> I wonder why giant market-making hedge funds would pay for that order flow... Another tech faustian bargain.
It really isn't. They're making money off the spread (eg. $105.01 bid vs $105.02 ask), which exists regardless of PFOF (regulation NMS mandates that the price be better or equal to NBBO). The reason why they want retail flow is that it's mostly "uninformed" and they're less likely to get run over.
>I like to tell a fairly textbook version of that story. Market makers stand ready to buy or sell stock from or to customers; they try to buy for a bit less than they sell at, and pocket the spread. If you go out into the market and say “hey I’ll buy anyone’s stock for $10,” and a really smart hedge fund comes to you and sells you stock for $10, that’s probably bad. You’ve probably made a mistake. The hedge fund is selling you the stock for $10 because it knows it’s worth $8. This is called “adverse selection.”
>More subtly, if a really big mutual fund comes to you and sells you stock for $10, that also may be bad. The mutual fund is probably selling lots of stock, because it’s so big; it sells you a little, then sells a little more, then a little more, until it pushes the price down to $8. The mutual fund isn’t necessarily smart, but by virtue of being big and doing big trades, it moves the price; if you are on the other side of its trades, you get run over. This is also a kind of adverse selection: You buy at $10 and are stuck selling at $8. Part of the spread that market makers earn in public markets—the difference between their buying and selling prices—compensates them for adverse selection, the risk of being run over by a counterparty who knows something they don’t.
>Market makers, the textbook theory goes, would much rather trade with retail orders. Retail investors generally don’t know much, so if you buy stock from them you’re probably not making a mistake. And retail orders are generally small and uncorrelated: One investor buys a little, another comes along a moment later and sells a little, it’s all pretty random, and you’re not facing an avalanche of steady sell orders that push the price down. Trading with retail is so nice that market makers—wholesalers—will both give retail orders a tighter spread (pay more to buy their stock, charge less to sell stock to them) and pay their broker for the privilege of doing it.
What you are describing is market makers, not Robinhood. Quoting directly from the S-1:
> Our PFOF and Transaction Rebate arrangements with market makers are a matter of practice and business understanding and not documented under binding contracts. For the three months ended March 31, 2021, 59% of our total revenues came from four market makers.
So 59% of Robinhood's revenue comes from selling PFOF to market makers. I promise you that there isn't some magic altruism on the part of market makers buying the PFOF and then routing 40-60% of trades off-exchange. If it was simple matter of profiting off bid-ask spread: Force those orders through the exchanges instead of through dark pools.
This is precisely why there's an entire section dedicated to PFOF regulatory risk in their S-1. It's increasingly a rigged game and rightfully deserves deep Congressional intervention.
> What you are describing is market makers, not Robinhood. Quoting directly from the S-1:
Did i claim that robinhood is a market maker and/or participates in market making? I simply explained how market makers are making money in a non-nefarious way and why they might pay robinhood for order flow.
> I promise you that there isn't some magic altruism on the part of market makers
As explained in my prior comment there's no altruism involved. Retail orders are valuable because they're uninformed/non-toxic
>and then routing 40-60% of trades off-exchange. If it was simple matter of profiting off bid-ask spread: Force those orders through the exchanges instead of through dark pools.
Why bring in dark pools and "off-exchange"? The whole point of buying orderflow is to execute it yourself rather than letting anyone execute them.
> If it was simple matter of profiting off bid-ask spread: Force those orders through the exchanges instead of through dark pools.
When the broker gives the order to a market maker directly, they can't offer a worse price than the exchanges (national best bid offer). Typically, they offer a better price (=price improvement, a "discount").
Robinhood takes a cut of that discount (and a larger cut than other brokers).
The other poster explained why the market maker gives a discount on "uninformed" retail flow compared to the NBBO.
I agree that HFT is just a silly game of being faster, and is largely rent seeking and even destroying value. But market making per se is valuable, and pretty competitive, and the spread constitutes the necessary and benign payment for that service.
Here's a suggestion:
1. Restrict trading to, say, 4 hours a day, 2 in the morning, 2 in the evening. You could maybe make it such that time zones have partial overlap.
2. During these hours, have an auction every minute, instead of continuous trading. Maybe with stochastic end time (to negate HFT techniques/sniping).
3. Impose a Tobin tax of, dunno, 1 basis point on every trade.
Measures such as those might limit the opportunity for profit from silly HFT (like replacing the cable from Chicago to NY by a slightly straighter cable to shave off a few milliseconds).
Former CBOE market maker here. You and the person you're responding to both have valid arguments. The problem is you are choosing only to present the ones that make this situation look good, and he's only insinuating the bad ones.
The fact is that this "spread" you speak of is a much more theoretical concept than Matt Levine understands. There is ample liquidity between the bid-ask in 99.9% of markets, and by selling order flow to someone who will internalize it at the worst legal price possible, they are intentionally failing to fill an order at the best possible price.
RobinHood also features various dark patterns that are designed to remove money from the pockets of their users and put it into their own pockets. Off the top of my head, I can list the following:
(a) Very difficult access to bis-ask spread information across multiple options. This keeps users ignorant of the fact that some options may be better priced than others, and gives market makers more opportunity to make more than a fair market spread on the transaction.
