Paul Schmelzing Shares Five lessons From Seven Centuries of the International Financial System

ELI5/TLDR

A finance professor spent seven years digging through old debt contracts — some literally from the Vatican Library — to rebuild interest rates, inflation, and government debt going back 700 years. His big finding: interest rates haven’t been falling since the 1980s, they’ve been falling since the 1300s. Most of what we call “weird” about today’s economy — negative rates, peace, short working hours, no financial repression — is actually the unusual part of history, not the normal part. And the 500-year stretch where governments could borrow ever more cheaply and safely looks like it’s finally cracking.

The Full Story

Paul Schmelzing studies dead investors. As he puts it, 93% of all the investors who ever lived are dead, and his job is to ask what the surviving 7% can learn from them. The raw material is things like a 1546 debt contract between the Fugger Bank and the English crown — short-maturity debt in Flemish currency at 12%, secured by London property and crown land. He claims this contract sits “at the cusp of perhaps the most important moment in the international financial system over the last 700 years” — more important, he says, than 1914, the 1980s, or 2008.

“93% of investors that ever lived, well, they are dead. And so the whole approach of financial history is really to ask what can the 7% alive today learn from the other 93%.”

The 700-year decline in interest rates

The project started during his PhD, when everyone was puzzling over “secular stagnation” — the idea that something broke in the 1980s and dragged real interest rates below zero for the first time in a millennium. Schmelzing was uncomfortable with grand pronouncements drawn from short, messy data. So he rebuilt the data itself, splicing together term structures from sources nobody had assembled before.

The result: interest rates have been on a downward trend not for 40 years but for roughly 700. Nominal rates slope down; inflation slopes up; and because nominal rates don’t fully compensate for that inflation, real rates slope sharply down across countries, maturities, and issuers. His provocative way of framing it — an investor in 1700, armed with the prior 250 years of data, could in principle have predicted the arrival of negative interest rates in the early 21st century. There’s noise around the trend line, but the data always returns to it, and the “half-life” of that return is between five and ten years — fast enough to matter to anyone investing today.

“If interest rates have been downward trending for more like 700 years rather than 40 years, maybe something other different is going on after all.”

Only two real breaks: the Black Death and 1557

When he runs formal statistical tests for genuine regime shifts — the real “eras of shifts and breaks,” to borrow Christine Lagarde’s phrase — most of the usual suspects (1914, World War II, the 1980s) don’t survive. Only two do: the Black Death around 1348, and a break in 1557. That 1557 break is the last time the cost of capital reset onto a new trend line, and we are still on that exact line in 2026.

What happened around then? Partly, violence fell. A researcher at LSE reconstructed homicide rates, and the chance of an elite investor dying a violent death within 20 years was once around 37%. It dropped sharply in the early 16th century — roughly when the feud (settling disputes by burning your neighbor’s castle with hired mercenaries) was banned and the Spanish Habsburgs centralized global finance into a single anchor currency for the first time. Schmelzing’s explanation is patience: when you’re less likely to be stabbed on the street, you save instead of consume, and rates fall. People started leaving larger endowments for their children — a measurable shift toward optimism about the future.

Geopolitics: real but always resets lower

For the typical investor across history, 91% of their investing life had a war going on between rich economies. The chart of war intensity goes back to 1495, and there’s barely a three- or four-year gap. Interest rates react sharply to each escalation — but the black line of rates keeps resetting at a lower level after every conflict, through the Thirty Years’ War, through the Napoleonic Wars. So markets “shrug off” geopolitics over the secular horizon even as they panic in the moment.

The catch: the period since 1945 is the longest stretch of peace between rich nations since 1270, when the Italian city-states exhausted themselves and stopped fighting. No wonder markets handle today’s geopolitical risk so badly — they’ve forgotten what it feels like. To scale the historical violence: the average rich-economy war killed about 290 people per million population. Apply that to the US today and you’d be talking roughly 87,000 deaths in a typical year.

How we spent the peace dividend

The post-1945 peace bought us three things, and Schmelzing is skeptical we spent them wisely:

Welfare and redistribution spending went near-exponential, after being essentially zero for 600 years. Working hours hit a 575-year low. And financial-crisis interventions — measured in joint work with Andrew Metrick by treating bailouts as a proxy for distress — have climbed relentlessly, making recent years the most financially distressed in recorded history (SVB being the latest near-miss). The first two he calls transitory; the crisis-firefighting one he thinks is permanent.

An audience exchange complicated the working-hours point. One questioner invoked Brad DeLong’s argument that we work less because we got more productive — innovation, not complacency. Schmelzing half-conceded but noted the pattern that big technological breakthroughs tend to be preceded by long stretches of working harder and harder.

The end of the 500-year government bull run

This is the throughline of all five lessons: governments have had an extraordinary 400-to-500-year run, borrowing ever more cheaply and safely, and it’s cracking. The US just crossed 100% debt-to-GDP and is on track to beat its World War II record within 18 months; the CBO projects ~180% by 2055. But debt-to-GDP, Schmelzing argues, isn’t the right measure — R minus G (interest rate minus growth rate) is. We’ve only had R-minus-G data for 30-40 years; the book extends it to the 1500s.

