The Dollar Early-Warning Indicator System

The Dollar Early-Warning Indicator System is a trend-oriented crisis radar built in four layers aligned with the monetary pyramid’s four circulation tiers (reserves / domestic dollars / offshore dollars / shadow dollars). It comprises 16 indicators designed to provide tiered escape windows up to 6 months before bank failures and shadow-banking liquidity crunches — standing in methodological opposition to mainstream official early-warning systems’ “volatility-oriented” approach (which can only provide hour-level response time).

The Framework As It Stands

This section is compiled from research drafts: it retains the original framework’s structure, terminology, and key formulations, including editorial bridging and external factual annotations; diagrams are drawn by the compiler following the original text’s structure.

This section is compiled from sessions 4 through 8 of the 2023 Mid-Year Main Course — “The Dollar Circulation Early-Warning System (I–V)” — totaling 5 sessions. The core problem diagnosis: all early-warning research frameworks publicly published by the IMF / World Bank / Fed / ECB have operated on “essentially the same approach from a decade ago to now: volatility-oriented — it doesn’t work” — when VIX / various stress indexes spike, the crisis has already occurred, leaving only hours of reaction time. By analogy, the Wenchuan earthquake bureau could only warn 10-odd seconds to half a minute in advance — copying this approach leaves no escape window at all.

Methodological Breakthrough: Rather than starting from volatility, the approach starts from the monetary pyramid’s layered structure — dividing into four layers of “reserve circulation (core) / domestic dollar circulation / offshore dollar circulation / shadow dollar circulation,” extracting 3–5 “trend-based” (not “volatility-based”) indicators per layer, and constructing a deep-defense radar that can detect crisis signs 6 months in advance.

Two Pervasive Subthemes

Subtheme A — The Four-Layer Defense Methodology (Trend ≠ Volatility): The methodological soul is the “circulation-volume school” (fed funds trading volume, GSE deposit balances, money market fund size, foreign savings share, triparty repo volume, TIC net inflows) standing against the mainstream “volatility school.” These indicators begin showing slope anomalies months before a crisis erupts, and one need not monitor thousands of indicators — just 16 suffice. Decision rule: the four defense layers are graded “inside-out” — problems in the reserve circulation = aircraft-carrier deck anti-aircraft gun (no escape); problems in domestic dollar circulation = close-range defense perimeter (hours-to-days); problems in offshore dollar circulation = mid-range intercept (weeks-to-months); problems in shadow dollar circulation = long-range warning (months-to-quarters).
Subtheme B — The Money-Pendulum Phenomenon + Long-Bond Rates as the Ultimate Risk Source: Total systemic risk does not disappear just because one layer is temporarily safe — it oscillates back and forth among the four layers of banks ↔ shadow banks ↔ offshore dollars ↔ reserves, like a pendulum. The 2023-03 Silicon Valley Bank crisis actually rescued shadow banking (money fled from banks into money market funds) is a prime example. The only ultimate risk source capable of simultaneously piercing all four defense layers is a single one: rising long-bond rates — which simultaneously devalue the medium/long-term bonds held by banks (enlarging unrealized losses on the asset side), devalue the repo collateral held by shadow banks (chain devaluation), and devalue the US Treasuries held by overseas central banks (accelerating outflows). This is the Achilles’ heel of the entire system. “The crux of US-China negotiations, the crux of every problem, the crux of whether shadow banking can survive.”

