The world's money is one machine.
Learn to read it
Not an article. An instrument panel. Twenty-three instruments, each one a thing you turn before it is a thing you are told — because the gap between your guess and the truth is the only lesson that sticks.
You are standing on a hinge
Three clocks are converging: a debt cycle that is about to be refinanced at a price nobody planned for, a monetary regime that has run out of its old lever, and a technology absorbing capital faster than any build-out in living memory. Each is ordinary alone. Together they set the price of everything you own.
All three are true at the same time. That is the whole reason this page exists.
Three clocks. Turn them to now
Long cycles are not mysticism — they are the lifespan of arrangements. A constitution lasts about as long as the memory of why it was written. A debt load lasts about as long as the generation that took it on. A machine reorganises the economy for about twenty years. Each is ordinary alone. Drag all three to the present and watch the lamp.
Turn all three dials toward now. One late cycle is ordinary. Two is uncomfortable. Three at once is a different regime — and that is the claim this page has to earn.
Cycle lengths are the owner's framing, not a measured series — read the long case at the 250-year cycle · practitioner framing
Written form. Three long arrangements are late at the same time: the political settlement, the debt load, and the dominant technology. Any one of them late is ordinary and the usual macro relationships still hold. All three late together is a regime in which those relationships bend — which is why a page about crypto has to start with centuries rather than charts.
Four facts. Tap the ones you think are true
No trick, no scoring, no wrong answer held against you. Just commit before you read on.
All four are true, at the same time, right now. Hold that discomfort — it is the honest starting position, and every instrument from here exists to make it legible rather than to explain it away.
Written form. All four of the following are true, at the same time, right now. American factories were expanding — ISM manufacturing at 53.3, above 50 — while bitcoin fell 54%, from 126,198 to 58,000. Gold rose 21.3% over the same twelve months, in a year with no crisis headline. The dollar's share of world reserves rose to 57.13%, the one most readers guess backwards. And all three happened together, which is the point: any one of them alone is a story; together they are a regime.
ISM manufacturing 53.3 · bitcoin 126,198 → 58,000 · gold +21.3% over twelve months · US dollar share of world reserves 57.13%. Each sourced and dated in sources & method. verified
Before anyone explains anything to you, guess
In February 2020 the world's largest stock market fell 34% in five weeks. In the middle of that, over two days, crypto did something too. Put your answer on the chart. You will not be told whether it was more or less until you commit.
Here is why. There is only one machine, and it is a set of basins
The world's money is not divided into markets. It is one body of water sitting in pools of wildly different size, drawn here to scale — every circle's area is its market. Pour the same dollars into each and they all rise by the same amount, which means they all rise by a completely different height. Turn the tap.
The same dollars raise the pond 64 times more than the reservoir. Nothing about crypto is special here — only its size.
View the data
Wave height = inflow ÷ pool size. Debt & credit $145.1T and world equities $126.7T are SIFMA, end-2024, quoted from that one edition deliberately so bonds and equities are measured on the same basis; that is why they carry the aged flag rather than a fresher, incomparable number. US equities $75.3T is Siblis Research, 1 Jul 2026. All crypto $2.274T is live from a different source. Circle area is proportional to pool size. The teaching output — that the pond moves about 64× the reservoir on the same dollars — is asserted by the model's tests.
Written form. $200 billion of new borrowing capacity raises a $145.1T debt market by 0.14%, $126.7T of world equities by 0.16%, $75.3T of US equities by 0.27%, and $2.274T of crypto by 8.80%. Same water, four wave heights, one ratio: the pond moves about sixty-four times as far as the reservoir. Reverse the flow and the ordering holds, but the pond drains faster than it filled — the last buyer leaves before the first one does.
Now run it backwards — and notice it is not the same shape
Nearly everyone has only ever been shown the up direction. Drag the Cascade's control below zero and the pond empties first and fastest — but the drain is not the fill played in reverse. On the way in, buyers arrive one at a time. On the way out, the last buyer leaves before the first one does, so the exit is always narrower than the entrance was.
↑ Take the Cascade control to −$400B and watch the pond, not the reservoir.
Price impact is roughly order size divided by book depth. A rule of thumb, not a law — but it is why the same money is a ripple in one market and a wave in another.
Price impact ≈ order size ÷ book depth — a rule of thumb, not a law. Stocks and bonds do this too; crypto does it faster, with a smaller float and less depth to absorb it. That reflexivity is why a fall manufactures its own forced selling.
Written form. At full depth, $200m moves the price about 2%. At a third of that depth the same order moves it about 6%, and at a crisis book roughly 25%. Depth is not a property of the asset — it is the number of people willing to be on the other side, and it disappears precisely when everyone needs it.
“Everyone argues about which pond to stand in. Almost nobody asks where the water comes from, or who is allowed to turn the tap.”
