By LARRY SWEDROE
In October 2019, five major U.S. brokerage firms — Charles Schwab, TD Ameritrade, E*TRADE, Ally Invest and Fidelity — eliminated trading commissions for their retail clients. Vanguard followed in January 2020. While touted as a victory for retail investors, concerns have been raised because the zero-commission environment obfuscates the execution costs retail traders still bear, and removing commissions increased brokers’ reliance on payments for order flow (PFOF) as a source of revenue that could lead to more expensive transactions costs.
The retail brokerage industry has embraced the PFOF model for the better part of two decades. Rather than directing marketable retail orders to public stock exchanges, brokers typically route them to wholesalers for off-exchange execution. In return, wholesalers pay brokers a fraction of a penny per share for these orders. For example, E*TRADE received 20 cents per 100 shares for market orders it routed to Citadel Securities in January 2020.
In order to determine the impact of payment for order flow on the trading costs of retail investors, Samuel Adams, Connor Kasten and Eric Kelley, authors of the December 2021 study Do Investors Save When Market Makers Pay? Retail Execution Costs Under Payment for Order Flow Models, compared the spreads paid on PFOF orders to similar-size trades that were routed to off-exchange venues with benchmarks derived from exchange executions. The benchmark is based on the “Post-Zero Period” of November and December 2019.
They began by noting: “Market makers who pay broker-dealers for their retail order flow take the opposite side of orders aiming to earn the spread. Two necessary conditions give market makers the ability to maintain a stable bid-ask spread, consistently and confidently price improve, and maintain profitability. First, purchased order flow cannot be correlated with common signals. In other words, it should not be informed (market makers face less risk of adverse selection). Second, shares purchased must roughly balance shares sold over the course of a trading day. If order flow is balanced, a market maker can easily and quickly fill orders while reducing carry risk.” And finally, wholesale market makers may extract valuable information about sentiment from retail order flow.
Following is a summary of their findings:
The difference between retail and exchange spreads narrowed slightly for small and medium-sized trades and widened slightly for large trades when they compared the Base Period (August and September 2019) to the Post-Zero Period; however, the relative increase in spread was a small fraction of the commissions that brokers formerly charged.
Across market cap subsamples, retail trades had effective spreads that were roughly 30-50 percent smaller than those for comparable exchange trades during the two months prior to the zero-commission shift.
A retail trader would incur a half-spread of about $1.76 for a single 250-share trade in a $40 stock. Commissions prior to October 2019 were typically around $5.00 per trade — the elimination of commissions meaningfully reduced trading costs.
Consistent with retail trades being less informed, the price impact for retail trades, while still positive, was less than half that for exchange trades — market maker profits are partly driven by the uninformed nature of retail trades relative to benchmark exchange trades.
Unsurprisingly, trading costs were typically larger for smaller stocks, and larger order sizes had a larger price impact.
The magnitudes of retail execution cost savings approximately doubled in the Covid Period (March-April 2020).
Spreads tended to fall throughout the day.
Adams, Kasten and Kelley explained why retail spreads are uniformly lower than spreads for comparable exchange trades: “A likely explanation is that (1) retail traders are less informed than other traders, (2) brokers’ order-routing mechanism correctly segment this uninformed flow, and (3) competition in the market for wholesale executions allows the traders themselves to reap some of the cost savings.”
Their findings led Adams, Kasten and Kelley to conclude: “In the wake of the Congressional GameStop hearings that criticised PFOF, our results suggest that widespread calls to ban PFOF are premature. If anything, our results indicate that the move to zero commissions, which was facilitated by the PFOF model, proved beneficial to retail investors in terms of overall costs of trading.”
They did add the caveat that the move to zero commissions could lead retail investors to trade more, as the evidence from such studies as Trading Is Hazardous to Your Wealth: The Common Stock Investment Performance of Individual Investors has found that the more retail investors trade, the worse their results.
Adams, Kasten and Kelley also discussed the December 2020 SEC settlement with Robinhood: “The SEC’s recent settlement with Robinhood exemplifies the relevance of exactly how much savings retail traders achieve on their trades. According to the settlement, ‘at least one principal trading firm communicated to Robinhood that large retail broker-dealers that receive payment for order flow typically receive four times as much price improvement for customers than they do payment for order flow for themselves — an 80/20 split of the value between price improvement and payment for order flow… Robinhood negotiated a payment for order flow rate that was substantially higher than the rate the principal trading firms paid to other retail broker-dealers — which resulted in approximately a 20/80 split of the value between price improvement and payment for order flow. Robinhood explicitly offered to accept less price improvement for its customers than what the principal trading firms were offering, in exchange for receiving a higher rate of payment for order flow for itself.’”
Adams, Kasten and Kelly’s findings are consistent with those of Pankaj Jain, Suchi Mishra, Shawn O'Donoghue and Le Zhao, authors of the October 2021 paper Trading Volume Shares and Market Quality: Pre- and Post-Zero Commissions, who found that overall market quality improves as zero-commission traders switch to odd lots and smaller order-size buckets. The result is that effective spreads decline because retail investors post many more standing orders inside the bid-ask spread. They also found that the new orders are relatively uninformed.
Investor takeaways
The evidence suggests that investors are able to benefit from the elimination of brokerage commissions, as their elimination has resulted in lower total trading costs on average. However, you should also be aware that lower trading costs should not lead you to increase trading because the evidence also shows that the more retail investors trade, the worse their performance tends to be. In other words, the average retail trader is “uniformed” and trades on what economists call “noise”. Unfortunately, an all-too-human trait is overconfidence, with most people believing that they are better than average (a mathematical impossibility).