A good chunk of today's very silly Senate subcommittee hearing on "Conflicts of Interest, Investor Loss of Confidence, and High Speed Trading in U.S. Stock Markets" was taken up by:
- Senator Ron Johnson pointing out that, when he trades stock in his personal account,1
he gets instantaneous execution of his trades, at the price he wants, for $10 a trade, and
- Professor Robert H. Battalio telling him, no, the problem is, you don't get your stock, the price moves away from you and your order is never filled.
Now this was obviously insane.2 Johnson knows when he gets his stock, and Battalio's insistence that Johnson's trades were just a fantasy seemed, at first glance, just baffling.
But it can be explained. Battalio's testimony was about maker-taker fees, and his research is about how retail brokers route nonmarketable limit orders. If you send your broker a nonmarketable limit order -- that is, if a stock is trading at $100, and you offer to buy it for $99 -- then your retail broker will probably send your order to whatever exchange gives it the biggest kickback. And this, Battalio found, makes your order less likely to execute: Brokers look out for their own pockets, not your order.
On the other hand, if you put in a market order -- "just buy me the stock at whatever price" -- or a marketable limit order -- "just buy me the stock, but not for more than $101" -- then you'll get your stock instantly, at the quoted market price (or better), and for $10 a trade.
So now the problem is easily resolved: Johnson doesn't trade using nonmarketable limit orders. It's true that his experience is that he trades instantly at the current market price for $10 a trade, but other small investors who trade using nonmarketable limit orders are losing out because the market is rigged against them.
This conversation felt surreal because Johnson's experience is obviously the normal one. Most retail trades -- particularly trades by people with day jobs, such as being senators -- aren't nonmarketable limit orders.3 Most retail trades are: I want to buy some stock, let's buy some stock, oh look I bought some stock, that was easy. Cheap, too.
This was a dumb exchange, and it won't surprise you to learn that it was one of many.4 But it was symptomatic of the tone of the Senate hearing, which was entirely about the retail stock trader experience. So much so that the committee invited Joseph P. Brennan, the head of the index-fund group at the Vanguard Group, and then ... asked him about retail brokerage routing decisions! The man runs $2.6 trillion of index money, including most of my investments. That is rather more than Johnson runs.5 His perspective on how maker-taker fees, or whatever, affect institutional investors might be worth listening to.6 So Carl Levin asked him why his retail brokerage colleagues route orders the way they do and, when he didn't know (why would he?), demanded that he go find out and come back with the answer.
Here is the problem with most public discussion of modern market structure:
- Modern market-structure developments have been in large part about segregating uninformed retail order flow from informed institutional order flow, and competing to trade with retail investors by offering them better prices and cheaper execution.7
- Therefore those developments have mostly been good for retail investors, possibly at the expense of institutional investors.8
- That doesn't mean these developments have been entirely good: They pose some risks to market stability, for one thing, and at least some aspects of modern markets have been bad for institutional investors, who can't buy stock as efficiently as they could in the absence of high-frequency trading.
- Institutional investors are way more important than retail investors. They manage my retirement money, for one thing. Also they manage everyone's retirement money, except for a small group of wealthy hobbyists like Johnson.
- But it is hard to rile people up by talking about predators taking advantage of multitrillion-dollar investors like Vanguard, or of multibillion-dollar hedge funds for that matter.
- So you rile people up by talking about predators taking advantage of little retail investors.
- But that's mostly not true: The retail investors get cheap fast trades, while their brokers collect a few pennies in fees from exchanges and internalizers.
- So all of your hearings get confusing and surreal.
You could see Brad Katsuyama struggling with this in his Senate testimony. Katsuyama runs IEX, a dark pool that doesn't use maker-taker fees and has built-in delays in order execution. He pitches IEX to institutional investors who want to avoid predatory high-frequency traders. He's also the star of my Bloomberg View colleague Michael Lewis's book "Flash Boys," and a very reluctant poster boy for the view that equities markets are "rigged" against retail investors. Anyone who actually read "Flash Boys" -- which John McCain praised repeatedly at this hearing -- would notice that everyone complaining about high-frequency trading in the book is a hedge fund manager, and that Katsuyama started IEX after his experiences as a broker for institutional investors.
