Publications
Tick Size, Lot Size, and Liquidity in Futures Trading (Journal of Financial Markets, Early View)
with Lars Nordén and Caihong Xu
Futures are standardized and heavily regulated contracts, and these features make futures trading possible at liquid secondary markets. However, regulations constrain futures trading at discrete prices and quantities by imposing a minimum tick size and a minimum lot size. We show, theoretically and empirically, that the tick size and the lot size are important for futures trading costs. In our model, we express the futures bid—ask spread, given lot size and tick size restrictions, as a function of futures volatility and trading activity by informed and uninformed traders. Our empirical results support the theoretical model.
Working papers
Exchange Competition, Fragmentation, and Market Quality
with Michał Dzieliński and Björn Hagströmer
Fragmented trading of the same security across multiple venues is prevalent in modern stock markets. This paper investigates how the market quality effects associated with such fragmentation depend on the type of exchange competition. Using a natural experiment where trading goes from fragmented to centralized overnight, we show that market liquidity falls, but surprisingly, the effect is unrelated to the degree of fragmentation. Instead, the liquidity reduction is concentrated among stocks that before the event had the most intense exchange competition on liquidity supply. The results imply that competition in other dimensions, e.g., speed and compliance, does not benefit liquidity.
Market Fragmentation and Trading Aggressiveness
Prior theoretical models predict that when the same asset is traded at multiple exchanges simultaneously, the time priority and the impact of the price weaken across exchanges, motivating traders to trade more aggressively in a fragmented market than in a centralized market. This paper tests this prediction based on public quotes and trades data. Employing a reinforcement learning approach, this paper estimates the optimal level of order splitting to achieve the minimal transaction costs in fragmentation and centralization. The results suggest that a liquidity trader will split the same parent order into child orders of similar sizes under a fragmented market compared with a centralized
market, and the execution takes shorter time under fragmentation. The results imply that the liquidity trader acts more aggressively under fragmentation.
Latency Arbitrage and Market Liquidity
There is an increasing concern that fast arbitrageurs cause market making more expensive by picking off quotes before the liquidity provider was able to revise. This paper shows that adverse selection and liquidity improves after restricting aggressive proprietary trading. Bid-ask spread declines by 11% and adverse selection costs declines by 21%. Liquidity providers earn higher realized spreads and quote higher volumes at better prices. I further identify cross-exchange latency arbitrage and find that the restriction eliminates more than half of toxic arbitrage trades. Market makers benefit from a lower likelihood of being sniped when public information arrives, and are hence subject to lower adverse selection costs. The findings suggest that restricting latency arbitrage improves liquidity by reducing toxic arbitrage.