Liquidity Crisis Signal: Systematic Interpretation from SOFR Loss of Anchor to BTC’s Plunge

In the third week of November 2025, global financial markets experienced significant turbulence, risk assets generally pulled back, and liquidity stress indicators deteriorated across the board.The main phenomena include:

1. Bitcoin’s largest single-week decline exceeded 22% (November 10-16), falling from around US$108,500 to a maximum of US$83,900;

2. The U.S. stock market experienced its most violent weekly fluctuations since 2022, with the Nasdaq 100 index falling nearly 7% at one point;

3. SOFR (Secured Overnight Financing Rate) continues to lose its anchor, with the weighted median reaching 5.34% on November 14, 4bp higher than IOER (Interest on Excess Reserves) of 5.30%, and the 95th percentile rose to 5.60%;

4. The use of the Federal Reserve’s Overnight Reverse Repurchase Facility (ON RRP) plummeted by more than $120 billion from November 13 to 14, showing that some traditional cash holders were forced to leave the safest parking lots;

5. The WTI crude oil futures curve briefly appeared a “super contango” with the front month to the far month being completely flat (November 11-12). The front-end M1-M2 price difference was once only +0.03 US dollars/barrel, the flattest state since April 2020;

6. The pressure on the offshore U.S. dollar financing market (Eurodollar) is also obvious, with the three-month U.S. dollar Libor-OIS spread widening to 38bp, the highest since the banking crisis in March 2023.

These phenomena are not isolated, but the “classic combination” of the three liquidity crises in September 2019, March 2020, and March 2023 has completely reappeared, and the intensity has exceeded the previous three times.

1. The true meaning of SOFR’s loss of anchor: not insufficient reserves, but denial of loans

The current excess reserve balance of the U.S. banking industry is approximately US$3.28 trillion (data as of November 12, 2025), which is 126% higher than the US$1.45 trillion during the repo crisis in September 2019, and is still 17% higher than the US$2.8 trillion during the crisis in March 2023.Simply looking at the “total reserves”, the system is not short of money.

However, the SOFR quantile chart published daily by the New York Fed shows:

• The interest rate in the 5th quartile has risen to 5.20% (still lower than the IOER);

• The 95th percentile interest rate has risen to 5.60%-5.75%;

• The volume-weighted dispersion (95th-5th percentile) expanded to 55bp, the highest since June 2022.

This means that the entire market has experienced a systematic rightward shift: the number of institutions willing to lend cash at low interest rates has significantly reduced, and a large number of cash holders would rather put their money in the Fed’s ON RRP (5.30%) or not lend at all, rather than lend to most counterparties at 5.30%-5.35%.The failure of SOFR was not a “lack of reserves” but a “denial of loans” – the market had widespread doubts about the quality of collateral and the creditworthiness of counterparties.

2. The fundamental reason for loan rejection: the credit cycle has entered the late stage of de-leveraging

The ultra-low interest rates and the “toughness narrative” from 2021 to 2024 have jointly given birth to the extreme irrationality of this round of credit expansion:

1. The size of the U.S. private credit market (Private Credit) has expanded from US$850 billion in 2019 to US$2.1 trillion in the third quarter of 2025, a large number of which are “covenant-lite” and PIK (payment of interest in kind) structures;

2. Among the leveraged loans issued in 2024-2025, more than 72% are “no maintenance clause” loans, much higher than the 48% in 2007;

3. Among commercial real estate (CRE) loans, the LTV (loan to value ratio) of office assets generally exceeds 85%, and even reaches 110% in some cities (including mezzanine debt);

4. In the field of consumer finance, the 60+ day serious delinquency rate for U.S. auto loans in the third quarter of 2025 has reached 7.9%, exceeding the historical peak of 7.6% in the fourth quarter of 2009; the 90+ day delinquency rate for credit cards is 5.8%, which is also close to the 2008-2009 level.

More critically, delinquency rates have reached their post-crisis peak ahead of schedule, even before the unemployment rate has worsened significantly, which is completely contrary to all previous credit cycles.Usually the delinquency rate lags the unemployment rate by 6 to 12 months, but this time the delinquency rate is already ahead, which means that once the unemployment rate rises to 5.5% to 6.0% in the first quarter of 2026, credit losses will be exponentially amplified.

3. The failure and re-verification of Bitcoin as the “canary” of global liquidity

Bitcoin will gradually evolve from “digital gold” to a high-beta risk asset from 2021 to 2025, and its correlation coefficient with the Nasdaq 100 will remain above 0.75 for a long time.From November 10 to 16, 2025, the correlation coefficient between Bitcoin and the Nasdaq 100 further rose to 0.91, showing that it has completely become a “liquidity thermometer.”

