May 2, 2018
My recent paper titled “High-frequency Trading and Extreme Price Movements” examines liquidity provision by modern market makers (MMs for short) during episodes of market stress, when prices jump significantly. Liquidity is a phenomenon that describes the ability of market participants to exchange cash and securities in a seamless and efficient manner. Liquid markets are good; illiquid – bad. Not only does illiquidity hurt the process of asset exchange, it hurts individual investors, pension funds, and even the companies, whose shares are being traded. In the past, market makers had an obligation to provide liquidity when it was needed. Modern markets have eliminated this obligation, and market observers worry that liquidity supply is the weakest when liquidity is most needed.
My research shows that modern MMs are pretty good at supporting the market in moderately stressful times, when one stock undergoes an extreme price movement, but when multiple stocks experience price swings MMs retreat. My paper is the first to document this phenomenon, but it is unable to fully explain it. Several papers scheduled to be presented at the Bank of Canada-Laurier Market Structure Conference on May 7, take the next step in trying to understand how liquidity in modern markets is supplied and under what conditions this supply is stable. They link liquidity supply to technology, trader type, order flow toxicity, and the granularity of the pricing grid. One particularly noteworthy finding suggests that although the more granular grid benefits liquidity overall, it is detrimental in periods of stress.
The most surprising result in my paper was the immediate switch from liquidity provision to non-provision when it came to price jumps in one stock v. multiple stocks. For jumps in one stock, the modern market structure seems to work well, but as soon as we move to two simultaneous jumps, the MMs withdraw. Moreover, as soon as we move to three jumps, they begin to consume liquidity instead of providing it.
My paper and the papers presented at the Bank of Canada-Laurier conference tell us that there is a lot to learn about modern liquidity provision. In a new project, I am looking at the behaviour of MMs under the proverbial microscope, digging into the most detailed data about their activity and holdings. I am meticulously decomposing every MM’s account information, trying to understand what drives their decisions to participate in liquidity provision or withdraw. This is a sizeable data exercise; modern markets produce multiple terabytes of data. This is probably the most data-intensive study of my career; it will take quite some time to finish, but the issue is too important not to look into it, as it affects every aspect of the market and the investments of individual investors, pension funds, and Canadian and international companies.
We see you are accessing our website on IE8. We recommend you view in Chrome, Safari, Firefox or IE9+ instead.×