Breaking down the 'payment for order flow' debate
a16z.com
a16z.com
https://www.bloomberg.com/opinion/articles/2021-02-05/robinh...
Hint: PFOF is fine.
> Now of course I am oversimplifying. For one thing, the wholesalers don’t have to fill every order out of inventory; they will do that with some orders and pass others on to the exchange to execute. They do not provide price improvement on 100% of orders, though they do compete to provide price improvement and are evaluated by brokers based on how much they provide. They will often lay off risk on the public markets rather than trading exclusively with retail customers; often they will be in the business of market making on the exchanges too, and will manage that business and the retail business in some interacting way. Anyone who trades a lot of stock benefits from having information about order flow, and a wholesaler who sees a lot of retail orders will have some informational advantages in its public trading.
Or TL;DR: the NBBO benefit pointed to by OP isn't absolute, order flow is valuable information, and rather than being some kind of exchange utility wholesalers are actually trying to maximize their own profits.
I'm fine with this? I'm not fine with how OP characterizes them, which doesn't seem aligned.
However, retail order flow is (pretty much by definition) not particularly valuable information: if it was, wholesalers would be unwilling to consistently take the other side at such good prices.
Of course order flow in general is highly valuable information, it's just that retail order flow isn't because it's very non-toxic (there is little to no predictive signal in the flow).
OP dismisses this out of hand because it would be a "monopoly" and jack fees. But has that been the case for the NYSE in other areas? I bet it hasn't been.
What it seems like is that these trading houses are padding their percentages by being market makers, and padding their models with order flow. Retail investors can do neither, so I think it's reasonable to say this isn't super fair.
Of course trading isn't fair for a multitude of reasons, but I wouldn't say wholesaling even makes the top 10 reasons. What is much more relavent is the large balance sheet, access to cheap debt, top tier talent, world class technology infra, etc.
Also most trading firms are market makers. it's not really a hard and fast line, anyone can provide liquidity and (try to) collect the bid ask spread (it's not easy/free money)
That's an excellent point yeah. I do buy that this is a tempest in a tea cup relative to all that.
> anyone can provide liquidity and (try to) collect the bid ask spread (it's not easy/free money)
Yeah I buy that too I mean, we can refer to it as "the modeling" but, I'm sure it's pretty complicated.
I learned a lot in our little thread! Thanks for talking w/ me :)
If that uninformed flow is going to market, there is a good market design to encourage price competition among MMs who fight each other for time and price priority. That mechanism is obstructed under payment for flow. Under this, the MM only needs to give at or better than NBBO. The primary competition among MMs here is to pass on as much rents to brokers as possible, instead of minimizing rents in the first place. The market is also more prone to consolidation due to scale economies and informational advantages that private flow access provides.
I also dispute the clean distinction between retail and institutional. Much institutional flow is just funds (e.g pension funds) that hold retail money. If they're getting maximally screwed by this, then so are their investors.
How is that not a short squeeze...
Sounds about right...
Why do you trust the SEC more than short interest numbers?
3.4 Short Selling and Covering Short Positions GameStop at the time was notable for its significant short interest (the ratio of shares currently sold short to shares outstanding).74 Figure 5 shows GME’s short interest over time, along with average levels of short interest among other non-financial common stocks. In the past, GME had several periods of high short interest, but none as high as the levels achieved from 2019 to mid-January 2021. GME short interest hit 50% of shares outstanding first in 2012 and then again in 2015, 2016, and 2018, before rising even further in 2019. From then until early 2021, GME short interest hovered around 100%, hitting its high of 109.26% on December 31, 2020. Some commentators have asked how short interest can get as high as it did in GameStop. Short interest can exceed 100%—as it did with GME—when the same shares are lent multiple times by successive purchasers. If someone purchases a stock from a short seller and subsequently lends the stock out again, it will appear as if the stock was sold short twice for the purpose of the short interest calculation.75 Short interest ratios tend to be quite low; for large non-financial stocks, they are often less than 2.5% whereas for small non-financial stocks they still tend to be less than 13%. Few stocks, if any, have short interest greater than 50% on a given date.76 Until recently, short interest of more than 90% was observed only a few times—in 2007 and 2008. When examining short interest as a percent of shares outstanding, GME is the only stock that staff observed as having short interest of more than shares outstanding in January 2021.
What’s the difference between a short squeeze and short sellers are buying to close their positions 2021, GME short interest hovered around 100%, hitting its high of 109.26% on December 31, 2020. Some commentators have asked how short interest can get as high as it did in GameStop. Short interest can exceed 100%—as it did with GME—when the same shares are lent multiple times by successive purchasers. If someone purchases a stock from a short seller and subsequently lends the stock out again, it will appear as if the stock was sold short twice for the purpose of the short interest calculation.75 Short interest ratios tend to be quite low; for large non-financial stocks, they are often less than 2.5% whereas for small non-financial stocks they still tend to be less than 13%. Few stocks, if any, have short interest greater than 50% on a given date.76 Until recently, short interest of more than 90% was observed only a few times—in 2007 and 2008. When examining short interest as a percent of shares outstanding, GME is the only stock that staff observed as having short interest of more than shares outstanding in January 2021.
Second, the real estate thing is a weird tangent. I just don't buy that adding an intermediary to real estate transactions is the win-win-win claimed. This seems like a weird detour to justify Opendoor (which, it should be noted, A16Z was an investor in).
