Hong Hao: AI Pullback Is a Pause, Sector Rotation Far from Over
Taylor Wilson
Economist Hong Hao sees the current AI sell-off as a mid-correction, not a trend reversal, with capital rotating out of overheated semiconductor plays and back into neglected old-economy names like Tencent and Alibaba — a style shift he says is only halfway done.
Has the AI sell-off run its course?
Using South Korea's KOSPI index — a benchmark tracking Samsung, SK Hynix, and other core AI-chain stocks — Hong Hao argues the AI sector correction is far from over.
Some individual names have already fallen sharply, but he expects the adjustment to last well beyond one or two weeks, with the near-term trend still tilted downward.
This means → investors trying to bottom-fish AI stocks may face an extended choppy period, not a clean V-shaped rebound.
He adds that in high-volatility conditions, neither bulls nor bears earn stable profits — chasing momentum in either direction carries outsized risk.
Why is money flowing from AI into Tencent and Alibaba?
Over the past few weeks, capital has rotated visibly into previously neglected names — Tencent, Alibaba, and even Meituan (despite its food-delivery price war) have all posted notable gains.
Hong Hao's core logic: these sectors were cold for nearly a year; the reflow takes time, and investors heavily positioned in semiconductors need to gradually accept the style shift.
In plain terms = if AI truly drives broad economic growth, old-economy companies like Tencent and Alibaba should benefit in tandem — the rally should be broad-based, not confined to chip stocks alone.
Does valuation support the rotation?
Hong Hao cites Alibaba as a case study: the stock earlier fell below $90, below its IPO price — a level he calls "clearly irrational, as if a decade of effort counted for nothing."
This reflects excessive pessimism in old-economy pricing, which paradoxically provides fundamental support for the rotation trade.
The CSI Value Index has begun outperforming the CSI Growth Index, and Hong Hao expects that relative performance to persist for some time.
This means → the style is shifting from "growth premium" to "value repair" — portfolios skewed toward growth may continue to underperform in the near term.
Will China unleash large-scale stimulus in the second half?
Hong Hao expects the policy tone to be "stability first", focused on shoring up weak spots and boosting consumption — not a major stimulus push.
He notes that Q2 retail sales significantly beat expectations, industrial output held up, and exports kept outperforming despite a strong renminbi — but real estate and related sectors remain weak, making the economic picture notably uneven.
In plain terms = the economy is not weak enough to warrant "big moves." As long as full-year growth stays near the ~5% target, policymakers are unlikely to add firepower. The PBOC's balance-sheet expansion capacity remains at historical highs, leaving little incentive to expand further.
How do the Fed and the renminbi affect asset pricing?
Hong Hao says that whether the Fed chooses to raise rates or shrink its balance sheet, either path puts downward pressure on risk assets.
From a monetarist standpoint, the Fed's excessive monetary expansion over recent years has not fully worked through — even with June inflation data coming in better than expected, inflation and inflation expectations may continue rising over the coming months.
On the renminbi, Hong Hao believes the long-term appreciation trend is most likely intact, but this year's pace has been too fast and momentum has reached extremes; the short-term trajectory may need to slow. The dollar is currently basing around 100, biased higher, which will temporarily cap renminbi gains.
This reflects a contradiction: the renminbi's real effective exchange rate is one of the most undervalued currencies globally, with appreciation pressure far exceeding depreciation risk — yet the short-term move has already overshot.
What signal would confirm an AI rally restart?
Hong Hao identifies late-July earnings from US cloud vendors — and whether they maintain capital-expenditure guidance — as the key checkpoint for judging if the AI trade can reignite.
This means → capex figures from Microsoft, Amazon, and Google will directly shape market confidence in AI demand. If guidance is cut, the correction could deepen further; only if guidance holds or rises can capital realistically rotate back into AI.
Content is for reference only, not financial advice.