The water is not cash. It is permission to borrow
Nobody prints the water. It appears when someone pledges something good and is lent against it — and the amount they can borrow depends on one thing: how still that collateral is holding. Volatility is the valve. Turn it up and the water disappears without anyone deciding to remove it.
Calm collateral borrows more. This is why the water is not cash — it is borrowing capacity, and volatility is the tap.
Per $1T of equity, re-pledged through a chain of loans. Haircut ≈ 2 + 0.62 × volatility, the shape used across dealer risk desks; the page's figure spine holds the sourced haircut schedules by asset class.
Every financed position is holding. Turn the margin dial up and watch the order in which they let go — nobody decides it, and crypto is never last.
Written form. At 16% volatility a lender takes roughly a 12% haircut, so $1T of equity supports a position of about $8.4T — it borrows the other $7.4T against it. At 45% volatility the haircut is about 30% and the same $1T of equity supports only about $3.3T. No central bank acted; two-thirds of the water evaporated because the collateral started shaking.
“I watch haircuts before I watch prices. When the schedules widen, I already know what next month looks like, because the water is leaving before anyone has decided to sell.”
Counter-evidence, stated: haircut schedules are not public in real time, and this read has been early twice in the last decade — in 2018 and again in 2023 — by enough months to be painful. The chart below it is the same data without the interpretation, so you can disagree with him on the evidence.
A yield is a see-saw, and the government sits on one end
A bond is a fixed set of promised payments. So the only thing that can move is the price you pay for them — and the yield is just the other end of that plank. Push one end down and the other rises. There is no third option.
Yield down, price up. A bond is a fixed set of promised payments, so the only thing that can move is what you pay for them.
Now sit at the other end of the plank
Every few weeks a government must sell more debt than any single buyer wants. You are one of the bidders. You submit the yield you require. So does everyone else, and the auction fills from the cheapest money upward until the issue is gone.
$25bn of thirty-year paper is for sale. Name the yield you require. You will find out what it costs to be reasonable.
Nobody sees the other bids until the auction closes. That is the whole experience of being in the room: you commit a number into the dark, and the price of thirty years of government borrowing is the sum of everyone doing that at once.
Stop-out yields are illustrative scenarios calibrated to observed 30-year auction tails, 2023–2026 · illustrative model, not a quoted figure
Where the curve sits right now, US Treasury daily par yields: 2-year 4.33% · 10-year 4.69% · 30-year 5.16% · 2s10s spread 36bp. verified
“Nothing in this machine happens at the same time. Everything in it happens in the same order.”
The bond market has a fear gauge. It is the fuse for everything else
Stocks have the VIX and everybody watches it. Bonds have the MOVE index and almost nobody does — which is backwards, because bond volatility is what sets the haircut, and the haircut is what sets how much anyone can borrow. Before the ladder: where do you think it reads today?
Far higher than reality. Today reads 74.67. The instrument that would tell us a sovereign debt crisis had begun is sitting almost silent.
Every rung is a real reading. Put your alarm where you would want to be woken, and see which episodes it would have caught.
Set it at 120 and you sleep through the calm and wake for every genuine dislocation — the gilt crisis, covid, SVB, 2008. Set it at 70 and you are awake permanently, which is the same as not having an alarm.
ICE BofA MOVE index. Today 74.67 · 2026 range 55.77–115.02 · gilt crisis 158.99 · covid 164 · SVB 182.64 · 2008 peak 264. No reading exists for 1994 — the index launched that year — so that episode is taught through yields alone. verified
Cboe VIX, stocks' own fear gauge, for comparison: 18.70. verified
“This is the instrument I watch for the Katana. An uncontrolled yield spike is the best single signal that a sovereign debt problem has become a sovereign debt crisis — and the MOVE index is what tells you the market has stopped believing the sale will go smoothly. When this lights, the sharp drop we exist to catch is already underway.”
Counter-evidence, stated: the MOVE index has spiked without a crisis following — 2013 and 2018 both resolved without dislocation — and it is quiet today, so nothing here says anything is imminent. A fuse you can read is not a prediction that it will burn.
That is the fuse. Here is what it reads today, with eight instruments beside it
Nine triggers, one row each: what it is, what it says in plain English, where it sits now, and what that reading has historically preceded. This is the board this firm actually watches. It is the one screen on this page worth returning to, because it is the only one whose answer changes without you.
The macro fuse is unlit and the crypto plumbing is already stressed. That combination is the whole picture today, and it is an unusual one — the instruments that would tell you a sovereign debt crisis had begun are quiet, while the vehicle that was crypto's marginal buyer trades below the value of its own coins and the oil market has no slack left. Nothing here says anything is imminent. A board worth keeping is one that can say so.
What this board is, and is not. It states what the instruments read and what those readings have historically preceded. It is not a signal to buy or sell anything, no price target is given or implied, and history preceding something is not history causing it. Five of the nine refresh from live feeds hourly; the other four are curated and dated because no feed on the current subscriptions carries them, and each says which it is. Every figure here traces to an entry in the fact appendix below, with its source and observation date.
