A Private Credit Cycle Inside a Sovereign Funding Cycle: The Synthesis and a Realistic Way Forward (Part 3 of 3)

• What grading could not settle
• What part two left standing, and what it could not confirm
• The structure, in one sentence
• The conditions that decide which layer binds
• The hinge: one number read in two directions
• Our own prior, put on the table
• The claims that cannot fail, and the versions that can
• The markers, and what would change the picture
• What a realistic way forward does and does not require

Two theses survived part one at full strength and part two’s grading left each guest holding one piece of machinery, so this final part stops asking which man is right and asks which layer of the same structure is binding now — and names the boring, published markers that will settle it.

DeepDives · 2026 · An AI–human collaboration that weighs every source, mainstream or alternative, under strict rules of truth and logic.

§What grading could not settle

Both guests handed us mechanisms, and both mechanisms are internally coherent. Neither is refuted by anything in the record we could reach, and that is the uncomfortable part. The honest opening for this final part is that the two theses may both be true, and true at the same time, because they are describing different layers of one structure.

A shock can begin on the private side of the balance sheet and be answered from the sovereign side, and a sovereign funding problem can travel into the real economy through the cost of private credit. The question is not which man is right. It is which layer is binding now.

Part one set the two systems against each other and let each man speak in his own terms. One guest came out of banking history and told us that the debt everyone watches is the wrong debt, that the crisis variable is private credit and the speed at which it grows rather than the level it reaches, and that government borrowing has not produced a financial crisis in a major developed economy since the Second World War. The other guest came out of the trading desk and told us that the sovereign funding position is the whole story, that the long end of the curve is repricing in front of us, and that the arithmetic of the interest bill leads to one destination.

Part two then took those claims one at a time and graded them against the record as it stands at the end of September 2026. Some held. Some needed narrowing. Four did not survive checking, and we said so rather than dropping them quietly. So this last part has one job left, and it is not to crown a winner.

§What part two left standing, and what it could not confirm

Each side came away from that grading holding one item that survived intact. On the sovereign side it is the arithmetic of the roll. The weighted average interest rate on all marketable Treasury debt was three point four seven five per cent at the end of August 2026, while the marginal cost of money is far above it. That gap is sourced, current, and quantitatively the strongest thing the sovereign case has.

On the private side it is the mechanism of money creation, and a central bank confirms it in its own words. The Bank of England’s Quarterly Bulletin, from the first quarter of 2014, states that the majority of money in a modern economy is created by commercial banks making loans, and that banks do not act as intermediaries lending out deposits placed with them. The same guest’s figures for the scale are consistent with it, because United States deposits went from five trillion dollars in 2000 to twenty-one trillion in 2025, with roughly ninety per cent of the increase coming from private banks lending to households and businesses and about ten per cent from the central bank.

Four claims did not survive checking, and we are carrying them as unverified rather than repeating them as facts. The first is the eight point three trillion dollars of refinancing this year, which no source we could reach confirms, though the order of magnitude is plausible because something like a fifth to a quarter of marketable debt matures inside a year. The second is the claim that only about one hundred billion dollars of a one point eight trillion dollar Asian trade surplus went into United States long-term bonds, where the net-flow figure could not be reproduced at all. The third is interest at four and a quarter per cent of GDP in 2027, which is a projection and sits above recent official projections. The fourth is the claim that more than a third of United States household wealth is now held by people over seventy, which was not checked against the distributional accounts and should not be repeated as a fact.

None of those four is load-bearing, and neither thesis needs any of them. Two smaller corrections are worth carrying for the same reason, because the direction of a point should survive the loss of its illustration. The private debt case puts household wealth at about one hundred and eighty trillion dollars, and the current figure is higher, with household and nonprofit net worth at one hundred and ninety-five trillion eight hundred and seventy billion dollars in the second quarter of 2026. The sovereign case offers Japan at about nine hundred billion dollars of United States debt and China at about eight hundred billion, against actual holdings of one trillion one hundred and eighty-five billion for Japan and six hundred and eighty-three billion for China.

§The structure, in one sentence

Here it is. A private credit cycle is running inside a sovereign funding cycle. The private cycle decides whether trouble arrives as an ordinary recession or as a credit crisis, because that cycle is about banks and households and businesses and the flow of new lending. The sovereign cycle decides the price of the backstop and the cost of rolling the debt that pays for it, because that cycle is about auctions and term structure and the rate paid on the stock. One sets the shape of the shock. The other sets the price of the answer.

Picture two clocks running at different speeds in the same room. The private clock ticks with the lending cycle, which historically runs in years to a decade and a half. The sovereign clock ticks with the maturity structure, which runs in months, because the debt is being refinanced continuously and the buyer has to be found again every time. When the fast clock goes wrong, the slow clock sets the cost of the rescue. When the slow clock goes wrong, the fast one transmits the damage into the real economy through the price of credit, because every business that rolls a loan pays the new rate whether or not anything has gone wrong at the business itself.

The price of the answer is where the policy rate sits. The effective federal funds rate was three point eight eight per cent on the twenty-eighth of September 2026, up from three point six three per cent in the middle of the month. That was a twenty-five basis point increase, and it was the first of its kind after a run of cuts, because the same rate was four point three three per cent a year earlier.

Now set that against the long end of the curve, where the ten-year sat at five point two four and the thirty-year at five point five six. The short rate is more than a point and a half below the long rate, which is the shape of a market that is not pricing an easy rescue.

§The conditions that decide which layer binds

The private layer dominates when private debt relative to income is high and rising faster than income, when the growth in lending is concentrated in one category, and when the marginal borrower is a household or a firm rather than a government. One high-rise condominium in South Beach is a market. Twenty of them in three years is something else, and the difference is the speed rather than the number.

The sovereign layer dominates when private debt relative to income is moderate, when the average cost of the debt stock sits well below the cost of new money, when issuance is parked at the front end and has to be refinanced again and again, when the foreign buyer has stopped adding to the stock, and when the interest bill is climbing towards the size of the receipts that pay it. When the marginal buyer moves from abroad to the domestic market, the constraint changes character. A foreign official buyer is price-insensitive for reasons of its own. A domestic buyer is price-sensitive, and it has alternatives, and it will charge for the privilege of funding a deficit that is running near two trillion dollars a year.

There is also a feedback path that belongs to the sovereign clock and not to the private one, and the trading guest sketched it as a doom loop. The central bank raises rates, the interest bill on the stock rises with a lag, the deficit widens, more bonds are issued, yields rise as the supply meets the market, and the central bank faces the same decision again from a worse position.

§The hinge: one number read in two directions

The private credit case removes its own fuse with its own evidence. The same guest who tells us that private debt is the crisis variable also tells us that United States private debt peaked at one hundred and fifty-three per cent of GDP and now sits around one hundred and forty per cent, which he describes as not great, but more moderate than it was in 2007. The level checks out. The Bank for International Settlements series on total credit to the private non-financial sector puts the United States at one hundred and forty point three per cent of GDP in the fourth quarter of 2025, which is his one hundred and forty. The peak we could not reproduce, because that series maxes out at one hundred and seventy-two per cent in the third quarter of 2008, not the one hundred and fifty-three he gives. But the current level is his number, and the current level is the one that matters here.

And that number does two things at once. First, it says that the private credit mechanism is not primed in the way it was primed going into 2007, when the stock was climbing towards a peak roughly thirty points higher than it is today. A fuse that has been partly drawn down is still a fuse, and the velocity test can still fire, but the specific condition that made the last crisis as deep as it was is weaker now than it was then. Second, and this is the part that decides the trilogy, the very moderation of the private debt stock is what leaves the sovereign funding cost standing as the binding constraint. The two claims are not two separate forecasts about the world. They are one fact read in two directions.

The banking historian treats one hundred and forty per cent of GDP as moderate, and means it as reassurance. The other man looks at a sovereign that is running a deficit near two trillion dollars a year and has to refinance continuously, and he sees a constraint that is already doing the work. They disagree about what the number implies for priority, and that disagreement cannot be resolved by checking the figure again, because both of them are reading it correctly.

The two theses are both structurally sound, and they are mutually exclusive in their priority. Neither can be the primary explanation at the same time as the other, because they put the binding constraint in different places, and only one place can bind at a time. That result is stronger and more interesting than naming a winner, and it is also less comfortable, because it means the answer is not in the material. It is in the sequence.

Two mechanisms can both be described correctly and neither can be dated. A map is not a clock. Being right about the shape and useless about the timing produces the same outcome for the listener as being wrong.

§Our own prior, put on the table

This series has argued across several of its own examinations that fiscal dominance combined with reserve currency status means the adjustment lands on the currency rather than on an outright default, and that hard assets reprice against it. That is the house view, and it is not neutral in this argument, because it takes a side on which layer binds: it says the sovereign layer is the decisive one and the currency is where it shows.

Where it is strong, it is strong on three points that do not depend on today’s prices. First, the debt is funded domestically rather than abroad. Foreign holdings stood at nine trillion two hundred and sixty-seven billion dollars in December 2025, about twenty-three per cent of gross debt and about twenty-nine per cent of debt held by the public. A sovereign that owes most of its debt to its own residents and institutions has a far smaller forced-seller problem than the popular version of this story imagines.

Second, the weighted average interest rate on all marketable Treasury debt was three point four seven five per cent on the thirty-first of August 2026, while the marginal cost of money sits far above it. That gap is what makes the roll a slow mechanical transfer to the holders of the debt rather than an event.

Third, a sovereign that borrows in its own currency cannot be forced into nominal default, because it can pay the amount in the unit it issues. That is not a reason for comfort about the value of that unit. It is a reason to expect the adjustment to arrive in the unit rather than in a missed payment, which is exactly what the house prior claims.

All three of those are structural rather than priced, so the case against has to be as concrete as the case for. Start with the prices as they stood on the twenty-eighth of September 2026. The ten-year real yield from inflation-protected securities was two point nine per cent, up from one point eight four per cent in October 2025. The ten-year breakeven inflation rate was two point three four per cent, and the five-year five-year forward breakeven was two point three six. Read together, the market is charging a high real rate while expecting about two and a third per cent inflation for a decade. That is the opposite of a debasement price, because a debasement price looks like low or negative real yields with breakevens climbing. What is on the screen instead is compensation for being repaid in money that is holding its purchasing power, at a rate above inflation.

The second problem is precedent, and it points the other way from the forecast. The largest recent inflation in the United States did not arrive with the central bank financing the deficit. It arrived with direct fiscal transfers to households. Federal debt as a share of GDP went from one hundred and six point eight per cent in the first quarter of 2020 to one hundred and twenty-four point one per cent in the third quarter of that year, and consumer inflation followed it up to about nine per cent in 2022. Debt exploded and inflation exploded, and the mechanism ran through transfers to households rather than through a bond purchase programme, which is a problem for the tidiest version of the debasement story.

The third problem is the decade after 2008, which is a live precedent for the opposite of the forecast. Debt went up, and inflation went down and stayed down. Yields fell, breakevens stayed contained, and the adjustment did not arrive.

Being early costs drawdowns, and that is not a rounding error. A structural argument that cannot survive an expensive decade is not a structural argument. That is a statement about what it costs to carry a view rather than an instruction about what to carry.

§The claims that cannot fail, and the versions that can

The claim that every road ends in a weaker dollar is a claim that cannot fail. If the currency weakens, the claim is confirmed. If the currency strengthens sharply, then the same argument says the stronger dollar makes the rest of the world’s dollar obligations heavier, forces foreign holders to sell United States assets, and that pressure is what delivers the weaker dollar in the end. The trade figures are there for it: the rest of the world owes about fourteen trillion dollars in dollar debt and owns about twenty-two trillion dollars of United States assets, a little under half of that in government debt, so a stronger dollar does put pressure on those holders. But a stronger dollar is not a counterexample. It is the route by which the prediction arrives. A statement that is confirmed by an outcome and confirmed by the opposite outcome is not a forecast.

The version of that claim which could be shown wrong is narrower and far more useful. It would be wrong if the dollar weakened only modestly and slowly while real yields stayed high, breakevens stayed near two and a third per cent, and hard assets failed to reprice against the currency over a defined window. It would also be wrong if the currency rallied hard, foreign holders did not become forced sellers, no disorderly move followed, and the dollar then simply stayed strong for years while the deficit ran on.

The private credit claim has a version that cannot fail as well, and the same courtesy is owed in that direction. If the next crisis is a private credit bust, the claim is confirmed. If the next crisis is a sovereign funding crisis, the answer will be that the country borrowed in its own currency, so the event was not really a sovereign event, or that the private balance sheet was the transmission channel. That is the same caveat the guest applies honestly and in advance, and the euro-area episode of 2010 to 2012 is the case it covers, where five countries faced sovereign funding crises and Greece restructured its debt, in a currency none of them issued. So the caveat is real. It is also elastic, and elasticity is the problem.

The version of the private claim that could be shown wrong is this. It would be wrong if a crisis arrived in a major developed economy, the trigger was plainly a failure at a government auction or a funding event in the sovereign market, and at that moment private debt was well below its historical peak with lending growth under the fifteen per cent threshold that the same guest proposed as his own warning rule. It is the number we will be watching.

§The markers, and what would change the picture

Start with the auctions, because they are the most direct test of the sovereign layer and they arrive on a fixed calendar. Watch the coverage and the high yield at each thirty-year and twenty-year sale, and specifically whether the bid-to-cover slips toward the low twos or below. For context, the September 2026 sales cleared at two point six one for the thirty-year on the tenth, two point five seven for the twenty-year on the fifteenth, and two point seven one for the ten-year on the ninth. A cover drifting into the low twos while the tail widens would be the first sign that the buyer is setting terms rather than taking them, and that is the marker that would confirm the sovereign thesis.

Second, watch the average cost of the stock moving toward the marginal cost of money. The weighted average rate on marketable debt was three point four seven five per cent at the end of August, against a policy rate of three point eight eight and a thirty-year near five and a half. If that average drifts up through four and then towards four and a half, that is not a market event. It is the claim most easily falsified in reverse, because an average that stays flat while the long end sits above five would mean the stock is being refinanced somewhere other than the market.

