Testing Mean Reversion in Synthetic Forex Pair Imbalances

Summary

The article studies discrepancies between a quoted EURGBP rate and a synthetic rate derived from EURUSD and GBPUSD, treating deviations as possible trading signals. It reports statistics from two months of five-minute data, including negative skew, heavy tails, and positive lag-one autocorrelation. The proposed approach waits for a persistent imbalance and seeks to trade its return toward equilibrium, with discussion of liquidity differences, spreads, latency, and session effects as possible sources of divergence.

The author reports that European-session conditions appeared more favorable in the examined data and describes a MetaTrader 5 backtest, but the provided text is incomplete around the performance results. The analysis covers a short sample and does not establish robust out-of-sample profitability. Transaction costs, execution timing, extreme negative deviations, and distributional departures from normality are material risks; the article’s confident profitability claims should not be taken as independently verified evidence.

Key ideas

• A synthetic EURGBP rate can be compared with the quoted rate to measure cross-pair imbalance.
• The reported sample has strong negative skew and heavy tails, so normal-distribution assumptions may be unsuitable.
• Positive lag-one autocorrelation motivates waiting for persistent deviations and trading toward mean reversion.
• Liquidity, spread differences, latency, and asynchronous news responses may contribute to temporary discrepancies.
• The short sample and incomplete performance details leave profitability and out-of-sample robustness unproven.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.