Acutis CapitalAcutis CapitalInvestment Research

Insights

Gold fair value models

1 Nov 2026The Acutis Digest

By Gareth David HoweAcutis Capital Founder

How does Gold’s price compare with the amount of money available across economies? One combined chart starts in January 1971. It uses the full saved Acutis Capital Global Liquidity series from 4 June 2007 onwards, and a five-region historical broad-money proxy with explicit estimates before that date.

Global Liquidity describes the availability of money and funding. Easier funding can support demand for assets including Gold, but it does not guarantee a higher Gold price. Interest rates, currency movements, central-bank demand and investor preferences can all change the relationship.

One combined history · Full Global Liquidity from 4 Jun 2007; historical M2/M3 and estimates before then · 1:1 response.

Gold price · USD/oz
$3,898.90
Liquidity model value
$2,534.68
Latest deviation
+1.43 SD
Saved through
30 Oct 2026
━ Gold price━ Global Liquidity model value┄ +1 SD┄ −1 SD

Gold · Global Liquidity fair value

30 Oct 2026

Gold price
$3,898.90
Model value
$2,534.68
Deviation
+1.43 SD
Versus model
53.8%

Full Global Liquidity · Daily Gold futures

Deviation from model · standard deviations

30 Oct 2026

Gold price
$3,898.90
Model value
$2,534.68
Deviation
+1.43 SD
Versus model
53.8%

Full Global Liquidity · Daily Gold futures

31 Oct 2006 – 30 Oct 2026 · 4,893 saved observations

1:1 liquidity response · December 2019 full-proxy anchor: $1,726.28 · Earlier money rebased at the June 2007 handover. Illustrative dollar level, not independently established fair value.

Historical data coverage · before June 2007

RegionRecorded money / FX overlap in source archive
USA1959-01 – 2026-09
China1998-12 – 2026-09
Euro area1980-01 – 2026-09 · pre-1999 FX estimated
Japan1980-01 – 2026-09
UK1986-12 – 2023-11

January 1971 is 67.3% estimated. Missing regional USD money levels are anchored to their nearest available record and scaled by the US M2 ratio. Estimates do not use Gold prices. Historical broad money is used only before the full Global Liquidity series begins.

Data definitions and sources
  • USA/China M2; euro-area/Japan/UK M3. Definitions and vintages differ.
  • Missing component USD levels scale with US M2 relative to the nearest available recorded anchor. No Gold-price inputs are used in money estimates.
  • Euro-area conversion before Jan 1999 is a DEM-based approximation, flagged estimated. Euro-area composition changes over time.
  • UK M3 ends Nov 2023 in this snapshot; subsequent UK values are estimated from US M2.
  • Official World Bank monthly Gold prices through Dec 2024; rounded repository continuation through Sep 2026. No futures splice.
  • Full-sample fit, historical revisions and future anchors: not point-in-time trading evidence.
January 1971 – October 2026. Before 4 June 2007: monthly average Gold and historical M2/M3 with estimates. From 4 June 2007: daily Gold futures and the full Global Liquidity proxy. The source handover is marked; monthly dates represent full-month averages, not closing prices.

The Global Liquidity model

The combined model responds 1:1 to liquidity: model value = December 2019 reference value × combined liquidity ÷ December 2019 liquidity. A 10% increase in the input produces a 10% increase in model value, without fitted dampening. The dollar reference is the previous full Global Liquidity fit’s first December 2019 observation; it is an illustrative anchor, not independently established fair value.

Before the full series begins, historical broad money is rescaled to match the full proxy’s first liquidity value on 4 June 2007. The historical level at that handover is linearly interpolated between the surrounding monthly observations. This aligns the dollar model at the join without changing either input’s percentage growth. It does not make the two liquidity definitions equivalent.

The extended input sums USA and China M2 with Euro area, Japan and UK M3, converted into US dollars. These broad-money definitions are not identical, and this series is not the live Global Liquidity gauge.

Recorded component values are retained wherever available. Missing regional dollar amounts are estimated as the nearest recorded regional dollar amount × US M2 at the target month ÷ US M2 at the anchor month. This assumes the regional-to-US money ratio remains constant across the missing interval. It does not use Gold’s price, but it cannot reconstruct actual historical regional money growth.

Before 1999 the Euro area’s currency conversion is approximated using Deutsche Mark exchange rates and the fixed euro conversion. Earlier unavailable regional balances are estimated from US M2; early 1971 is therefore predominantly estimated. The historical chart readings show the estimated share and which regions are recorded or estimated.

The chart preserves official monthly Gold prices before the handover and daily Gold futures afterwards. Its December 2019 anchor and residual standard deviation over the combined saved observations remain fixed when the displayed timeframe changes. The daily section contributes more observations than the monthly section; this is not an equal-month-weighted uncertainty measure.

Deviation and model bands

Deviation is the difference between ln(Gold price) and ln(model value), divided by the full-history standard deviation of those residuals, recalculated for the 1:1 model. A positive reading means Gold is above its model value; a negative reading means it is below. The dashed lines mark one standard deviation above and below the model, converted back to dollar prices.

These bands describe historical variation, not confidence intervals or buy and sell instructions. Changing the displayed dates does not refit the model or recalculate its standard deviation. The latest figures always refer to the latest saved observation.

What this research cannot establish

This is descriptive research, not a validated trading strategy. The 1:1 model assumes proportional growth and uses a dollar anchor derived from the previous full-sample fit. Its dollar level depends on that anchor. Neither this assumption nor a strong historical fit establishes causation, reliable future fair value or profitable entry dates. Both Gold and liquidity can trend over time, making an apparently strong relationship misleading.

The estimated early money history relies on later anchors and fixed regional ratios; uncertainty is not captured by the ±1 SD bands. Estimated European conversion, changing regional composition and the source handover add further uncertainty. Reporting delays, historical revisions, transaction costs and prior-only fits have not been reconstructed. Out-of-sample validation is required before assessing any strategy. From 2007, futures contract rolls can affect comparisons with spot Gold.

This article does not change the live Gold gauges, portfolio cash rules, holdings or model backtests. It is research, not personalised investment advice.