We asked Andrew Hunt last week whether, after years of "yes, but liquidity", we were finally coming to the end of something. His answer was an unambiguous "Yes, I think we are." The mechanism is the US current account. Hunt's long-standing rule is that once the American deficit approaches 1% of global GDP — roughly a tenth of world savings — funding becomes a problem. For 18 months it has been comfortably funded by extraordinary Asian surpluses. Korea, Taiwan and Singapore have been running surpluses near 20% of GDP, driven first by selling chips into the AI build-out and now, with capacity exhausted, by 60–70% price inflation in those same chips. The proceeds were recycled into US assets, principally through sovereign wealth funds.

That recycling is now stalling, maybe not reversing, but definitely losing momentum. Korea's fund is buying its own equity market rather than America's, Taiwan is politically fluid, and Singapore's GIC has doubts about AI's profitability and the durability of the stock market boom. Beneath these sits a monetary limit : the surpluses have grown too large to sterilise, so Asian authorities are giving up and letting their currencies appreciate. Less capital flows into the dollar, and the price of capital in the US rises. There may be increasing geopolitical overtones to this calculation as well.

What makes this dynamic bite now is that Washington is not offsetting it as it has done in the past. It could run down cash balances, do more QE, or let banks fund themselves wholesale. Instead cash balances keep rising, QE has been cut to a trickle, banks are discouraged from wholesale funding, and the Treasury buyback was, in Hunt's words, "a peashooter". He reads this as deliberate: a Warsh–Bessent tightening by quantity, which slows an overheated economy without the Fed funds rate and without provoking the White House too much (recent rate hike notwithstanding). Credit booms, he has always argued, end when the supply of credit is interrupted, not when its price rises.

On that logic last week's Fed decision mattered less than the market thought. Hunt put a hike at 60% against a market at 90%; the Fed went to 3.75–4.00%, with 16 of 18 participants pencilling in another. Tighter than his base case, but the real tightening was already underway.

The Bank of Japan is the more interesting one. Hunt argued Tokyo should not be raising rates at all: domestic inflation is below 1%, the economy is soft, and the monetary overhang is already being addressed through aggressive quantitative tightening. But Washington wants the yen carry trade unwound, because when Warsh first tightened, Japanese banks sitting on carry-trade dollars simply stepped into the gap left by American ones. He sees echoes of 1985, Japan tightening for someone else's reasons. In the event the BoJ hiked to 1.25%, a 31-year high, but split 7–2 with two reflationist dissenters, and the yen fell. A disorderly carry unwind is the clearest route to his bad scenario, so a hike with a dovish tinge is welcome.

Is this an adjustment or a crisis? It depends entirely on speed. A simulation Hunt ran with a major bank found that forcing the deficit to close inside six months would push US unemployment to 9% and collapse the housing market. That is not his base case. His base case is funding falling 10–20% over roughly 18 months, "choppier seas, certainly not flat water". Controlled looks like a US 10-year at 5.25%; uncontrolled is above 6%. The incentives favour the managed path: China cannot afford an export collapse, and 3% of South Korea's population has just been margin-called on equity trading accounts.

That 18-month timeline is the opportunity. If the US must close its deficit, investment spending falls and that spending is AI. Hunt would reduce the Asian chipmakers "providing the shovels to the goldmine" and buy the domestic side of the same trade: money that stops leaving Asia stays in Asia, favouring Korean retailers over Korean exporters. Europe gets a stay of execution, since Asia will fund European deficits before US AI capex, though as the surpluses shrink, European yields rise and domestic Europe weakens.

Fixed income is where the repricing has already happened. At 5% the maths looks very different from 4%, and US duration is entering territory where it starts to work, with the added attraction for Australian investors that hedging currently adds around half a point a year rather than costing anything. That said the case for turning point in this tidal surge in yields is not yet clear cut so we wouldn't call this a conviction call quite yet.

One signal governs the rest: dollar–Asia. Andrew suggests that violent moves would be the cue to cut risk sharply; orderly ones mean there is time to reposition. Credit spreads, notably, still have not repriced, the one market not yet corroborating this whole story. The conclusion from the discussion was not to run for the hills, rather to build a portfolio that fits this world rather than fights it.