(b) Forced close-outs for reasons that no other legitimate brokerages use. Even worse than being ill-infomed about what to trade is to have all of your agency removed. It's situations like these where the gap between a fair market edge and the edge that market makers take becomes offensive.
(c) Disallowing option exercise before expiration. There are many situations where an option owner should exercise his option prior to expiry. Not only does RobinHood keep its users ignorant of this fact, they actually don't even allow their users to do it. In some circumstances, this can give market makers a massive arbitrage opportunity.
While you are right that one thing that makes RH flow more valuable is the smaller average account money size, this is actually far less of an issue than just the average account financial IQ size. Citadel loves trading with pensions just as much as RH users (i.e. similar financial IQ, but far different sizes). It's just that the type of trading that happens with each is a little different.
Add in the aforementioned reasons for keeping them not only ignorant, but handcuffed, then the more market makers will pay RH for access.
>and by selling order flow to someone who will internalize it at the worst legal price possible, they are intentionally failing to fill an order at the best possible price.
AFAIK they have duty of best execution, so they're supposed get the best price irrespective of PFOF. Obviously this conflicts with their own incentives, but that's what the laws are for.
>RobinHood also features various dark patterns that are designed to remove money from the pockets of their users and put it into their own pockets. Off the top of my head, I can list the following:
I'm not a user so I didn't know any of these. Thanks for bringing these up. Informed complaints like these are far better than the "they're front running you!" complaints that people seem to repeat endlessly.
> AFAIK they have duty of best execution, so they're supposed get the best price irrespective of PFOF. Obviously this conflicts with their own incentives, but that's what the laws are for.
There is a duty of best execution. I honestly don't even know if its a crime or a licensing requirement or what. Reg NMS seemed to have obviated it, and judging by the failures to execute properly on the parts of major banks (e.g. my family had a major bank execute a bond trade for them that another bond trader friend of mine said was 10% below a competitive market price. That is, they paid 90, when you could have paid 100 in the competitive market), my impression is this law is totally unenforced.
Also, it looks like no one will be able to see my criticisms because YC wants to protect their investment going into the IPO by crushing this comment thread
>The fact is that this "spread" you speak of is a much more theoretical concept than Matt Levine understands. There is ample liquidity between the bid-ask in 99.9% of markets, and by selling order flow to someone who will internalize it at the worst legal price possible, they are intentionally failing to fill an order at the best possible price.
I remember reading about this (can't find the original source) and I thought there was a subtle distinction here, where the regulation only requires you to fill at a price at or better than the best quoted price, which isn't the same as the best possible price. Which does allow a broker to make their money (directly or indirectly) from the difference.
And this is an interesting little wrinkle, but one reason for the proliferation of high-frequency trading is to make the quoted spread to be wider than it needs to be. In the presence of predatory HFT techniques people who just want to provide honest liquidity (but who cannot afford 30 FPGA engineers) are forced to either leave the market all together or to provide quotes extremely wide to prevent the HFTs from picking them off (i.e. using information that everyone acknowledges will changes the prices, but doing it half a miscrosecond before you can, and thus forcing you to make a trade at a bad price).
It really isn't. They're making money off the spread (eg. $105.01 bid vs $105.02 ask), which exists regardless of PFOF (regulation NMS mandates that the price be better or equal to NBBO). The reason why they want retail flow is that it's mostly "uninformed" and they're less likely to get run over.
matt levine explains this in detail: https://www.bloomberg.com/opinion/articles/2021-01-29/reddit...
>I like to tell a fairly textbook version of that story. Market makers stand ready to buy or sell stock from or to customers; they try to buy for a bit less than they sell at, and pocket the spread. If you go out into the market and say “hey I’ll buy anyone’s stock for $10,” and a really smart hedge fund comes to you and sells you stock for $10, that’s probably bad. You’ve probably made a mistake. The hedge fund is selling you the stock for $10 because it knows it’s worth $8. This is called “adverse selection.”
>More subtly, if a really big mutual fund comes to you and sells you stock for $10, that also may be bad. The mutual fund is probably selling lots of stock, because it’s so big; it sells you a little, then sells a little more, then a little more, until it pushes the price down to $8. The mutual fund isn’t necessarily smart, but by virtue of being big and doing big trades, it moves the price; if you are on the other side of its trades, you get run over. This is also a kind of adverse selection: You buy at $10 and are stuck selling at $8. Part of the spread that market makers earn in public markets—the difference between their buying and selling prices—compensates them for adverse selection, the risk of being run over by a counterparty who knows something they don’t.
>Market makers, the textbook theory goes, would much rather trade with retail orders. Retail investors generally don’t know much, so if you buy stock from them you’re probably not making a mistake. And retail orders are generally small and uncorrelated: One investor buys a little, another comes along a moment later and sells a little, it’s all pretty random, and you’re not facing an avalanche of steady sell orders that push the price down. Trading with retail is so nice that market makers—wholesalers—will both give retail orders a tighter spread (pay more to buy their stock, charge less to sell stock to them) and pay their broker for the privilege of doing it.