|
Macro · Hard Assets · Monetary History
Nixon called it temporary. It has been temporary for fifty-five years, and the bill went to everyone holding the currency.
The war did not break a sound monetary system. It exposed one that was built to break, at exactly the moment a decade of deployments came due.
Wall Street Logic · 8 min read
Start with the objection, because it is the honest one and every serious critic of this asset makes it. Gold does nothing. It pays no coupon, files no earnings, employs nobody, funds no research, and has invented not one thing in the five thousand years humans have been digging it up and stacking it. A share of a business grows, a bond pays you to wait, a building collects rent, and gold sits in a vault costing money to guard. Every word of that is true, and it misses what the metal is for entirely, because a ruler is only useful if it cannot be adjusted by the person doing the measuring. Gold's supply grows at roughly the rate the human race can pull it out of the ground, which is slowly, expensively, and on a schedule that no central bank, treasury or legislature can accelerate to meet a political deadline, and that single unglamorous property is the whole of what sound money has ever meant. Not that gold rises. That gold cannot be made to rise on command, which means it does not measure itself, it measures everything else, and Issue 003 gave the modern alternative its name when it called the fiat system a ruler engineered to shrink. This is the piece about who shortened it, when, and what the shortening was used to pay for.
Before the history, one claim needs handling carefully, because it is the single most repeated line in gold writing and it is not quite true. You will hear that every fiat currency in history has eventually gone to zero. Plenty have, and the list is long enough to be sobering: the Song dynasty paper that Issue 003 walked through, the French assignat, the Weimar mark, the Zimbabwe dollar that needed a hundred trillion note printed before anyone admitted the experiment was finished. But sterling has been in continuous use for more than twelve hundred years and still trades this morning, and so does the dollar, and a claim that ignores the survivors is a claim that will be dismantled by the first reader who checks it. The survivors are the better argument anyway, provided you are careful about the comparison, because sterling held its value tolerably well from 1717 to 1914 and did so under a gold standard, which rather makes the point rather than refuting it. Judge currencies only across the era in which they have been pure fiat, which for practical purposes means the last fifty-five years, and the record is unbroken and needs no exaggeration: not one has held its purchasing power. The question was never whether the ruler shrinks. It is only ever how quickly, and who collects the difference.
The popular account of how that era began is that Nixon took the dollar off gold to pay for Vietnam, and that account is part of the truth in the way that a diagnosis of exhaustion is part of the truth about a heart attack. What Bretton Woods had built was a system in which foreign governments could exchange dollars for American gold at an official price of thirty-five dollars an ounce, and it functioned for as long as the United States held most of the world's gold and spent roughly within its means. It stopped functioning when it did neither. Through the nineteen sixties, military spending abroad and foreign aid pushed dollars out into the world considerably faster than the gold sitting behind them could grow, and the Federal Reserve's own published history of the period states the arithmetic without decoration: eventually there were more foreign-held dollars in existence than the United States had gold to redeem them with. At that point convertibility was a promise that could not be kept if it were ever called, and the holders of those dollars worked that out roughly as fast as you would expect. France converted and walked out of the London Gold Pool. Britain came asking for cover. The queue formed at the window, and on the fifteenth of August 1971 Nixon closed it, describing the measure to the country as temporary.
The part that usually gets left out is the part that makes the story worse rather than better, and it belongs to an economist rather than a president. Robert Triffin had described the trap a full decade before it closed, and what he described was structural rather than moral: a country whose currency serves as the world's reserve is obliged to supply that currency to the world by running deficits, and those same deficits eventually destroy confidence in the currency being supplied. The reserve issuer is required, by the mechanics of the job, to undermine itself. Vietnam did not create that contradiction and neither did any individual in the room that August. What the war did was set the date, by piling a decade of foreign deployments and domestic programmes on top of a design that had a failure built into it from the first day, and by making the failure arrive while there was still a queue at the window rather than quietly over another generation.
The war did not break a sound system. It exposed a system that was built to break, and a decade of deployments simply chose the date.