And here the long run flips the optimistic story. R-minus-G looks like it’s been trending down (good for sustainability) — but if you start the trend at the last real structural break in the 1930s, it actually slopes up. Growth has been falling since 1914, not the 1980s, masked by the volatility of world wars and oil shocks. So Blanchard’s COVID-era “we can afford anything, the debt pays for itself” was, in this longer frame, a bad idea. Rich economies as a whole are about to cross the all-time debt ceiling set by the Spanish Habsburgs in the 16th century — around 120%, the level where reserve-currency empires historically start to crumble.

His prediction: get used to financial repression — central banks buying up government debt, negative real rates, heavy balance sheets. The clean, independent central banking of the 1950s-70s was the outlier, not the norm.

China and the reserve-currency paradox

China’s share of global GDP is reverting to its 700-year mean; by the mid-2030s it should be back at its long-run average. The 20th century was the outlier, not China’s rise. But — and this is the optimistic close — China has always been the largest economy by size and was never a reserve currency. Reserve currencies (Italian city-states, Spanish Habsburgs, Holland, Britain, the US) were often tiny by comparison. So size isn’t what makes a reserve asset; something else does — institutional quality, ideas, credibility.

And the safety premium governments earn over private assets — which first turned positive around 1546 and rose for 400 years — started reversing in the 1950s-60s. Before 1557, the anchor of the financial system was private assets: land, the Corporation of London. Schmelzing’s parting suggestion is that we may be in one of those once-every-500-years moments where the signs flip back, and private markets — not government bonds — become the reserve asset of the decades ahead.

Key Takeaways

Real interest rates have declined on a near-continuous downward trend for roughly 700 years (since the 14th century), not since the 1980s — the core “suprasecular” finding that built Schmelzing’s reputation.
Rates revert to the long-run trend line with a half-life of 5-10 years, making the trend investable rather than a mere curiosity.
Only two genuine structural breaks survive formal testing across 700 years: the Black Death (~1348) and 1557. The cost-of-capital trend line set in 1557 still holds in 2026.
The typical investor across history spent 91% of their investing life with a war ongoing between rich economies, and faced a ~37% chance of violent death over any 20-year window.
The period since 1945 is the longest peace between rich nations since 1270 — markets are structurally unprepared for renewed geopolitical risk.
Average war intensity in rich economies was ~290 deaths per million population; scaled to the modern US that’s ~87,000 deaths in a typical war year.
Welfare/redistribution spending was near zero for 600 years before going exponential post-1900; Schmelzing tags this as transitory, likely reversing.
Working hours just hit a 575-year low (Bank of England data, adjusted for part-time/leave).
Rising challengers to a reserve currency consistently work more intensely than the incumbent (US vs Britain ~1914; China vs the West today) — correlation, not claimed causation.
Financial-crisis interventions (bailouts as a distress proxy, with Andrew Metrick) have risen to make recent years the most financially distressed in history.
The right debt metric is R minus G, not debt-to-GDP; extended back to the 1500s, and dated from the 1930s break, R-minus-G slopes upward — sustainability worsening, not improving.
Rich economies are about to cross the ~120% debt-to-GDP ceiling set by the Spanish Habsburgs — historically the level where reserve-currency empires start to fail.
The government “safety premium” over private assets turned positive ~1546, rose for 400 years, and began reversing in the 1950s-60s.
Reserve currencies have historically been small economies (Italian city-states, Habsburgs, Holland, Britain), while China has always been the largest by size yet never a reserve currency — size alone doesn’t confer reserve status.
China is reverting to its ~700-year average share of global GDP, projected to arrive there by the mid-2030s; the 20th century was the anomaly.
The downward rate trend holds across 100+ polities and even in private/mortgage rates (older than government bonds), suggesting a deep common cause rather than survivorship bias.
Book: A New History of the International Financial System, Yale University Press, out December 2026.

Claude’s Take

This is a genuinely good talk, and Schmelzing is the real thing — the eight-centuries-of-interest-rates work has been chewed over by serious economists since 2020, and it holds up better than most viral economic-history claims. The single most defensible idea here is the one he leads with: the secular-stagnation story implodes once you zoom out, because a 700-year downtrend doesn’t need a 1980s explanation. That reframing alone is worth the hour.

Where I’d apply the BS filter is the chain of inference from data to narrative. Reconstructing rates back to the 1300s is heroic but fragile — sparse early sources, a handful of issuers, judgment calls on weighting and splicing. To his credit, the Q&A pushes exactly there (dispersion, sparse data, survivorship bias) and he answers honestly, conceding that financial history is “more an art than a science” in picking the right precedents. The patience/violence story for 1557 is a plausible mechanism dressed up as a finding; correlation between homicide rates and interest rates is suggestive, not causal, and he mostly knows it. The “private assets become the reserve asset again” close is the speculative part — elegant, book-selling, and resting on a single 500-year analogy.

Two recurring tells of careful work: he repeatedly says “this is not a causal statement” (working hours, reserve-currency challengers), and he distinguishes transitory from permanent factors rather than declaring everything a regime change. That restraint is what separates this from the Lagarde/Tooze “polycrisis” rhetoric he’s implicitly critiquing. Score 8 — high-quality, primary-source-driven, intellectually honest, and the data is genuinely novel. It loses a couple points only because it’s a book teaser, so the load-bearing methodology stays in the “buy the book” black box.