Crossing Letters

No. 001 · June 1, 2026

Today, global capital and China’s tech evolution no longer run on parallel tracks. They intersect. Crossing Letters works at that intersection.

I read the architecture under the story, deciphering the structural pivots and hidden velocities where global capital meets China: who regulates what, how the machinery actually works, and what a given move signals about where policy is heading.

The news is only the thread to pull. The value is the trend behind it, the multi-year arc set out in the handbook, Inside China’s Financial Markets, which is what turns a headline into something you can use.

May was a “stable money, firm regulation, structural tuning, capital top-up” month. Four moves and a handful of smaller ones tell what that means.

Cross-Border

1. China’s endgame on the cross-border brokers

On May 22 the securities regulator and seven other agencies issued a plan to comprehensively rectify illegal cross-border securities, futures, and fund activity. The same day, cases were opened against three offshore-facing brokers with pre-notified penalties: Futu (FUTU) faces around 1.85 billion yuan in fines and confiscation, Tiger (TIGR) around 411 million yuan, and Longbridge a confiscation-and-fine order, with a 1.25 million yuan personal penalty on each founder or chief executive. Combined, the three exceed 2.2 billion yuan. Futu and Tiger shares fell sharply in premarket trade, with lows reported near 40 and 45 percent, and Futu has been buying back stock, roughly 418 million of an 800 million dollar program.

The headline reads as a sudden crackdown. The plumbing says otherwise. This is the closing stage of a five-year arc. The central bank first labeled these platforms “unlicensed driving” in 2021. New mainland onboarding was barred in 2022. The apps left domestic stores in 2023. The firms then spent 2024 and 2025 running down their mainland books, and Futu’s mainland clients were down to 13 percent of its total by the first quarter of 2026. May was the last step, not a surprise.

Pull the lens back two decades and the move fits a single, consistent pattern. China has steadily built a lattice of compliant, quota-bound, state-visible channels for cross-border capital: the qualified-investor scheme from 2002 (merged in 2020), Stock Connect in 2014 and 2016, Bond Connect northbound in 2017 and southbound in 2021, Swap Connect in 2023, interbank bond repo opened to foreign investors in 2025, and the Cross-Border Wealth Management Connect in 2021 and 2024. Each one is metered and visible to the authorities. Seen against that build-out, the broker enforcement is the mirror image: as the governable channels mature, the ungoverned ones are closed. These brokers were the last large retail conduit operating outside the lattice.

China runs a managed capital account, and the brokers routed mainland money offshore outside those channels. The new plan sets a two-year wind-down (May 2026 to May 2028) in which existing clients may only sell or transfer out, never buy or transfer in, after which domestic sites, software, and servers close; existing investors’ assets are protected. The action is notably coordinated: the securities regulator leads case handling, the cyberspace administration clears illegal information and accounts, the industry ministry pulls the apps, the market regulator pursues illegal advertising, the central bank and the foreign-exchange administration police cross-border funds and underground banking, and public security handles the criminal track.

The quiet beneficiaries are the channels in the map above: Stock Connect Southbound, the qualified-investor quota, and the Cross-Border Wealth Management Connect, all positioned to absorb the redirected demand, with Hong Kong’s regulator tightening look-through checks on mainland account openings. The readable conclusion is not that China is cracking down. It is that China keeps consolidating cross-border access into instruments it can meter, and has done so for twenty years.

In the handbook: Chapter 10 maps these compliant channels in full; Chapter 1 explains the regulatory architecture behind the eight-agency action.

Monetary Policy

2. Why Beijing is holding, and the margin underneath

The one-year loan prime rate stayed at 3.0 percent and the five-year at 3.5 percent in May, unchanged for a twelfth straight month, with the seven-day reverse repo rate steady at 1.4 percent since May 2025. The first-quarter monetary policy report, released May 11, kept the “moderately loose” label while stressing precision and forward guidance. With first-quarter GDP at 5 percent and prices firming at the margin, the case for cutting is weak. As one widely cited economist framed it, holding consolidates the effect of earlier cuts, supports bank soundness, and preserves room to move in step with fiscal policy.

The mechanism the report states plainly: because the policy rate, the seven-day repo, did not move, the loan prime rate has no room to fall. The benchmark is quoted as the policy rate plus a spread, so the spread, not goodwill, sets the floor.

