I wanted to make this thread because I think this is a topic many traders (particularly retail like us) could benefit from knowing more about, and because we now have a recent study that actually lets us quantify their effects on our fills.
This is important because I feel that the discussion/debate around Payment-for-Order-Flow (PFOF) often occurs at an emotional level: retail traders feel like they’re being taken advantage of, like the system is “rigged” against them. And you know what, I think this is a valid sentiment: these processes are largely opaque to us, and although we might “feel” like we get better fills with certain brokers, there’s been no real way to quantify that.
Which is why I found this study so interesting. For the first time we actually get a detailed comparison of price execution between brokerages. I’ll summarize some of the major points below, and if you’re interested in more detail I’d point you to Matt Levine’s column on it (if that link is paywalled you can try this one).
To collect their data, the researchers opened accounts at five of the top retail brokerages and made 85,000 trades in 128 stocks. Notably they did not study options, which I think would be fascinating for a future study. Mostly these trades were worth about $100, and they would just buy the stock and sell it later that day. By submitting the same trade for the same stock at the same time of day, the researchers could compare the fills that they received from the different brokers.
-
The fills they received were actually better than the national best bid-offer (NBBO). Compared to execution at NBBO, the brokers on average “saved them $0.03 to $0.08 per share, or 0.17% to 0.55% of their money.”
-
This price improvement didn’t correlate with PFOF. IBKR and Fidelity, who do not accept PFOF, actually did not substantially improve their customers trades as compared to TDA, RobinHood, and ETrade (who do accept PFOF).
-
The best average price improvement, for both S&P 500 and non-S&P 500 stocks, was achieved by TD Ameritrade. Below is the graph, where you can see average price improvement for each broker, along with how much they pay per trade in PFOF:
[center]
[/center]
In my opinion this is the most important chart in the whole article. MMs like Citadel and Virtu and Jane Street provide vastly different price improvement for identical orders of the same stock at the same time - just for different brokers. IBKR Pro, which also charges commissions, actually performed the worst.
In terms of reasons why this might be, there are a few ideas. One is that the MMs want to trade less with more “informed” traders. As an MM, you’d rather trade against a retail trader buying memestocks than a seasoned professional trader who might know what they’re doing. From the article:
The fact that the professors’ Interactive Brokers Pro account performed relatively poorly might have something to do with who Interactive Brokers Pro is designed for: pros. If you are a market maker trading with pros, you will charge a higher spread than if you are trading with amateurs.
Still, I’m impressed by differences between the commission-free brokers. I’d imagine most of the amateurs would be using RobinHood (since its app is most beginner friendly), so you might expect that they’d provide the best price improvement, but they don’t. It’s an interesting question.
Anyway, hope you learned something, and perhaps this post could start some kind of discussion on this important topic. Definitely comment below if you have any questions/comments.