What Do Low Frequency of Transaction Costs Really Measure?

• Activity Number: 500 - Statistical Challenges and Recent Advances in Finance and Business Analytics
• Type: Topic Contributed
• Date/Time: Wednesday, July 31, 2019 : 10:30 AM to 12:20 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Filip Zikes* and Mohammad Jahan-Parvar
• Companies: Board of Governors of the Federal Reserve System and Federal Reserve Board
• Keywords: Liquidity; Transaction costs; High frequency data; Realized volatility ; Effective spread

Abstract

A number of recent studies published in top finance journals argue that market liquidity and transaction costs can be accurately estimated from daily data. In this paper, we challenge these claims. We docent empirically, and corroborate by simulations, that: (1) when transaction costs are small relative to volatility, the low-frequency measures are severely upward biased; (2) the bias is a function of volatility and it is primarily volatility, not transaction costs, that drive the time-series dynamics of the low-frequency measures; (3) the bias and dependence on volatility have important implications for asset pricing applications. In summary, out answer is no, low-frequency measures do not generally deliver accurate transaction costs estimates for US stocks and FX rates.