So who paid, and how much? Take the official measure first. Cumulative inflation since 1971 runs to something close to seven hundred and twenty-five percent, which is to say that a dollar from that year buys today what eight dollars and twenty-five cents buys now, and that the currency has surrendered roughly eighty-eight percent of its purchasing power inside a single working life. That is the number the index admits to, and it is large enough on its own. Now run the same fifty-five years against the ruler nobody administers. The official gold price was thirty-five dollars an ounce when the window shut. Gold is a little over four thousand dollars an ounce this morning. Measured that way, the dollar has given up not eighty-eight percent but something closer to ninety-nine, and both of those figures are correct, and the interesting work in this entire piece sits in the distance between them. Part of that gap is inflation the index was never designed to capture, the compounding erosion that a basket of consumer goods does not register because the basket itself keeps being redefined. But this letter would rather be right than flattering, so the other part has to be said plainly: some of that gap is gold itself re-rating in real terms, because thirty-five dollars was a suppressed official price inside a two-tier market and today's price carries a monetary premium it did not carry then, earned over the past four years as a great many institutions decided that an asset with no counterparty was worth owning again. That re-rating is not a distortion to be netted out before the argument can start. It is the argument. What the chart is showing is fifty-five years of the world slowly repricing the difference between a promise and a thing.
Which brings us to how the promise kept its buyers after the gold was gone, and this is where most writing on the subject goes badly wrong, so it is worth being careful. Closing the window solved one problem and created another, because nothing now compelled the deficits to stop and nothing obliged foreigners to keep holding dollars they could no longer redeem for anything. What happened next was not designed by anyone. In October 1973 the Yom Kippur War produced an Arab oil embargo aimed at the United States, decided in Kuwait City rather than at any strait or shipping lane, and crude went from around three dollars a barrel to roughly twelve within a matter of months. The immediate consequence was a global recession and petrol queues that a generation still remembers. The lasting consequence was accidental: Saudi Arabia was abruptly earning far more dollars than any economy of its size could possibly spend, and that money had to be put somewhere. In June 1974 Kissinger and Prince Fahd issued a joint statement on economic and military cooperation, and a month after that Treasury Secretary William Simon flew to Jeddah, and out of his trip came a separate understanding under which Saudi surpluses were invested in United States Treasury securities. No signed text of that understanding has ever been declassified. What is fully documented is the concealment: Saudi holdings were folded into an anonymous grouping labelled oil exporters and kept out of the published country-by-country accounts for forty-one years.
Now the part where this letter parts company with almost everything else written on the subject, because the popular version is wrong in ways any reader can verify in about five minutes and this newsletter would rather lose a dramatic sentence than print one that will not hold. There was no secret treaty that forced oil to be priced in dollars. Oil was already priced in dollars well before 1974. The historians who have actually worked through the declassified record find no pricing agreement of any kind, and economists including Brad Setser at the Council on Foreign Relations argue that the causation runs in the opposite direction to the popular telling: oil sells for dollars because the dollar already dominated global trade, not the other way around. The story that went round the internet in 2024, that a fifty-year petrodollar agreement had quietly expired and nobody in the West had noticed, was simply false. What actually lapsed in June of that year was the United States and Saudi Arabia's Joint Economic Commission, an agreement containing no currency clause whatsoever, and PolitiFact had rated the claim false within a fortnight. A great many newsletters printed it regardless. Almost none of them corrected it afterwards, which tells you something useful about the standard of evidence in this corner of financial writing, and is the reason this section reads more like a retraction than a revelation.
What is left after the embellishment is stripped out is still the important part, and it is a story about an absence rather than a conspiracy. Convertibility had been a hard constraint with a very specific enforcement mechanism: spend beyond your means for long enough and somebody eventually turns up at the window wanting the gold. After August 1971 that constraint was gone, and nothing was put in its place. Recycling gave the deficits somewhere convenient to go, and oil exporters did buy Treasury securities with money the world had paid them in dollars, and that did help finance the American government. But do not lean on it harder than the evidence allows, because the Gulf was never the main channel through which this worked. Setser's point, and it is the correct one, is that the large sources of offshore dollars were the Asian manufacturing exporters rather than the oil monarchies, and Gulf recycling was one stream among several and by no means the biggest. The shape matters more than the size. A country that must borrow to fight a war eventually meets a limit, because someone eventually declines to lend. A country that issues the world's reserve asset meets that limit very much later, and by the time it arrives the cost has been distributed in thin slices across everyone who holds the currency, which is a far more comfortable political arrangement than a tax and requires no vote.
