Following Stocks That Institutions and Private Equity Funds Are Accumulating — Tested with Ranking Correlation IC
Machine-readable: Markdown · 한국어 원문
In Part 1 we confirmed that there is no information in the direction of foreign investors' net buying. So what about institutions — and private equity funds in particular — the so-called "smart money"?
Test 1 — Following Monthly Rankings (IC)
- Method: The rank correlation coefficient (rank IC) between the monthly net-buying amount ranking of institutions and private equity funds and the next month's return ranking. An IC above +0.05 is treated as a signal worth using in practice
- Sample: 43 months
- Result: IC +0.010 — effectively zero. No tendency was observed for stocks heavily bought by private equity funds to rise more the following month
Test 2 — Chasing Stocks That Spiked After Institutional Pre-Buying
Buying at the close of the trading day after a spike (same-day gain of +13% or more) in stocks where institutions net bought 6% or more of trading volume over five trading days, and holding for 20 days:
| Signal | Sample | 20-day excess vs. market | Beat-the-market rate |
|---|---|---|---|
| Chasing spikes after institutional pre-buying ≥6% | 1,095 | -4.1%p | 41% |
Test 3 — Entering at the Accumulation Point Without Waiting for a Spike
Since the problem might be that "chasing after the spike is too late," we also tested entering immediately at the point where institutional five-day cumulative accumulation of ≥10% is observed, with no spike condition:
| Signal | Sample | 20-day excess vs. market | Beat-the-market rate |
|---|---|---|---|
| Entry at the point of institutional five-day accumulation ≥10% | 155,828 | +0.9%p | 55.5% |
+0.9%p is positive in direction, but given transaction costs and the dispersion within the sample, it is not large enough to count as a practical edge. The verdict was recorded as "negligible effect."
What These Numbers Say
By the time net buying by institutions and private equity funds is observable in disclosure data, there is almost nothing left to be gained from that information. If institutions really did buy on information, the data is consistent with the view that the information was already reflected in the price by the act of buying itself.
Limitations
- The IC test was run on a monthly basis over 43 months; effects at shorter frequencies (weekly, daily) were not tested separately
- "Private equity funds" refers to the private-fund investor category, not tracking at the individual fund level
Data Sources
- Korea Exchange daily trading data by investor type (per-stock investor flows can be checked on each stock page)
This article documents tests on historical data for informational purposes only. It is not investment advice or a recommendation to buy or sell any security. Past test results do not guarantee future returns.
Comments
Comments on methods, data and interpretation are welcome. Buy/sell recommendations for specific securities may be removed.