SEC set to propose rules that would squeeze stock-market middlemen
wsj.com
wsj.com
https://www.bloomberg.com/opinion/articles/2022-12-15/the-se...
Former equity derivatives trader here. The Levine article linked to is one of the denser accessible discussions on the topic I’ve seen. Also, as a current English speaker, the “truth searching” the comment you’re responding to cites refers to the general process of learning, and nothing specific to market structure.
That's the point, just read the Levine article instead of going through half-correct HN comments.
Also, I'm a native english speaker (and lived in 2.5 english speaking countries) and have never heard the phrase truth searching
Yes, it's annoying. From conspiracy theorists to supposedly smart people with PhD's, I'll catch them in a fiction. "You just made all of that up," I'll say, to which they reply with handwaving and equivocation. Why are we so reticent to be comfortable with our own ignorance and hold our tongue?
Too many people have their ego and sense of self worth wrapped up in being correct on the internet, regardless of the topic.
When nobody has a cite, we're just going to noodle and be wrong about stuff and that's fine, even salutary. But sometimes the truth-seeking we do is kind of superfluous.
(I'm just writing this so it doesn't sound like I was just snarking about HN.)
In my view they should also get rid of reg NMS, which entrenches HfT arbs and gives a permanent advantage to people who pay to colocate at the exchanges. Europe has shown that reg NMS is not needed and doesn’t benefit investors.
[1] noone should be crying for robinhood. Their entire business model is about disadvantaging small investors to benefit themselves and Ken Griffin.
interesting, i've not heard about this.
Which venues are you referring to that don't just do price time broker priority matching?
If you think about this variance since there is a party on either side of the trade, and one will benefit and one will lose out. The downside of reduced variance is you will never gain a “lucky fill” that is much better than you expect. But you will also never get a really bad fill either.
Since the whole purpose of hfts In market making is essentially to try to always be the ones yo get the lucky fills, they will on average do this and everyone else will on average lose out (from high execution cost variance). That’s why I said on net retail investors should benefit from this (I think).
I hope that explanation makes sense.
The playing field in this context is basically latency arbitrage. Who cares what hedge fund is collecting what alpha?
We live in the most golden age for retail trading. Retail electronic trading inside a Roth is basically perfection for the individual right now. I think people are confusing the market micro structure for the account returns in a bear market.
My first Scott Trade brokerage account years back cost $7 a trade per side Now I pay absolutely nothing. Every trade I make I think about how this probably won't last.
An individual has no problem getting near infinite liquidity on their limit orders with no transaction cost. If someone is getting clipped a tick on a market order, oh well. Don't use a market order.
Liquidity is a good, non-technological reason that we may not want trades to be immediate.
Consider how the price is found at market-open: buy and sell orders are batched into the opening auction, then the exchange finds the single price that results in the most matching orders. All execute at that price.
Contrast that with the binary, point-in-time matching that happens during the day. Here, the order book sits there waiting for someone to match an offer. It leads to perversities like "iceberg" orders (that disguise the true size of the order by posting only a small amount) and market-making firms that will really buy a little bit above the bid and sell a little bit below the ask but hide that for strategic reasons.
Trade settlement times might be silly, but that's a back-end implementation detail relevant mostly to day traders; it has little to do with the price discovery part of the market.
For instance, lets say I bought a bunch of stock in ACME for 10 dollars a share on say, Wednesday. Lets then, take for example, I notice there was a huge spike on Monday at 12 PM and suddenly, its 15 dollars a share for say, a few minutes, before going back down to 10, so I execute (or, perhaps smartly, have some automation on the account that auto-executes the sale if its at or above N price). Reasonably, I'd expect to get my 15 dollars a share because I sold within the correct window (you can see the timestamp!).
And yet, this isn't what happens. I remember, quite distinctly, being bitten due to the delay in settlements, where in fact, you end up right back in the 10 dollar a share sale because of the delay. This is why real time access for everyone matters, IMO, because in any other market, you get to buy or sell at the time of point agreed price, right?
