[wrsh07]

Pfof is woefully misunderstood

In general, citadel wants to pay to trade with retail investors because it knows it isn't going to face adverse selection. So it will give them tighter bid/ask ratios (this is better for the customer) than they would get if they were trading in the open market, citadel isn't going to get hosed by one of them (because there's no adverse selection)

It's win win win

[taway789aaa6]

> PFOF and excessive off-exchange trading persist because so many trading platforms rely on the revenue it generates, essentially productizing their clients. Defenders of PFOF have claimed that retail brokers who route to high-speed traders (in exchange for PFOF) provide better price execution for investors and that it’s a net positive, despite creating an inherent misalignment between these platforms and their customers, and despite public evidence to the contrary. Leaning on the flawed argument that they categorically provide retail customers with best price execution quality, there is little by way of self-regulation to foment change or prevent applications designed to optimize transaction volume (i.e. speculation and day trading) and risky activity (i.e. margin and options trading). Further, their ability to claim best execution is part of the flaw of the system, as even within the current structure better outcomes are possible on an order-by-order, and aggregated basis.

https://advocacy.urvin.finance/advocacy/we-the-investors-pfo...

Not a win win.

[gruez]

>and despite public evidence to the contrary

Sounds serious, I wonder what it is...

>"410 The author deleted this Medium story".

doesn't look promising. The rest of the paragraph fails to state any concrete harm, instead focusing on abstract issues like "misalignment between these platforms and their customers", and "little by way of self-regulation ".

[Anon1096]

There is nothing in that statement that actually shows negative effects of PFOF.

> creating an inherent misalignment between these platforms and their customers

is just speculative harm, and as to the other part about preventing risky trading - this is literally what Robinhood et al customers want!

Meanwhile PFOF actually does have proven benefits in that it reduces spread for retail investors.

[Ntrails]

> Meanwhile PFOF actually does have proven benefits in that it reduces spread for retail investors.

To be fair, some of that is because its existence changes the pool of people trading on lit and thus increases spreads there. There are systemic effects that are a function of pfof that make it look better, and ofc there are a wide range of actors of varying quality...

[highwaylights]

This.

It seems very much like that bogus stat that HR departments were peddling 20 years ago about how they only hire the top X% of people because they reject (100-X)% applicants - it tells you nothing about the quality in the gap.

These systems don't have to actively attempt to front-run you or pro-actively make bad trades, they can just optimize for deal flow, which is enough to cause the customer to get a sub-optimal price.

[freeone3000]

You’re getting a price as good or better than if you had routed it through the backing exchange directly. National Bid or Best Offer continues to be the rule.

[wrsh07]

I think you're only highlighting my point that it's woefully misunderstood

The fact that 70k people signed a statement making a bunch of strong but vapid claims is umm telling

Let's take a longer money stuff excerpt:

>>> Some retail brokerages seem to make a lot of their money from payment for order flow. Others make less. Some big retail brokerages do not accept any payment for order flow at all: They still use this system (routing their orders to market makers), but they take 100% of the value in the form of price improvement for their customers instead of payments for themselves. Intuitively, you might think that the brokerages that get a lot of PFOF would get worse price improvement.

But, nope! Here is Bill Alpert in Barron’s:

Critics of retail brokers like Robinhood Markets condemn those companies for routing customers’ orders to market makers like Citadel Securities in exchange for payments. ...

The suspicion is that greater payments to brokers must be offset by less favorable execution prices. But that isn’t what a new study finds.

In an Aug. 13 working paper, five finance professors analyzed 85,000 stock trades they made through five leading retail brokers. They did get significantly different pricing through different brokers for identical orders to buy or sell at the current market price.

But their best pricing came from a broker that takes payment for order flow, namely TD Ameritrade, now a unit of Charles Schwab. Fidelity, which takes no order payments, got worse prices on the professors’ trades than did TD Ameritrade. And its prices were no better than those from the E*Trade unit of Morgan Stanley, which does take payments. Robinhood, which used revenue from order-flow payments to subsidize the industry’s first commission-free trading, delivered middle-of-the-pack pricing. Interactive Brokers ranked last in the execution pricing of the professors’ orders.

That's from https://news.bloomberglaw.com/mergers-and-acquisitions/matt-...

Excerpted Barron's: https://www.barrons.com/articles/payment-for-order-flow-sec-...

Paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4189239

[throwacomment]

Yes, PFOF is woefully misunderstood but its very much not win win win.

The reason its bad is because its anti-competive and gives them information that no-one else has access to.

By trading against you, Citadel prevents any other potential market maker from trading with you. With less competition, the spread widens and even after price improvement, you're paying more.

PFOF also tells them who they are trading against but anyone else who just sees a quote doesn't know that.

Generally, things are very zero sum so wins all around are very unlikely. But some thinking is needed to track where the value loss and gains are.

[wrsh07]

Please see my other comment that provides links to a study that shows: yes you get the best price from a broker using pfof

Your argument seems reasonable but isn't borne out empirically

https://news.ycombinator.com/item?id=42378516

[throwacomment]

I'm not talking about you getting a worse price today.

Suppose in some other industry, some monopolist consistently sells goods at a loss to drive out all the competition. In the last moments when they are doing this, yes its cheaper for you to buy from the monopolist at that moment. But after everyone is driven out of the market, you'll be paying more. Even though the monopolist is still the cheapest amongst all options.

I'm saying you're already in the "after" scenario here. You're saying that you can save a few cents with PFOF when you cross that 50 cents spread and yes that's true. But I'm saying that spread should be 25 cents and no-one is offering that because they've been driven out.

Now that I think about it, the more immediate consequence to you is that some of your order will not fill because PFOF exists, rather than you getting a worst price. Say you put a bid to buy at 100. And then I come along and want to sell at 100. Normally, you'd get to buy from me. But because my order is PFOF'd, Citadel decides that buying from me at 101 is a good deal so they do. This happens a few time with different sellers then you get fed up and/or the market moves. So you raise your bid to 150. Citadel sells to you at 149. You saved 1 off that 150 but lost out 49 from the trade you'd have gotten from me without PFOF.

[wrsh07]

This example is apples to oranges

Imagine you are a market maker: you offer 2 APIs. The first, you allow anyone to trade on. The second, you only allow traders who are doing less than 100k in volume per day (and don't allow users to have multiple accounts)

Which API are you able to offer tighter bid/ask spreads on? Why?

That's the point. Pfof is saying: the second API is so valuable to me that I'm willing to pay to obtain customers. In the worst case, there will always be the open-to-all API.

Your second example continues to show the lack of understanding. You're saying: without the market segmentation, somehow I have a wider bid ask. That's not right at all. The entity that gets bad spreads is going to be the entity that would take advantage of good spreads. That's the whole point. Maybe there's a point that vanguard ends up getting worse execution because it gets lumped in with the rest of the market, but the counterargument to that is essentially just volume: is the retail market big enough that if you didn't segment them onto a better spread that the overall market would end up with better pricing. The answer: maybe! In some things! Is that really what the people who hate pfof want though?

My understanding is that people who hate pfof are actually the ones benefiting the most from it. (ie because unsophisticated investors get better execution)

[throwacomment]

PFOF does two things and you're only focusing on half of it.

1. It segments the counterparty they trade with.

2. They get dibs on new orders arriving.

You're only talking about 1. I'm talking about 2.

1 is also bad because this segmentation also gives them inforamtion no-one else can get. But the chain of reasoning to concretely show why its bad (for someone getting their orders PFOF'd) is less obvious and longer.

> Imagine you are a market maker: you offer 2 APIs.

This is so wildly different from how market works. You'll have to clarify what you mean. If the only way to trade is through the API, then you'll offer infinite spread on both. If normal markets exist alongside, then I don't bother using either.

[wrsh07]

I'm not sure what to say. Your arguments are extremely hypothetical and there's no evidence of the claimed badness today.

I don't find them convincing - why is it bad that someone paying for exclusive access to data gets exclusive access to that data? There are so many exclusive data vendors in financial markets, this one seems relatively low value

[wbl]

Dude, I want the market to see I'm a moron! I'm not buying BH because I've observed private jets between Omaha and Washington but because I'm saving after having been paid.

[throwacomment]

Then find a way to tell everyone this in the open, not just Citadel. Then anyone else is free to trade against you. There can even be a micro-auction to get you the best price among all counterparties that want to trade with you. There's already auction mechanisms at some exchanges so I'm thinking attaching a voluntary "this order came from Robinhood" tag to your order shouldn't be too hard?

