The Kengly Letter
Neither side can exit — the tariff equilibrium is a Nash steady state and the routing trade is live

Neither side can exit — the tariff equilibrium is a Nash steady state and the routing trade is live

7 May 2026 · 7 min · issue

The thesis in one line. The US-China tariff standoff has reached a Nash equilibrium under Managed Friction: both sides have locked in strategies where unilateral exit is more costly than continued friction. The market prices tariffs as a negotiating chip that will be traded away. The Nash read is that they won't — and the structural beneficiaries of a durable tariff wall have not finished repricing.

The on-ramp: the US effective tariff rate on Chinese goods escalated sharply under the Trump 2.0 administration through 2025, with rates exceeding 40% on most goods categories by mid-year (Reuters, multiple dispatches, Jan–Apr 2025). China retaliated in sequence. Both sides have been here before — but this time, two structural shifts make the equilibrium stickier than 2018-2019.

Three pillars

One. Bessent has wired the tariff commitment into domestic political necessity. Treasury under Bessent is running a fiscal-dominance strategy: revenue from tariffs partially offsets the deficit pressure from Trump's domestic tax agenda. That is not a negotiating position — it is a budget dependency. Rolling back tariffs now costs Bessent a revenue line he has already pencilled in. This is the classic credible-commitment device: you remove your own exit option by wiring the commitment into the fiscal structure. Trump has signalled tariff permanence across multiple press cycles. The bond market has partially repriced this (Bloomberg, Apr 2025 (free)). The equity market has not finished doing so.

Enforcement of third-country routing rules is selective — Bessent's Treasury applies secondary pressure on obvious tariff washing (goods routed through Vietnam with de minimis value-added) but lacks the enforcement capacity to shut the routing valve entirely. That is not a bug; it is a feature. Selective enforcement maintains the political signal (tariffs are real, we are enforcing) while preserving the import-price relief that American consumers need. The game theory here: [Aggressive, SelectiveEnforce] dominates [Aggressive, FullEnforce] because the latter breaks American importers and creates domestic political blowback faster than it pressures China.

Two. Xi and Pan at the PBOC have made the routing infrastructure permanent. China's adaptive response is not to negotiate down the tariff wall — it is to route around it. Chinese FDI into Vietnamese manufacturing estates accelerated sharply through 2024, with CBRE Vietnam tracking a step-change in industrial park occupancy rates driven by Chinese-backed tenants (CBRE Vietnam, 2024 Industrial Market Report). The PBOC simultaneously expanded RMB bilateral swap arrangements with ASEAN central banks, building the settlement infrastructure for trade that bypasses USD clearing. Wang Yi has positioned this as China's "high-quality BRI" — the diplomatic language for institutionalising the detour. Xi's domestic position cannot absorb a visible concession on tariffs without reciprocal US rollback. The Adapt strategy available to him — and the one he is executing — is to make the routing so structurally embedded that the tariff wall becomes a feature of the landscape rather than a temporary condition.

Three. Vietnam is the bridge economy that cannot afford to lose. Pham Minh Chinh's government has actively solicited Chinese manufacturing FDI while maintaining preferential US trade status — a Proxy / Force-Multiplier role in the tariff game. Vietnam's exports to the US have grown consistently through the tariff escalation period because Vietnamese-manufactured goods carry a Vietnamese certificate of origin, not a Chinese one (Reuters, Dec 2024). The US has applied pressure on specific categories (solar panels, steel), but general electronics and consumer goods routing through Vietnam has continued. Vietnam cannot risk losing preferential US access (it would crater GDP growth) and cannot risk losing Chinese FDI (it would crater industrial expansion). It plays both sides precisely because neither side can afford to punish it too hard. That structural position is the bridge economy moat.

What the consensus is mispricing

The dominant market narrative — that tariffs are a temporary negotiating chip that will trade away in a Grand Bargain — requires one of two things to be true: either Trump trades tariffs for concessions in another domain (Taiwan guarantees, IP enforcement), or Xi offers enough of a concession to let Trump claim a win without triggering domestic Chinese backlash. Neither condition is currently met, and the structural wiring described above makes both harder, not easier, over time.

The structural beneficiaries of a durable tariff wall — ASEAN manufacturers, Mexican nearshoring operators, logistics and bonded warehouse operators along the routing corridors — have repriced but not finished repricing. They are still priced with a discount for "deal risk" (the probability of tariff removal collapsing the routing arbitrage). That discount is excessive given the Nash stability of the equilibrium.

The structural losers — companies with deep China supply-chain or revenue exposure priced for tariff normalization — carry the inverse risk. Apple's China manufacturing concentration, Qualcomm's China revenue dependence, and analogous names are priced with an implicit tariff-removal option that the Nash read says does not exist.

