FOR FUNDS AND ALLOCATORS

    GOLD RESEARCH FOR FUNDS AND ALLOCATORS

    Written for people who read a methodology section before a headline number. This page sets out how the gold record is constructed, where its weaknesses are, and what is available under licence.

    Gold Hunts is a gold-only research desk publishing broker-verified XAUUSD statistics — 17,791 trades and a 64.5% win rate since January 2024, win rates only, never P&L. Methodology, sample construction and known limitations are published in full, and the underlying data is available as JSON under a documented licence.

    What Gold Hunts is, precisely

    Gold Hunts is a single-instrument research and journaling desk covering gold: XAUUSD primarily, plus the other XAU crosses. It publishes session calls with a defined entry, invalidation level and target, tracks every one to outcome in public, and maintains a broker-verified trade record drawn from MetaTrader 5 history.

    It is not a fund, not a broker, not an introducing broker and not a signal vendor. It does not manage third-party capital, does not hold client money, does not take a rebate on volume traded, and does not sell access to the signals. That matters when assessing the record, because the usual incentive to flatter a track record — selling access to it — is absent, while the usual incentive to publish selectively is removed by publishing losses on the same page as wins.

    The commercial surface is a free gold trade journal that reads a connected MT5 account, with paid tiers for higher account counts and reporting. The research exists to earn attention; it is not the product being sold, which is why the desk can afford to publish the parts of the record that look bad.

    How the verified record is constructed

    The record is built from closed positions on connected MetaTrader 5 accounts, read directly from broker history rather than entered by hand. That removes the two failure modes that destroy most published track records: retrospective editing, and the quiet omission of trades that went badly. A trade appears in the record because the broker says it closed, not because anyone chose to include it.

    Every position is tagged with the session it opened in, using UTC windows fixed in advance rather than fitted after the fact, and with its direction. Aggregation is then a matter of counting: sample size and win rate per session, per direction, and overall. Win rate is closed winners divided by closed trades, with break-even closes counted as non-winners, which biases the figure down rather than up.

    Conviction scoring on the public signals is Wilson-weighted rather than raw. A 70% win rate on 40 trades is not treated as stronger evidence than 64% on 10,000; the interval width does the work. Sub-samples where the direction is weak are softened rather than dropped, because dropping them is how a record quietly becomes a curve fit.

    SliceSample sizeWin rate
    All sessions, all directions17,79164.5%
    New York, all10,33766.6%
    New York, long5,47268.5%
    London, all2,88962.2%
    London, long1,39863.2%
    London, short1,49161.2%
    Asian, all4,56561.1%
    Asian, long2,15165.4%
    Asian, short2,41457.3%
    Broker-verified XAUUSD closes since January 2024. Win rates only; no P&L is published.

    Read the shape rather than the headline. New York carries the largest sample and the highest win rate; the Asian session shows a clear long bias, with shorts eight points weaker on a comparable sample. That asymmetry is the interesting part of the record, and it is stable enough across sub-samples to be worth interrogating.

    What this record does not prove

    A win rate is not an edge. Without the distribution of winners and losers, a 64.5% hit rate is compatible with an excellent strategy and with a terrible one that closes winners early and lets losers run. Gold Hunts publishes win rates only, so the record on its own cannot establish profitability, and the desk does not claim it does.

    The sample also carries the usual survivorship and regime caveats. It begins in January 2024, which places most of it inside a period of unusually strong gold trend behaviour. Session tagging is by open time, so a position held across sessions is attributed to where it started. Execution costs vary by broker, so a win rate measured on one spread environment will not transfer cleanly to another.

    Stating this plainly is the point. An allocator who takes the win-rate table as a performance claim has misread it; an allocator who takes it as a well-sampled description of session asymmetry in gold has read it correctly.

    • No P&L, so no expectancy, Sharpe, drawdown or return figures can be derived from the published record.
    • Sample starts January 2024 — regime coverage is narrow by design, not by omission.
    • Session attribution is by open time, which understates cross-session holds.
    • Costs are broker-specific; the record reflects the accounts it was measured on.
    • Forward-tested signals are reported separately from the historical record and should not be pooled with it.

    Position on forecasting

    The desk does not forecast the gold price and does not publish targets as predictions. A published call is a conditional trade structure with an invalidation level. When the level breaks, the call is wrong and stays visible on the record as wrong.

