Regime Change: The End of the Old Bond Regime

In 1971 the United States closed the gold window. The mechanism was clear-cut: the end of convertibility. The global monetary regime changed over a weekend.

Markets took a decade to absorb the full weight of that event. The analytical category was wrong: operators were reading a cycle when a structural transformation was actually at work.

Today the same confusion dominates bond portfolios. The phrase "regime change" reappears in macro reports with excessive casualness. It deserves surgical precision.

Regime Change: Two Meanings, One Structure

The term "regime change" originates in geopolitical vocabulary. It describes the forced replacement of a government, often by external action. Iran 1953, Chile 1973, Iraq 2003: three precedents are enough to call it a pattern.

Financial logic shares the same architecture. A regime remains stable as long as its internal relationships hold. Then an external factor shifts the coefficients, and the previous structure ceases to describe reality.

Confusing the two levels leads to costly mistakes. Global capital reacts to geopolitical power transitions and monetary regime transitions with the same slowness. Both dynamics reward those who read the structure before the price does.

The Precedent Portfolios Have Forgotten

The bond bull market lasted thirty years. 10-year Treasury yields slid from the peaks of 1981 toward the lows touched in 2020. Every active manager today built their career inside that cycle.

That cycle closed in 2022. The dollar's share of global currency reserves has fallen from 71% in 2001 to 58% in 2024, in a context free of acute crisis events.

This figure carries weight. It describes a structural drift, far removed from any cyclical oscillation. Three decades of model calibration rest on a world that has since been archived.

What the Current Pattern Is Saying

Since 2020 two indicators have diverged. Nominal yields are rising, and confidence in the stability of Western sovereign debt is eroding in parallel. This divergence has clear historical precedents.

In 1946–1951 the United States compressed real yields through explicit financial repression. The 1951 accord between the Treasury and the Federal Reserve closed that phase. Resolution came through inflation, taxed onto the saver.

Today's pattern rhymes with that dynamic. G7 public debt has reached levels that make rate normalisation politically costly. The divergence between yields and perceived solvency always resolves. The question remains how.

The saver pays the bill in real terms. Positive nominal yields coexist with losses in purchasing power when inflation exceeds the coupon. That mechanism transfers wealth from the creditor to the sovereign debtor.

The Mechanism: When Causal Coefficients Change

An academic paper offers precisely the right conceptual tool. Published on 21 August 2026, it formalises the concept of "effect-defying root causes": variables whose causal coefficients change between a normal and an anomalous regime, as documented in the study by Ruer and colleagues^[1].

The distinction matters for risk managers. In a cycle, variable values oscillate while the relationships between them remain stable. In a regime change, the causal structure itself shifts.

Applied to macro, the message becomes sharp. Correlations that held for thirty years, equities versus bonds, dollar versus gold, can reverse sign when the underlying coefficients move. That is what distinguishes noise from signal.

The real-world case treated in the paper concerns IT monitoring and intensive care. Domains far removed from finance, and precisely for that reason useful. The same mathematics that locates a fault in an IT system locates the breaking point in a market system.

Why VAR Models Miss the Signal

Value-at-Risk models use historical variance as their primary input. They assume that the future distribution will resemble the past one. Inside a stable regime, that assumption holds.

During a transition, that same assumption becomes the source of disaster. The model measures the volatility of the old regime and ignores the ongoing restructuring of causal relationships.

History provides the evidence. In 2008 bank risk models were measuring correlations between credit tranches calibrated on a decade of rising property prices. The causal structure changed, and the tools kept reading the old world. The cost was systemic.

The Chief Risk Officer who relies on thirty-year historical series is calibrating their instrument on the wrong world. The difference graph discovery methodology points to the alternative: seek the coefficients that change, rather than the variance that widens.

My Position, and What Would Falsify It

My thesis is direct. The rate cycle of 2020–2026 has ended the thirty-year bond bull run, and the succeeding regime differs from everything for which current models were calibrated. This constitutes a regime change, not a cyclical phase.

Most asset managers treat 2022 as a severe correction. They await mean reversion. That reversion belongs to the archived regime.

What would change my reading: core inflation returning below 2% across the G7 for eight consecutive quarters, accompanied by a rebound in the dollar's share of global reserves above 62%. Those two facts together would revive the cyclical thesis. I am watching for them closely.

Three Implications for Capital

Capital allocated on the premise of the old regime faces a slow and painful repricing. Here are the three operational consequences.

1. Family Offices and SWFs (36-month horizon): reduce nominal duration, increase exposure to real assets and currencies tied to commercial balances in surplus.
2. CFOs and Investor Relations (18-month horizon): the "return to rate normalcy" narrative being presented to investors risks looking wrong. Revise it now.
3. Chief Risk Officers (12-month horizon): integrate a causal-coefficient-shift scenario into models, beyond the classic variance shock.

The Forecast

A precise statement. By 31 December 2027 the average annual yield on the US 10-year Treasury will remain above 3.5%, marking a definitive break with the compressed-yield regime of the previous decade.

Confidence: 70%. Horizon: 31 December 2027. Verification: annual average of the 10-year yield published by the US Treasury.

The signal that would falsify the thesis: an average annual 10-year yield below 3.5% in 2027. That figure would bring the old regime back into play.

What to Watch

Three leading indicators will confirm or refute the reading in the months ahead.

• The dollar's share of global currency reserves in IMF quarterly COFER data.
• Long-maturity Treasury auctions: bid-to-cover ratio and yield tail.
• The pace of gold accumulation by emerging-market central banks.

These three signals lead structural repricing by several quarters. Those who monitor them see the regime changing before consensus does. The rest will discover it from the price.