Securities Market Design: Exchange Access Fee, Tick Size, and Dynamic Price Limit

Abstract

The impact of market structure designs on market quality is of interest to academics, practitioners and regulators. Using three exogenous events and with the benefit of proprietary data, this thesis builds on theoretical models and examines the casual impact of three important market structure designs - exchange access fee, tick size, and dynamic price limit - on market quality. First, I provide a theoretical model of the exchange fee structures to show that in a competitive environment, traders optimize the degree of information in their trades to fully exploit fees/rebates. I examine the ‘natural’ experiment of a unilateral Nasdaq exchange fee reduction, and find evidence consistent with the theoretical model which indicates that the expected ‘washing-out’ of the fee changes is offset by the flight of highly-informed market orders to the remaining highest rebate venues. Far from fees washing-out, there is a redistribution of informed traders across venues. Second, I extend the theoretical model to examine the impact of the 2016 U.S. tick size pilot. I show that the information content of trades increases more in markets that subsidise liquidity providers. Moreover, markets subsidising liquidity consumers and off-exchange trades are the major beneficiaries of the sizeable rise in the tick size that acts as an additional transaction tax paid especially by liquidity traders and a corresponding subsidy to limit orders. This sizeable increase in transaction costs means that those uninformed traders that remain in the lit market during the pilot are far more price sensitive, encouraging them to flee venues subsidising liquidity providers in favour of cheaper venues subsidising liquidity consumers.Third, I analyse the impact of the intraday dynamic price limit rule – Limit Up Limit Down (LULD) and High Frequency Trading (HFT) behaviour around price limits on markets with different fee structures. Using difference-in-differences and propensity score matching, I find LULD interferes with trading activity but curbs short-term volatility without delaying the price discovery process. The magnet effect exists and the impact is stronger when approaching the upper price limit. Also, HFT trading activity decreases on market subsidising liquidity providers after the LULD halt which is driven by liquidity taking.