One consequence of increasing order internalization (because it's a good business, read the article) has been the remaining order flow on the public market is getting more toxic over time, which is part of the reason some (generally higher frequency) hedge fund strategies have done less and less well over the years.
I have only one (minor) issue with it. While it's true that the Citadels of the world have zero interest in using "a machine learning algorithm on your neighbor’s GM trades in their IRA to inform their thinking of the true value of GM as a business," I suspect there is value to them in being able to model and predict the aggregate behavior of a mass of retail investors, e.g., for use as an input to macro investing strategies. Retail investors have a reputation for being the last to join every party; there is value in knowing what they are doing in the aggregate.
(There’s a paper on this somewhere that I think I remember reading when doing level design for Starfighter.)
> If you could somehow make markets without informed traders—if you were guaranteed that Goldman could never trade with you—your risk (of being run over) would be lower and you could, therefore, provider tighter spreads. There is a way to do this; it is to trade exclusively with retail traders. They can’t run over you; they don’t have the informational edge or the capital to do so.
You collect, approximately, half the spread less your losses to adverse selection and hope to cover the (pretty formidably high at this point) essentially fixed costs of operating in the market by taking that small margin and parallelizing it over an _extremely_ high number of trades.
Robinhood solves a collective action problem for its users. Unbeknownst to some people, they are unsophisticated and cheap to provide liquidity to: market makers can trade with them and reliably offload their positions, reducing the risk premium necessary to make market making profitable. Robinhood advertises in a fashion that enables self-selection of unsophisticated buyers. Once the platform is sufficiently large, they can auction off the liquidity business. By regulation, the users can never be worse off than they would have otherwise been, because liquidity providers must run their bid-ask spread within the larger market's spread.
So Robinhood earns the difference in risk premium between the larger market and that of its user base, minus operating costs and profits of the liquidity provider.
I suppose the question is: what is the mechanism by which that profit margin can be transferred to users?
But TFA clearly demonstrates how they are. By removing retail investors the order flow to public market makers is much riskier. Those market makers will respond by increasing the spread to account for the increased risk. So yes, the retail investors get a price that is (very slightly) better than public price but the public price is dramatically worse than it otherwise would be. It's a net loss for the retail investor and public market maker while being a win for the internalized market makers.
Could we construct a reasonable scenario where the liquidity from segregated order flow would have displaced the core business of liquidity providers, because the volume is so large?
In other words, I don't understand if or why it is the case that HFTs would like to, in general, offer tighter spreads than looser ones.
Edit: is it just that volume decreases super linearly in relation to the price of liquidity?
how can one as a developer get into that field? I dream of front office positions at those places but the reality seems grim
Ninja edit: the business is very much based on personal connections and trust. Unless you have a stellar CV, it's your personal network that will get you to places. People who have worked together and have come to trust each other will tend to run into each other. E.g. you work for a PM who then moves on to another place, chances are they'll ask you to join them.
Out of all the industries, trading is the one where you rely most on others not to sink your career through ineptitude or malice. Hence the extreme emphasis on trust. If you are a PM, it doesn't matter if it's one of your traders and not you directly who fucks up, ultimately you are solely responsible for the PNL. Hence PM's are extremely picky who they let in their circle, but then if you are in good graces, the world is your oyster.
If you are the researcher/trader, you want the PM to be good and support you, because they are the primary multiplier of your performance. The PM's job is to devise the overall strategy and get money for your team. The more money the PM can secure, the more you can play with.
This results in tight knit teams who tend to stick together through jobs. In case someone in your inner circle is starting a new venture, they'll by far prefer to fill in the positions among the most trusted connections before even thinking of looking outside.
What should I begin studying to put myself in that position in finance?
The well kept secret of the industry is that complicated stuff rarely works, and if it does, people often don't exactly know why. Put differently, a good algorithm has obvious and intuitive failure cases. That is, it is (relatively) easy to reason under which market conditions will it work well and when it will not. Extremely complex blackboxes are tricky because it's not obvious when and why they stop working. They are also very easy to overfit. Every single additional parameter you add to a model is an additional degree that you are likely to overfit just because of survivorship bias.
It is for this reason that the absolute vast majority of systematic trading relies more on intimate market knowledge than very secret sauce and super complex ML strategies. Sure, many funds run some token ML crap (which is often quite vanilla, like gaussian processes or decision trees) with very minor impact on PNL, but this is purely for marketing reasons.
There is a small handful of companies who have the manpower, expertise and infrastructure to run more complicated models, but even there, the fundamental ideas tend to be relatively simple, and the complexity arises from the amount of data and market idiosyncrasies.
