WP303_tradingdata

Abstract

We use proprietary investor transaction-level data from two large German retail banks. For a large random sample of clients at the banks, we observe all their trades, including information on the securities traded, the type of trade (buy or sell), day and time of the trade execution, price and units of each transaction, and, importantly for our analysis, the platform used for each trade. This data covers about sixty-five million transactions during the years from 1999 to 2017 by more than two hundred and twenty-five thousand investors. At the investor level, we observe monthly snapshots of portfolio holdings and demographic characteristics such as gender, age, wealth, and income.