Terrance Odean on the $2.8bn catch in zero-commission trading

Berkeley professor’s research on the true cost of retail trading has caught the eye of regulators


Terrance Odean might have become a psychologist if not for Daniel Kahneman, one of the pioneers of behavioural economics, who convinced him it would be easier to make fresh discoveries in this newly emerging field – and that if the research failed to take off, he could easily land a higher paying job in finance.

“I’ve thanked him many times,” the Berkeley finance professor says.

Parts of the industry may have less cause to be grateful. Three decades on, Odean’s most recent research is raising tough questions about how well retail investors are being served by market players. And regulators are showing interest.

Competitive markets don’t appear by “magic”, Odean tells Risk.net, speaking about his research from his home in California. “There are elements that have to be in place for competition to work properly. In retail trading, I’m not sure those elements are in place.”

In December 2021, Odean – the Rudd Family Foundation Professor of Finance at Berkeley’s Haas School of Business – and fellow academics at the University of California at Irvine, UC Davis and Washington University, began an experiment to test execution quality at different retail brokers.

Their idea was to place identical stock orders at identical times at different brokers – E*Trade, Fidelity, Interactive Brokers, Robinhood and TD Ameritrade – and compare prices.

Millions of small investors have been enticed by the zero-commission trading offered by today’s retail brokers. But the true cost of transactions for investors – there is one – comes from the price at which trades get done compared with the best price available for public exchanges.

On the face of it, retail investors get a good deal: a price improvement on average of a third of the spread. What stood out, though, was how much of a better deal they got in certain cases versus others. Differences in prices on identical trades were “astonishing”, the academics wrote in a study released in August.

Market-makers are going to say they give more price improvement when the flow is less toxic
Terrance Odean

Sixty-nine percent of trades through TD Ameritrade, now owned by Charles Schwab, were executed at the mid-price or better, implying zero true cost of trading. Between 16% and roughly 40% of trades through other brokers reached the same threshold. Round trip costs – for trading in and out of a position – ranged from seven basis points to 45 basis points for the same stocks bought or sold at the same time. For every basis point of price execution difference, the researchers estimate an annual cost to retail traders of $2.8 billion.

Some divergence might be reasonable, Odean tells Risk.net. But the recorded differences were a big surprise.

What’s going on? A prior target of criticism in relation to retail trade execution has been payment for order flow (PFOF) – where wholesalers pay brokers to send them retail orders. The suggestion is that PFOF might incentivise brokers to send orders to venues where investors get an inferior price.

But PFOF practices did not explain the variation in executed prices Odean and his co-authors observed, he says. Fidelity takes no PFOF, the researchers note in their paper, but had worse order execution than TD Ameritrade.

Nor did brokers’ choice of where to send orders account for the variance. Instead, different wholesalers – firms such as Citadel Securities, G1X, Jane Street, Two Sigma, UBS and Virtu – seem simply to offer better or worse prices to different brokers, even on identical trades.

Terrance Odean
Terrance Odean

That might seem odd, but in fact makes sense, Odean says. Market-makers offer different execution to different brokerages to compensate for ‘toxic flow’ — trading that comes in directional waves, such as lots of buy orders at once – which adds to the market-maker’s risk exposure.

What makes less sense is the size of the differences. “Market-makers are going to say they give more price improvement when the flow is less toxic. That’s almost certainly true,” says Odean. “I don’t know, though, about the magnitudes.”

More data might help, the academics suggest.

Competition depends on buyers paying attention to price improvement, Odean says. That’s to say, brokers will care more about price improvement if customers care. But retail traders — even if they monitor trade execution closely, which they often don’t — lack the information needed to know how different brokers are performing.

It took the academics nearly a month to get trade routing data from brokers that shows which venues the trades went to. Brokers are required to supply such data to customers under SEC rules but aren’t used to doing so. “They weren’t holding out,” says Odean. “They just hadn’t been asked before.”

In their paper, the academics call for wholesalers to disclose stock-specific execution statistics by individual broker so that differences would be obvious to end-customers. Odean says he and his colleagues have discussed their research with representatives from the Securities and Exchange Commission.

Thinning herd

Over the past 30 years, Odean has documented much of what we know about retail investor behaviour: their tendency to underperform the market (even though their individual stock choices are sound), to sell winners and hold on to losers, to pay too much attention to recent returns, to make over-concentrated bets, and so on.

In a 2020 paper, he and his colleagues showed that features of the Robinhood app push its users into trading more heavily in certain securities.

Some things about how retail investors are trading have changed. In nature, herds follow leaders. Today the investor herd sees more easily where its leaders are headed.

In the past, investors might have reacted to the same signals in the same way. They might all have paid attention to the same stock that hit the news. But they were not observing and following a leader, Odean says.

“It might have looked like herding. But it was not the same as going on Robinhood and seeing which stocks had the biggest increase in Robinhood users that day, or seeing what most users are doing on Reddit, or having someone say: ‘Alright, here’s what we’re gonna do’.”

The internet makes it easy for investors to follow what other retail investors are doing. “You can communicate with them or see what they’re saying,” Odean says. “That probably facilitates a higher degree of correlated trading.”

Some habits, though, remain constant, he says.

“We’re human. We were human 20 years ago. We may have different opinions. We certainly have different market structures. But a lot of the psychology stays the same.”

Which may help explain why retail trading volumes have fallen back this year. “Retail investors like to buy stocks that have been doing well. There are fewer such stocks in a down market,” Odean says.

“Retail investors like to buy stocks that catch their attention. When the market is going up, they pay a lot of attention. They pay less attention when it’s not doing so well. So they trade less. They also trade for fun. It’s a lot less fun when you’re losing money.”

Correction, September 22, 2022: An earlier version of this article referred to hedge fund Citadel rather than Citadel Securities, which is a separate entity. We have made the correction.

  • LinkedIn  
  • Save this article
  • Print this page  

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact [email protected] or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact [email protected] to find out more.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here: