Cross-Impact of Order Flow Imbalance in Equity Markets (2112.13213v4)

Abstract: We investigate the impact of order flow imbalance (OFI) on price movements in equity markets in a multi-asset setting. First, we propose a systematic approach for combining OFIs at the top levels of the limit order book into an integrated OFI variable which better explains price impact, compared to the best-level OFI. We show that once the information from multiple levels is integrated into OFI, multi-asset models with cross-impact do not provide additional explanatory power for contemporaneous impact compared to a sparse model without cross-impact terms. On the other hand, we show that lagged cross-asset OFIs do improve the forecasting of future returns. We also establish that this lagged cross-impact mainly manifests at short-term horizons and decays rapidly in time.

• The paper introduces an integrated OFI variable via PCA that more accurately captures cross-impact on asset prices than using best-level measures alone.
• The paper employs OLS and LASSO regression models on NASDAQ LOBSTER data for the top 100 S&P 500 stocks to assess both contemporaneous and predictive impacts.
• The paper finds that lagged cross-asset OFIs significantly predict short-term returns, offering actionable insights for algorithmic trading strategies.

Analysis of Order Flow Imbalance in Equity Markets and Its Cross-Impact

The paper "Cross-Impact of Order Flow Imbalance in Equity Markets" by Cont, Cucuringu, and Zhang offers an empirical investigation into the implications of order flow imbalance (OFI) on asset price dynamics within a multi-asset framework. The authors present a nuanced exploration of both contemporaneous and predictive facets of OFI's impact, integrating analysis across multiple levels of the limit order book (LOB) and multi-asset settings.

The primary contribution revolves around quantitatively integrating OFIs from different levels of the LOB into an integrated OFI variable. The authors demonstrate that this integrated OFI, constructed through principal component analysis (PCA), can capture a more comprehensive impact on prices compared to using only best-level OFIs. Empirical evaluations reveal that while contemporaneous cross-impact models with multi-level OFIs do not enhance explanatory power for immediate price movements beyond a sparse model, lagged cross-asset OFIs show significant predictive capabilities for short-term future returns. This finding underscores the transient nature of cross-impact effects, which dissipate over extended horizons.

Methodologically, the paper utilizes an array of linear regression models, blending ordinary least squares (OLS) and LASSO to estimate the parameters of both price impact and cross-impact models. The paper tests these models on NASDAQ's LOBSTER data for the top 100 S&P 500 index components between 2017 and 2019. In-sample and out-of-sample metrics, including adjusted R-squared, are meticulously reported, providing robust insights into the models' performances.

The research unveils that a sparse cross-impact model including only statistically significant cross-asset interactions can slightly enhance explanatory power in certain scenarios. This aspect is primarily significant for assets with significant intraday OFI variability. The network analysis of cross-impact coefficients elucidates a sectorial structure, where orders from certain sectors (e.g., Information Technology and Communication Services) exert a stronger predictive influence on peers. The singular value decomposition (SVD) of the coefficient matrices further suggests a low-rank structure, indicative of the concentrated nature of cross-asset interactions.

From a predictive standpoint, OFI-based models exhibit noteworthy economic gains in out-of-sample tests, when used to inform trading strategies, surpassing models based on returns alone. This points to the practical relevance of incorporating cross-asset OFI variables in high-frequency trading strategies, especially in markets characterized by latent information diffusion across correlated assets.

The implications of this paper for equity trading and portfolio management are multifaceted. Theoretical advancements center on the dynamic modeling of LOB data, proposing that integrated OFIs might serve as a critical tool for understanding and predicting market movements. Practically, these insights can inform algorithmic trading designs that capitalize on identified short-term cross-impact opportunities, optimizing execution costs and improving market timing.

Future research avenues could explore the temporal dynamics of cross-impact, particularly in after-hours trading scenarios where liquidity constraints might accentuate such effects. Moreover, exploring the adaptivity and extensibility of these models in illiquid markets or under varying macroeconomic conditions would be beneficial. Enhancing predictive models through machine learning techniques that capture nonlinear relationships may also uncover further empirical regularities missed by linear models.

In closing, the paper enriches the existing literature by elucidating the complexity inherent in order flow dynamics and cross-asset interactions, providing a sophisticated lens through which both market participants and researchers can better interpret equity market behaviors.