Subsidizing Liquidity with Wider Ticks: Evidence from the Tick Size Pilot Study

• Published date: 01 June 2020
• Date: 01 June 2020
• Author: Justin McCrary,Robert P. Bartlett
• DOI: http://doi.org/10.1111/jels.12252

Journal of Empirical Legal Studies

Volume 17, Issue 2, 262–316, June 2020

Subsidizing Liquidity with Wider Ticks:

Evidence from the Tick Size Pilot Study

Robert P. Bartlett III and Justin McCrary*

Using data from the 2016–2018 Tick Size Pilot Study, we examine the efficacy of using

wider tick sizes to subsidize market-making in small capitalization stocks. We demonstrate

that realized spreads decay quickly within the initial microseconds of a trade. The effect

reduces the subsidy offered by wider tick sizes, particularly for non-HFT market makers.

The profit subsidy from wider tick sizes is also compromised by a significant shift in trad-

ing to “taker/maker” exchanges and to midpoint trading in non-exchange venues.

The pilot’s exception for midpoint trades also accounts for the fact that nearly a third of

trading remains in non-exchange venues despite the inclusion of a trade-at rule. Overall,

these findings point to considerable inefficiencies in the pilot study’s goal of using wider

tick sizes to subsidize liquidity provision in small capitalization stocks.

I. Introduction

In October 2016, the Securities and Exchange Commission (SEC) commenced a two-year

pilot study that widens the minimum quoting increment—or “tick size”—from $0.01 to

$0.05 for nearly 1,200 small capitalization stocks. Crafted in the wake of a steady decline

in the number of listed companies on U.S. exchanges, the pilot study aimed to examine

whether widening the tick size can enhance the liquidity and trading of small capitaliza-

tion stocks.

1

A central premise behind the Tick Size Pilot Study was that “a widened tick incre-

ment could increase market maker profits and that the increased profits could foster a

more robust secondary market for small capitalization stocks (and ultimately a more

robust primary market) by, for example, increasing liquidity, enhancing the attractiveness

*Address correspondence to Justin McCrary, 521 Jerome Greene Hall, Columbia University, New York, NY 10027;

email: jrm54@columbia.edu. Bartlett is at University of California, Berkeley; McCrary is at Columbia University

and NBER.

Statement of Financial Disclosure and Conflict of Interest: This research did not receive any specific grant from

funding agencies in the public, commercial, or not-for-profit sectors.

1

The pilot study itself was a product of a mandate contained in the 2012 Jumpstart Our Businesses and Startups

Act (JOBS Act) for the SEC to study the effects of the decimalization of stock prices in 2001 on IPOs and small

and middle capitalization companies (Nallengara & Ramsay 2013).

262

of acting as a market maker, and possibly increasing the provision of sell-side research”

(SEC 2015). In this regard, the pilot study was consistent with the theoretical literature

on tick sizes that evaluates the regulatory choice of tick size as involving a tradeoff

between minimizing transaction costs for investors and subsidizing liquidity providers to

make a market in a security (Angel 1997). According to this theory, a wider tick size

(such as existed prior to decimalization in 2001) should encourage dealers to make a

market in a security because the market maker “pockets the spread” in a stationary mar-

ket: the tick represents the minimum round-trip profit to a dealer who can buy at a lower

bid price and sell at a higher offer price. Yet by the same reasoning, larger tick sizes

increase transaction costs for liquidity takers who buy at the offer and sell at the bid. The

tick size can thus be thought of as taxing liquidity takers in order to subsidize liquidity

provision.

The notion that the market maker pockets the spread is a theoretical proposition,

as it posits a stationary market. Empirically, evidence from the decimalization of stock

prices in 2001 highlights how institutional features of the trading market can interfere

with this straightforward relationship between tick sizes and market maker profits. For

instance, while average quoted and effective spreads declined following decimalization

(see, e.g., Chakravarty et al. 2001; Bacidore et al. 2003), the effect was minimal for small

capitalization stocks, which had large spreads prior to decimalization (Bessembinder

2003). As another example, the smaller spreads of decimalization permitted dealers to

offset any drop in profits through participating in more trades (Ronen & Weaver 2001;

Coughenour & Harris 2004). These second-order effects complicate understanding how

tick sizes relate to liquidity providers’ incentives.

In this article, we use new data from the Tick Size Pilot Study to examine the cen-

tral premise in the theoretical literature on tick sizes and market making, namely: To

what extent does widening spreads increase market maker profitability? We focus on this

first-order effect of wider tick sizes for the simple reason that market maker profitability

is the primary channel through which wider ticks are believed to improve the liquidity of

small capitalization stocks. Evidence that wider ticks result in greater market maker profit-

ability is accordingly a necessary condition for wider ticks to achieve their stated policy

objective. Conversely, evidence that market makers capture only a fraction of the

enhanced transaction costs created by wider ticks would call into question the efficiency

of using wider ticks to subsidize liquidity in small capitalization stocks.

We begin with an empirical finding. As noted by Rindi and Werner (2019), daily

market maker profits reported to the SEC by FINRA increased by approximately 40 per-

cent in transactions involving securities that were treated with nickel tick sizes during the

pilot study.

2

This is a notable increase in profitability, but given that the tick size for these

securities was quintupled, the increase in market maker profits is in many ways more

modest than what proponents of the Tick Size Pilot Study might have envisioned.

2

The pilot study required broker-dealers to report daily market maker profitability to FINRA. The data, aggregated

across stocks by day, is available at http://tsp.finra.org/finra_org/ticksizepilot/tsp_index.html.

Subsidizing Liquidity with Wider Ticks 263

We posit that there are at least three reasons to question the efficacy of wider ticks

in enhancing market maker profitability in contemporary equity markets. The first con-

cerns the risk of informed trading in a context where liquidity provision is increasingly

performed by high-frequency trading (HFT) firms. Consider, for instance, the framework

of Aı

¨t-Sahalia and Saglam (2017), in which a strategic high-frequency trader receives a

signal about future order flows and exploits its speed advantage to optimize its quoting

policy. For instance, if the bid-ask in a penny quoting environment is $10.02 ×$10.08, a

market maker receiving an incoming buy order might fill the order at $10.08 by selling

short with an expectation of covering at the bid of $10.02, earning 6 cents per share. In

an Aı

¨t-Sahalia and Saglam (2017) framework, however, HFT liquidity providers observing

the transaction may update their own buy orders rapidly, moving the best bid above

$10.02 before the market maker can cover. Such a result will force the initial market

maker to quote at or better than this new price, reducing its expected profits.

The ability of HFT liquidity providers to adjust quotes in this fashion can jeopar-

dize the market maker subsidy ostensibly created by wider spreads. For instance, if the

original bid-ask spread were $10.00 ×$10.10 with nickel tick sizes, our market maker’s

ability to capture the 10-cent spread will be compromised by the incentives of HFT liquid-

ity providers to update bids to $10.05 upon observing buying interest. Price-time priority

rules observed by exchanges further complicate our market maker’s ability to profit from

the wider spreads: as HFT liquidity providers react to the buying interest, the wider ticks

should create longer queue lines at the best bid of $10.05, enhancing the challenge of

covering the short position (Yao & Ye 2014).

These long queue lines point to a second reason to question the efficacy of wider

ticks in enhancing market maker profitability. Using a regression discontinuity design,

Bartlett and McCrary (2012) and Kwon et al. (2015) find that, relative to the $0.0001 mini-

mum price variation (MPV) for stocks that trade below $1.00 per share, the penny MPV

for orders priced above $1.00 per share has the result of creating long queue lines at the

national best bid or offer (NBBO). Because the MPV regulates quotes but not trades,

however, traders seeking to avoid these long queue lines can trade in non-exchange

venues at sub-penny prices, generally by means of pegged midpoint orders (Bartlett &

McCrary 2019a). In the prior example, for instance, our hypothetical liquidity provider—

having gone short at $10.10—may seek to cover its position by submitting a midpoint buy

order to a dark pool rather than wait in line on a displayed venue at the best bid

of $10.05.

To the extent wider tick sizes result in longer queue lines, the migration of order

flow away from exchanges toward midpoint orders in dark venues should accordingly

place a further limitation on the effectiveness of wider spreads to subsidize liquidity pro-

viders, particularly those who display orders on exchanges. Indeed, concerns that wider

tick sizes might drive trading to non-exchange venues induced the SEC to divide the pilot

into three separate treatment groups. In group 1 (TG1), quotes were required to be

priced in nickels, but trade prices were left unconstrained (similar to the current penny

MPV rule.) In group 2 (TG2), quotes and trades were required to be priced in nickels.

Finally, in group 3 (TG3), quotes and trades were required to be priced in nickels, and

trading venues were also subject to a “trade-at” rule. The trade-at rule generally

264 Bartlett and McCrary