The Debasement Trade, Re-Tested: What the Stressors Do to Gold and Bitcoin, and Why the Fiscal Story Has Not Been Paid For

• The question this examination answers
• The debasement trade, and what actually moves it
• The scoreboard, in the trade’s favour
• The scoreboard on price, and the mechanism behind it
• The stressors, mapped onto the trade
• Gold and bitcoin are not one trade
• The distinction that cuts the other way
• The verdict, and what would confirm or break it
• The close

The debasement trade is right about the world and wrong about the price, because the cost of holding an asset that produces nothing is set by the real yield on government debt — and this year the real yield rose while everything the trade was supposed to fear arrived.

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

§The question this examination answers

Every series that makes arguments eventually owes its listeners a reckoning, and this is ours. For two years the debasement argument has been repeated so often that it has stopped being a thesis and become a position: governments borrow more than they can ever repay in money of stable value; therefore they dilute; therefore the scarce assets that cannot be diluted — gold, and for the newer half of the argument, bitcoin — must rise. The claim is stated in the present tense and with the confidence of an identity, and that is exactly the kind of claim a series that keeps score is obliged to test.

So the question here is narrow, and deliberately so, because narrow questions can be answered. Given everything the previous examination established about where the pressures on this economy actually sit, what do those pressures do to this trade — and what does the trade’s own record this year say about it.

That second half is what makes this examination different from the ones around it. It is easy to find commentary explaining why the monetary order is decaying, and easy to find commentary dismissing the whole idea as a doomsday story. What is rare is an account that takes the premise seriously and then holds it against its own scoreboard: the money supply, the deficits and the official-sector gold purchases on one side, and the actual prices of the assets that are supposed to express all of it on the other. Those two sides are telling different stories this year, and the reason they can both be true at once turns out to be the most useful thing in the entire examination.

And the debasement trade has an unusual property among market theses: it can be correct about the world for a decade and still ruin someone who holds it at the wrong size, at the wrong moment, with money they need. A thesis with that property has to be stated as observable conditions rather than as a conviction, and we will state it that way before the end.

§The debasement trade, and what actually moves it

Before any of the argument, the two definitions, because they are blurred constantly and the blurring is where most of the confusion lives.

The first is debasement proper: the growth of the quantity of money and credit against the goods and services it buys. It is a slow, measurable, almost mechanical process. The second is the price of the assets people buy to express a fear of the first thing, and that is not a process at all. It is a market price, cleared every second against everything else a person could own, and it moves on the cost of holding an asset as much as on the reason for holding it.

That is the hinge of this entire examination, so it is worth stating slowly. For gold and for bitcoin the cost of holding is unusually explicit, because neither pays any income at all. Modern gold pays no dividend, no coupon and no rent; bitcoin pays nothing and, by design, never will. Held either way, the return given up is the return on the safest debt in the world, which is inflation-protected government debt.

That real yield is the variable to watch in this entire examination, and it has moved a great deal this year: the ten-year inflation-indexed Treasury yield was one point eight eight per cent at the start of 2026 and is two point six two per cent now. The price of holding an asset that expresses a fear of currency dilution rose by three quarters of a percentage point, guaranteed, in real terms. Nothing about the currency had to change for the arithmetic of the trade to get worse.

§The scoreboard, in the trade’s favour

Now the scoreboard itself, stated plainly, because this is where the trade’s supporters and its critics most often talk past each other. Start with what has gone the trade’s way, and be generous about it, because it is a long list.

The quantity of money is growing again. The broad money stock is twenty-three point two trillion dollars, up five point one per cent over the past year, and it accelerated through the spring. The central bank’s balance sheet has stopped shrinking and is expanding again, up two point two per cent over the past year to six point seven four trillion dollars. Federal debt stands at one hundred and twenty-two point six per cent of the economy, and interest on it has reached one point two five trillion dollars a year, the highest share of national income in the modern record. The dollar has weakened about two per cent against a broad basket over the past year.