In this round of correction, Bitcoin once rebounded 8% in the US stock market on November 13, but plummeted again in the late trading and night trading of US stocks, showing the characteristics of “buy it and sell it immediately”.This is completely consistent with the trend of Bitcoin during the three liquidity crises in June 2022, November 2022, and March 2023: short-term rebounds in risky assets are often quickly extinguished by forced liquidations and redemption selling pressure.

Bitcoin’s leading decline illustrates the rapid withdrawal of marginal funds with high risk appetite around the world, and these funds are the core force that has driven the expansion of the valuations of all risk assets in the past four years.

4. WTI curve “super contango” recession confirmation signal

From November 11 to 12, 2025, the WTI front month-12 month price difference once narrowed to only +2.8 US dollars/barrel, and the front month-distant month curve was almost completely flat.This is the flattest state since April 2020. Historically, similar structures have only appeared in October-December 2008 and March-April 2020.

A flattened contango curve means:

1. Spot demand is extremely weak and traders are unwilling to hold physical inventories;

2. Forward demand expectations have been significantly revised downwards, and the market is no longer willing to pay storage costs for future delivery;

3. Financial investors are selling near-month contracts and buying far-month contracts for rollover arbitrage, further depressing near-month prices.

This is one of the clearest pricing signals for the global recession in the oil market, and is more reliable than traditional indicators such as the OECD leading indicator and the copper-gold ratio.

5. The Fed’s policy dilemma: QE cannot solve the credit contraction

New York Fed President John Williams and head of open market operations Roberto Perli both hinted in speeches in mid-November that “asset purchases may soon need to be restarted” (QE).However, both history and logic show that QE is almost ineffective in this crisis:

1. September 2019 crisis: After the Fed restarted “non-QE”, SOFR fell rapidly, but the real economy did not decline;

2. March 2020 crisis: Unlimited QE + fiscal deficit prevented collapse;

3. March 2023 crisis: BTFP (Bank Term Financing Program) instead of QE solved regional bank runs.

The essence of this crisis is the contraction of private sector credit, not the shortage of inter-bank reserves.QE can only increase bank reserves, but it cannot force banks or shadow banks to lend money to borrowers who are at substantial risk of default.On the contrary, QE will further push up U.S. bond prices and lower term premiums, causing an inversion between the value of collateral (U.S. bonds) and financing costs, exacerbating the risk of the remortgage chain breaking.

6. The hidden crisis in the offshore U.S. dollar market

SOFR only reflects the onshore U.S. dollar collateralized repo market, while more than 80% of global U.S. dollar financing occurs offshore (London, Cayman, Hong Kong, Singapore, etc.).Currently observable indicators of offshore pressure include:

• The 3-month USD cross-currency basis swap (CCS) has widened to -45bp, the widest since March 2023;

• Hong Kong’s offshore RMB Hibor-USD Libor spread rose to 280bp, showing signs of dollar shortage;

• Japan Trust Bank’s US dollar financing premium (TIBOR—SOFR) rose to 35bp.

Once the offshore U.S. dollar market is completely frozen, it will be reversely transmitted to onshore through foreign exchange swap lines, making it difficult for the Federal Reserve to quickly alleviate the global U.S. dollar shortage even if it launches QE.

Conclusion and Outlook: Systemic credit events may occur in the first half of 2026

Taking into account all leading indicators (SOFR losing anchor, Bitcoin plummeting, WTI super contango, delinquency rates peaking early, offshore US dollar basis widening), the global financial system is experiencing the most serious credit contraction since 2008.

Possible evolution paths:

1. December 2025-January 2026: A wave of corporate profit warnings + more large-scale layoffs → further deterioration of consumer credit;

2. First quarter of 2026: Private credit funds experience large-scale redemptions → are forced to sell assets → spiral downward in collateral value;

3. Second quarter of 2026: If a single large private credit fund or regional bank fails, it may trigger a full-scale liquidity crisis similar to March 2020.

By then, the Fed may be forced to restart unlimited QE and cooperate with the monetization of fiscal deficits, but since the credit contraction has entered an irreversible stage, the policy effect will be significantly lower than in 2020.Risk assets may experience a systemic correction of 20% to 40%, and the 10-year U.S. Treasury yield may briefly fall below 2.5% before rebounding due to the collapse of inflation expectations.

The most dangerous illusion in the current market is to continue to attribute liquidity pressure to “insufficient reserves” or “excessive Fed QT”, thereby expecting that a “dovish declaration” can reverse the situation.The fact is that the pendulum of the credit cycle has reached its end, and any attempt to cover up solvency problems with liquidity will eventually be ruthlessly exposed by the market.