Why should stocks be different? Why should only huge institutions be able to execute on point in time pricing?
Disclaimer: Its entirely possible that I'm missing something, but when I raised an issue with the broker (This was in the early 2010s, but I can't remember whom, exactly, I worked with) they pretty much made it clear it was because retail trades were cleared "in bulk" at the "agreed price" not the point in time price the sale was executed, if I recall correctly.
The problem with real time for everyone is that the tech required to do that is complicated and it's not necessary for most people.
Limit orders at the same price are filled in FIFO order, and it can be the case that that some limit orders fill while others do not simply because someone was only willing to buy 100 shares at that price and there were 200 available for sale - some sales won't happen.
You might be misinterpreting what the "listed price" actually is on a stock market. Prices that you see quoted are generally not prices that somebody is offering to sell at; they are the price that the last sale was made at. But that sale has already executed, so you cannot infer that there is someone else still willing to sell at that price. The only way to find out for sure what is available at what price is to place a buy order, and your buy order is not guaranteed to execute immediately because there might not be a matching sell order in the order book; it is not even guaranteed to execute as soon as there is a single matching sell order in the order book. It depends on the type of order and the exchange it is placed on and how orders and trades are reconciled. It's not like going into a store and taking an item to the cashier and buying it at the advertised price.
Not retail mistakes of course, but mistakes that might impact the people who count (unless your retail mistake impacts someone who counts, i.e. your broker who misfilled your order).
Over the past decade, many of the settlement times have been related to the feasibility of data retention capabilities, as opposed to the reality of it being all on top of a slower analog system for re-assigning shares and assets to different owners.
So in this decade I would say its over. 0 to 1 day settlement time.
Something the US could do to improve its situation would be to change from end of day novation to novation within five seconds of a trade. This might reduce the amount of capital that firms had to post.
US Cash equities already settle in T+2. Have for some time now. IT did used to be T+3
Why is this a good thing? There are consequences to actions, lets just live with that. If you mistakenly order something you didn't mean to the solution isn't to "catch the error" and prevent a transaction from happening, the solution is to make an new entry/transaction that reverses the previous one. If you end up having to eat some cost to reverse the "mistake" oh well, lesson (hopefully) learned you will be more careful next time.
There should be no settling, no clearing, every transaction should be immediate and final.
This would open up the possibility to not have any arbitrary market "open" and "close" times. There should be no need to settle up, clear, reconcile for the day/week etc. All transactions are immediate and final and the market can run 24/7 365.
I hear they're trying this idea on something called the "Blockchain" and that it's all going really well.
That is not what clearing and settlement are about. They are about liquidity and stability. In countries with well designed systems, equities settlement is an atomic action.
So, for a transaction involving a UK stock, the Bank of England moves money from one party to the other, and stock from the second party to the first as an atomic transaction.
Trading flows get complex. Some illustrations,
1. If your balance sheet is smaller than a global investment bank, you are regarded as higher risk to deal with. A clearing house won't onboard you on your own account. Instead, you have to get a kind of coverage from a major player who the clearing house does respect, at a cost to you, to clear against their balance sheet. The exchange membership might be yours, and the trades are yours, but for clearing and settlement you are represented by an IB who has a desk that specialises in covering trading risk for scrappy trading firms for a tidy profit.
2. Most participants in markets are not direct members - e.g. hedge funds and mutual funds tend not to be. Non-member access the market via a broker who is a market. Often a broker will outsource activity for some markets to another participant who is more effective at accessing a particular market.
You will have noticed that these arrangements create long chains of inter-dependencies. If there was only trading and settlement, a small error at one bank would cause a cascade effect that would affect hundreds of other participants and disrupt atomic settlement. The system would quickly and regularly stall. So between trading and settlement there is a round of "clearing". Here, everyone confirms that the details are right ahead of settlement. If a participant failed to send a transaction from their trading engine to their clearing system, that would be revealed. If a market maker closed the previous day with a short position they weren't supposed to have, they would get assistance to patch that up. There is lots of forgiveness and room for correction within the clearing round, but big-stick consequences if you don’t come to the right conclusions before it has closed. In this way, when the Bank of England goes to move positions, everyone ends up with what should have.
