Price Methodology

This page explains, in full, how every number on GoldRatesUAE is produced, so you can
judge for yourself whether to trust it, rather than take it on faith. If anything here
doesn’t match what you see on a rate page, that’s a bug worth
reporting, not an inconsistency you should assume is intentional.

Two separate numbers, never mixed

We track two distinct series and label them differently everywhere they appear, because
conflating them is one of the most common ways a gold-price site misleads people without
technically lying:

• Reference rate, the number shown in our main tables and used in
  calculators. Derived from the international gold spot price, updated on the source’s
  cadence, and stored with a source timestamp.
• Live spot-derived rate, a faster-moving, explicitly-labelled market
  tracker for people who want to watch the spot price itself move. It is never presented
  as an “official” retail rate, and it is never quietly substituted into a page
  title or a calculator as if it were the reference figure.

The conversion formula, step by step

We convert the international XAU/USD spot price to an AED-per-gram figure like this:

pure_aed_per_gram = spot_usd_per_oz × 3.6725 ÷ 31.1035
24K AED/g = pure_aed_per_gram × 0.999 (display purity, not a theoretical 100%)
22K AED/g = pure_aed_per_gram × 0.9167
21K AED/g = pure_aed_per_gram × 0.875
18K AED/g = pure_aed_per_gram × 0.75
 

Worked example: say spot gold trades at $4,300 an ounce. The pure AED/g value is 4300 ×
3.6725 ÷ 31.1035, which rounds to about 507.9. The displayed 24K rate is that
figure × 0.999, roughly 507.4 AED/g. It is not the raw 507.9, because 24K jewellery-grade
gold is quoted at 99.9% fineness rather than a theoretical 100%. 22K, at 91.7%
purity, works out to about 465.4 AED/g from the same base. That is why our 24K
figure can look slightly different from a naive spot conversion you do
yourself. The 0.999 adjustment is deliberate, not an error.

We use 3.6725 for the AED/USD conversion. That is the dirham’s
long-standing peg to the US dollar, not a floating market rate. The peg does
not move day to day, so a live FX feed would add noise without adding
accuracy. A troy ounce is defined as 31.1035 g and a tola as 11.6638
g. Both are held as exact constants internally. We never round mid-calculation, only
when displaying a final figure to two decimal places.

Where the spot price comes from today

At this stage the site’s spot feed is an interim source while a contracted
metals-price provider is set up, see the data sources page
for exactly what is active right now, its authority level, and its limitations. We’d
rather tell you plainly that a source is a temporary bridge than dress it up as something
more authoritative than it is. When the source fails or returns an implausible value, we
keep the last verified figure, mark it stale, and never show a “live” badge on
data we can’t currently verify.

Validation before anything is published

A new price point is rejected, and the previous verified figure kept, if any of the following is true:

• The price is missing, zero, or negative.
• The timestamp moves backward relative to the last accepted reading.
• The move exceeds a sanity threshold (currently 8%) versus the last accepted reading and
  cannot be confirmed from a second source. A genuine 8%+ intraday move in gold is rare but
  not impossible; when it happens, we’d rather serve slightly stale data for a few
  minutes than risk publishing a bad data point from a source glitch.

Every accepted or rejected point is stored, not just the accepted ones, so if a source
starts misbehaving, the pattern is visible in our own records rather than only in
complaints.

Freshness, not one blanket “last updated” timestamp

We deliberately don’t use a single “last updated” label for everything on
a page, because a rate table, a live-spot widget, and an editorial guide go stale on
completely different schedules. Instead we track a reference-price timestamp, a
separate live-spot timestamp, and a content-review date for editorial text, each shown
next to the thing it actually describes.

City retail premium, never a fabricated city discount

A city page’s local retail premium index is only published once we have at least 12
verified retailer quotes from at least 5 distinct retailers, with no single retailer
contributing more than 35% of the sample, all collected within the last 45 days. Below
that threshold, the page says so plainly instead of showing a number. See
Dubai’s rate page for this in practice, right now
it shows the “not enough data yet” state, which is exactly what should happen
for a newly built site rather than a suspiciously precise number with no real evidence
behind it.

Found something that looks wrong? See our accuracy log
or get in contact, reports like that are how the log gets written.