Let me use this as an analogy though. Imagine the NBBO-equivalent for house sales was a 10% broker fee split between buyer's broker and seller's broker. What this post is arguing is that the sale of order flow is fine because most people pay less than 10% this way.
Thing is, once you decide to buy or sell that house, you're doing so without necessarily knowing what the price will be or how much commission you'll pay other than "probably less than 10%". Whoever is processing that is free to take that buy or sell order and then act before processing it. In financial markets, this is called frontrunning. Some markets allow it, some don't.
Worse, you as an investor have less visibility into what's going on because these trades through a variety of different brokerages or exchanges can essentially occur off-market.
So I'm not buying whatever this is selling.
Side note: I honestly don't understand the popularity of Robinhood. It's amazing how willing many on HN are to castigate ad-supported models and the sale of user data while seemingly at the same time being completely fine with the sale of their order flow.
Getting too deep into the stock market/NBBO analogy breaks down because houses aren't undifferentiated fungible assets and so the details don't line up.
I happen to disagree with the article, but I didn't dismiss it just because it was written by someone with an interest.
Quite a lot of opinion about things in the world is written by informed people who have an interest in one side or another. You can read critically and decide which parts make sense to you and which do not.
Another problem is that market makers tend not to provide liquidity when it's most needed; when markets are most volatile and unpredictable. The GME spreads for example were very wide. When the market is very volatile they tend to increase the spreads until the market is predictable (as far as they can be) again. Effectively they're like bad insurance companies: When you need them they're nowhere to be found, when you don't they'll gladly take your money.
It's for everyone involved, except ones that make money of PFOF plus the market makers, probably better to have transparent inefficiency in pricing, then have defuse pricing which in some cases might be in your favour.
- One can typically get a better price than what is displayed or quoted
- The price of liquidity is higher when markets are volatile
While yes, nobody would argue those things are good, can't they just assumed to be the nature of markets? There doesn't appear to be anything about the micro structure of modern capital markets where HFTs thrive (stocks, some options, FX, etc.) that makes these particularly bad.
Do they? Or would they typically allocate more to the market makers that pay them the largest kick-back ("payment for order flow")? (Not that that is inherently wrong, but it is not quite as benign for the consumer as what a16z portrays it there.)
If bridging the "time divide" is a profitable business, why is it not done by normal market participants?
Instead of becoming a "market maker", why don't these players just go on the open market, buy shares and hold them until "demand surfaces" and sell them?
The business model which a16z describes seems no different from what every investor does. Except that the "market makers" get an unfair advantage. I doubt that this is for the better of the market. It seems like the only difference it makes to an open market is that some money is sucked out by these players because they get leverage over the other market participants.
Speculation: Buying a limited edition or already antique car, and never using it, because you believe someone else in the long future will want it more because its rare.
Trading: Buying a car today to sell in the next weeks because you have better information/trading connections/higher volume than the original seller and can source a better price, or you have the distribution channels to move it cheaply and take advantage of smaller profit margins in other areas.
Riiiiiiight.
https://www.sec.gov/files/staff-report-equity-options-market...
>By the end of January 2021, some funds had closed out their short positions in meme stocks, realizing significant losses. 62 In contrast, some funds that were long GME saw significant gains.63 Some investors that had been invested in the target stocks prior to the market events benefitted unexpectedly from the price rises,64 while others, including quantitative and high-frequency hedge funds, joined the market rally to trade profitably.65 Staff believes that hedge funds broadly were not significantly affected by investments in GME and other meme stocks. Staff did not observe that any advisers to private funds and registered funds experienced liquidity issues or difficulties with counterparties.
Their sources for this are various news articles that were told this by hedge funds or anonymous sources. It is mentioned somewhere in the report that the SEC does not have the data or capability of detecting naked shorting.
All we can know is that the short positions definitely existed at one point, but that's about it.
>GME short interest hit 50% of shares outstanding first in 2012 and then again in 2015, 2016, and 2018, before rising even further in 2019. From then until early 2021, GME short interest hovered around 100%, hitting its high of 109.26% on December 31, 2020.
https://www.sec.gov/files/staff-report-equity-options-market...
> "Wholesale market makers typically provide price better than the National Best Bid/Offer"
Just because I have a mortgage and a credit card doesn't mean I can't survive without them.
Now that we have computers, most of us are actually and literally creating something out of nothing.
> Just because I have a mortgage and a credit card doesn't mean I can't survive without them.
You could definitely survey without them, the same way you could survive without a car or a smartphone. A mortgage in particular is undeniably useful for many people.
Finance is a bandaid for deeper issues.
Not sure whether you're asking seriously, but while there is a lot of scamming and rent seeking in finance (which is why it is subject to heavy regulation), it does provide many functions:
1. transfer of purchasing power through time (by saving/credit)
2. proper valuation
3. facilitation of investments (for companies, infrastructure, etc.)
4. investment allocation
5. risk transfer and management
etc. All valuable functions.
> most of us are actually and literally creating something out of nothing
Your labour and your computer and the electricity used to run that computer etc. are not nothing. The economy has always been about creating something more valuable out of inputs that are less valuable. Nothing new here.
Except that isn't true because banks create money when they advance credit and always have done.
There is no transfer. There is a liquidity provision that aims to offset the drain of liquidity to financial savings. The two don't necessarily match, and mostly they don't, with a system tendency towards net financial saving.
Purchasing power is improved by banks because they take things that are not liquid and provide a more liquid instrument in exchange. In that sense they are not that much different from market makers.