This three-tier trading/clearing/settlement model allows entities to focus on types of risk that they are good at managing. Without them, the exchange would have to do it all (this is not realistic), there could be no activity between jurisdictions (e.g. french bank trading on swiss market) and it would be far harder to bring liquidity to market.
A system with fewer layers would be far riskier. The clearing houses act as a buffer if there is an insolvent participant. According to legend, when Lehmans was in collapse, the risk process at London Clearing House kicked in to unwind their positions and made a profit on that unwind. This mechanism reduces the possibility for contagion.
If a country tried to build a system that had less liquidity or worse stability, companies would move from that jurisdiction to another that suited them better. I am not sure of the details but understand Sweden blew up its finance sector through vindictive/populist legislation in the late nineties. Activity moved to London and the world moved on.
Clearing needs to be a day because if there are misalignments, you will need analysis and developers to have time to review what happened on the trading day, and have phone calls to discuss with teams at their other partners until a problem is uncovered. If this failed, the senior management would get involved on the T+1 evening. Hence, T+2 settlement is the optimal.
24/7 markets. Someone who wanted to build a trading desk would need to ensure coverage from risk and compliance officers, and to be on top of market events. With a traditional open/close market, it is feasible to do this with a single team. With a 24/7 market, you can't so you either miss liquidity, or are forced to put together an expensive, complicated global team for no increase in total trading opportunity. And on the exchange side, 24/7 makes the software far more complicated for no benefit beyond the bragging rights.
Well good news here for you, trades settle in T+2 and have for some time.
> That and let's also get rid of any special treatment for the investment industry that retail traders don't have
Well that is limited by only the deal you and your broker have. You can send all the complex order types that funds do, you are limited only by your broker.
Trading pre/post market still happens on exchanges so any delay is related to your broker. I can trade in "realtime" from my personal brokerage account with no issue or delay.
T + 2 is WAY too long. I click a button to buy/sell, money goes to one party, the stock goes to the other party, transaction is final and complete within milliseconds. That is how it should be.
2) What difference does it make how long it takes trades to settle? I tell my broker what I want. They have it show up in my account. Any 3 days for settlement is handled at no cost to me.
There's also a whole clearing industry that you're funding with your fees even though more efficient technologies could mostly eliminate it.
Like yeah, it's fun going to fancy restaurants with your salesperson from the clearing firm, but the only reason they can do that is due to massive amounts of rent-seeking. The more I work in finance, the more I think any high touch sales activity is an indication of a broken/corrupt market.
I'm also interested in this idea that large investment firms can trade out of hours - this has never been the case in any market that i've seen. There are some 24/7 markets though (CME is like this I believe?) so maybe there is a disconnect between the hours offered to customers vs the exchange hours? I suppose the other possibility is that the broker trades the opening auction and doesn't offer this to their customers, but i'd have thought they would simply bundle everything into the auction from overnight.
Back in the day, the UK settled twice per month, with the various brokers having tracked all of the trades for the fortnight then the accounts departments moving the difference once they had agreed the actual total (which of course would never match with a word of mouth + paper based system). It's amazing it worked so well for so long, all basically based on trust and hence why being an exchange member was so important for trading.
The move to a rolling settlement date part of the Taurus/Talisman project in the 1980s. This got cancelled, lots of IT companies had invested heavily to be ready for this got burnt, but what emerged was the less extensive CREST system which is still in use today.
The original move was to a T+5 rolling settlement, which moved to T+3 in the early 2000s and is now T+2.
No trades are guaranteed because you have to find someone willing to take the other side of the trade. This applies pre or post market. Some brokerages are happy to let you try and trade pre and post market, but that still doesn't mean you can find a buyer or seller to complete your trade, much less trades happen, so it's much harder to fill them. Large investment firms don't really get special treatment here, like you seem to think.
