August 5, 2024.
Global markets were bleeding.
Japan: -13%
Korea: -10%
Taiwan: -10%
Nasdaq Futures: -5%
I sent this note to my clients that day:
I understood exactly what was happening.
Japan had raised rates by 0.25%. The yen carry trade, a $2 trillion arbitrage that had quietly funded global risk appetite for two decades, was unwinding. Money was rushing back to Japan.
My read on the trigger? Spot on.
My call on timing? Completely wrong.
What actually followed:
Indian markets severely underperformed
Correction dragged on till March 2025
Six more months of consolidation
Over 12 months of pain
I told clients “next few days/weeks.”
It took over a year.
I understood the WHAT.
I completely misjudged the HOW LONG.
That failure has stayed with me for 14 months.
What did I miss?
Right Trigger. Wrong Duration.
Here’s what I got right:
The trigger (BOJ rate hike)
The mechanism (yen carry trade unwinding)
The strategy (avoid triple digit P/E stocks)
Here’s what I got wrong:
Duration
One word. That’s all it took to turn a good analysis into a bad call.
Understanding what is happening is easy now. X(Twitter), Bloomberg, NSE data, it’s all there.
But knowing how long the pain lasts? That’s where everyone struggles.
Hedge funds have systems for this. Real-time flows. Scenario models. Duration frameworks.
Retail investors get the news after the first move is done.
We know the “what.” We guess the “how long.”
That guess cost my clients 12 months of opportunity cost.
This is the gap I wanted to close.
Closing The Gap
So I built an AI workflow that does what I couldn’t do manually in August 2024.
It doesn’t just identify triggers. It models scenarios and maps how global events transmit to local markets.
What it does:
Pulls live BOJ policy data, yield curves, and FII flows
Runs institutional-grade scenario analysis (dovish, base, hawkish)
Maps which sectors break first and where money rotates
Outputs a visual dashboard anyone can actually read
Time to run: 13 minutes.
Cost: Zero.
The workflow has two parts:
1. Data gathering (Perplexity): pulls real-time market data with sources
2. Analysis engine (Claude): runs scenario modeling and generates the report
You can run it for any macro trigger. Any market. Any sector focus.
Let me show you how.
Before You Run The Prompt
This workflow has two steps:
Step 1: Gather real-time data (Perplexity)
Step 2: Run the analysis (Claude, Grok, or any reasoning model)
For Step 1, you don’t need to download anything. Perplexity pulls live data from BOJ, RBI, NSE, and other public sources.
Just run the data gathering prompt first. Copy the output. Paste it into the analysis prompt.
That’s it.
If you want more accuracy, you can optionally add:
Latest FII/FPI flow data from NSDL
Current Japan yield data from BOJ website
Recent RBI policy statements
But the prompt works without these. Perplexity fills the gaps.
Tool Demo: