How We Built Limit-Down Counting: DIP-A Strategy Deep Dive

[michaeljiangmingfeng-debug]

We built DIP-A (Dip Signal A-shares) on a simple observation: when the market hits the limit-down rule in China, it's a real-time panic gauge that doesn't exist anywhere else on Earth.

No black box here. The mechanism: CSI300 futures hit limit-down → extreme retail selling pressure → mean reversion setup. We count these occurrences and weight them as a contrarian signal.

Why this matters: retail behavioral bias in A-shares is severe. We quantify it, we expose the logic, and we let you audit it. We don't pretend the strategy is perfect—it has its own role, its own edge, its own drawdown.

The Philosophy

Every strategy is a quantitative expression of macro factor dynamics. We won't distort a strategy's character to chase dead benchmarks. We dare to be transparent because the logic is sound.

Key points:

• Limit-down counting is unique to Chinese markets (no equivalent in US/EU)
• The signal fires ~1x per year on average — most of the time it's in cash
• 0% return = cash position = the strategy working as designed
• When it fires, you buy GEM ETF (创业板) at peak panic

Why Not ML?

We could have trained a neural network on A-share crash patterns. We didn't. Because:

1. ML patterns break when regimes shift
2. Limit-down rules are structural — they don't change
3. Retail panic is behavioral — it's constant across cycles

This is the courage of transparency: we expose our logic so you can verify it with ChatGPT, Claude, or DeepSeek. We want you to audit us.

🔗 Full mechanism: AI Hall - DIP-A
🔗 Try it: Playground