My personal opinion is that this change will have lots of unintended consequences and it may even cause a shuffling of who are the big players. But it likely won’t mean the end of HFT microstructure trades. At least I hope not because I believe that would lead to higher spreads and less liquidity.
I don't think Europe has shown this, or should be seen as an role-model for execution quality. The European model benefits national champion exchanges and lazy brokers.
In Europe, all retail flow is forced to Regulated Markets where it directly faces more sophisticated players. There is nothing equivalent to the execution-quality guarantees that RegNMS delivers in the US.
RegNMS ensures a high quality of protection for the interests of retail investors, and is even-handed towards exchanges.
I gather that this arbitrage made large amounts of money a penny or so at a time. Which is to say, your own trades might net small numbers of dollars more -- money you're entitled to, to be sure, but not exactly changing the world.
I could be wrong about that, and either way I've got no love for middlemen providing nothing, not even liquidity. So I'm trying to get a handle on how big a change this actually is for everybody except the high-frequency traders.
> This proposal would address a controversial practice called payment for order flow, in which some brokers collect rebates for sending customers’ orders to wholesalers. Mr. Gensler has called the practice a conflict of interest and, in past statements, left open the possibility of banning it. The SEC’s best-execution proposal doesn’t go that far. But it imposes additional obligations on brokers that engage in payment for order flow to help ensure that customers are getting a good deal.
Shame. They’re skimming money straight off the top. Perhaps I should withhold judgment but I suspect they’re banking on the idea that Joe Schmoe, who downloaded an app to gamble with his future, isn’t going to seek out this financial disclosure data. I have a friend who can rattle off everything there is to know about option spreads and general technical analysis drivel. He initially didn’t believe me when I told him about PFOF.
After reading the article, my take is less harsh. It is really bizarre to me why instead of outright banning it, they just want to make it a really bad deal to do so. But if it works, great! We’ll see.
The tick size changes are interesting as well in the context of auctions. If a claim is "you can't have an auction with sub-penny pricing so there's no improvement", well then make the tick size smaller.
That’s what PFOF has brought to retail investors. It’s not like retail investors was ever put onto the public exchanges prior to robinhood and PFOF, they were sent to dark pools for institutional investors to trade against but retail investor never got the benefit.
The payments per share tend to be extremely small - on most symbols the broker isn't making $5-10 per trade from the MM, for reasonably sized trades.
Even if payment for flow was costing you that much, paying for an order vs giving more price improvement are indistinguishable. Payment for order flow just means that price improvement definitely goes to the broker instead of back to you.
> The customers think they have no cause for complaint, because they did better than the NBBO. The whole thing is an institutional equivalent of retail payment for order flow: The partner firms fill the customers at a better price than the NBBO, make some profit for themselves, and kick back some of it to the broker (Coda).
https://www.bloomberg.com/news/newsletters/2021-09-23/money-...
Are you saying that if makers' profitability at lit venues was greatly improved, they wouldn't compete to offer tighter spreads at all?
The main point of PFOF is that high frequency traders have much less risk when trading with single persons than they do trading with big players. They are willing to offer them much better prices than they will offer big players. Suppose there is 10cents difference. PFOF means they pay your broker, say 5c to give you a 5c better price. You are better off, your broker is better off, the high frequency traders are better off. The only people worse off are the big players. Because when HFTs can filter out the low risk trades, the high risk trades get more expensive for them.
Research has been done showing that different brokers split the price advantage differently. Robin hood, IIRC was one the worse side, allocating 80% of price advantage to PFOF, leaving only 20% to the customer. But the customer is still better off than paying the public exchange rate.
-----------------------------
The above was written before I found this report: https://www.afm.nl/~/profmedia/files/nieuws/2022/afm-paper-a...
I found it trying to find the source for the research I quote. That stat came from Matt Levine's newsletter. This report contradicts what I wrote above. It does seem that the report is about european brokers, whilst Matt Levine was talking about US brokers.
edit: I believe this is the orginal source regarding US markets: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4189239
But why should Robinhood or whoever else get part of the 10 cents that the algo is offering me, the little ol' retail investor? Just because they're in position to gobble some of it up without me knowing?