And central banks keep buying gold. They bought two hundred and eighty-nine tonnes in the second quarter, up sixty-two per cent from a year earlier, the strongest second quarter on record, with the year’s total expected to land between seven hundred and fifty and eight hundred and fifty tonnes. They are not buying because they think it is cheap, and not to trade; they are moving reserves out of a system whose rules were shown, in twenty twenty-two, to be discretionary after all, when the reserves of one country were frozen by the country that issued them.

On the money side, then, the premise of the debasement argument is being fulfilled. Anyone who claims the thesis is a fantasy has to explain the balance sheet, the debt ratio, the interest bill, the money growth and the record official-sector buying. Those cannot be explained away. They are the world.

§The scoreboard on price, and the mechanism behind it

Then the scoreboard on price, which is much less friendly.

Gold peaked above five thousand four hundred dollars an ounce in late January and has since fallen to around four thousand four hundred. That is a decline of roughly nineteen per cent, and the period of the decline is what makes it notable rather than routine: it included a war in the Persian Gulf, an oil shock, a record gilt auction in London and a five per cent American ten-year. Every one of those is the sort of event the trade is supposed to be paid for. Gold fell more than eleven per cent in June alone, its fourth consecutive monthly decline and its worst quarter since twenty thirteen.

Bitcoin tells the same story more sharply. It reached an all-time high near one hundred and twenty-five thousand dollars last October, traded near sixty thousand in the summer, and sits around seventy-six thousand now. Over the past year bitcoin has fallen while the American equity market rose eleven per cent.

That is the fact that should structure this whole examination: the reasoning behind the trade was not obviously wrong, and the trade still lost money. There is no version of this argument that requires the money data to be wrong, or the central banks to stop buying, or the debt to stop growing. All of those things happened. The price fell anyway.

The explanation is not mysterious, and it is the mechanism that turns a true observation about governments into an unprofitable position. Gold and bitcoin produce nothing; their competitor is the real, inflation-protected yield on government debt. When that yield goes from one point eight eight per cent to two point six two per cent in a year, the cost of holding a non-yielding asset rises by three quarters of a percentage point in real terms, paid with certainty, from the safest borrower in the world.

Meanwhile the expected inflation that is supposed to make debasement profitable has not risen. The ten-year breakeven is two point three three per cent and fell five basis points this week; the market’s five-year forward inflation expectation is two point three one per cent. The arithmetic of this year has been rising real returns alongside stable inflation expectations, the combination that hurts metals and digital metals together. The trade’s problem is not that the world failed to debase. It is that the insurance premium on the currency rose without any increase in the inflation the insurance is meant to cover.

In twenty twenty-two, inflation reached nine per cent, the highest in forty years. Gold finished the year roughly flat. Bitcoin fell about sixty-five per cent. That is the year in which the debasement thesis was most obviously true in the consumer price data, and it was among the worst years in the trade’s history. High inflation is not sufficient. What matters is the real rate.

§The stressors, mapped onto the trade

We scored fifteen stressors in the previous examination on transmission, severity and proximity. Each can now be interrogated with two questions the trade itself supplies: does this stressor push real yields up or down, and does it do that before or after it hurts the people holding gold and bitcoin.

The first group pushes real yields up, and it contains the stressors with the highest weights. The energy shock scores a hundred out of a hundred and twenty-five, and its first-order effect on this trade is negative, because an oil-driven inflation shock in a world where the central bank is already fighting above-target inflation produces tighter policy rather than looser. That is what is presently priced: a quarter-point increase at the meeting that concluded hours before this examination was recorded, with the policy rate held at three and a half to three and three quarters per cent all year. Anyone waiting for the debasement trade’s payout has to accept that, in the current regime, the thing the trade fears — an inflation shock — is also the thing that raises the rate it competes against.

Long-end yields and the term premium score eighty, and they matter twice over. The term premium pushes nominal yields up, because of a government issuing more debt than the market comfortably absorbs. Rising inflation expectations would eventually lift the breakeven component and help the trade; but the rise this year has been in the real component, which hurts it. Commercial property, also eighty, is the third name in the group and the hinge between the groups, because it destroys credit and collateral. That lowers real yields, which is what the trade wants, but only after it has forced selling. That timing is the second group.

The stressors most likely to produce the world the debasement trade needs are the credit event and the recession: property at eighty, the banks at sixty, private credit at forty-eight, the exhausted consumer at forty-eight, with employment as the hinge. Each of them, if it fires, pushes the central bank toward cutting, and cutting lowers real yields, which is the tailwind the trade has been missing for four consecutive quarters.

In a credit event, hard assets fall first, and the historical record is unambiguous about it. In October two thousand and eight, gold fell roughly thirty per cent from its March peak before beginning the rally that eventually took it to a record. In March twenty twenty, in the fastest liquidity event in modern markets, gold fell about twelve per cent in a fortnight and bitcoin fell more than fifty per cent in two days. The reason is mechanical rather than psychological. Leveraged holders sell what they can sell, not what they want to sell, and what they can sell instantly and in size is the liquid, twenty-four-hour, globally traded asset. Both episodes ended with the assets much higher. Both began with the assets being sold, and sold hard, at the moment when the case for holding them looked strongest.

So the honest statement of what the trade actually is: not a hedge against a credit event, but a leveraged expression of the policy response to one, which has to survive the event in order to be paid for the response.

The third group puts money in the trade’s favour without an intervening crisis, and it is smaller than its supporters assume. If inflation expectations un-anchor — if ordinary people stop believing the central bank can deliver the inflation it targets — then real yields fall passively even if nominal yields stay high, because the breakeven component rises while the nominal yield stays put. The trade then wins without anyone cutting rates and without any crisis having to be survived. That is the scenario implied by the one data point in the stressor inventory that points this way: households expect four point two per cent inflation over the coming year, the eighty-first percentile of the modern record, while the market prices two point three one per cent over the long forward horizon. The household view is the one part of the world that currently favours this trade.

The second, rarer path is the fiscal one. If the fiscal position deteriorates past the point where the central bank is forced to intervene in the bond market — buying the government’s debt because nobody else will at any tolerable price — the trade gets its strongest possible confirmation, because that is the moment when debasement stops being a worry and becomes a policy. Note how early that moment still is. The balance sheet has been growing again for two quarters, at two point two per cent. The deficit is six per cent of the economy. The doubled buybacks in August gestured at the door without walking through it.

§Gold and bitcoin are not one trade

The two assets then have to be separated, because lumping them together is the most common error on both sides of this argument, and this year provided the cleanest natural experiment yet.

According to data from the research house Bitwise, bitcoin’s thirty-day correlation with gold reached zero point eight in late August and its ninety-day correlation hit a record. Over the same window its ninety-day correlation with the technology-heavy Nasdaq index fell from above sixty per cent to thirty-three. Read that carefully, because it is a real event rather than a talking point: bitcoin stopped trading like a technology stock and started trading like a monetary metal, at least across that window. For an argument that bitcoin belongs in a debasement trade, this is the most favourable evidence in years, and it deserves to be acknowledged rather than dismissed.

And the bullish precedent is real. After the bitcoin-gold correlation reached roughly zero point six in late twenty twenty, bitcoin gained a hundred and seventy-two per cent. When it rose from about zero to half in late twenty twenty-two, bitcoin rallied nearly three hundred and fifty per cent over the following fourteen months. Those are large numbers attached to the exact signal now on the screen, but the honest reading of those episodes is narrower than it sounds. A correlation says two prices moved together over a defined window. It does not say one caused the other, and it does not say they will keep moving together. Bitwise’s own long-run chart shows bitcoin’s correlation with gold swinging between positive and negative since twenty sixteen. Treat the current reading as a description of a regime, not as a promise about behaviour — because a regime ends when the forces that created it stop, and nobody is told in advance when that is.

