the latest

Market Insight From Reg NMS

Tuesday, January 28, 2014

Jacob Bunge, has an interesting article in the Wall Street Journal that questions the costs and benefits of Regulation National Market System (Reg NMS): “A Suspect Emerges in Stock-Trade Hiccups: Regulation NMS”.  It is good to see some attention given to the complexity that regulation can introduce to the market. Based on the firestorm of conversation the article touched off on social media this morning, it is clear that there is a hunger for articles that examine our market structure.

The article is recommended reading for those interested in the conversation around simplifying our current market structure and improving our markets.  The information it presents and assumptions it makes allow us to revisit some commonly held misconceptions about electronic markets and high frequency trading (HFT).

First, an inaccuracy in the article. Mr. Bunge writes:

“Computerized firms called high-frequency traders try to pick up clues about what the big players are doing through techniques such as repeatedly placing and instantly canceling thousands of stock orders to detect demand. If such a firm’s algorithm detects that a mutual fund is loading up on a certain stock, the firm’s computers may decide the stock is worth more and can rush to buy it first.”

One is hard pressed to figure out how “placing and instantly canceling” orders garners any information about someone else’s intent. The only concrete information you gain is that your own orders were not filled.  Even if one accepted this vague claim at face value, to associate all high frequency trading with this activity is a gross mischaracterization that ignores the diversity of strategies and many benefits HFT brings to the markets.

The article gives us some interesting insight into cancel rates.  Mr. Brooks learned that one of his brokers “offered to buy 750 million shares of the stock while actually purchasing just 2.5 million”.  This means that 99.67% of the liquidity the broker sent to the market was cancelled.  Many are quick to point the finger at HFT for high messaging resulting from high cancel rates.  The information about Mr. Brook’s broker is a concrete example of high cancel rates being caused by participants other than HFT.

Another frequent complaint about today’s market is “phantom liquidity”: placing orders one does not intend to be filled in order to give the impression of fake interest in a security and price.  Here we have a broker, not a high frequency trading firm, doing just that: “the broker placed and canceled many smaller orders all across the stock market, creating a dense smoke screen of phantom interest in the security.”

Mr. Brooks is “bothered by the distortions to the stock market such practices may create.”  Many people who are bothered by these distortions are quick to blame HFT.  The information provided in this article, though, gives us other causes to consider.

We can also learn from Mr. Cronin’s order. The article contends that by sending an order to 10 separate exchanges and dark pools there was information leakage:

“A 1,000-share stock order, which is small for Invesco, traversed 10 separate exchanges and dark pools before it was filled. The order had also been sent to eight other venues where ultimately no shares were bought—but where other traders may have had a chance to catch wind of Invesco’s strategy.”

While we do not know exactly how the 1,000-share order was routed or executed, it is a good introduction to some basic points on possible information leakage:

• if the order is resting in a dark pool and there is no resulting fill, there is no market data that shows the order so information is not leaked through market data;

• if the orders are sent to exchanges as Immediate or Cancels (IOCs) or non-displayed orders and there is no resulting fill, then other market participants know nothing about the order. Thus, no information leakage.  (For the uninitiated, an IOC instructs the exchange to fill as much of the order as it can immediately and to cancel the remainder, with no resulting orders resting as passive liquidity on the book).

Mr. Bunge’s article concludes with two different points, both accurate and important.  Commissioner Piwowar says, “As a general matter, retrospective analyses of existing rules is just good government”. Mr. Campos — though he thinks things could be better —  says, “by and large, by a long way, the system works today”.

Mr. Campos is right, the system both works today and could be better. And Mr. Piwowar is right, we should apply retrospective analysis to the existing rules.  By embracing and acting on both of these thoughts, we will end up with a better market.

 

- MMI