EDGA Equities Exchange migrates to Maker-Taker Model: the impact on market microstructure

First published by TabbFORUM

By Tom Jardine, Client Facing Data Scientist, BMLL

From November 1, 2024, CBOE’s EDGA® Equities Exchange will shift from an "inverted" marketplace to a "maker-taker" (m-t) model. In the inverted model, price takers receive a rebate and pay to post passive liquidity, in contrast to the more traditional maker-taker model, where liquidity providers receive rebates and pay to take liquidity. For any firm making routing decisions, understanding the implications of this change will be critical in ensuring optimal execution in the US markets.

Why is this change being implemented?

Wait times can be quite lengthy for a maker-taker as participants line up to provide liquidity (and receive their rebate). An inverted market, on the other hand, typically has a shorter wait time, but participants have to pay to “cut the line”. CBOE has stated it is changing the market maker model to better serve the market by providing another maker-taker venue.

What is the expected market microstructure impact?

A quick glance at some of these metrics for the first three quarters of 2024 for the top 500 names in the US will provide an indication of the expected changes at EDGA going from inverted model to maker-taker.

Let’s start with a basic spread. While inverted markets have wider spreads in general, mainly due to the fact that it (quite literally) doesn’t pay to rest in these markets, EDGA has the tightest spread among inverted markets and is comparable to the larger maker-taker spreads (see Figure 1).

From November 1, we can expect EDGA to continue to have tight spreads.

Next, we look at “Sweep-to-fill” (STF). Using the BMLL Sweep-to-fill metric, or the spread one would encounter for a certain notional value sent to the exchange, we can see inverted markets have less liquidity around the BBO and as a result, higher SFT spreads (see Figure 2).

Result: Expect EDGA as an m-t exchange to tighten the depth spread.

Next, we look at probability of fill (the probability that a bid order placed on the 1st level of the book will be filled in 60 seconds). While the larger m-t exchanges dominate this metric simply because of their large volume, NYSE National (XCIS), an inverted exchange with relatively low market share, has a high probability of fill (see Figure 3). This intuitively makes sense if a participant rests in an inverted market (basically offering a rebate to a taker), the probability of fill, all else being equal, will be higher. This will be especially true in tick constrained stocks.

Result: As an inverted exchange, EDGA is still middle of the road for probability of fill. By converting to m-t and reversing the incentive again, all else being equal, we can expect a decline in the probability of getting filled after conversion. However, this would be offset by a tighter spread, as discussed above.

Finally, we take a look at time at CBBO or the percentage of time an exchange spends at the best price (including odd lots, a BMLL metric). As inverted markets are more of a “line cutting” mechanism, it does not encourage participants to spend most of their time providing liquidity, so time at CBBO for inverted markets is on the lower end of the spectrum (see Figure 4).

Result: A switch to an m-t exchange should enable EDGA to join the ranks of the other m-t venues and spend more time with resting liquidity at the best prices.

Bottom Line

These are all powerful and interesting metrics but why should a market participant care? Well, in a couple of days, EDGA is going to completely reverse its rebate structure after years of inversion and the above illustrations highlight the differing behaviors such a change can trigger.

If your SOR, algo engine or trading strategy is tuned to an inverted market and expecting a certain behavior from EDGA, you need to take into account these expected changes and factor those into your engines.

Using Level 3 historical market data, participants can better understand when and how to trade, make better trading decisions, improve algorithmic strategies and achieve best execution.