Zhihu Review: A-Share's Last Rally — M1 Framework Verification
The M1-based framework for timing A-shares is solid, but the core data is wrong: M1 peaked in Sept 2025, and the 'last rally' already played out
Original Thesis
The author’s chain of reasoning:
M1 growth rebounds → Liquidity drives A-shares higher → PBoC stops easing → M1 growth peaks → Inertial final rally → Slow grind down
The framework is sound. But the core data is wrong.
Data Verification
M1 Growth: Article Says “Still Rising”, Actually Peaked 7 Months Ago
Pulling PBoC data (via Eastmoney API), M1 YoY growth (new definition from Jan 2025):
| Month | M1 YoY% | M2 YoY% | M1-M2 Spread |
|---|---|---|---|
| 2025-01 | 0.4 | 7.0 | -6.6 |
| 2025-02 | 0.1 | 7.0 | -6.9 |
| 2025-03 | 1.6 | 7.0 | -5.4 |
| 2025-04 | 1.5 | 8.0 | -6.5 |
| 2025-05 | 2.3 | 7.9 | -5.6 |
| 2025-06 | 4.6 | 8.3 | -3.7 |
| 2025-07 | 5.6 | 8.8 | -3.2 |
| 2025-08 | 6.0 | 8.8 | -2.8 |
| 2025-09 | 7.2 | 8.4 | -1.2 |
| 2025-10 | 6.2 | 8.2 | -2.0 |
| 2025-11 | 4.9 | 8.0 | -3.1 |
| 2025-12 | 3.8 | 8.5 | -4.7 |
| 2026-01 | 4.9 | 9.0 | -4.1 |
| 2026-02 | 5.9 | 9.0 | -3.1 |
| 2026-03 | 5.1 | 8.5 | -3.4 |
| 2026-04 | 5.0 | 8.6 | -3.6 |
M1 growth peaked at 7.2% in September 2025 and has declined 2.2 percentage points since. The article claims M1 “has been rising from the late-2024 lows all the way to now.” That is factually incorrect.
CSI 300: Indeed Followed M1’s Lead
| Timeline | Event | CSI 300 |
|---|---|---|
| Sep 2024 | “924” rally launch | 4,017 |
| Sep 2025 | M1 growth peaks (7.2%) | 4,640 |
| May 13, 2026 | CSI 300 YTD high | 4,998 |
| Jun 10, 2026 | This article’s date | 4,748 |
M1 peaked (Sep 2025) → CSI 300 peaked (May 2026): ~7 month lag. This validates the “inertial rally” thesis — step on the gas then let go, car coasts forward.
M1-M2 Spread: Persistently Negative, Money Not “Activating”
M1-M2 spread has been negative throughout (2026-04: -3.6%), meaning:
- Corporate demand deposit growth < time deposit growth
- Money is moving from current accounts to savings, not the other way
- Liquidity is tightening, not easing
This contradicts the article’s claim that “water is coming.” Water came. Now it’s leaving.
Point-by-Point Verdict
| Claim | Verdict | Notes |
|---|---|---|
| A-shares are a liquidity-driven market | ✅ Correct | M1-CSI 300 correlation is high, 10+ year track record |
| M1 growth still rising | ❌ Factually wrong | M1 peaked Sep 2025 (7.2%), down to 5.0% by Apr 2026 |
| PBoC done easing | ⚠️ Oversimplified | PBoC is in wait mode, but may be forced to act if economy deteriorates |
| Inertial rally logic | ✅ Correct | 7-month lag from M1 peak to market peak matches history |
| Daily volume ¥4-5T = top signal | ❌ Made up | No structural basis; “abnormal volume” threshold scales with market cap |
| Liquidity turning → gold benefits | ❌ Logic error | Gold benefits from falling real rates, not tightening liquidity. Tightening may push real rates up, pressuring gold |
Conclusion
The author’s M1 framework is a good tool, but made a fatal error: wrong current data.
- M1 growth already peaked in Sep 2025, not “still rising”
- CSI 300 already peaked in May 2026 (YTD high 4,998); the “last rally” already happened
- The market is in the early stages of an M1 downcycle, not an “inertial rally” phase
Score: 5.5/10 — Framework is sound, but the core data error reverses the conclusion. If the author had written this in September 2025, the analysis would have been remarkably prescient.
Follow the data, not the wind. Including the wind from this article.