Payment For Order Flow is what makes trading for free possible. But free doesn’t mean without costs.

Here are the hidden costs.

What is Payment For Oder Flow?

It’s a way for a broker dealer to outsource execution for the many stock orders it receives, while getting paid on the way.

Let’s take an example:

• Let’s say you are RobinShodd Inc, a broker-dealer offering market execution services to a large retail audience of day traders. You have a large stream of individual retail orders to buy and sell shares continuously on all sorts of stocks.
• You are legally obliged to offer the best execution to each one of your clients.
• If you route the orders to this exchange, you have to pay execution fees.
• Let’s say, for the sake of example, that the NBBO (National Best Buy and Offer) on the stock exchange is $100-$100.50 for a given stock. Now, a well-established and reputable market-making firm comes to you and makes an improved spread of $100.10-$100.40, with NO execution fees. Even better, this firm gives you $0.05 for executing that order.
• What do you say? Hard to say no, right? Routing the orders to that market-maker is called “selling your order flow”.

The work of the market-maker

• Why would that market-making firm offer you this opportunity? Because the aggregation of all these trades is profitable.
• By collecting your orders and the orders of many similar broker-dealers, it receives at the same time instructions to both buy and sell the same stock. The market-maker is making money on the spread and passes on part of the improvement to both the retail traders and to RobinShodd.
• In our example, if trader A sells at $100 and trader B buys at $100.50, the market-maker makes $0.50. Even if it improves both the bid and the offer by $0.10, and gives you $0.05, it still makes $0.25 (minus the settlement costs).
• In practice, stock spreads are rarely $0.50 wide these days, and the market-maker’s income is far less than ~$0.25. The value of the order flow is far less than $0.05. RobinHood is actually making $260 per $million of trade. That would be a 2.6bp margin, or $0.013 for a $50 share, which is still 10x more than usual commission brokers.
• Now, the market-maker’s activity is not risk-free. If all the flow suddenly becomes buyer with no seller, it will have an increasing large short position. If the buyers keep on buying and lift all the other market-makers, as well as the stock exchange, the stock price will rise from $100.50, to $101, $103 or $105. The market-maker will be on the hook for the stock rise, like any market-maker should be.
•  So the important issue is IF all the orders are independent and bi-directional, or if they have a significant net direction. Well, it happens that a million of independent stock traders buying or selling one share have far less aggregate  direction than 1 institutional either buying or selling a million shares. The fact that the flow comes from retail, and is therefore disperse, is key.

The pressure on execution brokers

• Before Payment For Order Flow existed, the executing broker typically had to cover its costs by charging $5 for each of the client trades. With PFOF, executing brokers can charge no fee on the trade, and receive $0.05 on each of the shares executed by your client.
• Meanwhile, your client gets filled at a better spread, than if you were routing the order to the exchange, and the client has no exchange fees to pay.
• The business choice is obvious, and you sell your flow to the market-maker, and you charge $0 to your client.
• And now your client is so happy that he trades even more!
• Needless to say, once one execution broker started to offer free trades, all the other execution broker had to follow.
• It is hard to imagine retail traders opening their accounts at broker charging $5 for every trade and with larger spreads. The free trading model is here to stay.

Where is the problem?

Besides the exchange, which never sees any of the trades, everybody is happy. Well, not exactly:

