This paper provides a new empirical perspective on a long and continuing debate about the relationship between trading costs and financial market volatility. At least since Keynes stock market critique in 1936, stock price volatility has been related to low transaction costs which allegedly facilitate destabilizing financial speculation. In spite of the prominence of this idea in the general public, hardly any evidence exists on whether higher transaction costs foster or mitigate financial price volatility.
The question is interesting in at least three respects. First, regulatory, organizational and technological progress has considerably decreased transaction costs. Widespread financial market liberalization in the 1980s lowered trading commission and electronic trading in the 1990s further diminished stock trading cost. But at the same time individual stock volatility appears to have increased in the U.S. (Campbell, et al. (2001)). It is unclear if there is a causal link here or just coincidence. Second, transaction costs are influenced by the microstructure organization of the market. The introduction of smaller pricing grids (ticks) in the U.S. with price steps of 1/16th of a dollar instead of 1/8th appears to have reduced transaction costs for the majority of investors.
Does this regulatory transaction cost benefit come at the expense of higher stock price volatility or do we obtain more price stability at the same time? Third, transaction costs sometimes include a tax component. While security transaction taxes have generally decreased in the 1990s, they remain nevertheless important in a few countries like the U.K.2 Moreover, parts of the anti-globalization movement have elevated global security transaction taxes to one of their policy objectives. The policy debate about financial market stability seems to evolve around convictions rather than sound evidence.
To our knowledge this is the first paper which provides clean and strong statistical evidence on the nexus between transaction costs and stock price volatility. We use a large data set on the French stock transactions between 1995 and 1999 to show that higher transaction costs increase stock return volatility. Prior to 1999 and the introduction of the euro, French stocks were subject to an important transaction cost increase if their price moved above the French francs (FF) 500 price threshold. Above FF 500, the minimal tick size for quotes in the centralized electronic order book increased by a factor of 10 from FF 0.1 to FF 1.
The smallest feasible percentage spread for stock quotation therefore increased from 2 to 20 basis points. We document that the 20 basis point spread is indeed frequently binding for stock prices above FF 500 and therefore constitutes an exogenous cost component induced by the pricing grid of the electronic order book. The French stock market thus provides an ideal natural experiment on the role of transaction costs for stock return volatility.
Download
The Role of Transaction Costs for Financial Volatility: Evidence from the Paris Bourse