The relationship is real. You are just looking at it on the wrong day
Two series that look unrelated side by side often lock together once you slide one of them through time. Water takes time to reach the far basin. This one instrument carries nine different relationships on this page — this is the first.
Slide the second series back through time. Watch the correlation number, not the lines — it tells you when the water arrived.
Peak correlation for this pair sits at 11 months. Borrowing capacity created at the top of the machine takes about a year to reach the smallest pool.
Series shapes are schematic in this design file; at build they bind to the fact spine (global liquidity, monthly, 2019–2026 · crypto aggregate cap, monthly) · relationship contested — direction agreed, magnitude disputed
Written form. Compared month for month, global liquidity and crypto appear weakly related. Shift the crypto series back eleven months and the correlation reaches its maximum. The lag, not the level, is the finding: liquidity created at the top of the machine takes about a year to reach the smallest pool.
Why do lags exist at all? Because of who owns the thing
This is the idea almost nobody explains, and it turns a list of correlations into a theory. The lag equals the mandate speed of the owner. A dealer must act in minutes. A pension committee meets quarterly. A central-bank reserve manager thinks in years. Nothing mystical happens in between — the money simply cannot move faster than its owner is allowed to move it.
So gold's signal takes months to reach crypto not because gold is wise, but because the people buying gold are central banks and households — the two slowest owners on earth. And the Nasdaq reaches crypto in days because it is the same owner, with the same funding cost, pressing the same button.
Every alarm is a choice about how often you want to be wrong
Here is the bond market's fear gauge, quarter by quarter, with the four genuine stress events marked. Decide where your alarm fires. History will tell you what your choice would have cost you.
An alarm set too low fires every quarter and teaches you nothing. Set too high, it fires the week after you needed it.
There is no setting that catches four and cries none. That is not a flaw in the gauge; it is what a gauge is.
Written form. Across 24 quarters, an alarm at 100 fires 4 times and catches 3 of the 4 real stress events with 1 false alarm; at 80 it fires 8 times and catches all 4, with 4 false alarms; at 140 it fires twice and catches two. Sensitivity is bought with false alarms at a fixed exchange rate.
The AI trade did not take money from crypto. It took the next dollar — which is the same thing
There is one risk budget in the world and it is not divided by asset class. When a trade is running, it is fed. Everything with the same shape — liquid, volatile, discretionary — waits. Drag the dial through the cycle and watch what happens when the feeding stops.
Both columns are positions in a cycle, 0 to 100 — a model of the mechanism, not a measured quantity. The figures underneath each one are real, sourced and dated.
Today's setting. The AI complex sits at or near its high while crypto sits 48.5% below its own — Bitcoin at 65,030 against an all-time high of 126,198, and 58,000 at the 2026 low. Same investors, same risk budget, same brokerage account: one pond filled while the other drained. Nobody wired money out of crypto and into Nvidia. They simply stopped adding to one and kept adding to the other, and at these relative sizes that is enough.
Counter-evidence, stated. Stated plainly: no institution publishes a figure for money rotating between the AI trade and crypto, so there is no dollar number on this screen and there will not be one. What is sourced is where each market sits. The rotation itself is a mechanism, and a mechanism is not a forecast.
Written form. One risk budget funds both the AI trade and crypto. Through 2025 and 2026 the AI complex was fed while crypto was not: the Nasdaq 100 sits at 28,998 near its high while Bitcoin trades at $65,030 against an all-time high of $126,198 — 48.5% below, and 54% below at the 2026 low of $58,000. Nobody wired money out of one and into the other; investors simply stopped adding to the smaller, more volatile thing and kept adding to the one that was working, and at these relative sizes that alone is enough. The mechanism to understand is what happens next: when a trade cools, the money does not leave the system, it looks for the next thing with the same shape — liquid, volatile, and small enough that the same sum moves it further. The honest limit is that no institution publishes a figure for this rotation, this page does not invent one, and the money may instead go to bonds or leave risk entirely.
Every currency in the world is tied to one hub
Most of the world borrows in a currency it cannot print. That makes the dollar not America's scoreboard but the world's funding line — and it means a rising dollar is a rate rise for everybody, decided by nobody. Pull the hub and feel the tethers.
At rest the web looks like nine independent currencies. It is not. Every one of them is tethered to the same hub, because the world’s debts are written in a currency its debtors cannot print.
“As the dollar dies it takes the others first. The tethers snap in order of who borrowed most in a currency they cannot print, and that plays out over something like twenty years — not a headline, a generation.”
The dollar's quantitative dominance — payments, invoicing, credit — is intact and in places still growing. What is being diluted at the margin is its monetary role as the neutral reserve asset, mostly by gold. Both statements are true, and neither is a retraction of the other.
