Gold

XAUUSD broker-feed differences: why two charts may disagree

Two gold charts can show different values without proving either is wrong. Match the source, price basis and completed bar before interpreting the discrepancy.

Two XAUUSD charts can disagree because they describe different price sources, quote types or observation windows. Before deciding one is wrong, match the full symbol and provider, the instrument, the candle interval and the price basis. Compare the same completed bar, then record any difference that remains. A similar chart label is not proof of identical data.

You can review current ChartPilot signup options once you have a clearly identified Gold screenshot to examine. New accounts require a payment method for a seven-day Premium trial, then renew at $59 USD/month unless canceled before the trial ends; applicable taxes appear at checkout. There is no ongoing free plan for new accounts. See the current terms.

Keep the provider attached to XAUUSD

“TradingView gold” is not a complete source description. For example, TradingView’s OANDA XAUUSD page identifies OANDA beside the symbol. Record that provider along with the ticker. If you selected a different source, keep its name rather than shortening both records to “gold.”

Also check what the provider offers. OANDA Australia’s XAU/USD instrument page describes its Gold CFD. That does not make every gold-related chart the same contract. If one image is a gold futures contract or an ETF, stop treating it as a direct comparison of two XAUUSD feeds. Instrument identity is a separate difference to resolve first.

The general XAUUSD review checklist covers screenshot preparation. Here the narrower question is whether two observations are comparable enough to investigate a price mismatch.

Check what the price represents

A feed is the source supplying the chart’s price data. Its construction matters. OANDA Australia’s pricing explanation says its forex and metals CFD prices use liquidity-provider inputs processed by its pricing system to generate a midpoint. This is evidence about OANDA’s stated method, not a description of every broker or proof of why your two charts differ.

The plotted series matters too. TradingView’s OANDA help page describes those charts as midpoint-based and distinguishes that chart price from the broker’s bid and ask. Do not generalize that setting to every charting app or source. Check the selected symbol’s documentation and available settings.

For a particular quote pair, the arithmetic midpoint is halfway between bid and ask. If one view shows bid and another midpoint, their difference can exist without a disagreement about the underlying quote pair. A screenshot with an unidentified price basis leaves that comparison unresolved.

Build a comparison record before judging the candles

Use one row per image in your notes. Fill in these fields from the platform, not from how familiar the chart looks:

  • Identity: full symbol, provider, instrument description and displayed price units.
  • Price basis: bid, ask, midpoint, last trade or unknown, according to the actual source.
  • Bar definition: candle interval, ordinary or transformed candles, and the date and start/end time of the bar being compared.
  • Clock: displayed timezone and the corresponding observation time. Match the underlying interval, not merely a printed “10:00” label.
  • State: completed or still forming, capture time and any visible delayed-data or connection notice.
  • Observation: the specific open, high, low or close that differs, including every readable decimal.

Start with a completed bar so a later update to an unfinished candle is not mistaken for a feed discrepancy. A current quote panel is a different observation from an earlier close. If the two images use different timeframes, use the gold timeframe comparison guide before comparing their individual candle shapes.

A fictional check that changes one variable at a time

Every provider name, price and timestamp in this example is invented. These are not actual screenshots, current gold prices or broker measurements. Assume ordinary 15-minute XAUUSD candles covering exactly 10:00–10:15 UTC on the same fictional date. Each bar is complete and uses the same price units.

RecordProviderPrice basisClose
A1Provider ABid2,360.20
A2Provider AMidpoint2,360.40
B1Provider BMidpoint2,360.55

Suppose Provider A’s corresponding closing quote is bid 2,360.20 and ask 2,360.60. Their midpoint is (2,360.20 + 2,360.60) ÷ 2 = 2,360.40. The 0.20 difference between A1 and A2 follows from comparing bid with midpoint in this invented example.

Now align both Provider A views to the same midpoint series, interval and completed bar. In this example, both show 2,360.40. That controlled comparison removes the price-basis difference. It does not promise that every real platform comparison will resolve this way.

A2 and B1 already share the stated interval and midpoint basis, but their closes differ by 0.15. The record establishes that difference; it does not reveal its cause. Investigate the providers’ data and instrument documentation instead of calling one quote false. Do not combine A’s high with B’s low to create a candle neither source supplied.

The arithmetic above applies to the specified closing quote. It does not mean you can reconstruct an entire midpoint candle by averaging the separate bid and ask highs: those extremes need not occur at the same instant.

Keep an unexplained mismatch visible

If matching the provider, symbol, price basis and interval does not resolve the discrepancy, retain both original images. Record the exact differing value and consult the platform’s data documentation or support with those details. Mark the cause unknown until you have evidence. Two screenshots alone cannot establish deliberate manipulation, an execution failure or which feed is more accurate.

For the ChartPilot review, upload one readable image and keep its source record beside it. Compare the returned BUY or SELL explanation and reference levels with that same image before checking a different broker’s chart. The product can help organize visible evidence; it cannot use the screenshot to certify a broker feed or recover a hidden quote basis.

Review current signup options for an educational second opinion on your prepared chart. Confidence describes visible evidence, not a win probability. Screenshot analysis is not a live quote feed, does not execute trades and can be wrong.

Prepared with AI assistance and checked against the linked product information and sources. Product information checked November 1, 2026. Read our editorial policy.

Educational chart analysis. Not financial advice. Trading involves risk. AI can misread charts or miss context. Screenshot assessments are not live prices and do not execute trades.

Do not sell or share my personal information

ChartPilot Support

A little help, right here.

Getting started · Plans · Account help
LET’S GET YOU SORTED

How can we help?

Ask about your chart analysis, plans or account. We’ll help you find the next step.

Enter to send · Shift + Enter for a new line0/1200