The peak of AI capital expenditure is the day when gold makes its comeback.
```
This round of gold adjustment cannot be explained solely by "whether real interest rates are high." More crucially, there are two lines: one is dollar credit, corresponding to U.S. fiscal pressure and central bank allocation demand; the other is financial attributes, corresponding to global liquidity, dollar exchange rate, and risk-aversion trades. The current problem is that neither line is standing on the side of gold.
Zhejiang Securities financial engineering analyst Chen Aolin and others gave their core judgment in a July 1 research note: "Monetary attributes and financial attributes are the two most important factors driving gold prices, and both are currently facing headwinds, so the trend of gold adjustment may continue in the short term."
This framework incorporates AI capital expenditure into gold pricing. The logic is not complicated: if tech giants' capital expenditure supports the U.S. economy, the U.S. government doesn't need to rely on larger fiscal deficits to sustain growth, concerns about U.S. fiscal sustainability will ease, and the mid-term allocation logic for gold is weakened. Meanwhile, a strong economy and sticky inflation compress the Fed's room to cut rates, or even reinforce expectations of rate hikes, putting pressure on gold's financial attributes.
Therefore, for gold to regain a phase of opportunity, geopolitical risks are not necessarily the first to watch, nor is it only about the fall in real U.S. Treasury yields. A more immediate observation window may be the U.S. earnings season in July-August: storage and other segments have already seen substantial price hikes; if AI capital expenditure plans are not revised up in tandem, it means actual construction scale is declining, which could be a signal for gold to stabilize.
Gold Price Is Not Only Priced by Real Interest Rates, 2023 Already Gave a Lesson
In recent years, the place where gold is most easily misjudged is the market often compresses it into "the inverse function of real interest rates." As real rates rise, gold should fall; as real rates fall, gold should rise. This explanation works in many phases, but is incomplete.
2023 is a counter-example. The Fed was still in the tightening cycle, U.S. real rates were high, and by traditional frameworks gold shouldn't have been strong. But gold prices clearly diverged from real interest rates, driven by U.S. fiscal policy: during the rate hike cycle, the U.S. government still pursued expansionary fiscal policy, debt servicing pressure increased, dollar credit was damaged, and gold gained allocation strength.
Mid-to-late 2025 is another mismatch. The U.S. fiscal deficit shrank, central bank gold purchases also fell, and allocation logic weakened on the margin. But with warming Fed rate cut expectations and major inflows to gold ETFs, trading capital took over dominance, gold prices continued to rise. That is, gold can be pushed up by "dollar credit" or by "liquidity trades."
What to Watch for Medium-Long Term Allocation: How Long Fiscal Can Hold, Whether Central Banks Still Trust the Dollar
Zhejiang Securities' research framework clearly divides the factors affecting gold price into two categories: one is for medium-long term allocation logic, and the other is for short-term trading sentiment.
The core of allocation logic is the U.S. government's ability to pay back and global central banks' choices. Two key indicators:
U.S. Fiscal Pressure: The larger the fiscal deficit, the higher the ratio of government interest expenses to income; the more the market worries the U.S. cannot repay its debt, the more dollar credit is damaged, and the more prominent gold's allocation value becomes.
Global Central Banks' Attitude Toward the Dollar: If central banks everywhere reduce their dollar assets and increase gold holdings, they're voting with real money, showing distrust in the dollar system. Once this force becomes a trend, its support for gold prices is long-term.
This part explains medium-long term capital: it doesn't necessarily chase short-term moves, but once the direction forms, it often supports gold prices through lasting trends.
What to Watch for Short-Term Trading Capital: Dollar Direction, Safe-Haven Currencies, Global Liquidity Tightness
Gold's financial attributes are closer to the language of short-term trading capital. The core trading logic is whether the market is flush with cash and whether people are afraid. Three signals to watch:
First, safe-haven currencies (JPY, CHF). They share sheltering attributes with gold, and their moves reflect market risk sentiment better than the dollar index.