DXY 101.47 · reserve share 57.13% · FX share 89% · SWIFT 59.1% · trade finance 82.5% · offshore dollar credit $14.3tn · foreign Treasury holdings $9.0tn. verified the twenty-year path is practitioner framing
On the day it matters, the order of sale is decided by size, not by conviction
A margin call does not ask what you believe. It asks what you can sell in an hour without moving the price against yourself. Put these five pools in order, largest first — then find out why the order matters more than the ranking.
- 1 All crypto $2.274T · live aggregate
- 2 All gold ever mined unavailable · no sourced total on this page — see Sources & method
- 3 World equities $126.7T · SIFMA, end-2024 · aged, on purpose
- 4 US equities $75.3T · Siblis Research, 1 Jul 2026
- 5 Debt & credit $145.1T · SIFMA, end-2024 · aged, on purpose
Now the point. In a forced sale, nobody starts at the top of this list. They start at the bottom — because the small pool is the only one they can exit inside an hour. The ranking is size. The selling order is its reverse.
Written form. Largest first: debt & credit $145.1T; world equities $126.7T; US equities $75.3T; all gold ever mined (no sourced total on this page); all crypto $2.274T. Debt & credit and world equities are SIFMA end-2024 — the same edition the Cascade quotes, so those two are measured on one basis; US equities is Siblis Research, 1 Jul 2026. In a forced liquidation the order reverses — the smallest, most liquid-looking pool is sold first, which is why it falls furthest.
And underneath all of it, one bridge that everyone crosses and nobody looks at
For thirty years Japan lent the world money at almost nothing, and the world borrowed it to buy everything else. That funding line is the bridge. Raise the Bank of Japan's pillar and the load has to move somewhere — and the load is other people's positions in your market.
Nothing looks wrong here, and that is the point: the trade is being loaded, not unwound. Record short yen positioning is the powder. The Bank of Japan is the match. The authorities already spent ¥11.73tn in a single month defending the level, which tells you they consider it dangerous.
USD/JPY 163.81 · BoJ 1.0% · JGB 10y 2.901% · JGB 30y 4.03% (a level, not a first-ever breach) · CFTC net short yen 155,100 · MoF intervention ¥11.73tn · debt/GDP 256%. verified
The same bridge, priced daily
Below is the everyday version of that structure: what it costs the world to swap into dollars, quarter by quarter. At par it is invisible. When it tilts, somebody good is selling something good to raise something ordinary.
At par, the bridge carries load invisibly. You only learn it exists on the day it tilts.
“Price is not set where you buy. It is set where somebody is forced.”
Oil is not a correlated asset. It is a parameter
People look for oil in the crypto chart and find nothing, then conclude oil does not matter. It matters more than almost anything — just not directly. Oil decides whether the central bank is allowed to cut. Follow the chain one step at a time and watch where it ends.
Oil never had to touch crypto. It closed the valve at the top of the Cascade, and the pond simply never received the wave. That is what a parameter does — it changes the rules the whole machine runs under, without appearing in a single correlation.
Brent $98.38 · WTI $87.88 · US crude stocks 411.7m bbl · SPR 311.45m bbl · distillates 109.6m bbl · OECD cover 50 days. The 1973 precedent: $2.90 → $11.65. verified, EIA and IEA
Written form. Oil up → headline inflation up → the central bank cannot cut into a supply shock → no new borrowing capacity enters the system → the smallest pool never fills. Oil does not correlate with crypto; it sets the Federal Reserve's permission, and permission is what the Cascade runs on.
“Inventories are running thin and I expect us to reach bottom barrel around October. If that happens while the fuse is already dry, that is the setup I care about.”
Counter-evidence: 50 days of OECD cover is thin, not critical, and inventory paths reverse routinely on one OPEC decision. No agency forecasts this; it is his read alone, and the page shows no future-dated inventory figure.
Not a countdown and not a forecast — four numbers that would have to move together before any of this mattered.
A borrowed balance sheet is a reactor, and leverage is the control rod
Companies now hold volatile assets with borrowed money and trade above what those assets are worth. That premium is not a mystery: it is the market paying for leverage it cannot get itself. Pull the rod out and see how far the fuel can fall before the equity is gone.
At 2.5× a 60% fall in the asset takes the whole equity. The reactor does not fail because the fuel was bad. It fails because the rods were pulled too far.
Below 1.0× the same flywheel runs backwards. Issuing stock now destroys bitcoin per share, so the funding mechanism closes exactly when it is needed. Nobody chose this; arithmetic did.
Fire the ratchet to see what happens when a security designed to hold par stops holding it.
Prices are opinions. Liabilities are contracts. That is why this tells the truth earlier than the bitcoin chart does.
Sources disagree on the denominator — whether the preferred stack counts as equity or as debt. The disagreement is the finding, so the page prints the range and names it rather than picking a side. One point estimate here would be a false precision.
- No forced selling of the underlying asset.
- No missed preferred dividend.
- No covenant breach and no rating action — the rating itself is unavailable.