I don’t use Robinhood, but I don’t see anything wrong with that pricing model. Prohibiting it would be equivalent to banning no-fee trades for small time retail investors. It would be like banning ad-supported business models. Well, on second thought…
Regulations also mean that you have to get NBBO or better so your fill price is never worse than the market
"Payment for order flow sort of created the zero-commission retail-brokerage model, but it’s not really necessary anymore. You can run a profitable retail brokerage on net interest margin, without charging for trades at all."
So what everyone should be complaining about is not PFOF, but "net interest margin". In other words everyone should be complaining about the brokers not paying enough interest on the cash balances in the brokerage accounts.
I transfer cash in once I've decided to trade and the brokerage usually holds it for a max of a day or two until the trade actually executes (say I am buying an index fund where orders today are executed tomorrow). The only time there has been cash for longer than that was when I know I needed it (a week or two before buying my first place) so I didn't want to risk a stock market crash etc.
Some people use a brokerage account as a checking account. Most don’t. Interest can also be earned on lent securities.
I'm not sold that this move will actually result in much better prices for customers on average, but it will certainly "spread the wealth around" by allowing all brokerages to take advantage of a uniform price-improvement structure rather than negotiating individually with Citadel/Virtu. I have been an IBKR customer for a while, and I'll be happy because I don't get a lot of price improvement under the current system. Schwab customers, who get a lot of price improvement, may be less happy.
In options, there is the suspicion that the presence of a lot of auctions probably results in wider spreads, which we could see as auctions make it to the stock market, too. If all retail order flow goes into the market through special channels, the only trading remaining in the "normal" system will be from sophisticated traders, who you don't really want to trade against.
Of course, the SEC may then say "the experiment was a great success, retail traders are getting HUGE price improvement," ignoring that retail traders may be getting worse prices than the old NBBO from the widening of the spread.
The complexity and fragmentation as well - you can preferentially do auctions at venues you have an advantage on, and bully the venues away from fixing whatever you exploit by threatening to take away flow.
> In options, there is the suspicion that the presence of a lot of auctions probably results in wider spreads, which we could see as auctions make it to the stock market, too. If all retail order flow goes into the market through special channels, the only trading remaining in the "normal" system will be from sophisticated traders, who you don't really want to trade against.
Effectively all retail flow already goes through even more cutoff special channels? But I agree with the point, this seems like adding a tremendously complex game instead of just shrinking tick sizes. At least in options, you tend to have much wider spreads. But the fact that plenty of stocks trade at min tick is a sign the spreads could naturally be tighter.
This is true, except the key factor is that the price signal eventually reaches the "normal" market when you look at the trading patterns of the wholesalers, mixing it in with the price signals coming from the trades that the wholesaler does for other clients (including their own strategies).
Moving to an entirely separate public information channel means that 100% of the trading activity on the markets (aside from the separate channel) is driven by sophisticated parties. That makes a big difference - for example, during the GME fiasco, there was uncertainty about how much of the activity was retail vs institutional traffic. After the proposed rule, we will know exactly how much of that flow is retail.
Sounds like a sensible change to me.
https://deliverypdf.ssrn.com/delivery.php?ID=285088095002029...
Whether the total amount of price improvement in the market increases still does seem to be up for debate - I personally doubt that it will be much better for average retail investors than the old system.
While Citadel and Virtu are very large market makers, there are other equally large firms that don't jump through all the hoops to participate in wholesale, but would probably do on-exchange auctions.
The SEC also talks about non market makers using this for execution, but I don't think that's going to happen immediately...
1.Imagine a Robinhood user places an order to sell 100 $GME
2.Robinhood doesn't send it to an exchange first, it holds onto it for a few milliseconds.
3.Meanwhile, Citadel/Virtu or others execute orders below what the bid would have fetched a few milliseconds ago.
4.Now, the order is routed to the exchange above what the market was trading at and thus certainly goes unfilled.
