Sixty Billion Yuan Through Hong Kong: The PBoC's Bill Sale and the Synthetic CNH Problem

Hook

Here is the error.

A wire report dated September 23 states that the People's Bank of China will sell 60 billion yuan of central bank bills through Hong Kong. The same report adds that the issuance "enhances offshore renminbi liquidity." Four sentences in, and the direction of the operation is already inverted.

A central bank bill sale is a drain. The PBoC removes CNH from the offshore pool, issues a paper claim against itself, and pays a coupon for the privilege of holding that which it just absorbed. The pool contracts. The funding rate rises. The cost of borrowing the currency in order to sell it short rises with the funding rate. That is the mechanism, in full. If you want to know what the instrument does, watch the CNH HIBOR fixing. Do not watch the headline.

Over the past three years I have read more bad coverage of offshore RMB plumbing than good. It fails the same way every time: the offshore market gets described as a smaller copy of the onshore market. It is not. Separate pool, separate curve, separate institutions, separate legal regime. And since roughly 2023, a growing set of on-chain instruments that reference this market without holding any enforceable claim on it whatsoever.

Tracing the gas leak where logic bled into code.

Context

Offshore renminbi — CNH — is not a diaspora of the onshore currency. CNY trades inside the capital account, is priced against the MLF and the LPR, and is managed by the PBoC through reserve requirements, open market operations, and administrative guidance. CNH trades outside the capital account, in Hong Kong, Singapore, London, and Luxembourg, and is priced by supply and demand in an offshore pool that the PBoC can only influence indirectly.

The primary indirect instrument is the offshore bill. The PBoC began issuing these in Hong Kong in late 2018, shortly after the first prolonged CNH depreciation episode, and has returned to the market in irregular waves since. The mechanics are unremarkable. The PBoC auctions a bill to institutional bidders, typically at 3-month, 6-month, or 1-year tenor. Settlement runs through the Hong Kong Monetary Authority's Central Moneymarkets Unit — the CMU — which has been the settlement backbone for Hong Kong debt instruments since 1994. The bill is denominated in CNH, settled in CNH, and absorbed from the CNH pool.

The reason a 60 billion yuan number matters is context. Historical single-tranche offshore bill offerings have generally landed in a band between 5 and 25 billion yuan. A 60 billion figure therefore implies one of two things: either an extraordinary single-tranche operation, or an aggregate of multiple tranches packaged into one headline. The wire copy does not say which. It also does not disclose the tenor structure. It does not disclose the clearing or winning yield. It does not disclose the subscription multiple. All four are standard disclosures for this instrument, and all four are missing.

The wire also carries a date without a year annotation in the parsed copy. Every time-sensitive judgment that follows inherits that ambiguity. I am flagging it now rather than burying it, because the difference between a squeeze conducted into a depreciating tape and a squeeze conducted into a stable one is the difference between a meaningful signal and a routine rollover.

There is a second reason this instrument matters to people who do not trade CNH. Offshore renminbi is one of the reference legs behind a family of crypto instruments: CNH-quoted stablecoins, tokenized money-market wrappers that report NAV in CNH, and the offshore funding desks whose delta-neutral books are denominated partly in renminbi. When the PBoC moves the CNH curve, it moves the inputs to those instruments. Whether it moves the instruments themselves is the question this piece actually tries to answer.

Core

The transmission chain, stated as an invariant

The CNH bill transmission chain has five links. They are sequential, and the order matters.

Link one: the PBoC auctions the bill. Link two: successful bidders pay in CNH, which leaves the offshore pool and enters the PBoC's balance sheet as a liability. Link three: the overnight and 1-week CNH HIBOR fixings rise, because the pool of lendable offshore renminbi has shrunk while demand for it has not. Link four: the carry cost of a short-CNH position rises, because shorting requires borrowing CNH, and borrowing CNH is now more expensive. Link five: speculative shorts unwind, and USD/CNH faces mechanical upward resistance.

The invariant is that the bill is a price-of-money instrument, not a quantity-of-money instrument for the onshore system. It does not touch the onshore base money supply in any direction. Beijing's MLF, RRR, and LPR decisions operate on the onshore curve. The offshore bill operates on the offshore curve. The PBoC has deliberately separated the two, and the separation is the point.

This is the part the wire report missed. When the PBoC reaches for an offshore bill instead of an onshore tightening, it is telling you it does not want to raise domestic rates. It wants to defend the exchange rate without paying for that defense in onshore credit conditions. Three-impossible-trinity triage, executed in the offshore market precisely because the offshore market is where the cost stays contained.

The carry cost, written as a function

This is where the crypto transmission becomes concrete. A profit-seeking desk that wants to short CNH offshore — or that wants to run a delta-neutral basis trade with an offshore renminbi leg — pays a financing cost that is a direct function of the HIBOR curve. Write it out.