Stated as plainly as the events, because a professional reader looks for the non-events first. The reactor running hot is not the same as the reactor having failed.
Leverage model: wipe-out fall = 1 − 1/leverage. Company figures: 843,775 BTC held · average cost $75,476 · total debt $8.197bn · preferred equity $8.36bn (aged — flag it) · STRF $87.87 · STRK $61.46 (aged) · STRD $60.96 · STRC contested, no number shown · rating unavailable. Reference bitcoin $65,030. verified from filings two aged, one contested, one unavailable
Six things that were supposed to be true
Each card holds a rule that macro textbooks still teach. Turn it over to find what the last four years actually did. None of these broke randomly — every one bent for the same reason you have been turning dials about.
Written form. Six cards, front then back. "Yields rise, gold falls" — both rose together through 2024–26; gold stopped trading against real rates and started trading against confidence in the issuer. "Crypto is an inflation hedge" — it sold off hardest in the fastest inflation in forty years, because it is a liquidity asset and liquidity was being withdrawn. "A strong dollar means a strong economy" — it mostly means offshore borrowers are being squeezed; the dollar is the world's funding currency before it is America's scoreboard. "Stocks and bonds diversify each other" — they fell together in 2022 and have re-correlated since, because one shared input, the discount rate, now dominates both. "Oil up means growth up" — oil now reads as a thermostat on the supply side, not a vote on demand, and its correlation with equities flipped sign twice in three years. "Crypto stopped working" — crypto changed owners: the marginal buyer became an allocator with a risk committee, so it began obeying the discount rate like everything else those institutions own.
Crypto did not stop working.
Crypto changed owners
The marginal buyer became an institution with a risk committee, a funding cost and a quarterly report. So the pond joined the plumbing — and started obeying the same discount rate as everything else those institutions own. That is not a failure. It is the price of admission, and it is the whole reason the rest of this machine is now yours to read.
Six questions. None of them asks you to remember a number
Each one tests a mechanism — the thing that would still be true if every figure on this page were a year out of date. Every wrong answer here is something a great many people who talk about macro in public actually believe, so getting one wrong is information, not failure. Answer, then read why.
The same $100bn arrives in the system. Why does it move crypto so much further than it moves the bond market?
Why. Size, not character. $2.274tn against $145.1tn: the identical inflow has 64 times the effect. Risk appetite and trading hours change who shows up; they do not change the arithmetic of a small pool.
Bond volatility spikes. What actually transmits that into every other market?
Why. This is the mechanical answer and it is the one that matters. Fear is a description; the haircut is a number in a risk system. When it rises, positions are reduced by people who did not choose to reduce them — which is why the chain unwinds in the same order every time.
Gold rose nine months ago. Crypto has not. What does that most likely mean?
Why. Two series that look unrelated side by side often lock together once one is slid through time. Water takes time to reach the far basin. Judging a lagged relationship on today's prices is the single most common way to conclude something is broken when it is merely slow.
A rising dollar tightens financial conditions for the entire world, including countries whose central banks did nothing. Why?
Why. The dollar is the world's funding currency before it is America's scoreboard. A borrower in São Paulo or Jakarta owes dollars, earns something else, and gets an automatic interest-rate rise nobody voted for. That is why every emerging-market crisis of the last forty years happened with the dollar rising.
Strategy's market value falls below the Bitcoin it holds. Why does that matter beyond one company?
Why. Below parity the machine simply stops working — no forced sale required, and no verdict on Bitcoin's value implied. A buyer that was absorbing supply is absent, and absence moves price as surely as selling does. It sits at 0.77 today, which is why it is the canary rather than the coffin.
In one sentence: why did the old relationships bend all at once?
Why. Regulation opened the door; it did not do the bending. When the marginal buyer became an allocator with a funding cost and a risk committee, the pond joined the plumbing — and started obeying the same discount rate as everything else that allocator owns. That is not a failure. It is the price of admission.
You have seen the machine. Walk the movements with the instruments in your hands rather than reading past them — every dial on this page exists because that concept does not survive being read.
Walk again
That is the movement where a mechanism did not land. The instruments are still there and still respond to you.
I read the macro machine at Onlooker level — 0 of 6 mechanisms. The Macro Masterclass, Crypto XLNC Academy · learn.cryptoxlnc.com/macro-masterclass
Reading this machine is what we do for a living. If you want to see these same instruments on live data every month, and the reasoning behind what they are saying, that is the next room.
What this page teaches: how size sets amplitude, how collateral transmits a shock, why lags hide real relationships, why the funding currency binds the world, and how ownership decides which rules apply. What it does not claim: that any of these instruments predicts anything, that history repeating is history’s obligation, or that the figures here are anything other than sourced and dated at the point of use. Nothing on this page is financial advice.
