The Equation Series
Menu

Part II: The Markets and Players

Chapter 7

Market Makers and the Speed Business

On 10 March 2014 a New York trading firm that almost nobody outside the industry had heard of filed to sell shares to the public, and one sentence of its prospectus became the most quoted statistic in the history of high-frequency trading. Virtu Financial, founded in 2008 by Vincent Viola, a former chairman of the New York Mercantile Exchange, with Douglas Cifu and Graham Free, made markets in more than 10,000 securities and other instruments on more than 210 venues in 30 countries, with 151 employees at the end of 2013. From January 2009 to December 2013, the filing said, it had had one losing trading day in 1,238. Its average daily adjusted net trading income in 2013 was about $1.65 million (S-1). The timing was poor. As Michael Lewis's Flash Boys turned high-frequency trading into a public scandal, Virtu postponed the offering in April. It listed on Nasdaq a year later, on 16 April 2015, at $19 a share; the final prospectus updated the statistic to one losing day in 1,485 (424B4). Virtu then became the only large US electronic market maker whose results are audited and public, which is why this book uses it as the evidence for a business whose largest firms disclose almost nothing (companion's profile of Virtu Financial).

Two years before the filing, the firm that Virtu would eventually buy had shown the other side of the same business. On the morning of 1 August 2012, Knight Capital, a Jersey City market maker founded in 1995 that had built the largest US business in filling retail stock orders, deployed new code to the router that sent its orders to the exchanges. The release reached seven of the router's eight servers; on the eighth, a reused flag reactivated a retired function called Power Peg, which had lost the check that stopped it once an order was filled. Trying to fill 212 small retail orders, the router sent child orders continuously and obtained over 4 million executions in 154 stocks, for more than 397 million shares. Before the open an internal system had sent 97 automated emails referring to the fault; nobody acted on them. Searching for the cause in live trading, staff removed the new code from the seven correct servers, which spread the fault to all eight. By the time the system was stopped, about 45 minutes after the open, Knight held about $3.5 billion of unwanted long positions and about $3.15 billion of unwanted shorts. The loss was more than $460 million (SEC order, 16 October 2013); Knight's own early estimate was about $440 million pre-tax. Knight had made about $115 million of net income in all of 2011 (company figures, as reported).

Knight survived only because others rescued it. On 5–6 August a group led by Jefferies, with Blackstone, Getco, Stephens, Stifel and TD Ameritrade, put in $400 million, which gave the investors roughly 73% of the firm on conversion (reported). Within a year Knight had merged with Getco, a high-frequency trader, to form KCG Holdings (1 July 2013), and Virtu bought KCG in 2017 for about $1.4 billion. The SEC's order of 16 October 2013, its first case under the market access rule, Rule 15c3-5, which requires pre-trade risk controls, fined Knight $12 million (companion's profile of Knight Capital).

The preview ends here. The full chapter is in the book.

In this chapter

  1. 7.1Renting out immediacy
  2. 7.2Why so few firms
  3. 7.3Payment for order flow
  4. 7.4What speed buys
  5. 7.5A worked example: how a tiny edge becomes almost no losing days
  6. 7.6Options market makers and the flows they hedge
  7. 7.7When the machine breaks
  8. 7.8How an expert reads a market maker