forward validation record, not merely a backtest; findings are published as they stand, null results included. The statistical tests are pre-specified, and the conclusions are allowed to disappoint.
LAB MANDATE
Research first —
Demonstration over persuasion
The lab develops, tests, and validates systematic strategies under rigorous, pre-specified conditions, in partnership with AlgoX Intelligence. Its work is presented the way serious research should be: demonstrated, documented, and open to scrutiny. Qualified institutions and serious counterparties will find the record speaks in the register they expect.
RESEARCH METHODOLOGY
How every strategy is built
01
02
Walk-Forward Validation
In-sample optimisation followed by out-of-sample walk-forward testing. Survivors are published; the rest are documented as null results.
03
Regime Awareness
A strategy that ignores market regime is a strategy waiting to fail. All HKM models are regime-conditional — trained on labelled states, validated across transitions.
04
No P-Hacking
We do not iterate parameters until significance is achieved. Statistical tests are pre-specified. Results are reported as found, not as selected.
05
Null Results Published
Firms that only publish winners are selling survivorship bias. We publish what did not work and why. This is how intellectual credibility compounds.
06
Alternative Data Integration
Sentiment and dominant-emotion scores, prediction-market probability signals, and social volume dynamics feed strategy inputs alongside price. Signal diversity reduces fragility.
LAB STANDARDS
What we refuse to do
These are not preferences. They are structural commitments that define the quality of every output the lab produces.
RULE 01
No in-sample optimisation alone
A strategy that looks good only in the period it was trained on is not a strategy — it is an artefact. Every model exits with a walk-forward track record, not a backtest.
RULE 02
No overclaiming
We do not promise returns. We do not show cherry-picked equity curves. We show methodology, validation logic, and where the strategy's assumptions could break.
RULE 03
No regime-blind deployment
A momentum strategy deployed in a mean-reversion regime will lose. Period. All HKM strategies carry regime conditions — if the regime is wrong, the strategy sits out.
R&D PARTNER
AlgoX Intelligence
Industry-leading investment trading programme developer — our authorised R&D and demonstration partner for systematic strategy development, live demo outputs, and institutional showcase materials. All strategies demonstrated under the partnership pass HKM's internal validation standards before any public showing.
ALTERNATIVE DATA
Prediction-Market Integration
Polymarket prediction-market data is used within the lab as a macro risk proxy and money-flow signal — not as a speculative instrument. Probability price movements are treated as informed-money sentiment, cross-referenced against HKM's social signal corpus. This is proprietary intelligence input. It is not a bet.
LIVE VERIFIED RECORD
The record, third-party verified
Portfolio WC_HKM streams live to Myfxbook, an independent third-party verification platform — equity, drawdown, and open positions, updated as they happen. Historical performance is a record, not a promise; nothing here is investment advice.
OPEN TRADES · LIVE
A strategy is an argument about how markets behave. The lab's job is to state that argument precisely, test it out of sample, and report what survived