MONTHLY UPDATE
Quantitative research on Korean equities.
Factor signals, ML filters, and reproducible backtests on KOSPI50.
TQK STRATEGY · KOSPI200 TOP 50
Cumulative Return
Backtest: 2023.01 — 2026.04
KOSPI as of 2026-05-12
REGIME-ADAPTIVE SYSTEM
How the strategy adapts.
The system classifies the market into three regimes each month and switches scenarios automatically — aggressive when trending, balanced in chop, defensive when falling.
▲BULL
Trending up
Scenario
KOSPI200 · S3 · 3 · 3 holdings
CAGR-weighted — aggressive
Monthly avg (backtest)
+5.01%
vs KOSPI +6.30%▼ -1.3pp
━SIDEWAYS
Range-bound
Scenario
KOSPI200 · S4 · 7 · 7 holdings
Sharpe-weighted — balanced
Monthly avg (backtest)
+0.24%
vs KOSPI +0.51%▼ -0.3pp
▼BEAR
Trending down
Scenario
KOSPI50 · S8 · 5 · 5 holdings
MDD-weighted — defensive
Monthly avg (backtest)
+1.78%
vs KOSPI -4.24%▲ +6.0pp
METHODOLOGY
Backtested across 12 years.
8 years for design, 1 year for tuning, 40 months for live validation. We split the data this way to verify whether the system actually works on data it has never seen.
Training · 8Y
2014 – 2021
Model design and learning
Validation · 1Y
2022
Hyperparameter tuning
Out-of-sample · 40MO
2023.01 – 2026.04
Verified on unseen data
Data sources
KRX · DART · Kiwoom OpenAPI+
Universe
KOSPI200 · Top 200 by market cap
Rebalancing
First Monday of each month
FREE NEWSLETTER
Get the free monthly note
One email a month. Market commentary and a teaser of the premium signal.
No spam. Unsubscribe with one click.
WHY THEQUANTKOREA
What makes us different.
An item-by-item comparison of how we differ from most channels.
The comparison describes general patterns observed in the industry, not any specific channel. Every TheQuantKorea claim is verified monthly through real-money trading results.