Then there is the asymmetry that matters more than any correlation, and it comes straight from the demand data. Gold’s buyers are not one group. In the second quarter, gold exchange-traded funds had outflows of forty-five tonnes — Western investors selling — while central banks and other official institutions bought two hundred and eighty-nine tonnes, the strongest second quarter on record. Total first-half demand was two thousand five hundred and twenty-two tonnes, worth a record three hundred and eighty billion dollars. So gold has a buyer who does not care about the price. Central banks buy to hold reserves outside the dollar system, after watching the reserves of one country be frozen in twenty twenty-two. That buyer will keep buying whether the price is four thousand dollars or five thousand, and will keep buying through a bear market. Bitcoin has no equivalent base, and flows are precisely the kind of buyer that leaves when prices fall. That is the difference between a structural bid and a momentum bid, and in a liquidation, between an asset with a floor and one without.

§The distinction that cuts the other way

One more distinction cuts in the trade’s favour in a way that is easy to miss.

The reason gold fell this year while central banks bought record tonnage is that the two groups of buyers are different, and only one of them sets the price in any given month. Central banks buy physical metal in the over-the-counter market as part of a multi-year reserve strategy. Western speculators trade futures and funds against the central bank’s rate path; the CFTC’s data showed large speculators cutting their net long position by about fifteen thousand contracts in a single week, into a stronger dollar and firmer yields, and that flow moved the headline number.

Two things can be true at once: the monetary order is quietly repricing, and the price of the assets that express that repricing is falling. The debasement trade can be working in the structural sense while losing in the mark-to-market sense. Both statements are accurate about the same year. And for anyone who is not a central bank, only the second one pays the bills.

There is also a longer-cycle risk the trade’s supporters should be made to answer. Gold peaked above eight hundred dollars an ounce in January nineteen eighty, in a panic about inflation and monetary credibility far more extreme than anything priced today, and did not trade at that level again until two thousand and seven. Twenty-seven years. After the financial crisis it doubled to nearly one thousand nine hundred dollars by twenty eleven, and then drifted back toward a thousand while the central bank remained as accommodative as it had ever been. The story stayed intact and the price did not. Being right about the destination and wrong about the timescale is, for anyone who needs the money inside a decade, indistinguishable from being wrong.

§The verdict, and what would confirm or break it

So the verdict, stated in the form that can be argued with.

On the premise, the trade is right. Debt is one hundred and twenty-two point six per cent of the economy; interest costs a trillion and a quarter a year, the highest share of national income in the modern record; the deficit is six per cent; the money stock is growing five per cent; and the central bank’s balance sheet has begun expanding again. Nobody who looks at those numbers honestly can conclude that the currency is being managed for the benefit of its holders. The premise is not merely plausible; it is close to arithmetic.

On the price, the trade has been wrong for four consecutive quarters, and the mechanism is the real yield: one point eight eight per cent at the start of the year, two point six two per cent now, against inflation expectations that have barely moved. Being right about the world and being paid for it are different things, and this year demonstrates the distinction cleanly.

For the trade to be confirmed, we would want four of the following five: real yields falling back below one per cent while the debt ratio keeps rising, which is the condition that paid in twenty twenty and twenty twenty-one; inflation expectations moving up rather than sideways: ten-year breakevens above two point seven five per cent, and the long forward measure above two point five. The dollar breaking below its twenty twenty-five low, which for the broad index means below one hundred. The central bank buying bonds again in size, which is the difference between a balance sheet drifting up two per cent and the outright monetisation that turns a fiscal worry into a fiscal fact. And, slower than the rest, official-sector gold purchases above seven hundred and fifty tonnes a year for another two years.

