XAU/USD Technical Analysis: How Gold Is Analysed on This Site
The full method behind the gold analysis on this site — daily, 4-hour and 1-hour timeframes, pivots, RSI(14), EMA 20/50/200 and ATR(14) — with the Friday 6 November 2026 session as the worked example.
The gold analysis on this site is a daily and hourly read of verified candles — not a prediction and not a trading signal. It combines four ingredients: daily and weekly pivot points, RSI(14) on three timeframes, EMA 20/50/200 on daily closes, and ATR(14) for the size of a normal day. This guide explains each ingredient, how they are read together, and what the analysis deliberately does not do, using the Friday 6 November 2026 session as the worked example.
The four ingredients
- Daily and weekly pivots (Classic method primary; Fibonacci and Camarilla kept strictly separate). The pivots in force for a session are calculated from the previous completed session's OHLC, and are labelled by the session they were calculated from.
- RSI(14) by Wilder, computed separately from the 1-hour, 4-hour and daily closes — each from its own verified series of closes, never mixed.
- EMA 20, EMA 50 and EMA 200 on daily closes: the short, medium and long trend. Price above an EMA is above that trend; price below it is below it.
- ATR(14) on daily closes: the size of a normal day's move. It sets expectations for how far price can travel, and for how much a level break is worth.
- Session high and low, always named with their provider and price basis, because the BID candle low and the MID quote low are different numbers.
Timeframes: daily sets the bias, 4-hour the structure, 1-hour the detail
The daily timeframe sets the bias: where the close sits against the daily Pivot, the daily RSI and the three EMAs. The 4-hour timeframe shows the structure: a sequence of higher highs and higher lows is bullish structure, lower highs and lower lows is bearish, a mix is mixed. The 1-hour timeframe shows the latest moves and the level tests. A move on the 1-hour chart that contradicts the daily reading is a counter-trend move until a daily close confirms it.
Worked example: Friday 6 November 2026
Friday's session (Finnhub/OANDA, MID): open $4,087.84, high $4,107.085, low $4,029.83, close $4,070.59 — down 0.43% on Thursday. It was the Nonfarm Payrolls day: after the 18:30 PKT release gold dived below $4,100, broke S1 ($4,048.43) on a verified 1-hour close at 19:00 — the first 1-hour close below S1 after seven closes above it — then reclaimed S1 on the 20:00 close. The session closed above the new Saturday Pivot of $4,069.17 by $1.42.
| Indicator | Value | Reading |
|---|---|---|
| Daily RSI(14) | 48.30 | Below 50 — weak-neutral |
| EMA 20 (daily) | $4,065.93 | Close above — short trend intact |
| EMA 50 (daily) | $4,154.38 | Close below — medium trend weak |
| EMA 200 (daily) | $4,363.56 | Close below — long trend weak |
| ATR(14) (daily) | $83.94 | A normal day moves about $84 |
| RSI(14) 1-hour | 50.60 | Crossed above 50 on the final bullish candle |
| RSI(14) 4-hour | 53.68 | Above 50 — medium-term momentum firming |
| 4-hour structure | Bullish | Higher high + higher low on the 21:00–01:00 candle |
Reading the levels together
- The close above the Pivot and the bullish 4-hour structure are real, but the daily RSI is still below 50 and the close is below EMA 50 and EMA 200. The bounce is real; the daily trend is still weak.
- The verdict at the 03:30 daily report was that the bearish thesis was weakening, not invalidated — and conviction moved only 5 points from Bearish to Neutral. A single session does not overturn a daily trend.
- The S1 break at 19:00 and the reclaim at 20:00 show why a break needs a completed candle close: the spot price spent the hour below S1, but the level held on the close.
- NFP moved gold through several levels in minutes. On event days, levels break fast and the usual reading of a break is worth less.
What the analysis does not do
- It gives no buy or sell signal. Scenarios and their entry, invalidation and target levels are reference levels only, and they need price-action and timeframe confirmation.
- Its conviction percentages are technical conviction and decision preference — not statistically verified probabilities of profit or loss.
- A level touch is not a break, and a spot move is not a candle close. Only verified completed candles confirm anything.
- When data is missing — a 1-hour or 4-hour candle, an RSI value, volume — confidence is lowered and the physical-holdings allocation is kept, never increased in favour of selling.
Data sources: 1-hour and 4-hour candles from Dukascopy (BID); spot and daily OHLC from Finnhub/OANDA (MID). Every value on this site carries its timestamp in PKT, its provider and its price basis, and a value that cannot be verified is shown as unavailable — never estimated from a spot snapshot.