Goldman Sachs: China's K-Shaped Divergence Widens, Urgency for 924-Style Broad Easing Remains Low
nashnova research
Goldman Sachs says China's supply-demand gap is widening — August industrial output accelerated to 5.2% while retail growth slowed to 0.4%, and official data may understate the real shortfall. This means → unless employment deteriorates sharply, broad easing is unlikely in the near term.
Supply strong, demand weak — how big is the gap?
August industrial production grew 5.2% year-on-year, up from 4.5%. Retail sales growth slowed from 0.6% to 0.4%.
Goldman argues official data understates the weakness: large retailers saw sales fall nearly 4%, while small retailers grew 3%. Statistical coverage differences partly mask how soft domestic demand really is.
This means → industrial data alone paints a recovery picture, but the consumer side tells the opposite story. The supply-demand "scissors gap" keeps widening.
K-shaped divergence — which sectors are soaring, which are sinking?
Within tech manufacturing, extremes coexist: August industrial-robot output rose 34.6% and semiconductors 20.6%, while micro-computers fell 25.6% and smartphones 22.3%.
Retail mirrors the split: telecom-equipment sales jumped 27.3%; auto sales dropped 18.5%.
In plain terms = the AI capex boom is pulling one slice of tech onto a fast track, while consumer electronics and autos are sliding down a separate lane. That is what "K-shaped" means — two lines in the same economy, one up, one down.
When will property hit bottom — will the usual playbook apply?
By the historical pattern of large global housing downturns, China's market is approaching a typical trough — roughly 30% real-price decline over about six years, which would imply a bottom around 2027.
Goldman doubts China will follow that script. The key difference: when U.S. housing bottomed, the labor market was already recovering. China's labor market remains weak — rents were still down 0.6% year-on-year as of August 2026.
This reflects a critical shift. The old story was "property dragging the economy down." Increasingly, it is "the economy dragging property down." Outside tier-one cities and the high-end segment, further price weakness remains the likely path.
Credit slowdown — successful rebalancing or demand deficit?
Bank loan growth fell from above 10% in 2023 to 5% by August 2026. Total social financing growth — a broad measure of economy-wide credit — dropped from near 10% to 7.2% over the same period.
PBOC Governor Pan Gongsheng frames the slowdown as a natural result of structural transition: credit-heavy sectors like property and infrastructure are giving way to AI and advanced manufacturing.
Goldman flags the other side: the same industrial shift hurts employment. Per unit of value added, services are over 20% more labor-intensive than manufacturing; construction is over 100% more. This means → the further the economy tilts toward high-end manufacturing, the fewer jobs it creates — wages and consumer spending may stay under pressure.
Why isn't broad easing coming — what is policy waiting for?
Goldman judges that unless the labor market deteriorates sharply, policymakers see little urgency for additional broad easing. At the September 15 NBS press conference, officials highlighted industrial robots and semiconductors rather than smartphones and autos — reinforcing that read.
In plain terms = as long as the "good-looking numbers" hold, policymakers have cover to stay put. Consumer-side weakness is temporarily masked by structural bright spots.
Goldman expects sequential growth to pick up in coming months (faster government-bond issuance, more infrastructure spending), but full-year GDP is projected at 4.5% — right at the lower bound of the 4.5%–5.0% target range.
How will the growth target adjust — what is the path of least resistance?
Goldman highlights a policy dilemma: a major fiscal expansion to boost consumption runs into debt-sustainability concerns; shifting spending from investment to consumption runs into pressure from the growth target itself.
The path of least resistance may be to gradually lower the target — for example, from 2026's "4.5% to 5.0%" to "around 4.5%" for 2027 — while pursuing gradual yuan appreciation and further export-rebate cuts to address a large trade surplus and rising trade friction.
This means → the December Politburo meeting and Central Economic Work Conference will be the key signposts. Any shift in wording around next year's growth target is the earliest signal of a policy turn.
市场有风险，内容仅供研究参考，不构成投资建议。