I read China's macro numbers before I look at any equity screen. The habit started because China releases most of its data in one batch, so a single morning gives you the whole tape at once. The problem is the tape never says one thing. Here is what the July batch said when I pulled it on September 1.
cpi cooled faster than the run rate suggested
Consumer prices rose 0.5% year over year in July. In June the same gauge was 1.0%. Month over month prices fell 0.1%.
Going from 1.0% to 0.5% in one month is a big step, even for a volatile series. It tells me the disinflation impulse is not gone. Nothing here points to a hard deflation scare, but the momentum clearly stalled.
the trade surplus is still huge, just a bit less huge
China's trade balance with the United States stayed deeply positive in July. Exports were about $41.9 billion, imports about $14.0 billion, leaving a surplus of $27.9 billion.
That is still a very fat number. But June's surplus was $28.9 billion, so July shrank by a bit more than 3%. Exports themselves eased from $43.5 billion to $41.9 billion. One month is not a trend, but the direction flipped from widening to narrowing.
housing is split by city, not by direction
New home prices in July rose 3.0% year over year in Shanghai and 2.6% in Hangzhou. The same month, Xi'an fell 5.6% and Chengdu fell 5.1%. The second-hand market is weaker almost everywhere. Beijing second-hand homes are down 4.5% year over year, Shanghai down 2.0%.
The aggregate "China housing" story hides two markets. Tier-one coastal cities are stabilizing, while weaker cities keep sliding. Any single national number will mislead you.
the stock market sits right under the round number
The Shanghai composite closed at 3986.3, up 0.86% on the day. It is knocking on 4000. Kweichow Moutai, the largest stock by weight, trades at 1299.52 with a price to earnings ratio of 6.47.
A blue chip at PE 6.5 is cheap against its own history. Combined with 0.5% CPI, the equity market is pricing something between stagnation and slow recovery, and it has been climbing anyway.
the honest part
I need to be careful about what this batch can and cannot say. First, these are July numbers read on September 1. Monthly releases arrive with a lag, so this is a rearview mirror, not a steering wheel. Second, the GDP figure is nominal and annual: 2025 nominal GDP was 140.19 trillion yuan, up 3.9% from 2024. Nominal growth mixes volume and price, and with CPI near zero the real number is higher than 3.9%, but the exact split depends on deflators I cannot see from this feed. Third, the housing sample is ten major cities, not the full 70-city set, so the tier-one stabilization may not extend to smaller places. Fourth, official Chinese data gets revised, and different sources sometimes disagree on the same month. I quote the feed, not my own opinion of the feed.
What I actually take from this: inflation momentum has stalled, the trade cushion is still thick but no longer growing, housing has split into two markets, and equities are grinding toward 4000 on modestly priced leadership. None of that is a trade signal. It is a base layer of facts to update every month.
how i pull this
All of these figures came from one normalized endpoint instead of ten scattered portals. A plain ask with no API key looks like this:
curl -X POST https://agentdatum.com/api/v1/ai \
-H "Content-Type: application/json" \
-d '{"query":"give me the latest china cpi, china us trade balance, and tier one housing price changes"}'
The full machine-readable catalog of what is available is here: https://agentdatum.com/.well-known/ai-catalog.json
This article was written with AI assistance for drafting and editing.
Top comments (0)