Free Site Registration

The Data is Already Out There to Design Better Markets

Traders Magazine Online News, February 25, 2019

Phil Mackintosh

Data collection is no justification for increasing investor costs. 

There are plenty who call for the industry to be more “data driven” with market structure reform. But ever since SEC Chair Mary Jo White started to talk about this, what we’ve seen is data collection rather than analysis.

The SECs Tick Pilot is a case in point. Designed to deepen liquidity on the NBBO by widening spreads, the pilot was widely regarded as a costly failure. That’s mostly because artificially widening spreads added costs to investors, estimated at $300 million to $900 million, despite the pilot impacting just 2.5% of market liquidity.

But alas, before the tick pilot data has even finished being collected, we have a new Access Fee Pilot – which is also expected to widen effective spreads and reduce liquidity at the NBBO. What that should prove is that data collection alone doesn’t help design better pilots.

Golden Rule: Do No Harm

U.S. Equity markets are the cheapest, most efficient and also most transparent in the world. Our primary focus should be to protect those advantages—and do no harm.

Thanks to that transparency, we could start by using the data we already have. Did you see that FINRA recently boasted that it collected an all-time high of 135 billion records in just one day?  Even more impressive, FINRA recently used that data to compute opportunity cost, something we recently said was lacking from the Access Fee Pilot proposal.

This is important as it confirms we do already have the data required to identify routing conflicts and quantify opportunity costs. Sadly the final Access Fee Pilot ruling did not quantify any benefits—although it does quantify costs—if you’re interested, start at page 222.

What the Market Tells us about Tick Sizes

Even simple data can tell us a lot about trading economics.

Let’s start by plotting all corporate stocks across pricespreaddepth and market cap (Chart 1) and see what we can learn.

Chart 1: Spreads naturally widen as market cap and liquidity fall

Artificially wide spreads increase queue size

The upward sloping diagonal line shows how the 1-penny tick forces spreads wider as price falls. Note that the largest circles (depth of the NBBO or queue length) generally sit on this line. This shows that once spreads are capped, the trade-off between the cost of crossing a spread and the cost of waiting swings in favor of waiting—leading to longer queues and time to fill. These spreads are “artificially expensive” or “tick constrained.”

The second diagonal line supports this conclusion. Those are stocks in the tick pilot (the pilot ended on Sept 28, 2018), where a 5-cent spread was “artificially expensive,” incentivizing rational traders to capture spreads regardless of the longer queues that created.  Note that these are the same stocks that cost investors between $300 million and $900 million above.

Artificially high prices have worse spreads, too

For more information on related topics, visit the following channels:

Comments (0)

Add Your Comments:

You must be registered to post a comment.

Not Registered? Click here to register.

Already registered? Log in here.

Please note you must now log in with your email address and password.