8 ms·
This article isn't exactly convincing. The SEC identified $34mm in poor execution. If you did 65c each way per option contract & free stock trades that would on
by rwmurrayVT 6y ago
This article isn't exactly convincing. The SEC identified $34mm in poor execution. If you did 65c each way per option contract & free stock trades that would only be 52.3m option trades. That's miniscule. That's over 5 years. It would take a whole lot of unidentified trades for that to be substantial.
The spread on most trades is probably in the range of 50c on low liquidity shares and $1 on low liquidity options.
Perhaps the reason RH collects more in revenue from order flow is because they:
1) get paid more for their order flow on a per trade basis
2) they don't collect any fees from options driving down revenue
3) they charge less for margin fees with less users using margin
There are any number of reasons, but to say that you're getting royally screwed is a massive stretch in my mind. Prior to RH fees were outrageous. The fees are still outrageous on my self-directed 401k. I'm not getting $10 in losses from bad execution.
- ceejayoz 6y agoThe article mentions that may just be the amount the SEC can actually prove. > Overall, Robinhood’s investors may not have been harmed by their decision. The SEC identified $34 million that Robinhood customers lost by not receiving best execution, which is much less than the money they saved by forgoing trading commissions. (That said, their losses might also be much higher. That $34 million comes only from “certain” transactions, which might be the tip of the iceberg that the SEC possessed the ability to measure.) They were, however, duped.
- elliekelly 6y agoI think the author has misunderstood the SEC Order. The article says (emphasis added): > The SEC identified $34 million that Robinhood customers lost by not receiving best execution, which is much less than the money they saved by forgoing trading commissions. But paragraph 42 of the Order[1] says (emphasis added): > ... certain Robinhood orders lost a total of approximately $34.1 million in price improvement compared to the price improvement they would have received had they been placed at competing retail broker-dealers, even after netting the approximately $5 per-order commission costs those broker-dealers were charging at the time. [1]https://www.sec.gov/litigation/admin/2020/33-10906.pdf https://www.sec.gov/litigation/admin/2020/33-10906.pdf
- tyrust 6y agoI came here to post the same. The author of this article really missed the mark with this conclusion.
- matthewdgreen 6y agoAnytime you move fees from a regime where they’re fully visible to the customer into one where they’re hidden, the customer is eventually going to get fleeced. It may not happen instantly, but it’s inevitable. You’re simply removing one of the only competitive metrics that customers have to choose between brokerage firms and replacing it with something opaque. (The fact that Robinhood even exists is proof that explicit-fee brokerage models produce competition from lower-fee competitors.) Unfortunately, the degree to which customers are getting fleeced is hard to know. Which, let me repeat, is exactly why this model exists. The article mentions it’s at least $34M but even the SEC seems unable to calculate the exact amount. And if they can’t determine it, then how in the world is the typical customer supposed to? At the end of the day, the degree to which customers are getting fleeced is probably exactly proportional to Robinhood’s revenue from order flow. If RH is receiving vastly less revenue-per-customer from that source than explicit-fee exchanges make from fees and order flow, then customers are getting a possible bargain. If RH is doing as well or even better with this model than competitors, their customers are probably getting screwed and don’t even realize it. (And if by some miracle RH isn’t abusing this, then other brokerages will abuse it and try to outcompete RH.)
- gandutraveler 6y agoI agree that it needs to be more transparent but RH isnt the only one hiding the back office deals. There needs to be a larger inquiry into all the trading platforms like Fidelity, Schwab. On the other hand price improvement is so hard to enforce. I'm not sure how SEC can even regulate it.
- hinkley 6y ago> then other brokerages will abuse it and try to outcompete RH. If you make higher margins but don’t lower your prices, the only other way to outcompete is to offer a broader array of services. Otherwise “outcompete” just turns into “save up more money and buy them out”.
- rlucas 6y agoThere are an awful lot of businesses that charge huge margins and use that extra cash flow, plus their scale, to essentially fence out new entrants by a mix of sales/marketing, channel spiffery/bribery, and regulatory capture. It's often a fairly "bad for you"/"predatory" product, too, but the problem is that it's very hard to create a new entrant doing it the "good for you" way, because all that extra margin the bad guys make can be used to squeeze you out of the market. Legacy industry examples would include whole life insurance, title insurance, payday lending, and whatever it was that "Dun & Bradstreet Credibility Corp" was selling (not the historical D&B) a few years back. I consider it enough of a pattern of "bad money chases out good" that I identify it regularly among pitches I receive.
- janmo 6y agoI am not a lawyer, but the problem with this $34mm is that seems to be illegal. Quote: "In the United States, accepting payment for order flow is only allowed if no other trading venue is quoting a better price on the National Market System"
- jkhdigital 6y agoRight... so everyone quotes the same price, and then they compete based on paying more for order flow instead of the sticker price. It’s the difference between having a larger bid-ask spread or larger transaction fees: either way, the cost of execution will be paid.
- tptacek 6y agoRetail order flow directed to electronic market makers obtains better prices than the NBBO. That's what "price improvement" refers to. The entire controversy here is about how much better than the NBBO Robinhood's prices were.
- pliny 6y agoDon't have much to say about the rest of the comment but >The spread on most trades is probably in the range of 50c on low liquidity shares and $1 on low liquidity options. Is demonstrably wrong. The minimum spread on options contracts is $1 and the spread on liquid options contract (not SPY liquid but like TSLA or AAPL) is generally in the mid $10s and often hits $100+ on ATM contracts). Right now an ATM call on AAPL a month out quotes 4.75 to 4.85 which is $10 and the same on TSLA quotes 44.05 to 44.40 which is $35. Not a trivial amount, and it implies that the broker could get price improvement that is an order of magnitude more than the typical commission on options (~1$ per contract).
- rwmurrayVT 6y agoRH sets an initial default value for the limit order at the middle point between the bid/ask. You are in control of your limit order pricing there. If you wish to join the queue at the bid you can. There is no option for a market order. You can leave your bid and wait to be filled same as others.
- siruncledrew 6y agoEven prior to the potential news of RH routing orders to certain market makers for greater margins, I think any serious trader looking to optimize orders would have not been using RH to begin with. The feature set in RH is straight up heavily diluted compared to other brokers like IB or TD (TOS). There's no way to set exchange parameters in RH, users can't edit the charts for TA, and the functionality is heavily reduced on purpose for the sake of not overloading users that have basically no clue what they are looking at anyways. Not that any of this excuses RH. Given RH's user base, some random person trading PLTR and GME calls probably is not even looking into any of this stuff to begin with.