Albert J. Menkveld   My Two Cents   
About  CV  Published Papers  Working Papers  Media  Software  Advice  Contact 

HFT Good or Bad? A Literature Survey

Today the Dutch regulator AFM published a study on high-frequency trading (HFT). They characterize it as a data driven assessment of HFT trading strategies.

In parallel, I released a survey of the academic literature on HFT. AFM kindly let me use some of their graphs to illustrate it. I read 100+ manuscripts to identify the economic arguments for and against HFT. Does it benefit or harm market quality?

High-frequency traders run automated trading strategies that stand out by their dazzling speed. They consider non-zero positions costly and should therefore be considered intermediaries.

To navigate the waves of economic insights, I decided to group the theoretical studies on these new intermediaries according to seven themes:

  1. Speed dispersion in limit-order markets.
  2. Speed used to prey on large orders.
  3. Speed in a run game after (public) news.
  4. Speed to connect fragmented markets.
  5. Speed as a source of (endogenous) quote flickering.
  6. Speed to create productive intermediation chains.
  7. Speed to satisfy investors' appetite for faster trade completion.

For all themes I discuss the relevant studies, the economic channels they identify, and therefore the ways in which HFT affects market quality.

Useful perhaps, but where does this all lead us in terms of the larger question? Is HFT good or bad?

Somehow one should weigh all the pros and cons, add them up, and come to a reasoned, balanced, and informed conclusion.

I decided to at this point bring in the empirical studies on HFT. An argument gets more weight if either its assumptions or its predictions find support in the data.

You might disagree with where I come down on the value of HFT (and I do come down). One benefit of structuring the manuscript the way I did is that you can apply your own weights and come to a different verdict.

Anyway, I hope you will enjoy exercising your mind through this often complicated material. I hope you come out somewhat less confused.

P.S.: The HFT survey article is here.

Click [here] to see list of all posts.