It's a strange blind spot. Really, people should care if Vanguard is getting ripped off -- certainly more than they should care if Johnson is getting ripped off. There are not a lot of other areas of American business where the interests of hobbyists get priority over the effective functioning of the mass market. No one argues against airline safety rules because they make life harder for wingsuit flyers.9
But in financial markets, the public focus is on protecting retail investors in individual stocks. I've criticized the Securities and Exchange Commission for this focus, but in the recent market-structure debate the SEC has mostly avoided it: It's been sober and sensible, if perhaps a bit slow-moving. Congress, on the other hand, can never resist a bit of populism. Even if it makes no sense.
I don't even have to tell you how awful it is that a senator is day-trading single-name equities in his personal account, do I?
2 I'm obviously exaggerating; Battalio was trying to discuss normal market experiences, not literally Johnson's personal experience. I think. It got pretty muddled.
Though more than I would have expected. The way I tried to approximate it is:
- Look at Rule 606 reports for big retail brokers such as TD Ameritrade, E*Trade, Fidelity and Scott Trade.
- See what percentage of orders are limit orders. This ranges from 26 percent at Scott Trade to 85 percent at TD Ameritrade, but seems to run around 50 percent.
- See what percentage of limit orders are routed to stock exchanges, as opposed to wholesalers (who get the marketable orders). This ranges from 30ish percent for some stocks at Fidelity to as much as 60+ percent for Nasdaq stocks at E*Trade. Figure it's usually 40-ish percent.
- Multiply and you get that around 20 percent of retail orders are nonmarketable limit orders.
That seems high! I've never used a nonmarketable limit order, and neither apparently has Johnson, but I guess a lot of people have?
In fact Battalio's study covers 28 million nonmarketable limit orders over two months in 2012 -- again, more than I would have expected. But the costs that he finds are tiny: He estimates that the aggregate cost to investors from mis-routing of trades is about $6 million over those two months, or about $36.79 million per year. That's a little under 25 cents per order, with an average order size of several thousand dollars. Versus a $10 commission, that's basically nothing.
4 Some things that bugged me:
- Johnson telling Battalio to "forget the price movement" and just focus on fees. That's not how to regulate market structure!
- Battalio complaining that retail brokers compete on commissions and use maker-taker fees to subsidize low commissions. That's a good thing! Or, arguably a good thing, anyway.
- Nobody knowing what a "stop order" is, and using it as a synonym for "limit order" or possibly something else.
5 The $2.6 trillion is, I mean, not my investments. Johnson is plenty richer than I am.
6 His testimony mentions maker-taker only briefly, focusing mostly on bigger structural issues of Regulation NMS and market data feeds, which are more important for Vanguard.
7 Equity markets are dealer markets. If you are a dealer, you don't want to buy from, or sell to, someone who has better information than you do: You'll sell a stock just before it goes up, or buy it just before it goes down. So if you have some way of knowing that an order is uninformed, you will want to trade with that order, and you'll be willing to pay for the privilege. And if you have some way of knowing that an order is informed, you will want to avoid that order.
This drives internalization and payment for (retail) order flow. It partially drives maker-taker fees. And it drives high-freqency-trading behavior like fading in the face of a big order: If there's a big order, you don't want to trade with it, so your prices move away from it as quickly as possible.
8 I've discussed this in a bunch of places, including footnote 2 here, which has links.
9 I guess if you told me that people do argue against airline safety rules because they're bad for hobbyist small-plane flyers, I would believe you.
To contact the writer of this article: Matt Levine at firstname.lastname@example.org.
To contact the editor responsible for this article: Zara Kessler at email@example.com.