Eight Core Arguments

All existing official early-warning systems are “volatility school” and offer zero forward-looking value. Personally reviewing all early-warning research reports and frameworks publicly published by the IMF / World Bank / Fed / ECB, the conclusion: the volatility-oriented approach does not work — when VIX / various stress indexes spike, the crisis has already occurred, leaving only hour-level reaction time.
The monetary pyramid comprises four circulation tiers + five flow forms. Reserves (core) / domestic dollars / offshore dollars / shadow dollars — four circulations; within these, shadow banking also has two special forms: “direct collateral flow” and “derivative mutual flow,” totaling five forms. The four defense layers from inside out correspond respectively to “no escape / hours-to-days / weeks-to-months / months-to-quarters” escape windows.
Fed funds market trading volume is the first early-warning indicator for reserve circulation — trend-based, not volatility-based. Under normal conditions, the fed funds market consists of a stable structure of “three sellers (GSEs engaged in IORB arbitrage) + one buyer (foreign banks)”; from 2022-Q4, persistently rising volume indicates new players entering as borrowers — verified to be small and mid-sized regional US banks that “borrow even when unprofitable,” the only plausible explanation being severely insufficient reserves, “losing money to borrow.” This indicator sent a trend-based warning several months before the March 2023 Silicon Valley Bank collapse.
The Federal Home Loan Banks (FHLBanks) = the second Federal Reserve — its bond issuance and loan volumes are “short-term” early warnings (a bank failure within 1-2 weeks). Normal business volume approximately $1.2 t r i ll i o n; s p ik e d t o$ 1.5 trillion in 2023-05. “The money the Fed spent saving banks is far more than the announced figure, because a lot is hidden here.” They fund themselves by issuing agency bonds to money market funds, then lend to distressed banks — in essence using shadow-banking money to rescue traditional banks. Also, for every $3 F H L B ank s l e n d,$ 1 goes to insurance companies (particularly life insurance), unrelated to real estate — insurance is accumulating new risk.
$3 t r i ll i o nin reser v es i s t h e " ye ll o wl i g h t " w a r nin g l e v e l;$ 2.5 trillion is the “red light” crisis level — the Fed will not stop QT (unless there is a crisis). The Fed’s publicly stated QT target path: by 2024-07, cut interest-bearing liabilities (reserves + RRP) from $5.05 t r i ll i o n t o$ 2.5 trillion — halved. Reserves from $3.18 t r i ll i o nin 2023 - 07 t o$ 1.87 trillion in 2024-07. This framework estimated the moment reserves would fall below the $3 trillion warning level to be 2023-11. “Short-term Treasury rates must be high enough (> 5.05% RRP floor) for RRP money to shrink fast and reserves to shrink slowly” — this is the real reason T-bill rates have been running at 5.5%.
The core original indicator for offshore dollar circulation is “foreign-bank savings as a share of total US bank savings” — in every historical offshore dollar crisis, this ratio has dropped sharply beforehand. Cross-border dollar transfers are in reality a change in accounting labels between two US banks inside the US; not a single dollar actually crosses a border. So-called “dollar outflows” = US domestic savings account labels converting to foreign-bank savings accounts; “dollar inflows” the reverse. 99.99% of news coverage is wrong on this point. Lower foreign-bank deposit share = more depleted funding acting as “reserves” for the offshore dollar system = greater difficulty in cross-border clearing and transfers.
The two core indicators for shadow-banking long-range warning: money market fund size + triparty repo (including GCF) trading volume. Money market funds are the entire savings base of shadow banking; after the March 2023 banking crisis their size surged to $5.47 t r i ll i o n —" t h e bankin g cr i s i s a c t u a ll yresc u e d s ha d o w bankin g ." T r i p a r t yre p o (es p ec ia ll y GCF) v o l u m e d o u b l e df ro m$ 2 trillion to $4 t r i ll i o nan d i ss t i ll c l imbin g — d e a l er s^{'} co ll a t er a l ha ss i g ni f i c an tl ys h r u nk \to d es p er a t e t o b orro w f ro m e a c h o t h er v ia GCF \to v o l u m ecross in g$ 6 trillion is a clear signal that a major event is imminent.
The “ultimate risk source” of the entire four-layer defense is just one: rising long-bond rates. Long-bond rates↑ → medium/long-term bonds / MBS collateral devalued → shadow-banking collateral chain devalued (money fund NAV impaired) + bank asset-side unrealized losses enlarged (already $2.2 t r i ll i o n, e q u a lt o 96$ 1.7 trillion, 3× 2023 level) — Fed QT exit + US bank selling + declining overseas share → the only entities that could fill the gap are China / Japan / Gulf oil states. “The US will not pick a fight with China in the next two years” — this is the underlying financial reason.