Which is why the data of the past four years deserves more attention than it gets, because the arrangement is being unwound in the open by institutions that are not being remotely secretive about it. The foreign share of United States Treasury holdings has fallen from around half in 2010 to roughly thirty-one percent last year. Over the same stretch gold reached twenty-seven percent of global official reserves at the end of 2025, passing Treasuries at twenty-two percent for the first time since 1996. One caveat belongs here immediately, and the European Central Bank supplies it themselves rather than leaving it to critics: the greater part of that crossover is price rather than reallocation, because gold rose about thirty percent in 2024 and about sixty percent in 2025, which inflates its share of any portfolio mechanically without a single ounce changing hands. Hold prices constant at 2023 levels and gold sits at sixteen percent against Treasuries at twenty-six, and Treasuries still lead comfortably. Anyone telling you central banks have dumped American debt for bullion is selling you something, and this letter would rather hand you that sentence than have you find it somewhere less friendly.
What is genuinely real is the direction of new purchases, which is where large holders always reveal themselves. Central banks bought around two hundred and forty-four tonnes on a net basis in the first quarter of this year, up seventeen percent on the quarter, and China has now added to its reserves for twenty consecutive months. That pace has cooled from the extraordinary run of 2022 through 2024, when official buying cleared a thousand tonnes a year three years running, and it is still running at roughly double the rate of the entire 2010s. Institutions of that size do not sell what they already hold, because selling moves the price against them and announces the intention. They change what they buy next, quietly, and this is precisely what that looks like from the outside. The fund flows tell a more complicated story than most gold letters will admit, and it is worth having the complication rather than the slogan: North American gold funds saw about five and a half billion dollars of redemptions in June, part of roughly eighteen and a half billion pulled out between March and June, though April was actually an inflow month, while Asian funds posted their largest monthly outflow on record in that same June and global gold funds still took in around eight billion across the first half, a seventh consecutive quarter of inflows. Western selling into a falling price is real. The tidy story in which the East takes delivery of metal while the West dumps paper was a first-quarter phenomenon, and the second quarter declined to honour it.
Which leaves the drawdown itself, and Issue 008 already wrote the manual for reading one of these correctly, so this needs restating only once. Gold's price answers to the real interest rate and the dollar, this morning, this week, and with the Federal Reserve now priced at something like a one-in-three chance of raising rates on Wednesday and better than four-in-five by September, and two-year yields at seventeen-month highs, downward pressure on the metal is exactly what the textbook would predict and none of it touches the reason to own the thing in the first place. Gold's value answers to something considerably slower: the scale of the promises that have been made in a currency somebody can create, set against a supply of metal that nobody can expand on any timetable that matters. Those promises did not shrink in June. The interest bill on the American federal debt has passed a trillion dollars a year, the oil price is being set by drones over shipping lanes and pipelines rather than by any ministry, and the same Committee that may raise rates this week has no mechanism whatsoever for reducing the debt that its own rate rise makes more expensive to carry. The price answers to Wednesday. The value answers to the pen, and the pen has always kept a slower clock.
Half a century ago a war in the Middle East disrupted the oil supply, and a monetary system that had just cut its anchor loose reorganised itself around the consequences without anyone having to design the reorganisation or put it to a vote. It worked, on its own terms, for fifty years. It financed the deficits and the deployments, and it sent the bill to everyone holding the currency, in slices thin enough and slow enough that almost nobody wrote in to complain. This week a war in the same region put crude back around a hundred dollars, and the institutions that financed the last arrangement are steadily accumulating the one reserve asset that no government can issue and no treaty can dilute. Nixon told the country the measure was temporary. Fifty-five years later, the central banks appear to have concluded that he was right, just not about the part he meant.
|
