Contrarian macro · plumbing first

The Shadow Fed

Here you will see whether stablecoin supply simply follows Federal Reserve plumbing with a measurable lag — and what that plumbing says about next quarter's float.

Since 2019-12-11, weekly changes in dollar plumbing and weekly changes in stablecoin supply have moved together at +0.25 over the last 180 days, and the strongest alignment over 2019-12-11 to 2026-09-18 appears when the plumbing moved first by about 21 days (correlation +0.12 at that offset versus +0.01 with no shift). Over the last 30 days the plumbing expanded by $77.2B and stablecoin float grew by $3.3B.

Updated 19 September 2026 · all figures carry their own observation dates below

Net liquidity · 2026-09-16$5,869BFed balance sheet minus Treasury cash minus reverse repo · 30d change $+77.2B · 90d $+13.5B

Charlie's readThis is the dollar fuel left in the system after the Treasury's cash pile and money-market funds take their cut of the Federal Reserve's balance sheet. Over the last 30 days it expanded by about $77.2B (90 days: $13.5B). More fuel in the pipe usually means more room for stablecoins to grow a few weeks later — that delay is what this page measures.

Stablecoin float · 2026-09-18$310BAll tracked dollar-pegged issuers · 30d change $+3.3B · 90d $-3.2B

Charlie's readThis is all the dollar-pegged stablecoin money in circulation — the crypto economy's cash. It grew by about $3.3B over 30 days. A growing float means fresh money is actually arriving in crypto, not just shuffling between coins.

90-day rolling correlation+0.17Weekly plumbing change vs weekly float change · 180-day reading +0.25 · long-run middle +0.01

Charlie's readCorrelation measures how often two things move together, from −1 (always opposite) to +1 (always together). At +0.17 the two are barely linked right now — the float is dancing to its own tune (new issuers, yields), not the Fed's.

Measured lead+21 daysStrongest alignment over the scan, at +0.12 vs +0.01 with no shift

Charlie's readOf every delay we tested, the two series line up best when the plumbing is given a 21-day head start — so the Fed's balance sheet really does tend to move first, and stablecoins follow roughly 3 week(s) later. The alignment is weak, though — treat the lag as a hint, not a law.

The pipe and its shadow

Net liquidity is the dollar fuel left over after the Treasury and the reverse-repo facility take their cut of the Fed's balance sheet. Stablecoin float is what the crypto economy actually prints. If the second is just the first with a delay, the two lines should rhyme with an offset.

Net liquidity vs stablecoin supply, billions of dollars

Fed weekly balance sheet observed 2026-09-16 · Treasury cash observed 2026-09-16 · reverse repo observed 2026-09-18 · stablecoin float observed 2026-09-18

Takeaway: the two lines share their big turns. The question is not whether they rhyme — it is who moves first, which the lead-lag scan below answers.

Charlie's readingLiquidity expanded over 30 days; stablecoins grew. Net liquidity — the Fed balance sheet minus the Treasury's cash account minus the reverse-repo facility — stands at $5,869B as of 2026-09-18. It expanded by $77.2B over 30 days and by $13.5B over 90. Stablecoin float stands at $310B as of 2026-09-18, having grew by $3.3B over 30 days. The rolling 90-day correlation between the two change series reads +0.17, against a long-run middle near +0.01.

Does the shadow follow, and by how much?

Each week of plumbing change is tested against the float change 30 days before it through 30 days after it. A peak on the positive side means the plumbing moved first and the float followed — the shadow-Fed thesis, with a number on the delay.

Correlation by day-shift of the stablecoin series

7-day changes, 2019-12-11 to 2026-09-18 · 2,474 paired weekly moves

Shifting the stablecoin change series from 30 days early to 30 days late against the plumbing change series, the correlation peaks at a shift of +21 days (+0.12), versus +0.01 with no shift at all. A peak away from zero is what a lagged-derivative relationship looks like; a peak at zero would mean the two simply share the same news on the same day.

Rolling 90-day correlation of weekly changes

Same pairing, sliding window · 2019-12-11 to 2026-09-18

Takeaway: the link breathes. When the rolling reading sits near zero, the float is being driven by its own issuers and yields, not by the balance sheet.

What the plumbing says about next quarter

The verdict uses only the current direction of net liquidity and the measured historical lead. It is a plumbing statement, not a price target.

Verdict · 19 September 2026The plumbing gives no clean signal this quarter. The strongest historical alignment is weak (+0.12 at a +21-day shift), too weak to trust the apparent lead, so stablecoin supply is not reliably shadowing the balance sheet right now. Treat the float as driven by something else — yields, new issuers, or demand the plumbing does not capture — until the correlation rebuilds.

The contradictionThe least convenient fact for this page: even at its best offset the two series explain only part of each other (+0.12 correlation is not a lock), and the relationship strengthens and fades — the 90-day rolling reading has ranged widely around its middle of +0.01. Stablecoin supply also carries its own drivers: new issuers, yield competition with treasury bills, and non-dollar demand that never touches the Fed's balance sheet. The shadow follows the plumbing only while those drivers stay quiet.

How this could be wrongThis reading could be wrong because weekly Fed and Treasury figures are revised, the reverse-repo facility is now near empty ($1B as of 2026-09-18), so its drainage phase is mostly behind the data, and correlation of changes over 2474 paired weekly moves cannot separate plumbing causation from a shared reaction to the same rate expectations. If the peak lead collapses toward zero in coming months, the shadow-Fed framing should be retired, not defended.

The bottom line

Everything above, compressed into plain English: what the numbers say right now, what Charlie watches next, and what would prove the whole thesis wrong.

Charlie's read · 19 September 2026As of 19 September 2026: the Fed's plumbing has expanded by $77.2B in 30 days while stablecoin float has grew by $3.3B, and the link between them is weak (best at a +21-day shift, correlation +0.12). The historical link is too weak right now to lean on, so the honest call is no clean signal for next quarter's float. What Charlie watches next: the weekly Fed balance-sheet prints — sustained expansion should surface in the float within the measured lag. What would prove the thesis wrong: the lead-lag peak collapsing toward zero, or the float making a big move while the plumbing sits still — either means stablecoins answer to something else.

For the statistically curious

The full plumbing — exact data definitions, statistical choices, and honest limitations. Nothing below changes a number above; it only explains how the numbers were built.

Methodology, data sources and limitations

How the numbers are built

Net liquidity = Fed total assets ($6,747B, weekly, observed 2026-09-16) minus Treasury General Account ($877B, observed 2026-09-16) minus overnight reverse repo ($1B, daily, observed 2026-09-18). All in billions of US dollars.
Stablecoin float is the combined circulating supply of all tracked dollar-pegged issuers, daily, observed 2026-09-18.
Both series are compared as 7-day changes on paired dates from 2019-12-11 to 2026-09-18. Correlations are Pearson on those changes. The scan shifts the float series by ±30 days.
Weekly balance-sheet readings are carried forward between publication dates; daily readings within a Fed week are therefore partly mechanical, which softens short-window correlations.
Correlation of changes cannot prove causation; both series may react to the same rate expectations. Missing data is shown as missing — nothing on this page is interpolated across gaps or set to zero.