Reasoning Chain (Framework Skeleton)

flowchart TD
    A["Problem Diagnosis<br/>IMF/World Bank/Fed/ECB<br/>Volatility-oriented early-warning completely failed<br/>Escape window = hours"]
    A --> B["Methodological Breakthrough<br/>Start from the monetary-pyramid layers<br/>Track trend indicators (circulation-volume school)<br/>Not volatility"]
    B --> C["Four-Layer Defense Structure<br/>Reserves / Domestic Dollars / Offshore Dollars / Shadow Dollars<br/>Tiered from inside out"]
    C --> D["Layer 1: Reserve Circulation<br/>Aircraft-Carrier Deck Anti-Aircraft Gun"]
    D --> D1["Indicator 1: Fed Funds Volume↑<br/>= Small/mid banks borrow even at a loss<br/>Trend signal months ahead"]
    D --> D2["Indicator 2: BTFP/Discount Window Balance"]
    D --> D3["Indicator 3: TGA+RRP+Reserves Three-Way Balance<br/>Reserves below $3T = yellow light<br/>Below $2.5T = red light"]
    D --> D4["Indicator 4: GSE Deposits"]
    D --> D5["Indicator 5: FHLBanks Bond Issuance+Loan Volume<br/>Spike = bank failure within 1-2 weeks"]
    C --> E["Layer 2: Domestic Dollar Circulation<br/>Close-Range Defense Perimeter"]
    E --> E1["Indicator 6: Total US Bank Deposits"]
    E --> E2["Indicator 7: Uninsured Deposit Leverage Ratio<br/>$8.9T uninsured"]
    E --> E3["Indicator 8: CRE Delinquency Rate<br/>4.5% and rising"]
    E --> E4["Indicator 9: Junk Bond Spread + Charge-Off Rate"]
    E --> E5["Indicator 10: Personal Credit Default<br/>+\"pawn shop near me\" Google Trends proxy (original)"]
    C --> F["Layer 3: Offshore Dollar Circulation<br/>Mid-Range Intercept"]
    F --> F1["Indicator 11: Central Bank Swap+FIMA Usage<br/>= Signal overseas major banks are seeking rescue"]
    F --> F2["Indicator 12: 3M FRA-OIS Spread<br/>100% predictive in every historical crisis"]
    F --> F3["Indicator 13: Term SOFR<br/>Replacing LIBOR"]
    F --> F4["Indicator 14: Foreign Bank Deposit Share ★Original<br/>Earliest signal for offshore dollar crisis"]
    C --> G["Layer 4: Shadow Dollar Circulation<br/>Long-Range Early Warning"]
    G --> G1["Indicator 15: Money Market Fund Size<br/>$5.47T reverse warning"]
    G --> G2["Indicator 16: Triparty Repo+GCF Volume<br/>2→4T doubled; >6T = imminent blowup"]
    G --> G3["Derived 6: TIC Foreign Official Treasury Holdings<br/>= The real focus of Yellen's China visit"]
    G --> I["Subtheme B: The Money-Pendulum Phenomenon<br/>2023-03 Silicon Valley Bank Crisis<br/>actually rescued shadow banking<br/>Risk only transfers, never disappears"]
    I --> J["Ultimate Risk Source: Long-Bond Rate↑<br/>Simultaneously pierces bank+shadow+offshore three layers<br/>30Y Treasury breaking 5% = a once-in-a-lifetime opportunity"]
    J --> K["Geopolitical Cross-Border Embedding<br/>Yellen's July 2023 China visit<br/>= Shortage of medium/long-term bond buyers<br/>2024 net new medium/long bond demand $1.7T = 3× prior year"]

