The Gateway to Algorithmic and Automated Trading

How to align broker and customer interests and make exchanges more competitive

Published in Automated Trader Magazine Issue 44 Q1 2018

US regulators are focused on conflicts of interest caused by maker-taker rebates. Is there a way to align broker and customer interests and still incentivise liquidity provision?

Authors Bio

Michael Friedman

Michael Friedman is General Counsel and Chief Compliance Officer at Trillium, a proprietary trading and technology firm based in New York. Michael is also the lead designer of Trillium's trade surveillance platform Surveyor. Prior to joining Trillium, he was a securities litigation partner at a large multinational law firm.

The Securities and Exchange Commission's (SEC) new Chairman Jay Clayton recently announced that the agency will move forward with a pilot program testing new methods of how equity exchanges charge brokers for each trade.

The issue is that many exchanges have a pricing model for the access fees they charge brokers for each trade whereby the broker who got there first (the 'maker', because he 'makes' liquidity on the exchange) receives a subsidy or rebate, and the broker who gets there second and trades as counterparty to the maker (the 'taker', because he 'takes' liquidity away from the exchange) pays a fee. The taker's fee is always slightly larger than the maker's rebate (usually 0.0030 USD fee / 0.0029 USD rebate per share), and the exchange pockets the difference.

Figure 01: Prices before and after collection of access fees

Figure 01: Prices before and after collection of access fees

When displayed quotes are priced net of access fees, the narrowest bid-ask spread occurs where the venue collects the smallest fee.

The purpose of the maker-taker model is to encourage market makers to display more quotes, which aids in price discovery and reduces transaction costs by narrowing spreads. The concept was first introduced by then-newly-founded Island ECN in 1997 as a way to lure order flow away from incumbent venues. It has since been copied by most major exchange groups in markets where venues must compete for order flow.

Critics charge that the maker-taker pricing model may encourage brokers to send orders to the venue that is most likely to pay the broker a rebate rather than to the venue that is most likely to execute the trade at the best price for the broker's customer.

In response to these criticisms, the SEC's Equity Market Structure Advisory Committee recommended a pilot study (SEC, 2016) that would experiment with changing the sizes of the rebates and fees, but would leave intact the basic maker-taker model and the 0.0001 USD difference between fee and rebate. The theory behind the pilot is that if rebates were smaller, brokers would give greater weight to factors other than rebates (factors where the customer's and broker's interests are more aligned) in deciding where to route orders.

Nasdaq experimented (Hatheway, 2015) with a unilateral reduction in the size of its rebates and fees for 14 symbols over a four-month period in 2015. With a smaller incentive to post quotes the experiment predictably resulted in Nasdaq receiving fewer posted quotes, and Nasdaq losing market share to other venues with higher rebates.

IEX (IEX, 2017), and now also one of CBOE's venues (Bats, 2017), disclaim the maker-taker model entirely and instead charge both maker and taker the same flat fee. Essentially, both IEX and CBOE have very small market share relative to other venues who do offer rebates.

None of these experiments remove the maker-taker conflict entirely while preserving the liquidity incentive of rebates. There is a different approach which does both, and it was first formulated by a panel of professors over 13 years ago (Lehmann and Hasbrouck, 2004). The solution is to include the access fee owed by the taker in each quoted bid or offer. Figure 01 illustrates what this would look like.

The dotted lines represent the price the maker will pay if he buys (or will receive, if he sells). The solid lines represent the price the taker will pay if he buys (or will receive, if he sells) after the exchange collects its access fees. The quote displayed to the market is the solid line, not the dotted line. In all cases, the spread is wider after the fee is collected (solid) than before (dotted). Note that the narrowest bid-ask spread occurs where the venue collects the smallest fee.

From the maker's perspective, a market maker willing to buy at 20.33 USD and to sell at 20.35 USD will appear higher in the order book, and therefore get filled sooner, by routing to an exchange that only inflates that spread by 0.0002 USD than by routing to an exchange that inflates that spread by 0.0003 USD.

From the exchange's perspective, the exchange with the lowest access fee attracts the most quotes. This creates an obvious incentive - that does not currently exist - for exchanges to compete on price, which happens to be the stated purpose of the Securities Acts Amendments of 1975 (see Further reading), the statute under which Reg NMS was promulgated.

From the taker's perspective, the venue with the smallest fee is also the venue with the best price for the customer. Voilà, no more conflict between broker and customer.

At the time that quoting net of fees was first proposed in the comment period for Reg NMS in 2004, it was rejected partially because it would require hundredth-of-a-penny tick sizes to show access fees measured in mils (0.0001 USD per share) in quotes. Some observers questioned the technical challenges of displaying an extra two decimal places, but those challenges have since been overcome for stocks priced under a dollar, where hundredth-of-a-penny tick sizes are permitted. As Island ECN co-founder Josh Levine once observed (Levine, 2001), tick sizes cannot be smaller than a hundredth of a penny because "four decimal places is the smallest increment that results in a change in the net settlement amount of a single round lot (100 shares)", so there is no danger of a race to infinitesimal ticks if the current whole-penny floor is removed.

Hundredth-of-a-penny tick sizes would have another great benefit to market structure. Many of the most unsavoury aspects of modern trading arise from the race among market makers to have their orders executed first. The high-cost arms race to reduce latency with colocation, FPGA chips and microwave towers, as well as the proliferation of complex exchange order types, came about in order to help market makers claw their way to the top of an order priority queue that is often quite deep in liquid stocks with one-penny spreads. Hundredth-of-a-penny tick sizes would allow market makers to compete on price to get to the top of the queue, greatly reducing the relative advantages of being fastest.

It's a simple change that could fix a lot.

Futher reading