Second, global liquidity. Gold doesn't generate cash flow; in loose liquidity environments, willingness to buy gold increases. The liquidity factor is synthesized from global central bank policy expectations, global leverage ratios, and other indicators.
Lastly, financial stress. The OFR financial stress index is used to proxy global financial system stress. When the index is above 0, gold prices tend to rise more easily. This is not the whole of gold's long-term pricing, but under shock events, it brings pulse-like impacts.
As Long as One Line Sparks, Gold Isn't Bad; Both Lines Facing Headwinds Is Trouble
Backtesting results are clear: gold doesn't need both allocation and trading factors to improve at the same time. As long as one is activated, gold usually performs decently.
In the sample from January 2007 to June 2026, when both allocation and trading factors are above 0, gold's cumulative return is highest. With allocation factor positive and trading negative, monthly win rate is still 60.3%. With allocation negative and trading positive, the monthly win rate is higher, reaching 76.2%.
The truly unfavorable state is when both factors are below 0. This interval occurred for a total of 62 months, with average annualized monthly return of -8.6% and win rate of only 41.9%, below the full sample win rate of 57.1%.
Based on this, the integrated timing strategy is very simple: as long as either allocation or trading factor is above 0, hold gold; if both are below 0, stay out. From January 2007 to June 2026, this strategy's annualized return is 13.1%, compared to London gold's annualized return of 9.9%. It avoids sustained drawdowns in 2013-2015 and 2022, and participates in several gold bull markets.
But boundaries must be clarified: looking at allocation or trading factor alone can capture part of the upside, but neither can beat the benchmark. The challenge of gold is that its pricing is often handed off by different types of capital at different stages.
Gold's Current Problem: AI Lets the U.S. Economy "Run Without Reliance on Fiscal"
Models show that from March 2026, both gold's allocation and trading factors turned negative. In other words, both monetary and financial attributes are under headwinds.
The key variable here is AI capital expenditure.
From 2020 to 2025, U.S. economic performance correlated obviously with fiscal deficits, with expansionary fiscal policy a major force in the economy's strength. But since 2026, the two diverged: the U.S. economy improved, yet fiscal deficits are shrinking. This means U.S. growth is increasingly driven by the private sector, rather than the government.
More specifically, domestic private investment has become the core factor driving U.S. GDP growth. Once tech giants' capex supports economic growth, the U.S. government no longer needs to rely on aggressive fiscal stimulus to maintain resilience. For gold, this undermines the logic of "unsustainable U.S. fiscal position—damage to dollar credit—rise in gold's allocation value."
On the other hand, AI capital expenditure supporting the economy also narrows the Fed's space to cut rates. Strong economy and stubborn inflation push up rate hike expectations, or at least suppress rate cut expectations, so gold (a zero-yield asset) is pressured.
Gold's Short-Term Opportunity May Wait for AI Capex Expectations to Peak
If what suppresses gold is AI capex, then stabilization signals for gold may also come from changes in AI capex expectations.
The near-term observation window is U.S. equity earnings season in July-August. Storage and other segments have seen large price hikes recently. If tech giants do not raise AI capex plans significantly, it means that with rising costs, the actual construction scale may drop.
Once the market starts to trade the "short-term peak in AI capex," the logic supporting the U.S. economy with private investment might be reassessed, and the Fed's space to cut rates could reopen. For gold, this may bring a short-term allocation opportunity.
Risk is, the conclusions here are from historical data and modeling. Historical backtests do not represent the future, parameter settings affect outcomes, and gold's price is volatile. Even if both factors turn negative, it does not mean gold prices will only fall, but at the very least, it means that a short-term reversal needs stronger catalysts.
Risk Warning and DisclaimerThe market has risks, investment requires caution. This article does not constitute personal investment advice, nor does it take into account individual users' special investment objectives, financial circumstances, or needs. Users should consider whether any opinions, views, or conclusions herein are suitable for their specific situation. Invest at your own risk. ```