And here is what would break it, which the same precision demands. Sustained real yields above two per cent with inflation falling toward target: that is the nineteen nineties, and gold answered those conditions with twenty years of decline. A credible fiscal consolidation — a deficit back below four per cent of the economy — would remove the premise rather than the price, which is the more thorough defeat of the two, because there would no longer be anything for the trade to be right about. A central bank that refuses to monetise: the current chairman has said he wants a smaller balance sheet, and if he holds that line through a recession, the trade’s central scenario never arrives. Or a liquidity event with no policy response. And for bitcoin specifically, the one risk gold does not carry: access, because central banks buy gold in the over-the-counter market and store it in their own vaults, while bitcoin depends on exchanges, custodians, banking on-ramps and regulatory tolerance, all of them third parties that can be withdrawn at the worst possible moment.

That is why the next genuine credit event is the most informative test available, and not one to be relaxed about. It will settle what correlation data cannot: whether bitcoin behaves like a monetary asset under stress, as its August correlation with gold suggested, or like the highest-beta expression of liquidity that it was in twenty twenty and twenty twenty-two.

Then the markers, dated, so the argument can be settled rather than continued. If the ten-year real yield is above two per cent at the end of this year, the headwind is still in place. If it falls below one per cent without a recession, the transition has arrived earlier than the inventory implied. If ten-year breakevens cross two point seven five per cent, expectations are un-anchoring and the trade wins passively. If the central bank’s balance sheet grows more than five per cent over the coming year, that is monetisation by another name. If official-sector gold purchases come in above seven hundred and fifty tonnes for the year, the structural bid is intact regardless of the price. If bitcoin’s correlation with gold stays above zero point five while its correlation with the technology index stays below forty per cent for another two quarters, its behaviour has changed in a way that matters rather than merely describing a month.

§The close

The examination found a split, and the split is the finding. On the premise the debasement trade is right, and no honest reading of the debt ratio, the interest bill, the deficit, the money growth or the record central bank gold purchases can say otherwise. On the price it has been wrong for four consecutive quarters, and the mechanism is not a mystery: the real yield on inflation-protected government debt rose from one point eight eight per cent to two point six two per cent while inflation expectations stayed where they were, and the cost of holding an asset that pays nothing is fixed against exactly that number. The reason both things can be true at once is that a structural bid and a market price are not the same measurement, and only one of them shows up in a portfolio.

What would falsify this reading is checkable rather than rhetorical. The premise fails if a credible fiscal consolidation arrives — a deficit below four per cent of the economy — because then there is nothing left to debase. The mechanism fails if real yields stay above two per cent while inflation falls toward target, which is the nineteen nineties repeating and gold’s twenty-year answer to it. The sequence fails if a credit event arrives with no policy response, or with a response too small to lift these assets after the forced selling. And bitcoin’s inclusion in the trade fails if, in the next genuine liquidity event, it trades like the highest-beta expression of leverage rather than like a monetary metal — the question its August correlation with gold raised and did not answer.

The fiscal arithmetic is real. The real yield is real. They are currently pulling in opposite directions, and the real yield is winning because it is the one that pays every day. The debasement trade’s own record this year is the evidence for that sentence, and it belongs in a series that keeps score rather than in a series that makes calls.

Sources: this article names its sources inline — the inflation-indexed Treasury yield and the breakeven and forward inflation-expectation series; the broad money stock and the central bank’s balance sheet; federal debt, the deficit and the interest-cost series; the dollar’s broad trade-weighted index; the official-sector and exchange-traded-fund gold demand data, and total first-half demand with its dollar value; the commodities and positioning data, including the CFTC’s weekly commitments series; the correlation data from the research house Bitwise, including its own long-run chart; the twenty twenty-two freezing of one country’s reserves and the official-sector response to it; and the historical episodes of nineteen eighty, two thousand and eight, twenty eleven, twenty twenty, twenty twenty-two and twenty thirteen, examined for sequence rather than for narrative. Where an estimate is contested, the contest is stated. 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.