Aside from reading books about market segments and biographies of companies, what are some things you can do to gain "intimate market knowledge"?
Also, doing it on your own is tricky. You have to constantly talk and mingle with others or you quickly lose insight into current developments. Having a group of people to bounce ideas off of is incredibly important. On your own, every single small topic is a potential rabbit hole that you can spend endless time researching. In a crowd, someone has already explored that rabbit hole most likely and can give you practical pointers. At risk of sounding like a marketing prospectus, there's real synergy in being part of a team. I dare say the age of lone wolf traders is gone. They've been arbitraged away. Successful people have teams behind them.
So I'd say the best way to gain market knowledge is to join a systematic trading shop (in almost any technical capacity) and once inside, you just talk with the others. People in systematic firms tend to have a solid academic approach and are usually more than happy to teach and enlighten anyone interested.
What other name do they go by? Like what should I google if im looking for one in my city?
Is there no way for me to get a leg up on gaining that "intimate market knowledge" no insider books or reading/newsletters?
There are only a few big systematic funds. I think it's just Two Sigma, Jane Street, Citadel, Point 72/Cubist and AQR. Maybe a few others. You can apply to those yourself, but they're picky. Then you can look on LinkedIn for hedge funds looking for programmers, although it may not be obvious which are discretionary versus systematic.
Hedge funds are only in a few cities. About half are in NYC and Connecticut, then some are in Chicago, Boston, Houston and the Bay area. And maybe a few small ones elsewhere on the West coast, I'm not familiar with that area.
I'm considering trying to get into the field when I finish my Comp Sci PhD in a few months. I'm a little worried that my sole explicit financial qualification is that I really enjoyed a course on Black-Scholes towards the end of my undergraduate in Physics, although I probably would word that slightly differently on a CV.
The other thing you get with a typical brokerage that you don't get, as far as I know, with Robinhood is customer service.
(emphasis mine)
7bp is an average. The article also points out that the yield occasionally goes as high as 100% per year.
If there's some savings account or bond that pays you a bit, you can stick your cash there.
Might be easier to access via IB, which seems to be a more complete offering, but I haven't actually checked what you can access there.
Also, you don't get paid interest on uninvested cash balances. Cash in my Schwab checking and brokerage accounts automatically earns interest. Vanguard sweeps cash balances into one of their money market funds. But Robinhood pays nothing, and pockets the money instead.
Obviously nothing stops them from paying interest, but they did try that whole Federally insured deposits with interest stunt a while back before the commissioner stepped in to say 'na bro.'
IB is more international. But in the end you will need at least 3-4 Brokers for a decent international exposure.
- TOS (TD Ameritrade's ThinkOrSwim platform for those unfamiliar) has a much nicer platform but the commissions are so absurd that I just don't consider it.
- IB is cheap.
- IB is great for international exposure. I really don't think you do need another broker. What other broker is superior to IB in which markets?
- IB is amazing at handling multiple currencies. You can deposit and withdraw in any currency I've had need to care for and they're usually local accounts in the relevant country (e.g. USD in US, EU in EU, GBP in UK, CHF in Switzerland etc.). You can also easily and extremely cheaply convert between them.
- IB has an API you can use with low effort and zero cost. TDA does not.
This is a really good article.
The net interest thing has a parallel in the HF world. CTAs, a type of hedge fund, used to do substantially the same thing. The margin requirement for holding a load of futures is quite low, but the customers of the fund have put cash into the fund. So back when interest rates were higher, a fair bit of a CTA's returns would be interest on the cash.
He's right about Robinhood. From the HFT side, the problem is if you're in the "real" institutional market, someone will run you over occasionally. With retail, that never happens, and the spread you can offer is thus tighter. There's no front running to it, just the ease of non toxic flow. So RH is trying to put forward zero comms as a way to get some deposits.
He's also right about options. It's tempting to think that you can just sell some options to collect premium, or do a combination of some sort (iron condors, butterflies, etc) to save yourself from wasting money. But keep in mind the costs when you're doing this, and the fact that you are simply paying up to squish some risk into an unfamiliar shape. It also has a gambling-like tendency to create occasional wins to keep you interested.
As for investment advice for retail people, if you are a coder there doesn't seem to be any reason to use anyone other than IB. They are the cheapest, and if you can code you can use their APIs to trade whatever it is you want. I'm not saying anyone can make a great trading strategy, but you can certainly manage some passive investments with barely any fees. Or just take advantage of RH and do it free I suppose.