On April 15 I published The Final Rally in China A-Shares. It made two calls about what would follow.
First, this advance was fundamentally liquidity-driven. M1 growth had recovered from its late-2024 low and lifted the market from the 2,600s to above 4,000. But the central bank had already made clear that another large liquidity injection was unlikely. Liquidity was near its peak.
Second, liquidity and equity markets do not peak at exactly the same time. Momentum could carry prices higher for one last rally, much as a car keeps rolling after the driver lifts off the accelerator.
As positioning became more extreme, I warned about the risk on June 25. Last week's decline confirmed the adjustment.
M1 confirms the turn
Start with the source of the water.
M1 growth peaked at 7.2% in September 2025 and then eased. It was around 5% when the April article was published, rebounded to 5.49% in May, and fell to 3.97% in June. Subsequent policy signals told the same story: no new wave of water arrived.
Sources: People's Bank of China and Shanghai Composite. Latest available data.
The final rally came and went
The Growth 100 Total Return Index bottomed at 10,947 during the early-April shock. By June 22 it had reached 18,492, a 68.9% gain. Even measured from the April 15 publication date, the advance was 50%.
That was the final rally. It arrived, and it ended.
Source: China A-Share All-Share Index. Latest available trading day.
From the June 22 high through July 17, the Growth 100 fell 24.8%. The broad A-share index lost 14.4% over the same period. The correction was market-wide, but growth absorbed the heaviest damage. Both April signals, the M1 turn and the final rally, had played out.
Growth versus value reached a historic extreme
Source: CNIndex. Growth 100 TR (480080) and Value 100 TR (480081), 31 Dec 2012 = 1,000; through 17 Jul 2026.
The ratio divides the Growth 100 Total Return Index by the Value 100 Total Return Index. It measures the relative strength of growth against value. After reaching 2.43, the highest reading since the data began at the end of 2012, it reversed quickly. That exceeded even the 2.31 peak of the 2021 bull market.
After the 2.31 extreme on November 9, 2021, growth underperformed value by 33.4 percentage points over three months, 35.0 points over six months and 36.7 points over twelve months.
| Window | Growth 100 TR | Value 100 TR | Relative |
|---|---|---|---|
| 3 months | -18.5% | +14.9% | -33.4 pts |
| 6 months | -24.7% | +10.3% | -35.0 pts |
| 12 months | -26.0% | +10.8% | -36.7 pts |
| Window | Growth 100 TR | Value 100 TR | Relative |
|---|---|---|---|
| 3 months | -48.0% | -29.1% | N/A |
| 12 months | -41.1% | -28.7% | N/A |
In the current cycle the ratio rose from 0.94 on April 8, 2025 to 2.43 on June 25, 2026. From June 30 to July 17, Growth 100 fell 22.3% while Value 100 gained 6.5%. This was not merely a broad decline. It was a rotation.
This was not merely a selloff. It was a change in market leadership.
Turnover distribution and ETF flows are still needed to confirm whether money was leaving the entire market, but the style rotation closely resembles the aftermath of the previous two extremes. In the data, the selloff was not unexpected.
Why clear data still traps investors
If the April article called for a final rally and the late-June data were flashing red, why did so many investors stay?
Because narratives near a market top always give people a reason to keep buying: this is a technological revolution; old valuation rules no longer apply; every dip is an opportunity; the best strategy is simply to hold.
The language barely changes from one cycle to the next. Only the favored sector does: Internet Plus in 2015, core assets in 2021, and AI this time.
The issue is not a lack of forecasting ability. It is discipline. When the data say the odds have become poor, can you stop adding risk? Forecasting is a skill. Acting on deteriorating odds is discipline, and that is what most investors lack.
What may come next
There are two broad paths.
First, the growth-to-value ratio rebounds and forms a double top. Growth has fallen rapidly, so a tactical bounce back toward the previous high is possible before the ratio turns down again.
Second, leadership rotates directly into value, making the latest ratio high the peak of this style cycle.
For the broad market I retain April's view: a slow, grinding decline is more likely than a single crash. The central bank is not adding water, but it is not draining it abruptly either. For individual investors, a slow decline can be harder because every pause feels like the bottom and every bounce invites another wait.
A final note
My rule in April was simple: hold while the indicator rises; reduce exposure when it turns; leave when the decline is confirmed.
M1 has now confirmed its turn, the growth-to-value ratio has retreated from a historic extreme, and the data support a change in leadership.
Markets rise when money arrives, fall when it leaves, and rotate when one style has been pushed too far.
Review your own decisions over the past two weeks. Which narratives persuaded you to buy despite poor odds, and which hopes stopped you from acting when the data turned?
Understanding that is far more important than predicting tomorrow's close. Follow the data, not the crowd.