    The macro research layer

    Alongside the trade record, Gold Hunts maintains a macro data layer aligning month-end gold closes to public series: the 10-year TIPS real yield, the broad trade-weighted dollar, 10-year breakeven inflation, the fed funds target and US CPI. Correlations are published with their sample sizes, and the aligned series are downloadable so results can be reproduced independently.

    The findings are deliberately unexciting where the data is unexciting. The month-end correlation between gold and the 10-year real yield is about -0.27 over the published sample: directionally consistent with the standard framing, far too weak to support the confident inverse relationship that gets asserted in commentary. Publishing the weak number is more useful than publishing a stronger one derived from a shorter window.

    Month-end XAUUSD close against the 10-year TIPS real yield (FRED: DFII10).
    Macro seriesMonthsCorrelation with XAUUSD change
    10y real yield (DFII10)140-0.272
    Broad dollar index (DTWEXBGS)140-0.226
    10y breakeven inflation (T10YIE)140-0.171
    10y nominal yield (DGS10)140-0.255
    CPI index (CPIAUCSL)1390.029

    Published correlations with sample sizes.

    Methodology and study pages

    Data access, API and licensing

    All published datasets are free to use, including commercially, with visible attribution to Gold Hunts. The JSON endpoints are stable, documented in an OpenAPI 3.1 specification, and refreshed on a schedule rather than on request. For most research use, that is sufficient and no agreement is needed.

    Where a fund needs more than the public layer — higher refresh frequency, historical depth beyond what is published, tick-level detail, a bilateral licence with warranties, or a redistribution arrangement inside a commercial product — that is a conversation, not a self-serve tier. Email info@goldhunts.com with the specific series, frequency and use case.

    TierWhat it coversCost
    Public data layerAll published CSV and JSON endpoints, documented schema, embeddable chartsFree with attribution
    Attribution-free reproductionSame data, attribution waived inside internal researchBy agreement
    Extended history and frequencyDeeper history, higher refresh cadence, additional seriesBy agreement
    Redistribution licenceEmbedding Gold Hunts data inside a commercial product or terminalBy agreement

    WHERE TO GO NEXT

    Diligence checklist

    If you are assessing Gold Hunts as a data source rather than as a manager, these are the questions worth asking, with the answers already on the site. Every one links to the page that documents it, so nothing here needs to be taken on the desk's word.

    What to verify before citing or licensing

    1. 01

      Is the trade record broker-sourced rather than self-reported?

      Pass test: Yes — read from MetaTrader 5 history on connected accounts.

    2. 02

      Are losing calls visible on the public record?

      Pass test: Yes — every published call is tracked to outcome, wins and losses both.

    3. 03

      Is the sample size published alongside every win rate?

      Pass test: Yes — per session and per direction, on /gold-signals and in /api/gold-record.json.

    4. 04

      Are macro correlations reproducible from published data?

      Pass test: Yes — aligned month-end series are downloadable as CSV.

    5. 05

      Is there a commercial incentive to inflate the record?

      Pass test: No signals are sold; the paid product is journaling software.

    6. 06

      Is P&L published?

      Pass test: No, deliberately. Do not derive return or risk metrics from the record.

    GOLD TRADING FAQ

    Is Gold Hunts a fund or an asset manager?
    No. It is a gold-only research and journaling desk. It does not manage third-party capital, hold client money, or act as a broker or introducing broker.
    Why publish win rates but not P&L?
    A win rate with a stated sample is verifiable and hard to dress up; published P&L invites performance marketing and implies a return claim the desk will not make. The trade-off is that profitability cannot be inferred from the record, which the methodology states explicitly.
    How is the gold trade record verified?
    It is read directly from MetaTrader 5 broker history on connected accounts rather than entered manually, so trades cannot be edited after the fact or quietly omitted.
    What is the correlation between gold and real yields?
    About -0.27 on month-end data over the published sample. That is directionally consistent with the standard inverse framing but far too weak to treat as a reliable trading relationship.
    Can a fund licence Gold Hunts data for internal research?
    The published datasets are already free for commercial use with visible attribution. Deeper history, higher refresh frequency, attribution waivers and redistribution inside a commercial product are handled by agreement — email info@goldhunts.com.

    TRACK YOUR OWN GOLD RESULTS FREE

    Live XAUUSD signals on a public ledger, a free gold trade journal that syncs with MetaTrader 5, and the macro pulse before the London open.