Paper: arXiv 2309.06711
Authors: Jérôme Busca, Léon Thomir
Abstract
It is a well-documented fact that the correlation function of the returns on two “related” assets is generally increasing as a function of the horizon $h$ of these returns. This phenomenon, termed the Epps Effect, holds true in a wide variety of markets, and there is a large body of literature devoted to its theoretical justification. Our focus here is to describe and understand a deviation to the Epps effect, observed in the context of the foreign exchange and cryptocurrency markets. Specifically, we document a sharp local maximum of the cross-correlation function of returns on the Euro EUR/USD and Bitcoin BTC/USD pairs as a function of $h$. Our claim is that this anomaly reveals the activity of short-term momentum traders.
Complexity vs Empirical Score
- Math Complexity: 5.5/10
- Empirical Rigor: 7.0/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper presents an interesting empirical discovery by documenting a localized deviation from the classic Epps effect in FX and crypto cross-correlations. It combines sound statistical time-series analysis with a plausible behavioral market-microstructure explanation linked to short-term momentum trading.
Research Flowchart
flowchart TD
A["Research Question: Investigate cross-correlation anomaly (Epps Effect deviation) in FX & Crypto"] --> B["Data Inputs: High-frequency EUR/USD and BTC/USD return data"]
B --> C["Compute Cross-Correlation Functions across varying time horizons (h)"]
C --> D["Analyze Lead-Lag Relationships & Market Microstructure"]
D --> E["Findings: Sharp local max in cross-correlation at short horizons reveals short-term momentum traders"]