Rate Hikes Are Validating Gold’s Long-Term Thesis: A Debt and Interest-Cost Stress Test of Five Countries
In the short term, rate hikes weigh on gold. In the long term, rate hikes are themselves the reason we’re bullish on gold.
The Fed raised rates by 25bp on September 16, its first hike since July 2023, and the market prices roughly a 70% chance of another hike at the October 27–28 meeting. That same week, the 10-year TIPS real yield rose to 2.79%, an 18-year high. Spot gold has fallen 24% from its record high of $5,589 on January 28. Shanghai gold (Au99.99) dropped from 968 yuan/gram in August to 929 yuan/gram in mid-September.
My view has two parts.
Short term: If US–Iran talks fail and the US and China make no substantive progress, an October hike is almost certain. Gold will stay under pressure, and a return to around 900 yuan/gram for Shanghai gold is a reasonable target.
Long term: This round of hikes is validating our gold thesis. Governments must refinance at higher rates, so interest bills balloon rapidly. No government is willing to truly tighten, which leaves only one path: inflation.
To ground this view in data rather than intuition, we ran a debt and interest stress test through 2030 for five countries: the US, France, the UK, Japan, and Germany. Below I lay out the process and the conclusions.
Starting Point: How Much of the Budget Interest Already Takes
Table 1 shows the five countries’ 2025 starting points. The US is the only one whose debt ratio, deficit ratio, and interest ratio are all the highest or second-highest of the five:
The deficit is 6.8% of GDP, and the IMF says bluntly that it “sees no consolidation plan whatsoever.”
Net federal interest in FY2025 was $970 billion, or 13.8% of outlays and 18.5% of revenue.
In the first 11 months of FY2026, $1.05 trillion has already gone to interest.
Japan looks the most anomalous. Its debt is 206% of GDP, yet interest is only 1.5% of GDP, because the average coupon on outstanding JGBs is just 0.98%. It is a spring that has yet to be released.
France and the UK are the two European issuers where market pressure is most concentrated:
France was downgraded to A+ by three agencies within one year, and the OAT–Bund spread broke above 100bp for the first time since 2012.
The UK 30-year gilt yield is 5.85%, the highest since 1998.
Germany is the only country with fiscal room, but it is using up that room faster than anyone.
Official Optimistic Forecasts Have Already Fallen Behind the Market
Every country’s official forecasts use interest-rate assumptions that lag the current market:
US: The CBO’s February ten-year outlook assumed a 3-month T-bill rate of 3.3% and a 10-year yield of 4.1% for 2026. On September 23, the market was at 4.19% and 5.11%.
UK: The OBR assumes a 10-year gilt yield of 4.5%, while the market is already at 5.36%.
France: The average rate on new issuance in the first half was 3.14%, while the 10-year OAT has reached 4.5%–4.7%.
Germany: The fiscal plan was built on an ECB rate of 2.4%–2.7%, while the 10-year Bund is already at 3.57%.
Japan is the only exception. The 3.0%–3.6% used in the Ministry of Finance’s projections is in line with the market.
The 2030 Bill: What the Numbers Mean
When rates rise, only maturing debt that is rolled over and newly issued debt pay the new rate. So we estimated the share of each country’s debt that “reprices” each year. That estimate draws on each country’s share of short-term debt, the share maturing annually, and net issuance. We then added all the extra interest to new borrowing, which creates a compounding feedback loop between interest and debt.
US: Under Scenario 1, where today’s rates stay unchanged, FY2030 net federal interest is $1.86 trillion, or 20.1% of outlays and 28.1% of revenue. Under Scenario 2 it reaches $2.30 trillion, or 34.8% of revenue. In other words, for every $3 collected in taxes, $1 goes to interest. The historical peak, in the mid-1990s, was about 15% of outlays, when rates were 7% and debt was only half of GDP.
Japan: Interest payments rise from ¥10.5 trillion in FY2025 to ¥39 trillion under Scenario 2. That is 24.5% of general-account expenditure and 39.4% of tax revenue. Including redemptions, debt service would consume more than a third of the budget, something never seen in the postwar era.
Germany: Federal interest rises 170% over five years. Its share of federal spending goes from 6.1% to 12.7% in the baseline and 18.5% under Scenario 2. Half of the fiscal room Germany spent a decade building would be eaten up by interest within five years.
UK: Debt interest under Scenario 2 reaches £179 billion, more than double the defense budget.
France: In 2027, interest will overtake education and defense to become the single largest item in the state budget. The finance minister himself said so on September 21.
Three Ways Out, and Why It Ends in Inflation
Once the interest bill grows this large, a government has only three options: austerity, default, or inflation. Developed countries won’t default on debt in their own currency, so the real contest is between austerity and inflation. My view is that austerity has no political majority, and inflation is the path of least resistance. That view rests on what each country is actually doing, not on sentiment.
US: Congress just passed a law locking in tax cuts for ten years, and the Supreme Court then cut tariff revenue by 60%. The FY2026 deficit is $2.1 trillion. Whichever party wins the midterms, there will be no majority for cutting Social Security and Medicare.
Japan: The new government has gone further than anyone:
It formally abandoned its primary balance target.
It folded supplementary budgets into the initial budget.
It set up an investment framework with no cap.
It made a record ¥143 trillion budget request for FY2027.
It plans to cut the consumption tax on food from 8% to 1%.
All of this is happening while the BOJ is hiking, the 10-year JGB yield has broken above 3%, and super-long JGBs have sold off three times in a year.
France: The 2026 budget was forced through using Article 49.3 of the constitution. Three rounds of extra cuts within a year still failed to hold the deficit to 5%, and the presidential election comes in spring 2027.
UK and Germany: These are the only two with the institutional conditions for austerity: the UK has its fiscal rules and the October 28 Budget, and Germany has its debt-brake tradition. But the UK’s headroom under its fiscal rules has shrunk from £23.6 billion to around £10 billion. Germany just amended its constitution to exempt defense spending from the debt brake.
Historically, policy has never stood still when interest burdens approached these limits:
When US net interest neared 18% of revenue in the mid-1980s, the Reagan administration raised taxes and passed the Gramm–Rudman Act.
When Japan’s interest payments neared 20% of expenditure in the late 1980s, it introduced the consumption tax.
When UK debt interest neared 11% of spending in the early 1980s, Thatcher imposed austerity.
All three cases shared one precondition: government debt was only 40–50% of GDP, so the pain of austerity was bearable. Today the US is at 124%, Japan at 206%, and France at 116%. At these debt levels, the same amount of austerity lowers the interest bill but not the debt. The recession it causes stalls nominal GDP, so the debt ratio actually rises. That’s why the US and UK did not use austerity to pay down debt after 1945. Instead, they relied on more than a decade of negative real interest rates, known as financial repression, to bring debt ratios down from 120% to below 50%.
To spell the logic out fully: rate hikes are not the opposite of inflation. In a high-debt environment, rate hikes are the prelude to inflation. Each hike shortens the time public finances can hold out. It also brings forward the clash between central bank independence and fiscal sustainability.
When interest takes 30% of US federal revenue and a quarter of Japan’s budget, central banks will come under pressure to stop hiking, restart bond purchases, or tolerate a higher inflation target. They won’t do this because they want to, but because the alternative doesn’t exist politically. The Fed has already bought more than $300 billion of T-bills this year. The stated purpose is “reserve management,” but in substance it is a buyer for the Treasury’s short-term debt.
What This Means for Gold
The short term comes first, because its logic runs opposite to the long term. Over the past two decades, gold’s most stable pricing anchor has been real interest rates. For every 50bp rise in the 10-year TIPS real yield, gold falls roughly 10%. This round’s 24% decline from the January high matches the rise in real yields from 1.8% to 2.79%. Another 25bp hike in October will keep real yields and the dollar weighing on gold, so a return to around 900 yuan/gram for Shanghai gold is reasonable.
Two caveats apply:
The market already assigns a 70% probability to this hike, so most of it is priced in. The additional shock will depend on the dot plot and Warsh’s guidance, not on the hike itself.
Central bank gold purchases in H1 2026 were only 345 tonnes, the lowest since 2022. Official buyers do hesitate at high prices, so short-term demand for gold is not as unconditional as many people assume.
The long term is our core view. Historically, gold’s biggest bull markets did not happen because debt was high. They happened when high debt was combined with suppressed real rates. That was true in the 1970s, in 2009–2011, and in 2020. The data above show three things:
The interest burden will rise sharply before 2030, and official forecasts already underestimate it.
The more rates are raised, the faster that burden grows, and the sooner governments hit their limits.
Given current debt levels and politics, austerity is not the majority option. Inflation is.
Put these three together, and this round of hikes is not weakening the case for gold. It is making the case play out faster. Every day real rates stay high, the fiscal clock ticks faster. At some point the market will start pricing in that central banks have to stop. Real rates will then turn down while the interest burden is still rising. That moment is the signal that the long-term logic has taken over from the short-term logic.
So our operating framework is this: short-term weakness is a buying window, not a break in the thesis. We use real rates, not the gold price itself, to judge timing, and we watch three signals:
The 10-year TIPS real yield turning down within the 2.5%–2.8% range.
The Fed halting hikes or expanding T-bill purchases while inflation is still above target.
The BOJ’s hiking path starting to run into political constraints.
If any one of the three appears, the second stage of the thesis has begun.
I believe the best time to buy gold will be between the end of this year and early next year. By then the Fed will have hiked another 25–50bp and reached its projected peak. That is where the cycle turns. After that, the Fed can only hold or cut, which puts no further pressure on gold.
When This Would Be Wrong
US: Suppose the dot plot turns after the October hike, and a US–Iran deal brings oil back below $80. Then the US Scenario 1 falls back to the baseline, interest levels off at around 16% of outlays, and the long-term thesis is delayed.
Austerity in other countries: Austerity would prove possible in these countries if the following happen:
The UK’s October 28 Budget rebuilds more than £20 billion of headroom.
Germany trims the defense-exempt borrowing during parliamentary review.
The market accepts France’s 2027 budget.
Japan’s December budget keeps JGB issuance under ¥40 trillion and postpones the consumption tax cut.
Positive real rates (the most important one): Central banks might keep real rates positive even as interest burdens rise, with the US TIPS real yield holding above 2% and Japan’s real policy rate turning positive. Then the 2022–2023 pattern would repeat, with high debt and low gold prices at the same time.
We don’t think all these conditions will hold at once. But each one that does would weaken how strongly the interest burden feeds through to gold.