Key Data Anchors

2023-03-09/10 Silicon Valley Bank collapse: FDIC takeover, assets $209 billion (second-largest US bank failure in history, after 2008 WaMu). The trigger was unrealized losses on medium/long-term Treasuries / MBS becoming realized losses as uninsured depositors ran.
2023-03-19 Credit Suisse absorbed by UBS: The Swiss National Bank conducted a USD swap with the Fed for $100 billion (largest single-instance volume in history), simultaneously using the FIMA Repo Facility (Foreign and International Monetary Authorities Repo Facility).
2023-06 IORB / RRP / Fed Funds Target Rate: IORB 5.15% / ON RRP 5.05% / Fed Funds target 5.00–5.25%, forming the “ceiling and floor” of the reserve market interest rate.
2023-06 US Treasury debt ceiling: The ceiling was raised from $31.4 t r i ll i o n (p re - J u n e 4 b re a c h) t o$ 32.5 trillion; TGA replenishment from roughly $500 bi ll i o n t o w a r d$ 600 billion (September) → $800 billion (year-end), dependent on money market funds in the overnight reverse repo account absorbing the issuance.
**2023-04 US banking system total unrealized losses $2.2 t r i ll i o n * * : E r i c a J ian g e t a l . s u r v eye d 4, 844 U S bank s —$ 2.2 trillion in unrealized losses ≈ 96% of total pre-rate-hike US bank capital. If half of uninsured deposits were to flee, approximately 190 banks with over $300 billion in assets would fail.
2023-06 Fed QT path: Reserves at $3.18 t r i ll i o na tt h e t im e; i d e a lpl an t ore d u ce t o$ 1.87 trillion by 2024-07; estimated moment reserves would fall below the $3 trillion warning level = 2023-11.
2023-08 Fed research: 37% of listed US companies under severe financial stress (ICR < 1), the most severe or second most severe level since 1970.
2023-07 30Y mortgage rate broke 7%: More than doubled from the 2022 low of 2.8%; MBS secondary market prices correspondingly devalued.
2023-06-30 USD LIBOR officially discontinued: 1M/3M/6M/12M USD LIBOR ceased publication, replaced by SOFR + CME Term SOFR; current monitoring should switch to Term SOFR + SOFR-OIS spread.
2023-07 Yellen’s China visit: Surface topic “international debt relief”; underlying real demand is medium/long-term bond buyers — China / Japan / Gulf oil states are the only possible buyers for the $1.7 trillion net new medium/long-term demand starting 2024-Q1.
**Money market fund size $5.47 t r i ll i o n * * : I C I J u n e 2023 d a t a, s u r g e df ro ma pp ro x ima t e l y$ 4.85 trillion after the banking crisis. Short-term securities as a share of total publicly held US Treasuries: 18.1% (approaching the internal 20% ceiling).
FHLBanks data anchor: Normal loan volume approximately $1.2 t r i ll i o n; s p ik e d t o$ 1.5 trillion in 2023-05. For every $3 l e n t,$ 1 goes to insurance companies (particularly life), unrelated to real estate.
Triparty repo + GCF trading volume: From a steady $2 t r i ll i o n p os t - 2021, d o u b l e d t o$ 4 trillion+ by June 2023; the framework sets a threshold of “breaking $6 trillion = imminent major event.”

The 16 Observation Indicators Assembled Across the Four Defense Layers

Decision Rules (Subthemes A/B Applied):

Escape-window tiering: Anomaly in Layer 1 indicators = hour-level response; Layer 2 = days-to-weeks; Layer 3 = weeks-to-months; Layer 4 = months-to-quarters
Money-pendulum cross-check: Single-layer indicator anomalies must be checked against “whether risk has shifted to another layer” (bank deposit outflows should be cross-checked against whether money market fund size is rising in tandem)
Ultimate alert: 30Y UST yield + MBS rates rising together + triparty repo volume spike → triggers Subtheme B ultimate alert

Layer 1: Reserve Circulation (Aircraft-Carrier Deck Anti-Aircraft Gun · Hour-Level Escape Window)

