This is the live gold signal board. Every XAUUSD call is published before the session it covers with an entry level, an invalidation level and a target, timestamped before the outcome is known, and kept on the public record whether it wins or loses.
A call is not a tip. It is a written statement of what the gold desk expects a specific session to do, in enough detail that it can be marked right or wrong afterwards without argument. Each one names the session it applies to, the direction, the range the desk expects the session to trade inside, the liquidity sweep it thinks comes first, the level that invalidates the idea, and one or two targets. Because all of that is fixed before the session opens, nothing can be quietly reinterpreted once price has moved.
The fields mean the following. Direction is the session bias, long or short. Entry, or the entry zone, is the area where the idea becomes actionable rather than a price you must pay. Invalidation, shown as the stop level, is the price that says the read was wrong; if it trades, the call is closed as a loss on the record. Target one is the first objective, usually the nearer edge of the expected range, and target two is the extension if the session keeps going. Expected range is the high-to-low span the desk thinks the session produces, which is what tells you whether a target is realistic at all. Session names the UTC window. Conviction is a confidence label, not a probability of profit.
Conviction on each gold call is anchored to the broker-verified gold win rate across a disclosed trade count, shown separately from the live forward test so the two are never conflated. No profit figures are published anywhere on the public record. Win rate on a stated sample size is the only performance claim made here, and the sample size is always shown next to it so a small run of results cannot be mistaken for an established edge.
Scoring happens after the session closes, not while it is open. The desk records whether the direction was correct, whether the invalidation level traded first, and which targets were reached, then adds the result to the running record. Direction hit rate is the share of scored sessions where the stated bias matched what the session actually did. It is deliberately separate from target hit rate, because a call can read the direction correctly and still miss an ambitious target, and blending the two would flatter the record. Open calls are shown as pending until they are scored, and amendments are logged rather than overwritten.
Session context is the reason the same call performs differently for two traders. Gold behaves like a different instrument in each window. The Asian window, roughly 23:00 to 07:00 UTC, is thin and range-bound and often builds the high and low that London later sweeps. The London open, 07:00 to 09:00 UTC, expands the range and frequently takes out one side of the Asian range before a direction holds. The London and New York overlap, 13:00 to 17:00 UTC, is the deepest liquidity of the gold day and produces the largest directional legs. The New York afternoon, 17:00 to 21:00 UTC, tends to fade momentum into the close. Rollover, 21:00 to 22:00 UTC, is a cost window where spreads widen and swap is applied, and no new gold risk belongs there. A call written for the overlap simply does not have the available range to work in the Asian session, which is why every call states its window. The gold session clock explains each window in detail.
Macro releases sit on top of that. At 13:30 UTC the United States publishes CPI, PPI, retail sales and Non-Farm Payrolls, and gold reprices instantly on those numbers, with slippage in the seconds around the print capable of exceeding the risk a trader believed was on the ticket. FOMC decisions and Federal Reserve speeches move the metal the same way. Calls state when a release falls inside their session so the trader can decide whether to stand aside, and the gold macro pulse and the gold economic calendar list what is due.
What a call does not decide is as important as what it does. It does not set your position size, and it cannot, because size depends on your account balance, your risk per trade and your broker leverage. It does not choose your stop distance beyond naming the level that invalidates the idea. It does not decide whether the trade suits you at all: a call in a window you cannot watch is information, not an instruction. Signals are research. Position sizing, stop placement and execution remain with the trader, and leveraged gold carries a high risk of rapid loss.
Instruments are gold only. XAUUSD is the primary market, where a standard lot is 100 troy ounces so a 1.00 dollar move per ounce is 100 dollars per lot, and 0.10 lots is 10 dollars per dollar of movement. XAUEUR, XAUGBP and XAUAUD are also covered, and they can diverge from XAUUSD when the dollar itself is doing the moving, which is worth checking before assuming a dollar-denominated call transfers. Every time on the board is UTC.
The honest way to use the board is to record what you actually did with it. Take the calls in one window, log each one in the free gold journal with the session and the reason you took or skipped it, and read your own win rate after a meaningful sample rather than after a good week. That comparison, your record against the published record, is worth more than any single call, and it is the reason the journal, the calculators and these signals are free.
GoldHunts is an alternative trading interface for gold that works with a supported MetaTrader 5 broker account. Orders are routed to that broker account through the MetaAPI connection, so the trader keeps their existing broker and funds. It covers gold instruments only, and live order execution requires a GoldHunts Live subscription. Start free: the signal board, the journal, the calculators and demo trading stay free. GoldHunts Live is 7.99 pounds a month or 49.99 pounds a year, about 4.17 pounds a month, with a 7-day trial. Nothing here is a promise of returns, there is no autonomous trading, and no order is placed without the trader.
A worked example shows why sizing has to stay with the trader. Suppose a call gives an entry near 2,410 dollars, invalidation at 2,398 dollars and a first target at 2,434 dollars. The risk is 12 dollars per ounce and the first objective is 24 dollars per ounce, so the idea is offering roughly two to one before costs. On a standard 100-ounce lot that 12 dollar stop is 1,200 dollars of risk; at 0.10 lots it is 120 dollars, and at 0.01 lots it is 12 dollars. Which of those is appropriate depends entirely on your balance and the share of it you are prepared to lose on one gold trade, and no published call can know that. The same call is prudent for one account and reckless for another purely because of the lot size chosen, which is why the board publishes levels and never lot sizes.
Costs belong in that arithmetic too. The gold spread widens around the 21:00 to 22:00 UTC rollover and around data releases, the overnight swap is applied on positions carried through rollover, and slippage on a market order in the seconds after a print can be several dollars an ounce. A call that looks like two to one on the levels can end up closer to one and a half to one once the spread paid on entry and exit, the swap on a multi-day hold and realistic slippage are included. Every call is written against the levels the desk expects to see, not against your broker's execution, so checking the numbers on your own account before deciding is part of using the board properly rather than an optional extra.
Amendments deserve a plain explanation, because they are where most public records quietly go wrong. If the desk changes a level after publication, the change is logged with a timestamp alongside the original, and the call is still scored against what was published. A call whose invalidation trades is recorded as a loss even where price later reverses to the target, because the level said the read was wrong at that moment. Calls that never reach their entry are recorded as untriggered rather than as wins, so the direction hit rate is not inflated by ideas that were never actionable. Reading the record with that in mind is the difference between judging a method and being reassured by one.
GoldHunts — the AI-powered gold trading operating system.