← Home

How the market bias is calculatedtwenty-one checks, each one named

The card on the home page runs the list below against several hundred H1 candles, straight from our own MetaTrader terminal. Each check votes up, down, or neither. The percentages are the share of the decided checks pointing each way, which is why they always add to a hundred.

It is a count, not a forecast. If seventeen checks have decided and the card says 62%, then eleven of those seventeen point up — the undecided ones are not in the percentage at all. It says nothing about what price will do next, and there is no version of this list that would.

The newest candle is the hour currently forming, so the readings move during the hour and can change their mind before it closes. That is what makes the card live; it also means a reading taken two minutes apart is not the same reading.

It is also not order flow. Spot gold has no central exchange and no consolidated tape, so the share of the market that is buying is not a number anyone can publish — ours included. Any site showing you a buy-versus-sell split on a spot instrument has inferred it from price, exactly as this page does, and most of them do not say so.

The three headline readings

Trend

  • Bullish when the close is above the 200-period exponential average and the 50-period average is above the 200-period. Bearish when both are the other way round. Neutral whenever they disagree — so price can be above the 200 and the trend still read Neutral.

Signal

  • Taken from the vote only: Buy at 65% or more of the decided checks pointing up, Sell at 35% or less, Neutral in between. It also stays Neutral whenever fewer than eight checks have decided at all, because a majority of four is not a majority. It is a count, not a recommendation.

Volatility

  • ATR(14) — a simple average of true range, not Wilder's — against the median of its own last 100 readings. High at 1.25× or more, Low at 0.8× or less, Normal in between.

The twenty-one checks

Moving averages

  1. Price against its 10-period simple average
  2. Price against its 20-period simple average
  3. Price against its 50-period simple average
  4. Price against its 100-period simple average
  5. Price against its 200-period simple average
  6. Price against its 10-period exponential average
  7. Price against its 20-period exponential average
  8. Price against its 50-period exponential average
  9. Price against its 100-period exponential average
  10. Price against its 200-period exponential average

Averages against each other

  1. 50-period simple average against the 200-period
  2. 20-period exponential average against the 50-period

Oscillators

  1. RSI(14) against its midline of 50
  2. Stochastic %K(14, 3, 3) against its midline of 50
  3. CCI(20) against zero
  4. Williams %R(14) against its midline of −50
  5. MACD(12, 26, 9) line against its signal line
  6. Close against the close ten bars ago

Bands and channels

  1. Price against the middle Bollinger band (20, 2)
  2. Price against the Keltner centre line (EMA 20, ATR 10)
  3. Price against the midpoint of the 20-bar Donchian channel

The live column appears when the terminal is sending. Nothing is shown from a reading older than 15 minutes, and gold is closed from Friday evening to Sunday evening.

Why some checks are undecided

A reading sitting exactly on its own reference is not pointing anywhere, and counting it as a vote would make the percentage jump about on noise. So every check has a dead band, and anything inside it is counted as undecided and left out of the percentage.

Price comparisons count as undecided within 5% of the instrument's own ATR(14). The MACD check, which lives on a much smaller scale, uses 2% of ATR. The oscillators use a few points either side of their midline: 2 for RSI, 5 for the stochastic and Williams %R, and 20 for CCI.

One further reading is computed and not shown on the card, because three columns is all the card has room for: condition. ADX(14), Wilder's, at 25 or above is read as Trending, below 20 as Ranging, and in between as Mixed.

Trading carries risk and most retail accounts lose money. Nothing on this page or on the card it describes is financial advice or a recommendation to trade. The readings are arithmetic performed on recent prices.