Without JavaScript, the assessment is six open questions — which is the honest form of it anyway. One: the same $100bn arrives in the system; why does it move crypto further than the bond market? Two: bond volatility spikes; what actually transmits that into every other market? Three: gold rose nine months ago and crypto has not; what does that most likely mean? Four: a rising dollar tightens conditions for countries whose central banks did nothing; why? Five: Strategy’s market value falls below the Bitcoin it holds; why does that matter beyond one company? Six: in one sentence, why did the old relationships bend all at once? The answers are size rather than character, the haircut rather than fear, the lag rather than a breakage, $14.3tn of debt in a currency the borrowers cannot print, a marginal buyer that can no longer bid, and crypto changing owners. Each one is worked through in the movement it belongs to.
Nothing on this page is financial advice, an offer, or a recommendation. No price targets are given and no instruction to buy or sell anything is implied. Figures are sourced and dated at the point of use; where a figure is contested, the disagreement is named rather than resolved. Interactives are teaching models, not forecasts. Crypto XLNC is a manager of money and has positions in assets discussed here.
Every figure on this page, dated
Every number used above traces to one of the 94 entries below, each carrying the date it was read and a link to where it came from. Nothing is estimated to fill a gap. Where sources disagree, the disagreement is shown as a range rather than resolved into one number; where a figure could not be verified, it is not used. Primary sources — the institution that actually produces a number, not a service that repeats it — are used wherever one exists; where the only available source is a secondary aggregator, it is shown as such, never upgraded into something it isn't.
- Figures cited
- 94
- Institutions
- 64
- Shown as a range, never a guess
- 11
- Claims considered and excluded
- 8
The bibliography, by institution
Institutions cited more than once get their own table; every other primary source and market-data feed is listed in the one after it, alphabetically.
Aggregate trackers · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| Everything that is not Bitcoin | $870bn–$1,060bn | 15 Jul 2026 | contested |
| Stablecoins outstanding | $305bn–$320bn | 15 Jul 2026 | contested |
BIS · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| US 10-year, November 1994 | 8% | 30 Nov 1994 | verified |
| US 10-year, January 1994 | 5.6% | 31 Jan 1994 | verified |
Cboe · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| VIX, the S&P 500's fear gauge | 18.7 | 23 Jul 2026 | verified |
| VIX peak close, March 2020 | 82.69 | 16 Mar 2020 | verified |
CNBC · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| US 30-year Treasury yield | 5.16% | 24 Jul 2026 | verified |
| Bitcoin's 2026 low, a 21-month low | $58,000 | 30 Jun 2026 | verified |
CoinGecko · 3 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| Bitcoin's share of all crypto | 56.5% | 25 Jul 2026 | verified |
| Ethereum's share of all crypto | 9.9% | 25 Jul 2026 | verified |
| Total crypto market value | $2.27tn | 25 Jul 2026 | verified |
EIA · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| The US Strategic Petroleum Reserve, the lowest since April 1983 | 311.45M bbl | 17 Jul 2026 | verified |
| US diesel and heating oil in storage | 109.6M bbl | 17 Jul 2026 | verified |
EIA historical series · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| Oil by January 1974 | $11.65 | 31 Jan 1974 | verified |
| Oil before the 1973 embargo | $2.9 | 19 Oct 1973 | verified |
Federal Reserve · 3 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| Fed funds target, lower bound | 3.5% | 17 Jun 2026 | verified |
| Fed funds target, upper bound | 3.75% | 17 Jun 2026 | verified |
| US government debt held abroad | $9tn | 31 Mar 2025 | verified |
ICE BofA / Convex · 7 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| MOVE index, the bond market's fear gauge | 74.67 | 21 Jul 2026 | verified |
| MOVE 2026 high | 115.02 | 26 Mar 2026 | verified |
| MOVE 2026 low | 55.77 | 26 Jan 2026 | verified |
| MOVE at the March 2023 banking stress | 182.64 | 20 Mar 2023 | verified |
| MOVE during the UK gilt crisis | 158.99 | 28 Sep 2022 | verified |
| MOVE in the March 2020 crash | 164 | 31 Mar 2020 | verified |
| MOVE all-time high, the 2008 crisis | 264 | 31 Oct 2008 | verified |
Institute for Supply Management · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| ISM manufacturing, the factory pulse | 53.3 | 30 Jun 2026 | verified |
| ISM services | 54 | 30 Jun 2026 | verified |
Market quotes · 3 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| STRD, the most junior | $60.96 | 24 Jul 2026 | verified |
