Stock exchange rebates: Do they help or harm market quality?

When NASDAQ slashed trading fees across fourteen stocks during a 2015 fee pilot experiment, it revealed insights about how exchange pricing shapes market behaviour

When NASDAQ temporarily largely removed its subsidy to liquidity provision paid for by a levy on market orders, it created a unique opportunity to examine how exchange pricing affects market behaviour. The four-month experiment involved 14 highly liquid stocks – seven listed on the NASDAQ and seven on the NYSE – with total trades reaching US$1.9 trillion, making it one of the largest controlled market studies ever conducted.

The experiment focused on stocks with significant trading volume and high off-exchange activity. These stocks were chosen by NASDAQ to improve the quality of the experiment from its perspective. To do so, stocks had to be very liquid with a high volume of trades, especially in off-exchange and dark pools, as NASDAQ was primarily concerned about whether fees and rebates discouraged its lit market in which the limit order book is displayed, according to the authors of new research which found that stock exchanges that pay traders to provide liquidity create deeper but slower markets, even when regulators require trades to occur at displayed prices.

The scale of this NASDAQ experiment provided insights into the effects of exchange fees across the entire U.S. trading landscape. This is by far the biggest market for equity securities in the world. During this period, the net fee remained constant at one-hundredth of a cent per trade ($0.0001) before, during, and after the experiment. This constancy in net fees helped isolate the impact of the fee structure changes, according to the research article, Does Maker-Taker Limit Order Subsidy Improve Market Outcomes? Quasi-Natural Experimental Evidence, which was co-authored by Peter Swan, Professor in the School of Banking and Finance at UNSW Business School together with Dr Yiping Lin, FinTech Program Coordinator at UNSW and Co-founder at Alt Data Tech and Frederick Harris, Professor of Economics and Finance at Wake Forest University.

While Yiping Lin was engaged in writing his UNSW PhD thesis under the supervision of Prof. Swan, he was invited by NASDAQ and its Chief Economist, Frank Hatheway, to assist NASDAQ in analysing the world’s largest-ever experiment, known as the Fee Pilot, to be carried out by an individual company. The research he undertook on behalf of NASDAQ while at NASDAQ’s headquarters in New York formed the basis of an important component of his PhD thesis and the article published by the A* Journal of Banking and Finance.

The researchers were fortunate enough to have all of NASDAQ’s unique resources at their disposal. They utilised proprietary NASDAQ trader-level data that formed part of the FINRA/Nasdaq Trade Reporting Facility system. This data is far superior to what is known as US TAQ data, which is ordinarily the only data available to academic researchers. The data is based on order submission and trades, including price, volume, and a unique identifier for the trader submitting the order. The analysis employed a difference-in-differences approach, comparing the 14 stocks in the experiment against a carefully matched control group of stocks with similar characteristics. This methodology helped isolate the specific effects of the fee changes from other market factors that might have influenced trading patterns during the study period.

Understanding exchange trading fees

Stock exchanges traditionally operated under a “maker-taker” model where they charged fees to traders who “took” liquidity (executing existing market orders) while paying rebates to those who “made” liquidity (placing new limit orders). As noted in the research, “In the late 1990s Electronic Communication Exchanges and in the early 2000s stock exchanges in the United States introduced a subsidy to one side of the market (namely, non-marketable limit order buyers or sellers called makers of liquidity) paid for by a revenue-neutral tax (fee) on marketable orders (called takers of liquidity) to encourage liquidity provision.”

Read more: Navigating the noise: how ambient sound affects trading performance

This fee structure influenced market behaviour in significant ways. The exchanges that provided the highest rebates for placing orders attracted more liquidity providers, creating deeper markets but also longer queues. Think of it like throwing bait into the water – it attracts both desirable fish and unwanted predators. Similarly, high rebates drew in both beneficial liquidity providers and sophisticated traders trying to profit from the fee structure, fundamentally changing how markets operated.

