Method notes behind a public, scored data record. We publish the coefficients that survived, the hypotheses that did not, and the conditions under which we would be proven wrong.
A language model writes our event results and a script checks every number against the official release. We replayed the checker on 16 releases from August to November. It rejected 10 correct figures and let wrong ones through. Here is what we fixed.
We tracked K-pop attention daily for 2.7 years. After 65 spikes, agency stocks beat the KOSPI 29 times. Six of the ten biggest came after a rally.
Korean retail money is leaving stocks and the kimchi premium is back in the news. Two months of half-hourly records say the two never moved together.
The press blamed US rates and the yen. Our daily ruler says both were weak — the extra drop vanished in a day and the real move started the day after.
Nvidia beat, and everyone wants the next Asian chip trade. We ran 113 tests on 20 Taiwan stocks. Verdict: nothing to pick — the market is the stock.
Seoul crashed 5.8% and got it back in a day. Our ledger says only the index returned — the money is still sitting by the exits, four sessions on.
Washington funded domestic transformers on 13 August. Three days later our ledger crossed the other way — the first state change since we began.
China is still first in US battery imports, despite the tariffs. But in chips it is eighth. The split happened to a product, not a country.
A 51% markup on Nasdaq taught crypto’s oldest lesson to the stock market. We measure the kimchi premium every thirty minutes — all July it stayed inside a tenth of one percent.
The KOSPI lost a fifth of its value in four sessions. In the same week, three chipmakers printed records — and the customs ledger set a new high.
Tokyo follows the US close with beta 0.55. But the invoices go to Greater China — and the US share of Japan’s chip exports is about two percent.
The transmission coefficient works in semiconductors. It does not work here — and the reason turned out to be the mechanism itself.