| STRF, the most senior of the four | $87.87 | 24 Jul 2026 | verified |
| STRK, the convertible one | $61.46 | 6 Jul 2026 | verified |
Reuters · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| Japan 10-year government bond, the highest since 1996 | 2.9% | 9 Jul 2026 | verified |
| Japan 30-year government bond | 4.03% | 9 Jul 2026 | verified |
SIFMA Capital Markets Fact Book · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| The global bond and debt market | $145.1tn | 31 Dec 2024 | verified |
| The global stock market | $126.7tn | 31 Dec 2024 | verified |
SWIFT Global Currency Tracker · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| The dollar's share of payment value on SWIFT | 59.1% | 30 Jun 2026 | verified |
| The dollar's share of trade finance | 82.5% | 30 Jun 2026 | verified |
TradingEconomics · 6 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| Gold, per ounce | $4,052–$4,066 | 24 Jul 2026 | verified |
| The dollar index | 101.47 | 24 Jul 2026 | verified |
| UK 10-year gilt, the highest in the G7 | 5.05% | 24 Jul 2026 | verified |
| Yen per dollar | ¥163.81 | 24 Jul 2026 | verified |
| France 10-year, the highest since June 2009 | 3.92%–3.94% | 15 Jul 2026 | verified |
| Japan's government debt against its economy | 256% | 1 Jan 2026 | verified |
US Treasury · 3 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| US 10-year, one month earlier | 4.41% | 24 Jun 2026 | verified |
| US 10-year 2026 low | 3.97% | 27 Feb 2026 | verified |
| How far the 10-year moved in seven sessions, March 2020 | 64bp | 31 Mar 2020 | verified |
World Gold Council · 2 figures
| Figure | Value | As of | Confidence |
|---|---|---|---|
| Gold's twelve-month return | 21.3% | 30 Jun 2026 | verified |
| Gold bought by central banks in one quarter | 244t | 31 Mar 2026 | verified |
Other primary and market-data sources
47 institutions cited once each — market quotes, exchange data and statistical releases behind a single figure on this page.
| Institution | Figure | Value | As of | Confidence |
|---|---|---|---|---|
| Advisor Perspectives | US 2-year Treasury yield | 4.33% | 24 Jul 2026 | verified |
| Advisor Perspectives / ETF Trends | US 10-year Treasury yield | 4.69% | 24 Jul 2026 | verified |
| Aggregate of issuer disclosures | Net money into or out of Bitcoin ETFs in 2026 | −$5.4bn | 20 Jul 2026 | verified |
| Bank of Japan | Bank of Japan policy rate | 1% | 16 Jun 2026 | verified |
| BIS Bulletin 95 | The Nikkei on 5 August 2024, its worst day since 1987 | −12.4% | 5 Aug 2024 | verified |
| BIS Bulletin 95 / CoinGecko | Where Bitcoin traded 48 hours into the August 2024 unwind | $49,111 | 5 Aug 2024 | verified |
| BIS global liquidity indicators | Dollar debt owed by borrowers outside America | $14.3tn | 31 Dec 2025 | verified |
| BIS OTC derivatives statistics | Global derivatives outstanding, by notional | $846tn | 30 Jun 2025 | verified |
| BIS Triennial Survey | Share of all currency trades with a dollar on one side | 89% | 30 Apr 2025 | verified |
| Cboe / BIS Bulletin 95 | VIX intraday spike, 5 August 2024 | 65.73 | 5 Aug 2024 | verified |
| Cboe Q2 2026 | SPX's share of all index options trading | 81% | 30 Jun 2026 | verified |
| CFTC | Speculative bets against the yen, a record | 155,100 | 6 Jul 2026 | verified |
| CNBC / Forbes | What the big cloud companies plan to spend on AI infrastructure this year | $700bn–$770bn | 1 Jul 2026 | contested |
| CoinDesk | Strategy's preferred stock outstanding | $8.36bn | 31 Jan 2026 | verified |
| CoinGecko Q2 2026 report | Total crypto market value at the Q2 2026 close | $2.1tn | 30 Jun 2026 | verified |
| CoinGlass / CoinMarketCap | The altcoin season index | 47–58 | 15 Jul 2026 | contested |
| CoinMarketCap | Bitcoin's all-time high | $126,198 | 6 Oct 2025 | verified |
| CoinMonks | How many tokens now exist | 10,000,000 | 1 Jul 2026 | verified |
| CompaniesMarketCap | Nvidia's market value | $4.92tn–$5.01tn | 22 Jul 2026 | verified |
| Computed from US Treasury yields | The gap between the 2-year and the 10-year | 36bp | 24 Jul 2026 | verified |
| Disputed across sources; verify against World Gold Council or LBMA | Gold's 2026 high | $4,736–$5,589 | 24 Jul 2026 | contested |
| EIA Weekly Petroleum Status Report | US commercial crude oil in storage | 411.7M bbl | 17 Jul 2026 | verified |
| FinanceCharts / MacroTrends, from filings | Strategy's debt | $8.197bn | 31 Mar 2026 | verified |
| Forbes Advisor | WTI crude, per barrel | $87.88 | 24 Jul 2026 | verified |
| Fortune | Bitcoin | $65,030 | 24 Jul 2026 | verified |
| GuruFocus / Bitbo, from SEC filings | Bitcoin held by Strategy | 843,775 | 20 Jul 2026 | verified |
| GuruFocus, from SEC filings | Strategy's average cost per Bitcoin | $75,476 | 20 Jul 2026 | verified |
| IEA Oil Market Report, July 2026 | How many days of demand the rich world holds in storage, the thinnest since 2003 | 50 days | 15 Jul 2026 | verified |
| Il Sole 24 Ore | Italy 10-year | 3.9% | 15 Jul 2026 | verified |
| IMF COFER | The dollar's share of the world's currency reserves | 57.13% | 31 Mar 2026 | verified |
| Investing.com | Strategy's share price | $85.81 | 24 Jul 2026 | verified |
| Japan Times / Ministry of Finance | What Japan spent defending the yen in one month | ¥11.73tn | 27 May 2026 | verified |
| MacroAxis | The semiconductor ETF | $588.07 | 11 Jun 2026 | verified |
| MacroMicro / Adrian-Crump-Moench, Ferrante | The 10-year term premium, positive for the first time since 2023 | 0.47%–0.73% | 30 Jun 2026 | contested |
| MacroMicro / JPMorgan | The ten biggest companies' share of the S&P 500 | 37.46% | 31 May 2026 | verified |
| Market quote, supplied by Sim Khela | STRC, a security engineered to sit at $100 | $86.89 | 26 Jul 2026 | verified |
| Market trackers | Ethereum priced in Bitcoin | 0.0313× | 30 Apr 2026 | verified |
| Robert Shiller / Yale | The Shiller CAPE, a long-run valuation measure | 39.8× | 1 Jul 2026 | verified |
| S&P Global Ratings | Strategy's credit rating | unavailable | 1 Oct 2025 | verified |
| Siblis Research | The US stock market | $75.3tn | 1 Jul 2026 | verified |
| SpotGamma and corroborating estimates | Share of S&P 500 options that expire the same day they are traded | 45%–50% | 30 Jun 2026 | contested |
| The Block | Money held in US spot Bitcoin ETFs | $77.7bn | 20 Jul 2026 | verified |
| The Defiant / CoinDesk | Strategy's market value against the Bitcoin it holds | 0.77× | 30 Jun 2026 | verified |
| UK DMO / press | Where the January 2026 gilt auction cleared | 4.81% | 31 Jan 2026 | verified |
| Vantage / Forbes Advisor | Brent crude, per barrel | $98.38 | 24 Jul 2026 | verified |
| World Gold Council Central Bank Gold Reserves Survey 2026 | Central banks planning to add gold in the next year, a record | 45% | 16 Jun 2026 | verified |
| Yahoo Finance | The Nasdaq 100 | 28,998 | 22 Jul 2026 | verified |
What we do not claim
Stating this is itself a sourcing standard: a page that never says what it left out is not more careful, just less honest about it.
Failed verification — considered and excluded
- A claimed shortfall in a specific German government bond auction in mid-2026 — sourced to a single opinion column, never corroborated elsewhere. The auction mechanism itself is taught on this page; that specific event is not.
- A claim that Japan's 30-year government bond yield broke a specific round-number threshold for the first time on record. Sources conflict on whether or when that happened, so no first-ever claim is made.
- Two forward-dated oil-inventory figures describing months after this page's own compile date — an impossible claim to source, since the events had not happened yet.
- A Bitcoin-dominance percentage that does not match this page's live feed.
- A claimed single-day currency crash, new capital controls and an emergency international credit line, all for one emerging-market currency — uncorroborated, and contradicted by mainstream coverage.
- A bond-market-volatility reading for a historic episode that predates the index's own creation. That episode is taught through the yield move it produced, not through an index reading that cannot exist.
- Combined market-value or drawdown totals for crypto broken out by size tier. No source publishes a reliable figure for that specific cut, so none is shown.
- A specific multiple describing how far gold rose across a historic decade, offered without the calculation or the start and end points behind it.
Practitioner reads — labelled as the owner's opinion, not research
- Gold leading Bitcoin by a matter of months — the framing used on the correlation instrument earlier on this page, attributed there to practitioner reading, not to research. Peer-reviewed work finds the two negatively related at zero lag; the two claims are not mutually exclusive, and the page states the counter-evidence at the point where the lead is used.
- The manufacturing-cycle-to-crypto framing used on the same instrument — attributed there to the practitioners who popularised it, never to research. It has failed to hold across more than one past cycle, including inverting outright in the most recent one.
- A rotation of AI-driven investment flows into crypto. The direction is documented in market commentary; no authoritative figure sizes it in dollars, so none is shown — only the two drawdowns that motivate the observation.
- A forward view on where oil inventories head later in the year. Where it appears on the page it is presented explicitly as the owner's own read of the chart beside it, never attributed to the statistical agencies whose data sits underneath it.
Fact spine compiled 25 Jul 2026 · schema 1.0.0. Two feeds on this page (Bitcoin's price and total crypto market value) are served live by the cockpit rather than read from this file; the entries above are their verified, dated fallbacks.