Noise-Robust Ornstein–Uhlenbeck Estimation for Intraday Pairs Trading

Summary

The paper studies estimation of mean-reverting price dynamics from high-frequency observations, where market microstructure noise can bias parameter estimates. It proposes a maximum-likelihood estimator for an Ornstein–Uhlenbeck process that accounts for noise and accommodates irregularly spaced observations. It also establishes that an OU process with independent Gaussian white noise, sampled at equal intervals, has an ARMA(1,1) representation.

The authors apply the estimator to an intraday pairs strategy using mean-variance optimization. In an empirical study involving seven Big Oil companies, they report that using the proposed estimator increases strategy profitability. The supplied description does not give the sample period, implementation costs, detailed benchmark comparisons, or the size and statistical reliability of the improvement. The reported result therefore motivates noise-aware estimation but does not establish that the approach generalizes to other assets or trading conditions.

Key ideas

• High-frequency market microstructure noise can bias estimates of mean-reverting price processes.
• The proposed maximum-likelihood OU estimator handles noise and irregularly spaced observations.
• A discretely observed OU process with independent Gaussian noise has an ARMA(1,1) form.
• The paper combines the estimator with mean-variance optimization for intraday pairs trading.
• A study of seven Big Oil companies reports higher profitability, though the summary omits implementation and robustness details.

Tags

# 1811.09312


# Estimation of Ornstein-Uhlenbeck Process Using Ultra-High-Frequency Data with Application to Intraday Pairs Trading Strategy









When stock prices are observed at high frequencies, more information can be utilized in estimation of parameters of the price process. However, high-frequency data are contaminated by the market microstructure noise which causes significant bias in parameter estimation when not taken into account. We propose an estimator of the Ornstein-Uhlenbeck process based on the maximum likelihood which is robust to the noise and utilizes irregularly spaced data. We also show that the Ornstein-Uhlenbeck process contaminated by the independent Gaussian white noise and observed at discrete equidistant times follows an ARMA(1,1) process. To illustrate benefits of the proposed noise-robust approach, we introduce a novel intraday pairs trading strategy based on the mean-variance optimization. In an empirical study of 7 Big Oil companies, we show that the use of the proposed estimator of the Ornstein-Uhlenbeck process leads to an increase in profitability of the pairs trading strategy.

Shown in full with attribution under the source's licence. Licence: abstract CC0

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