Position implied

Three concrete positions, each tracking a different facet of the same equilibrium:

One. Long ASEAN manufacturing arbitrage — VNM (Vietnam ETF) on TradingView and EWW (Mexico ETF) on TradingView — over an 18-month horizon. The routing infrastructure is now embedded at the industrial-park level; unwinding it requires a US-China deal that the Nash equilibrium makes unlikely within that window. Risk: selective US enforcement escalates to category-specific prohibitions on Vietnamese-origin goods with Chinese beneficial ownership.

Two. Short "reunion narrative" bellwethers — names with majority China supply-chain or revenue concentration priced for tariff normalization. The option value of tariff removal is embedded in these prices and will erode as each passing quarter without a deal updates the base rate. This is not a short on the companies themselves but on the tariff-removal premium.

Three. Long USD on import-price persistence. Tariff-driven supply-chain restructuring reprices import costs upward on a multi-year lag as sourcing shifts from cheap Chinese production to slightly more expensive ASEAN or nearshore production. That is a structural, persistent import-price tailwind that the Fed cannot ignore and the OIS market has not fully priced. Long USD via DXY on TradingView against a basket, or via rate-differential positioning.

Red-team responses

Attack 1: The 100-day deal shock — Trump trades consumer-goods tariff rollback for an optics win with Xi

The Grand Bargain doesn't need to be complete. Trump rolls back consumer-goods tariffs (phones, apparel — optically large, politically visible) in exchange for a non-economic concession: a Xi visit, a large LNG or aircraft purchase agreement, or Taiwan status-quo language. Bessent keeps steel, aluminum, and chips tariffs and claims fiscal neutrality. Consumer tariffs disappear. VNM and EWW sell off sharply in a session as the routing arbitrage collapses.

Response: The attack is valid but misfires on the fiscal math. Consumer goods tariffs account for the bulk of the Bessent revenue line — steel and aluminum are noise relative to apparel and electronics volume. A "selective rollback" that preserves fiscal impact means keeping exactly the tariffs that Chinese manufacturers are currently routing around via Vietnam. If Trump rolls back consumer electronics tariffs entirely, the routing arbitrage does compress — but VNM and EWW holders are long the manufacturing infrastructure, not the tariff delta per se. Vietnamese factories building iPhones components do not disappear because Chinese-made phones re-enter the US at lower rates. The position's 12-18 month horizon also matters: a deal shock on day 90 of Trump 2.0 would require political conditions that do not currently exist — no visible US concession that Xi can present domestically, no Taiwan language that survives Republican Senate scrutiny. Deal risk is real; deal probability within the 18-month window is not the consensus narrative suggests.

Attack 2: CBP tightens certificate-of-origin rules and closes the gate unilaterally

US enforcement is not permanently stuck at "selective and capacity-constrained." The executive branch has unilateral authority to tighten certificate-of-origin rules. A new executive order substantially raising the local value-add threshold for electronics would make light-assembly routing in Vietnam economically unviable overnight. No China deal required — a single executive order, within CBP authority. Solar panels and steel got hit this way already. Electronics could follow.

Response: This is the most structurally serious attack on the position. The counter: Chinese FDI into Vietnamese manufacturing since 2022 has been building genuine local value-add capacity — full production lines, not bonded warehouses. CBRE Vietnam's industrial park data shows Chinese-backed tenants occupying manufacturing estates, not relabeling sheds. The installed base in electronics already clears the current threshold by meaningful margin; a tightened rule would squeeze marginal routing operations, not the industrial-park anchor tenants. The 18-month position horizon matters again: enforcement policy changes require regulatory notice periods, industry comment periods, and implementation timelines — CBP has never moved from announcement to enforcement on a major COO rule change in under 12 months. The risk is real but the timeline asymmetry favors the long. Risk management: monitor CBP rulemaking docket for electronics COO notice; exit if formal rulemaking initiates.

Attack 3: Vietnam loses its bridge role — one side stops tolerating it

Vietnam's bridge position requires both the US and China to tolerate it simultaneously. China can weaponize FDI as leverage — threatening to redirect Belt-and-Road manufacturing to Indonesia or Bangladesh if Hanoi fails to push back on US enforcement pressure. US pressure on Vietnam's rare-earth export policies or IPEF compliance could also trigger Vietnamese policy shifts. The bridge breaks if either side stops playing along.

Response: This attack underestimates Vietnam's structural geography and moat depth. Vietnam's manufacturing draw is geographic (coastal access, proximity to Chinese supply chains) and demographic (young labor force, lower wages than China). Indonesia and Bangladesh lack Vietnam's cluster density — the Chinese FDI that built these estates took four years of continuous capital deployment. The moat is not easily reproduced. More decisively: the US-China game creates a shared interest in Vietnamese tolerance that neither side can easily override. American importers need the supply relief; Chinese manufacturers need the routing outlet. Both Beijing and Washington have a functional side deal with Hanoi that is load-bearing for their own domestic political purposes. Vietnam's leverage is precisely that both sides need it more than they can punish it. The bridge breaks only if one side accepts the economic cost of losing Vietnam access — and neither has demonstrated that willingness.

Sources

Get next week's issue

One equilibrium read per week. Free during the soft-launch.