#	Indicator	Data Source / Frequency	Anomaly Threshold	Escape Window
1	Fed funds market trading volume	NY Fed EFFR Volume [public]; daily	Sustained upward trend > 3 months with positive slope = new borrowers entering	Months-to-weeks
2	BTFP / discount window / other credit facility balances	Fed H.4.1 [public]; weekly	Total balance persistently rising = banks’ Fed-rescue dependence not declining	Months
3	TGA + RRP + reserves three-way balance	Fed H.4.1 / FRED WTREGEN + RRPONTSYD + WRESBAL [public]; weekly	Reserves below $3 T = ye ll o wl i g h t; b e l o w$ 2.5T = red light	Weeks-to-months
4	GSE deposits (Fannie / Freddie / FHLBanks)	Fed H.4.1 [public]; weekly	Persistent decline = fed funds supplier drying up	Months
5	FHLBanks bond issuance + loan balance	FHLBanks Office of Finance [public]; monthly	Spike to above $1.5T + simultaneous bond issuance = bank failure within 1-2 weeks	1-2 weeks (short-term warning)

Layer 2: Domestic Dollar Circulation (Close-Range Defense Perimeter · Days-to-Weeks Escape Window)

#	Indicator	Data Source / Frequency	Anomaly Threshold	Escape Window
6	Total US bank deposits (H.8)	Fed H.8 [public]; weekly	Weekly net outflow > historical mean + 1.5σ	Days-to-weeks
7	Uninsured deposit leverage ratio (per-bank)	FDIC Quarterly Banking Profile + Call Report [public]; quarterly	> 50% = high risk; outflows of 10% can trigger failure	Weeks-to-months
8	CRE delinquency rate	MBA Commercial Mortgage Delinquency Survey [public/paid]; quarterly	Office delinquency rate > 4.5% and rising	Quarterly
9	Junk bond spread + charge-off rate	ICE BofA US HY Index OAS / FDIC [public]; daily + quarterly	Spread at long-term low + corporate bankruptcies spiking = severe divergence	Months-to-quarters
10	Personal credit default + Google “pawn shop near me” trend	NY Fed Household Debt Report / Google Trends [public]; monthly	Google trend significantly above pre-2022-07 level	Months

Layer 3: Offshore Dollar Circulation (Mid-Range Intercept · Weeks-to-Months Escape Window)

#	Indicator	Data Source / Frequency	Anomaly Threshold	Escape Window
11	Central bank USD Swap + FIMA Repo usage	Fed H.4.1 [public]; weekly	Any single central bank’s swap exceeding $10 billion in a single day = overseas major bank already seeking rescue	Days
12	3M FRA-OIS spread	Bloomberg/Refinitiv [paid]; fallback: Term SOFR 3M − 3M OIS [public]; daily	Persistently widening > 25–30bp = appeared before every historical crisis	Weeks-to-months
13	Term SOFR (replacing LIBOR)	CME Term SOFR [public, 1-day lag]; daily	Persistently elevated + spread with IORB widening = offshore dollar tightening; LIBOR has been discontinued, switch to Term SOFR	Months
14	Foreign bank savings as share of total US bank savings ★original	Fed H.8 (foreign-related deposits / total deposits) [public]; weekly	Downward trend + falling below historical floor = offshore dollar crisis approaching	Months-to-quarters (earliest signal)

Layer 4: Shadow Dollar Circulation (Long-Range Early Warning · Months-to-Quarters Escape Window)

#	Indicator	Data Source / Frequency	Anomaly Threshold	Escape Window
15	Money market fund size (two-way reading)	ICI Weekly MMF Report / OFR MMF Monitor [public]; weekly/monthly	Size rapidly rising → bank deposits being drained (see Subtheme B money-pendulum); size turning down → shadow-banking liquidity crunch	Months
16	Triparty repo (including GCF) trading volume	NY Fed Triparty Repo Statistics / DTCC GCF Repo Index [public]; monthly	Total doubled to $4 T an d s t i ll r i s in g = d e a l ers a re d es p er a t e; * * cross in g$ 6T = imminent major event**	Months-to-quarters
Derived 6	TIC foreign official US Treasury holdings	US Treasury TIC [public]; monthly (2-month lag)	China / Japan / Gulf states net adding = US-China negotiations concluded; persistent net reduction = long-bond buyer pipeline cut off	Quarterly

Ultimate Alert (Subtheme B Triggered)

Any two of the following three conditions met → advance warning of a once-in-a-lifetime entry window:

30Y US Treasury yield breaks 4% (FRED DGS30) → begin preparing
30Y mortgage rate sustained > 7% (Freddie Mac PMMS or MBA) → MBS collateral devalued
Triparty repo / GCF volume breaks $6 trillion + money market fund size turns down → shadow-banking liquidity crunch imminent

Framework’s own words: “When the 30-year Treasury rate breaks 5%, one may perhaps go all-in — a once-in-a-lifetime opportunity is about to materialize.”