[lldb]

It is not a win. In a recent study, Robinhood with Citadel has the worst price improvement (execution quality) of any brokerage on the market. I’ve personally observed this - Robinhood might “improve” by 1/10 of a cent from NBBO while Fidelity is frequently closer to the mid.

[loeg]

How is that not a win? Robinhood customers still got better execution than NBBO. If you don't like Robinhood getting a tiny kickback here, you're free to go to another brokerage.

[wrsh07]

This is just noting that different brokers give different performance

That doesn't really have anything to do with pfof (TD Ameritrade gives better execution and receives pfof)

https://news.ycombinator.com/item?id=42378516

[loeg]

Presumably a market maker would pay (PFOF) slightly more to deliver slightly worse execution (keeping the spread).

[wrsh07]

Sure that sounds plausible but it's literally not what happens in practice (see the other comment I linked that discusses research on this very thing)

[loeg]

Yeah, I've seen the Levine column on it.

[wrsh07]

Here's the money stuff excerpt: https://marginalrevolution.com/marginalrevolution/2021/02/th...

> I feel like most of what I read about payment for order flow is insane? Otherwise normal people will start out mainstream explainer articles by saying, like, “Robinhood sells your order to Citadel so Citadel can front-run it.” No! First of all, it is illegal to front-run your order, and the Securities and Exchange Commission does, you know, keep an eye on this stuff. Second, the wholesaler is ordinarily filling your order at a price that is better than what’s available in the public market, so “front-running”—going out and buying on the stock exchange and then turning around and selling to you at a profit—doesn’t work. Third, because retail orders are generally uninformative, the wholesaler is not rubbing its hands together being like “bwahahaha now I know that Matt Levine is buying GameStop, it will definitely go up, I must buy a ton of it before he gets any!” The whole story is widely accepted but also completely transparent nonsense.

[1oooqooq]

it's already public that frontrunning is perfectly legal if you can do it with large volume as to not show intent of frontrunning one single person.

[slt2021]

yeah, Citadel's annual $30,000,000,000 profit is not coming out of thin air or just from bid-ask spread. Customers are being taken for a ride definitely

[nobodywillobsrv]

Adverse selection goes both ways. If PFOF leads to adverse selection against your flow then it's not win-win. You might say you are willing to trade of adverse selection up to the cost of the fees, but then you are trading a known fixed fee for an unknown stochastic penalty. And also who sets the fees?

The entire thing is adversarial and it's really just a choice of game you choose to play.

[WiSaGaN]

It's not. Centralization of liquidity is better for everyone. HFT thrives on fragmentation of liquidity. HFT is not wrong, but fragmentation of liquidity is.

[rjrdi38dbbdb]

Nope. It's not better for known uninformed traders. If you mix them in with informed traders, market makers must widen spreads.

This is very obvious in institutional FX. Pure "retail" flow will get quoted much tighter spreads by banks and market makwrs than you'll see on any ECN. Yes, it can get skweded against predictable flow, but a true "noise" trader won't be affected by that and will definitely be better off with tailored liquidity.

[WiSaGaN]

You don’t have to trade with market makers.

[rjrdi38dbbdb]

So you're hoping get price improvement by crossing with other trader orders in the book?

Unless you have a good high frequency predictor and low latency order management (you don't), you're going to experience adverse selection. Either because you're taking resting orders that HFTs are smart enough to avoid or because your resting orders get run over by informed traders.

[WiSaGaN]

So you are saying HFT will avoid your market order in this case, while HFT will provide better price when they are the sole counter party in separate liquidity pool? HFT will always maximize profit. To have multiple venues you are just paying HFT as middle man to transfer liquidity from one to another, where you can trade directly with each other if everyone is on one venue, e.g. one centralized limit order book. Transfering liquidity is not HFT's fault, but saying paying for order flow is better for retail is just disinformation. Without evenly discussing the function of HFT, you will get disinformation that demonize HFT as well, and common people won't listen to you later.

[rjrdi38dbbdb]

> So you are saying HFT will avoid your market order in this case, while HFT will provide better price when they are the sole counter party in separate liquidity pool?

Yes, absolutely. The best feeds (tightest spreads) are only given to specific clients who are requested to trade exclusively with them. If they detect you splitting your orders up between venues, they'll worsen your feed. The feed they'll send to public lit ECNs will generally be their worst (widest spread).

[neximo64]

Distorted incentives