The China Story: The 5% Illusion
How China counts its economy—and why the most important number in global macro is almost certainly wrong.
Every quarter, China’s National Bureau of Statistics releases a number, and every quarter, global markets move.
The number is GDP growth.
For 2024, it was 5.0 percent—precisely on target, as it almost always is. Analysts nod, fund managers rebalance, central banks update their models.
It probably shouldn’t.
This is not a fringe argument. It is the considered position of Rhodium Group, the Brookings Institution, the IMF’s own internal staff reports, and—most remarkably—a Chinese premier who once called his country’s own GDP statistics “man-made” and fit only “for reference.” The question is not whether China’s growth data has problems. The question is how large those problems are, how systematically they are structured, and what they mean for anyone attempting to price assets, commodities, or risk using Chinese data as an input.
The answer, assembled from four independent lines of evidence, suggests the divergence between official reporting and economic reality is not a rounding error. It is a structural feature of the Chinese statistical system—one that has grown wider than at any point in the post-2008 era.
I · The Identity Problem
Start with arithmetic. GDP growth has two versions: real (adjusted for inflation) and nominal (in current prices). The relationship between them is an accounting identity that cannot be violated:
Nominal GDP Growth = Real GDP Growth + GDP Deflator
Accounting identity · Non-negotiable
In 2024, China officially reported real GDP growth of 5.0 percent. But nominal GDP grew at only 4.2 percent. The implied GDP deflator is therefore negative—meaning the Chinese economy experienced deflation even as it supposedly grew at a pace that would be the envy of every developed nation.
In recorded economic history, no major economy has sustained 5 percent real growth while running consecutive years of deflation. Japan came close in the 1990s. The result was called the Lost Decade.
Evidence 01 · The Ministry of Finance Confession
Chart 01 · The GDP Illusion
II · What the Taxes Say
Tax revenue is the closest thing to a real-time audit of economic activity that a government produces. It is harder to smooth than output statistics because it flows through a separate administrative system—the finance ministry—that does not report to the same statistical bureau that produces GDP.
In 2024, China’s tax revenue declined 3.4 percent. Total fiscal revenue grew a nominal 1.3 percent only because non-tax revenue surged 25.4 percent—the result of local governments imposing fines and confiscations on businesses so aggressively that the central government publicly rebuked them.
An economy genuinely growing at 5 percent in mild deflation should produce tax revenue growth of roughly 3 to 4 percent. The gap between what taxes say and what GDP claims is approximately 8 percentage points.
III · Li Keqiang’s Confession
In 2007, a classified US diplomatic cable recorded a conversation between then-Liaoning Province Party Secretary Li Keqiang and the US Ambassador. Li, who would later become Premier, told the ambassador that China’s GDP figures were “man-made” and “for reference only.” He suggested three metrics were more reliable: electricity consumption, rail freight volume, and bank loans.
“GDP figures are man-made and for reference only.”
Li Keqiang, 2007 · Party Secretary of Liaoning · Later Premier of China · WikiLeaks diplomatic cable
The Li Keqiang Index shows far more volatility than official GDP—which is exactly what you would expect from a real economy subject to business cycles, rather than a statistic managed toward a political target. Physical flows of goods and energy are harder to manipulate than value-added accounting.
IV · The Provincial Accounting Problem
For years, the sum of all provincial GDP figures exceeded China’s national total—sometimes by 7 to 12 percent. Provinces over-reported because local officials’ promotions depended on economic performance metrics. The NBS acknowledged the problem and took over GDP calculation in 2014, but has never publicly explained how it reconciles local data with national totals.
Evidence 02 · The Provinces That Admitted It
In 2017–2018, Inner Mongolia cut its industrial output figure for 2016 by 40 percent. Liaoning and Tianjin made similar admissions. The NBS’s own audit of 2,051 companies found that 58 percent had severely falsified their data. These are not outliers. They are windows into a system that operates this way at every level.
V · The Deflation Impossibility
China has now experienced ten consecutive quarters of effective deflation. Producer prices have been negative for 39 straight months. This matters because deflation interacts with real growth calculations in a specific way: a falling price index mechanically inflates the real growth rate relative to what anyone actually experiences in terms of income, revenue, or wages.
Deflation flatters real GDP even as it devastates nominal economic reality. Workers’ wages grow slowly. Companies’ revenues shrink in nominal terms. Local governments’ tax bases erode. But the GDP deflator math makes all of this look like robust real growth.
Every independent methodology points below the official figure. None points above it. This is not noise. It is a directional signal.
VII · Why the System Can’t Stop
The system cannot easily stop overreporting because stopping would require admitting previous reports were false—which would trigger the confidence collapse the overreporting was designed to prevent.
VIII · What It Means
Monetary policy miscalibration. If the PBOC believes growth is 5 percent when it is 2.5 percent, policy will be structurally too tight. Real interest rates in China are above 4 percent in a deflationary environment—a pro-cyclical configuration that deepens the downturn.
Fiscal policy undershoot. Stimulus designed for a 5 percent economy is insufficient for a 2.5 percent economy. The medicine is dosed for a patient who weighs 80 kilograms when the patient weighs 40.
Commodity demand miscalculation. Iron ore, copper, and crude oil models calibrated to 5 percent Chinese growth carry a structural upward bias. The commodity supercycle narrative built on China’s growth story contains an embedded error that compounds over time.
Every independent methodology points below the official figure. None points above it. This is not noise. It is a directional signal.
ZTrader.AI Research · The China Story · Issue 01
Sources
[1] Rhodium Group — China’s Economy: Rightsizing 2025, Dec 2025
[2] Rhodium Group — China’s Financial and Fiscal Decay, Feb 2026
[3] CEIBS — China’s 2025 Economy, 2025
[4] IMF — 2025 China Article IV Consultation, Dec 2025
[5] St. Louis Fed — China’s Official Economic Data, 2021
[6] CSIS / Big Data China — Measurement Muddle, Oct 2023
[7] NY Fed — Is Chinese Growth Overstated? 2017
[8] Fortune — China’s Economic Data Is Famously Unreliable, Aug 2025
[9] Radio Free Asia — China Exposes Rampant Data Fraud, 2018
ZTrader.AI Research