KOSPI back-testing methodology, KRX execution mechanics, and Korean equity factor rotation — written for researchers who work with real market data, not abstractions. Articles document findings; they do not constitute investment advice.
Most retail back-testers ignore KRX lot-size constraints and model fills at mid-price. This article quantifies how lot-level slippage accumulates across quarterly KOSPI 200 constituent rebalances — and why 20–60 bps per fill is not a rounding error.
Read article →
Over 60% of KOSPI 200 배당락일 (ex-dividend dates) fall in the final two weeks of December. We examine how naive rotation strategies misalign with KRX T+2 settlement and inflate apparent annual returns by 30–80 bps.
Read article →
KOSDAQ's thinner liquidity and wider factor dispersion produce back-test outcomes that look very different from KOSPI 200. We compare momentum and value rotation signals across both universes using identical parameter sets.
Read article →
Retail order flow around KOSPI 200 quarterly 구성종목 변경 (constituent rebalancing) shows patterns consistent with recency bias and short-term herding. Systematic strategies need to account for this — or it negates the point of having a signal.
Read article →
With roughly 20 years of clean KOSPI data and KRX constituent history, even a 3×3 parameter grid produces out-of-sample degradation that looks like a regime shift but is just overfitting. We show where the line is.
Read article →
Public KRX data provides OHLCV but not full order book depth. This article documents the volume-weighted spread reconstruction methodology Finology applies to approximate historical bid-ask surfaces for slippage estimation — including where it breaks down.
Read article →
삼성전자 and the 2차전지 (secondary battery) cluster together exceed 35% of KOSPI 200 weight. We back-test momentum-based rotation signals between these two sectors across 2005–2024 and examine how lot-level slippage changes the result. Past back-test performance does not predict future returns.
Read article →