The US tick-size harmonization and optimization problem

The US equity markets have evolved considerably in the past 20 years, since decimalization and the ‘one size fits all’ tick-size regime. The existing US tick-size rules were designed to benefit dark market participants[1] over exchanges, but few would have guessed that off-exchange market share would be within eyeshot of beating the exchanges. We propose leveling competition between exchanges and dark participants with a revised Regulation National Market System (Reg NMS) Rule 612 designed to harmonize tick-size rules by banning sub-tick trading. And we propose a five-level (0.1 cent, 0.5 cent, 1 cent, 5 cents and 10 cents) tick-size regime, driven by four key variables: share price, quoted spread, effective spread and realized spread. We believe that this proposed tick-size rule update will not only promote competition in US equity markets but will also help ease concerns around other related issues.

Context: tick-size scrutiny

On 6 May 2021, the US Securities and Exchange Commission (SEC) chairman, Gary Gensler, in testimony before the ‘Game Stopped’ hearings by the US House Committee on Financial Services, provided early insights into the SEC’s plans to scrutinize a wide range of issues. The SEC’s aggressive agenda, termed the ‘everything crackdown,’ is the most ambitious in the organization’s history.[2]

Earlier this month, Chairman Gensler reiterated comments that the commission is considering ‘shrinking tick sizes’ and ‘leveling competition between trading venues and wholesalers.’[3] We believe that an updated Rule 612 will go a long way toward achieving the SEC’s fair competition goals.

Tick-size rule harmonization

Dark participants have a natural advantage over exchanges by providing order anonymity, reduced order information leakage and potentially lower market impact costs. In 2005, Reg NMS Rule 612, governing tick-size, provided another advantage to dark participants, empowering them to jump in front of the exchanges with economically insignificant $0.0001 (1/100^th of a penny) price increments. As a result, a significant amount of off-exchange trading volume is geared toward exploiting this benefit by simply pegging a $0.0001 ‘price improvement’ to the national best bid and offer (NBBO). This systematic ‘front jumping’ of the exchanges is called ‘sub-pennying,’ or ‘fractional ticking.’

Dark participants have grown in sophistication over the years since Reg NMS, and regularly provide significant value to investors in terms of innovation, technology and reduced market-impact costs. We believe that dark participants will continue to provide value to clients without the need to rely on the fractional ticking ‘race to the bottom’ benefit provided to them under Rule 612.

Although this proposed update to Rule 612 may lead to lower off-exchange market share, resulting from the loss of fractional tick ‘front-jumping’ benefits, we are not advocating lower off-exchange trading volume, per se, and we are not proposing a ‘trade-at’ rule or forcing dark orders into the light.

We believe that there is value in price discovery processes, but there is also value in minimizing order-information leakage. We see the on/off exchange trading market share balance to be a function of innovation, technology and investor preferences. We also believe that there is room in US equity markets for innovative pricing structures, such as ‘maker-taker,’ ‘inverted’ or ‘payment for order flow’ (PFOF), as long as sufficient disclosures are made and clients have a choice between ‘all-in’ and ‘cost-plus’ pass-through pricing options. But there needs to be fair competition across all market participants, and certainly dark participants should not receive preferential treatment over the exchanges.

Tick-size optimization

The Reg NMS Rule 612 ‘one size fits all’ tick-size approach is designed for simplicity, and there’s value in simplicity. But market participants are increasingly calling for more tick-size levels, as many low-priced stocks are tick-constrained while high-priced stocks typically trade multiple ticks wide.

Most exchanges globally allocate stocks to tick-size levels based on share price. In 2019, Nasdaq proposed categorizing stocks into tick-size buckets based on duration-weighted average quoted spread.[4]  We think that both these variables have predictive ability. However, we believe that tick-size optimization could be further enhanced with the addition of two more variables: effective spread and realized spread.[5]  Effective spread captures trades inside the quote spread and realized spread captures inventory effects.

We propose an ordered choice probit model that classifies stocks into tick-size level (0.1 cent, 0.5 cent, 1 cent, 5 cents and 10 cents) as the dependent variable, to the vector of independent variables that include share price, quoted spread, effective spread and realized spread. In our model, y^*_it is the unobservable latent variable that is positively correlated with the observed categorical variable, y_it, on a 5-point scale from 1 to 5, where 1 represents the lowest tick-size level and 5 represents the highest tick-size level, and gamma, g, is the threshold value. The observed y_i is related to the unobserved y*_i as follows:

                y_i   =   1  (0.1 cent tick)      if   g₁ < y*_i  < g₂

                          2  (0.5 cent tick)      if   g₂  < y*_i  < g₃

                          3  (1 cent tick)         if   g₃  < y*_i  < g₄

                          4  (5 cent tick)         if   g₄  < y*_i  < g₅

                          5  (10 cent tick)       if   g₅  < y*_i  < infinity

We think that the primary listing exchange would be well-suited to be in charge of tick-size calculations and that tick-size adjustments should be done on a periodic basis, perhaps semi-annually, rather than dynamically, to give programmers time to adjust their algorithms when changes are made.

We hope that this proposal leads to further debate on the subject, and we believe that development work toward harmonizing tick-size rules and tick-size level optimization will go a long way toward aligning stocks and exchange-traded funds (ETFs) with tick-size levels that optimally balance quote and trade spread characteristics with liquidity trade-offs.