Underneath the “moderately loose” label, the cost shows up on bank balance sheets, and the trend is the point. The system net interest margin fell from 2.08 percent at the end of 2021 to 1.42 percent at the end of 2025, then slipped another 2 basis points to a record-low 1.40 percent in the first quarter of 2026 as loans repriced lower. One-year top-rated bank certificate-of-deposit yields averaged a record-low 1.47 percent in April, the same compression from the funding side. The Q1 print is not a shock; it is the latest step of a five-year decline. Holding lending rates down while pushing credit growth squeezes the spread banks live on. Keep that in mind for item 4.

In the handbook: Chapter 6 covers the margin compression; Chapters 1 and 2 cover policy-rate-to-LPR transmission. The likely next easing is a structural tool aimed at innovation or small business, not a broad cut.

Liquidity

3. Shorten and lengthen: the liquidity tell

In May the central bank rolled 600 billion yuan of one-year medium-term lending facility (MLF) against 500 billion maturing, a net injection of 100 billion, while cutting outright reverse repos by 1 trillion yuan across two tenors. April had drained 200 billion through the same facility, ending more than a year of net additions. The pattern is “shorten and lengthen”: pull back short-dated funding, extend term funding. System liquidity moved from “extremely ample” to “reasonably ample” over March to May, as drains ran from 250 billion in March to 600 billion in April and around 1 trillion in early May.

The intent is a shift from aggregate easing toward structural fine-tuning. Trimming the short end discourages idle fund-churning and excess leverage and nudges over-low short rates back toward the policy rate. Adding term funding supports longer bank credit and a heavier government-bond calendar, which picked up sharply in late May as the finance ministry launched the first recapitalization bond for state financial institutions. The medium-term operation coming in larger than expected was itself a signal that the supportive stance is intact.

The deeper trend the handbook tracks sits underneath this: the ten-year government yield near 1.75 percent at the end of April is the floor of a multi-year decline, and bank balance sheets have been tilting steadily toward government bonds as loan demand softens. A range-bound bond market looks likely from here, with the shorter, more adjusted maturities arguably better placed.

In the handbook: Chapter 3 on the government-bond issuance this is funding, and on the secular path of yields; Chapter 1 on the central bank’s tool kit.

Banking

4. Recapitalizing the big banks, again

Also on May 22, the finance ministry launched the first tranche of a second round of special-treasury-bond injections into the state banks: 300 billion yuan in total, a five-year tranche on May 22 and a seven-year tranche on June 12, with annual coupons. Market expectation is that this round targets the two large banks left out last time. The 2025 first round put 500 billion yuan into four others, so the two rounds together, 800 billion yuan, build capital coverage across all of the large state commercial banks. One brokerage estimate puts the 300 billion as capable of supporting roughly 4 trillion yuan of asset expansion, and the first round lifted core tier-one ratios by 0.48 to 1.51 points. This round is smaller than last year’s 500 billion and started slightly later, giving the ministry room to fit it around the ultra-long special-bond schedule.

This is the answer to the squeeze in item 2. The big banks are not in trouble: as the figure shows, capital adequacy has held near 15.5 percent and core tier-one near 10.9 percent across 2020 to 2025, well above the 7.5 percent floor. The issue is forward capacity. When the margin is at a record low, banks cannot rebuild capital fast enough from retained earnings to keep expanding credit, so the state supplies it through sovereign bonds. Fiscal and monetary policy work the same lever from two ends: the bond funds the capital, the capital funds the lending.

In the handbook: Chapter 6 on bank capital adequacy; Chapter 3 on the sovereign issuance side.

Also In May

5. Four smaller moves that compound

The finance ministry is issuing up to 6 billion yuan of renminbi-denominated green sovereign bonds in Hong Kong, deepening the offshore renminbi market. The interbank market saw its first blue-bond credit derivatives, cleared through the Shanghai Clearing House, extending the green-finance instrument set toward the marine economy. Jiangsu issued 71.1 billion yuan of special bonds to swap out hidden debt, the local-debt refinancing treadmill grinding on. And the Shanghai exchange tightened pledged-repo rules under a “reward the strong, limit the weak” principle, adding a 30 percent single-bond concentration cap and barring high-risk bonds from negotiated trades. None of these is a headline on its own; together they are the slow deepening of the market’s plumbing.

In the handbook: Chapter 3 on bond-market instruments and risk controls; Chapter 10 on the offshore renminbi.