5.300 milliseconds per the regulation elapses and the order comes back to the HFT firm to fill.
The bigger issue here is the nepotism in providing wholesale prices to HFT firms. The stock exchanges do this and so does Robinhood. In an open and competitive market, the playing ground should be regulated to be equal for all.
Or infact, as motorsports participants know very well, the cost of access to markets should INCREASE with size, not decrease. . . If you're the present Formula1 team winner, you pay SIGNIFICANTLY more to enter next years championship than the last place team.
It has a revenue disaggregation on page 17. Just 15% of its transactional revenue last quarter comes from stock.
> ...the company has quietly been laying the groundwork to become a standalone market maker,
> One way to reduce its reliance on PFOF is for Robinhood to actually match its own order flow instead of selling it to other brokers, company executives said. They even hinted during the roadshow meetings with investors that they too will be looking at getting into the business…
https://www.foxbusiness.com/markets/robinhood-ipo-twist-busi...
Am I right in understanding that Robinhood took this and abused it by turning itself and its customers as the source for high frequency traders?
"Robinhood makes money in a number of ways, notably through a system known as payment for order flow. That is, Robinhood routes its users' orders through a market maker who actually makes the trades and compensates Robinhood for the business at a rate of a fraction of a cent per share." [0]
The market maker is usually a big player like Citadel. And with this information, they can front-run the market.
[0] https://www.investopedia.com/articles/active-trading/020515/...
I'm sure these nice American businesses are complying 100% with this rule.
OTOH, I'm not sure that it's easy/possible to really know what the 'NBBO' is, at any given 'moment'.
Pretty much all conspiratorial, "motivational" analysis of PFOF critics on HN are simply wrong. As Matt Levine explains it clearly, from the utilitarian perspective, PFOF is not a bad thing at all for retail investors. Whether it can be even better is a different question and clearly SEC is trying to find that with the new proposal, but as is always the case with a complex system with differently motivated actors, what exactly the new balance point for the new system will be is uncertain.
If someone can add that as a comment to the rules to the SEC that would be fantastic.
> The SEC charged the company with failing to inform customers about payments it received from trading firms to route customer orders through them, a move that resulted in customers paying higher prices to execute trades.
> Robinhood customers' orders were executed at prices that were inferior to other brokers' prices," the SEC found.
https://www.reuters.com/business/finance/robinhood-pay-65-ml...
Completely eliminate the idea of high frequency frontrunning.
It’s no different than telling your neighbor you’re willing to sell your house for $X without publicly listing it.
I imagine this is part of the meta-point you're making across the thread that people on HN talk out of their hat about finance (I plead guilty) but I feel like your comment about Money Stuff/"truth searching" was enough, and that asking people to defend their misuse of terminology is punching down.
EDIT: I suppose we can scratch the "finance-related" bit
Moreover, the PFOF story depends on 'price improvement'. If public stock markets aren't accurate, then 'price improvement' is a meaningless concept, because it compares the price to the public stock market price.
> It’s no different than telling your neighbor you’re willing to sell your house for $X without publicly listing it.
The difference is that the most accurate price for 'my house' is not of significant public value. Whilst the most accurate price for stocks is quite valuable to the wider world.
I don’t think there is an equivalent public reporting mechanism for private “dark pool” transactions of financial instruments.
Another way to say “without impacting the market” is “denying the market information it needs to set prices accurately.”
and have to be reported within 10s of being fully filled
I agree with this, as an individual investor, from my perspective the most likely thing that happens is that the SEC screws things up.
I started out paying what in today's money would be $50 a trade to buy or sell a stock. So for a $2000 trade, that's going to mean starting out down 5% on trading costs ($50 buy + $50 sell). Now I pay nothing.
Does it matter if I pay $50.01 or $50.005 per share on a trade? No, it does not. Leave it alone and let the brokerages compete, they're not monopolies. I much prefer the current system to even a small risk of having to go back to commissions.
Just ask the humans who trade with alpaca and robinhood.
Robinhood had citadel. Alpaca has the Japanese counterparty... where it’s legal to front run.