Compiler’s Perspective

Coordinates: Category = Observation Indicators and Signals · axis_h = Qi · axis_v = What It Is

The core operation of Crystal Ball of Clarity — Placing Your Energy in What Is Clear (Meditative Operational Method) is to place attention on a small number of clearly observable nodes rather than being consumed by an infinitely expanding signal stream. This framework’s 16 indicators are the financial instantiation of that operation: from thousands of market indicators, 16 trend-based (not volatility-based) nodes are extracted; each is assigned an explicit data-source formula and threshold; and the four-layer escape-window tiering enables graded response — rather than waiting for VIX to spike before reacting.

Perspective Test: Someone monitoring VIX / stress indexes received zero information in February 2023 (markets calm), whereas this framework’s Layer 1 indicator — fed funds market trading volume — had already shown a multi-month sustained upward trend by 2022-Q4, indicating that small and mid-sized regional banks were “borrowing even at a loss.” The difference comes down to a single operational choice: watch the slope of volume rather than watch price spikes. Using the same underlying data, Layer 5 (FHLBanks loan volume spiking to $1.5 trillion) provides a short-term warning of “bank failure within 1-2 weeks” — an entire order of magnitude more lead time than waiting for VIX.

Proprietary Assertion: Indicator 14 — foreign bank savings as a share of total US bank savings (the H.8 formula) is the only indicator in this framework marked “original,” and the framework makes explicit: 99.99% of news coverage understands cross-border dollar transfers as “money actually crossing a border,” so the mainstream “dollar outflows/inflows” narrative has no early-warning capability for crises at this layer. Using this ratio as a months-to-quarters long-range early warning for offshore dollar crises is an operation that can only be produced by someone who has read this framework — The Financial Anomaly Indicator System contains no such formula, and The VIX Fear Index is incapable of capturing this level at all.

This framework should be read alongside Repo and Shadow Money and The Hedge Fund Repo Crunch: the repo chain (GSEs/FHLBanks/money market funds/triparty repo) is the infrastructure underlying this framework’s four defense layers; understanding how repo transmits pressure between banks and shadow banks is what allows one to understand why rising money market fund size and rising triparty repo volume often appear in tandem — this is the structural diagram of “the money pendulum,” not two independent signals. Read alongside The Logic Review and Verification Method: the H1 2023 SVB/Credit Suisse crisis is a shared retrospective subject of both frameworks, but with different angles — the Logic Review asks “which logical chain broke,” while this framework asks “which layer first crossed a threshold.”

Sources

Compiled draft z-0132 · archived 2026-07; parent: 2023 Mid-Year Main Course, “The Dollar Circulation Early-Warning System (I–V),” sessions 4–8 (5 sessions total), de-noised transcript, compiled 2026-05-16 v1-final (only translation errors corrected: IORB/RRP rate labeling, BTFP tool naming, FIMA full name, LIBOR discontinuation framing — framework arguments and indicator structure unchanged). External public sources: NY Fed Federal Funds Data (EFFR Volume), https://www.newyorkfed.org/markets/reference-rates/effr-volume; Fed H.4.1 / H.8 Release, https://www.federalreserve.gov/releases/; ICI Weekly Money Market Funds Statistics, https://www.ici.org/research/stats/mmf; US Treasury TIC System, https://home.treasury.gov/data/treasury-international-capital-tic-system; NBER Working Paper, Jiang E. et al., “Monetary Tightening and U.S. Bank Fragility in 2023,” 2023-04.

Z-Finance

Explorer