• The broker has an incentive to push retail traders into trading. All is good:
  • “Look how rich you could become by trading!”.
  • “This trader has quit his job to concentrate on day trading”.
  • “Why don’t you trade options to increase your leverage?”
• Unfortunately, trading is not investing. If investing is already hard, making money on trading is virtually impossible without collecting the spread. Professional traders will also confide – knowing the institutional flow because you are on the trading floor of a large bank is critically valuable. Mr. Day Trader in his living room benefits from neither the spread nor the institutional information.
• And so the statistics are alarmingly clear: almost all day traders lose money. Sometimes a lot. Sometimes, they kill themselves over it.
• Is the market-maker front-running the client’s flow? It could be buying shares on the stock exchange before filling the clients at the increased price. The regulators have an eye on this, and have slapped a few fines already.
• Also, the market maker has an incentive to delay the execution. Simply said, there are 1,000x more buyers and sellers every second than there are every milliseconds. In our days where a millisecond for a computer is what a century is for us, delaying orders has costs – the market can move away. The SEC has studied this in detail, and the effect is definitely not negligible, especially for less liquid assets like options.
• If the retail flow has more value than the average, then the institutional flow is less valuable. Since the institutional flow is still routed to the exchange, and the exchange can’t mix-and-match both flows anymore, institutionals end-up with larger spreads and all the exchange execution costs.
• Meanwhile, the exchange are losing trading volumes to the market-makers, while their costs remain. The exchange fees will not compress.
• And some pesky regulatory issues: Rule 605 and 606 says that brokers have to disclose that they sell their order flow, how much they make, as well as their relationships with market-makers. Weirdly enough, those disclosures are hard to find on brokers’ websites. But if you find them, you would actually see that brokers have close to doubled their PFOF income just from Q1 to Q2 in 2020.

Could/should we regulate those downsides?

How could we regulate to avoid the downsides?

• Banning payment for order flow: There is a precedent – the UK has banned the practice in 2012. Here’s what would happen if the SEC does the same:
  • Broker-dealers would lose the benefits of selling the flow, and they will have to charge an execution fee again.
  • Unfortunately, 57% of Schwab’s revenue comes from the interest and stock lending fees on their traders’ accounts, not from the commissions or from selling the flow. Large broker-dealers can live without the charge.
  • The smaller execution brokers do not enjoy the same advantage, which means that many small execution brokers would disappear altogether.
  • Meanwhile, institutionals would not benefit – their flow would be still be separated from the retail flow and routed towards the exchange, unlike the retail flow.
  • The spreads would widen again, as more trades would be filled at the exchange’s NBBO, rather than in-between (65% to 90%, according to Barron and the CFA institute)
  • But it would force brokers to invest into how to improve their execution, like InteractiveBrokers does.
• Force brokers to route their orders to the exchange: The NYSE actually has built a “Retail Liquidity Program”, with discounted execution costs. Hopefully market-makers will be willing to contribute both sides of the spreads on the exchange.
  • The market-makers would be in competition. The big ones would lose market share and profits, while smaller ones would gain and multiply.
  • The brokers would lose the benefit of selling their flow and have some increased costs. Brokers will have to add back the charges on trades, or die, or both.
  • It’s unsure what would happen to the stock spreads and execution costs. Probably some deterioration.  Most likely, retail traders would not be very happy.
  • The institutionals would still lose, because their flow is still too big and routed to the main part of the exchange.
  • The big winner is the exchange.
• Banning the zero-commission model: That’s unlikely to work.
  1. Brokers can charge $0.01 instead of $5 per trade.
  2. Regulators really encroach into how companies should make their profits. It’s not the tradition. It’s probably not a good thing. And the lawsuits could pile on.
  3. What about democratizing finance? OccupyWallStreet and WallStreetBets will love that the government asks Wall Street to charge commission at the expense of Main Street.

Or we can also hope that retail traders will eventually understand that paying $400 for a bankrupt company might not be such a good idea after all? That will take some time, a few more headlines and few billion dollars of losses. Which is what the regulators want to avoid.

Credits and Sources:

• Investopedia: Payment for Order Flow
• Matt Levine at Bloomberg: People are Worried about Payment for Order Flow
• Logan Kane, on Seeking Alpha: Robinhood Is Making Millions Selling Out Their Millennial Customers To High-Frequency
• Kurt Drew, on Seeking Alpha: Payment for order flow. Think twice.
• Anonymous, on Warrior Trading: A Primer on Payment for Order Flow: Brokers Selling Orders to HFTs

Related Navesink articles:

A History of Daytrading: Regulators, Congress and Robinhood

Robinhood’s $65m SEC penalty and the ‘gamification’ of trading

Most Robinhood day traders lose money