The NASDAQ market structure experiment

The study examined how NASDAQ’s temporary fee reduction affected trading patterns and market quality. On 2 February 2015, NASDAQ implemented a maker-taker fee pilot for 14 traded stocks on the NASDAQ where the take fee was lowered from 30 cents per 100 shares to 5 cents for orders that removed displayed liquidity; the make rebate for adding displayed liquidity was simultaneously lowered to 4 CPS from an indicative 29 CPS. The reduction represented a significant change in NASDAQ’s approach to market making.

By dramatically lowering both the fees charged to traders who executed against existing orders and the rebates paid to those who provided liquidity, NASDAQ created conditions that let researchers observe how fee structures influence trading behaviour. The experiment revealed that even small absolute changes in fee structures could trigger substantial shifts in how traders interact with markets, where they choose to trade, and how quickly their limit orders get filled.

The results challenged conventional wisdom about market structure.  A number of top theorists published articles in major journals arguing that fee structures “washed out” due to cum fee prices adjusting to changes in fees. This was despite explicit Securities and Exchange Commission (SEC) rules prohibiting trades at cum-fee prices.

The research paper noted that NASDAQ’s market share decreased by 10% in response to the maker-taker fee reduction, which showed that traders actively respond to changes in fee structures. However, while market share declined, order execution speed improved and cum-fee costs decreased, suggesting that lower fees might benefit certain types of traders, such as smaller “liquidity” traders such as households, even if they reduce overall trading volume.

Practical implications for market participants

The findings demonstrated that exchange fees significantly influence trading behaviour and market quality. The withdrawal of the subsidy massively reduced the placement of limit orders on the NASDAQ platform, according to the researchers, who said this action improved the likelihood of a limit order being hit from 3.65% to 14.86%. This reduction in the placement of limit orders greatly thinned out the NASDAQ exchange listings and especially market depth, making it far more expensive for sizeable orders to be placed on NASDAQ as the decimated liquidity supply was rapidly eaten up.

For market participants, these results suggest careful consideration of fee structures when making trading decisions. The research showed that “maker-taker platforms are far more likely to be at the NBBO [national best bid and offer] than are inverted platforms” but also noted that “the cum fee spread is higher on the maker-taker platform than on the inverted platform.”

Read more: Why ANZ faces allegations of bond market manipulation

If your order size is relatively small and you are relatively uninformed, that is, trading for liquidity reasons rather than superior private information, you are better off trading on an inverted taker-maker platform as the cum fee cost to placing a market order is lower. The liquidity provision is sufficient for small orders. By contrast, large institutional traders require considerable liquidity and huge market depth. Otherwise, the market turns too much against them. Hence, they will always prefer the most extreme “maker-taker” venue with high limit order subsidy and access fees such as NASDAQ.

Doubtless influenced by false claims that exchange fees don’t contribute to liquidity and are costly for traders, the SEC capped exchange fees at one-third of a cent per trade. The new evidence provided by the study shows that the cost of large orders could be substantially lowered by relaxing the cap.

Key takeaways for financial professionals

The research provides valuable insights for traders and market participants. Fee structures matter significantly for execution quality and trading costs. While maker-taker platforms may offer better depth and greater likelihood of being at the best bid and offer, they can result in longer queues and lower execution probability for limit orders. Trading professionals should consider these trade-offs when choosing execution venues and developing trading strategies.

Subscribe to BusinessThink for the latest research, analysis and insights from UNSW Business School

One size does not fit all as traders are exceedingly diverse in their requirements. Contrary to claims in the literature, maker-taker and inverted venues are exceedingly different in terms of both market depth and the cum fee cost to market order placement. There is further scope for the SEC and FINRA to widen the gap between these two venue types by raising the fee cap.

The study indicates the importance of the SEC’s rules creating a very competitive national market with vanishingly small net transactional fees. The combination of fee reductions and low raw spreads has facilitated massive growth in trading volume despite considerable fragmentation and the creation of over 300 trading venues. The SEC must also selectively remove the one-cent minimum tick for liquid stocks.

The findings suggest that market structure continues to evolve, with fee models playing a crucial role in determining market quality and trading behaviour. As noted in the research, “mechanism design in the security markets is a continuous process of platform innovation, adaptation to rival initiatives, and regulatory dynamics.”