Getting rid of commission was actually probably a bad idea. It creates an expectation that a very complicated service should be free, and puts a lot of pressure on now free brokerages to find a new revenue source. There's a reason why RH is accused of helping Citadel front run their own customers.
This seems to me a lot like when VC backed startups artificially suppress prices below break even to push out competitors, with the intent to raise prices even more later. It's good for the consumer while they can get VC subsidized goods and services, then very bad for the consumer once the VC funds dry up.
> There's a reason why RH is accused of helping Citadel front run their own customers.
Are there serious accusations of front-running (a serious financial crime) and collusion to facilitate that crime? Or are you referring to people with a poor understanding of the interactions here making complaints that really boil down to privacy concerns, not actual front-running?
I haven't seen anyone post any evidence of actual front-running.
For why Citadel would pay for retail flow, if they're not committing financial crimes, market making is essentially profiting from separating pricing signal ("alpha") from pricing noise. If the prices are jumping around randomly, you make money by holding prices steady against that noise. If the prices are moving in one direction because new information ("alpha") has become available, you lose money if you try and hold prices steady against those moves. Empirically, in aggregate, retail flow has a lower signal-to-noise ratio (lower alpha) than the market as a whole, so market making on just that flow is more profitable, even though Reg NMS[0] requires Citadel to give RH customers' round-lot orders prices at least one cent per share better than available in the open market (price improvement over NBBO, combined with no sub-penny pricing).
By paying to exclusively trade against a low-alpha channel and damp out some of that noise before it affects the market as a whole, Citadel makes more money, at the expense of other market makers who would normally have exposure to that noise.
Also, there's a bias in execution called adverse selection: bad trades tend to get filled faster than good trades. For their non-market-making strategies, crossing against low-alpha flow has less adverse selection than trading those strategies on the public markets. Here too, RH customers trading round lots get prices at least 1 cent better per share than NBBO, unless the orders are passed through and placed on a regular public venue.
Disclosure: I work in financial services, but have never worked for Citadel.
This seems to me a lot like when VC backed startups artificially suppress prices below break even to push out competitors, with the intent to raise prices even more later. It's good for the consumer while they can get VC subsidized goods and services, then very bad for the consumer once the VC funds dry up.