Third, watch foreign holdings in the Treasury International Capital series. The total was nine trillion two hundred and sixty-seven billion dollars in December 2025, against nine trillion three hundred and forty-nine billion the month before. Flat is not yet a turn, and the marker is whether flat becomes a rapid fall, because that is the difference between a buyer standing still and a buyer leaving, and only the second one forces a repricing at the long end rather than a gradual one.

Fourth, watch interest expense as a share of federal receipts, and watch the deficit before interest, which was running about two point six per cent of GDP. Interest on public issues ran to nine hundred and eighty-two billion dollars in the fiscal year through the end of August, and the full year runs above one trillion. If that line keeps climbing as a share of what the government collects, the sovereign case is being paid. If the ratio stabilises because receipts rise, it is not.

Fifth, watch private credit growth against the fifteen per cent threshold that the private debt case set for itself. That is the guest’s own number, and it is the cleanest falsification test in the trilogy. Below the threshold and steady, his mechanism is dormant whatever else happens, and the ordinary recession is the ordinary recession. Above it, and concentrated in one category in particular, his mechanism is live and the velocity of lending becomes the thing that matters rather than the level of the stock.

Sixth, and this is the pair that decides the debasement question rather than the debt question, watch the ten-year breakeven against the ten-year real yield. Inflation expectations rising while real yields fall would be the first genuine confirmation of a debasement price, because that combination is what a currency discount looks like when it is being charged. Real yields staying high with breakevens contained would confirm the opposite, that the market expects to be repaid in money that holds its value, and that the pressure sits on the borrower rather than on the unit of account.

Then the events rather than the readings, because four things would actually change the picture, and each one maps to a thesis. A long auction that fails outright or tails badly enough that the Treasury has to clear well above the when-issued level is the event that moves the sovereign layer from a slow transfer into a live constraint. A decisive turn in foreign holdings, from flat to a fast decline, is the same event arriving through a different door. Private credit growth accelerating past the fifteen per cent threshold, in one lending category and at speed, is the event that arms the private mechanism and makes an ordinary cycle dangerous again. And a run of consumer inflation above four per cent while real yields fall is the first real evidence of a debasement price rather than an assertion of one. That last level is not exotic, because consumer price inflation was three point seven one per cent in August 2026 and peaked at four point two seven per cent in May, which is exactly why the test is worth naming now rather than later.

§What a realistic way forward does and does not require

The two implications for anyone who is not a trader are not the same implication, and they do not live in the same time zone. The private debt case implies that the sovereign argument is largely a distraction, and that what actually matters for a household or a business is the private credit cycle, because that is what decides whether ordinary firms and families get into trouble. It is about the flow of new lending, the cost of renewing a loan, and the health of the bank that holds it. The sovereign case implies that the currency and the term structure are the risk, and that is a question about what you hold and for how long, rather than about what you trade this week. Neither of those is an instruction and neither is a prediction. They are two different questions about the same balance sheet, and the answers arrive on different clocks.

So what a realistic way forward requires is smaller than either guest’s conclusion, and what it does not require is worth stating just as plainly. It does not require choosing a winner in advance, and it does not require a position on which man is right, because an argument between two sound mechanisms cannot be settled by preferring one description of the plumbing. It does not require a forecast with a date on it, and it does not require an instruction about what to carry, because sizing a view is not the same as reading the machinery. What it requires is the discipline of the six readings and the four events: auction coverage, the average rate on the stock, foreign holdings, the interest share of receipts, private lending growth against a threshold, and the breakeven against the real yield, all read on dates that are published in advance and scored by outcomes that do not care which of the two is talking.

And that is where the trilogy has to leave it, because neither man gets to have the next crisis be his by definition. So the markers are the test, and they are deliberately boring. They are checks that anyone can run, and they are the only way to settle an argument between two sound mechanisms.

This has been a deep dive into what to watch when two debt theses disagree.

Sources: this examination names its sources inline — the two theses stated across parts one and two by a banking historian and a trading guest, with their separate machinery and their commercial interests noted; the Bank of England Quarterly Bulletin article on money creation in the modern economy from the first quarter of 2014; the Bank for International Settlements series on total credit to the private non-financial sector; the household and nonprofit net-worth series and the distributional accounts; the effective federal funds rate and the market tape for the ten-year, the thirty-year and the seven-year; the Treasury auction results for September 2026; the Treasury International Capital series on foreign holdings; interest on public issues and the deficit before interest; the inflation-protected securities real yield, the ten-year breakeven and the five-year five-year forward breakeven; the consumer price inflation series; the federal debt and consumer inflation paths of 2020 to 2022; the decade after 2008; and the euro-area sovereign funding crisis of 2010 to 2012 with Greece’s restructuring. Every source is weighed with the same skepticism regardless of politics or business model. The audio version of this article, read by DeepDives, is available on Wavlake.