But that’s not enough - the fixed and variable costs to providing even decent execution are staggering.

I think the second bit, margin lending and API access, might be them selling sheeps’ clothes to the wolves. As a sophisticated trader, it may be of value for me to mix my flow with that of unsophisticated retail traders. It’s an interesting model and a careful balancing act. As a market maker I would keep a close eye on the information content of their trades and cut them off the moment they started looking too sensible. Compounding the problem would be that the informed traders will tend to dominate the un-informed, in terms of volume.

My alternate hypothesis is they're going to offer commission-free trading in a limited pool of symbols - you should be able to get sufficiently reliable retail flow for, say, AAPL to be able to internally cross most orders.

P.S. Not sure using FINRA affiliation to vouch for your security credentials is compliant

So, now I'm unsophisticated because I would like to own some individual stocks? You assume the execution is all that matters, but if you are investing in the long-term profitability of a company, then I certainly hope a few hours or even a day doesn't matter. But a $10 commission on a $1000 trade starts you at 1% in the hole.

Think of it this way: if you had to take the other side of a thousand random trades by individuals putting down a thousand dollars and one random trade by an individual putting down a million dollars, which one would you prefer? More to the point, if I gave you a dollar for every percentage point a random thousand-dollar trade outperformed the S&P 500 and every percentage point a random million-dollar trade outperformed the S&P 500, over the long run, which one would you rather take? Assume a liquid stock and so no significant market impact by either trade.

[1] http://www.investopedia.com/terms/s/sophisticatedinvestor.as...

My understanding is that the NBBO is, by definition, the best executable price. If you're able to execute for a better price then the price you paid was the NBBO.

Let's say the market is 147.05 at 148.30 (I, as the market maker, am willing to buy at 147.05 and sell at 148.30). That's the NBBO. If someone IMs me and says they're willing to buy at 147.10, that moved the market and thus the NBBO. I am required, by law, to report that trade to the consolidated tape.

But if someone IMs me and says they're willing to buy at 147, that's worse than the NBBO. I would then consider lifting at 147 and selling at 147.05. Why would someone buy at worse than the market? Maybe the market was 147 when they put in the order and their computers were slow to catch up. Or maybe they're George Soros placing a multi-billion dollar bet and know that if they put an order into the NYSE it's going to execute at much worse than 147. Either way, I functioned as a market maker by taking their problem and making it mine.

Also note, he said he was acting as a MARKET MAKER.

E.g. let's look at where a particular AAPL option closed on Friday:

AAPL131221C00550000 9.65 bid 9.85 ask

Another, bigger difference because it's further in the money:

AAPL131221C00500000 54.15 bid 55.20 ask

Another, bigger difference because it's further out in time:

AAPL150117C00550000 bid 65.50 ask 66.40

The market maker buys at bid, sells at ask. If he can do that thousands of times a day in a thousand different securities, the profits start to add up.

Doing this in options means the market maker needs to be quite nimble, even more than with stocks. This is because he's buying or selling a DERIVATIVE and the UNDERLYING security can change in price quite quickly.

But that's why these guys buy fast computers and why they pay smart people the big bucks to write software for those computers (and to e.g. program FPGAs for high speed trading).

It's not as easy to make money being a options marketmaker because you have to realize not all of your trades will be someone taking your bid/ask. Even so, you'll have delta-hedge your position with the underlying which is a imperfect hedge that is subject to jump risk. Typically a AAPL options marketmaker is taking orders from traders and has up to couple hundred of these AAPL with long name positions, and they have to balance out the greeks, delta and gamma. They may even have to buy some options themselves to hedge themselves and pay the market price.

1) If they're FINRA regulated as a brokerage, then it should be enough.

2) Your point on uninformed flow is very very interesting here. I wonder if having a marketmaking group on the other side of a chinese firewall that has flow rights justifies this model by itself.

With IAB, a small player (5k+ in IRA or 10k+ in normal account) can trade on terms comparable to what a mid-sized fund can get.

When it comes to financial assets, I'd rather pay an established player $1 per trade than a fly-by-night operation $0.

At my last check it was mostly geared towards C++/Java with less than ideal documentation. However there are some good open source wrappers available (Python): https://github.com/blampe/IbPy

As stated in the parent comment, the minimum account balance is $10k.

Although (I had a 404), I think it's actually: https://github.com/InteractiveBrokers/tws-api-public

Indeed the PDT requirement is a pain...

https://github.com/ib-ruby/ib-ruby

Specifically for options trading, they have incredibly low commissions, and instant execution. I was always impressed by their SDK and tools they exposed. I used the Java toolkit, I never made anything substantial in it though. They are extremely security focused as well.

Also good for options trading too.

Yeah, that's understating it. Here's the interest rate on a $100,000 margin loan:

   Interactive Brokers 1.09%
   Schwab              6.875%
   TD Ameritrade       7.25%
   Fidelity            6.575%
   Scottrade           6.50%
   Etrade              6.14%

     not that one should ever go on margin
     if you don't know what you are doing

   It was announced on October 10, 2008,
   that McClendon sold approximately 33.4
   million shares, approximately 90% of his
   stock in Chesapeake Energy (stock symbol CHK),
   for $16.52 per share to meet a margin call
   after the drop in the U.S. stock market
   that week. The stock had been worth as much
   as $74.00 per share in the year prior to the
   sale, a loss of nearly $1.92 billion.

For those wondering why they can offer $0 commission, it's because retail brokers actually make money on every trade you make. When you send your $10 through etrade or some other brokers, they don't actually send your order to the market, they sell it to getco or some other market maker.

The reasons are complex and linked to NMS regulation. In short, there are two types of traders, informed and uninformed. Market makers do not want to trade with informed traders, because they tend to lose money on those trades. However, reg NMS mandates that every one gets the same best bid and offer. So the spreads you'll see on the market reflect conditions where adverse selection is anticipated, which is why it's very profitable to trade at those prices against uninformed flow.

Now why don't brokers offer to pay rebates (less than $0) to customers? Well they're not legally allowed to do that. The SEC regulates the maximum rebate and it's not that big.

So why does Ameritrade for instance will charge you for trades? They don't execute anything after all... Well, they're just the middle men. What you're paying for is their advertising campaign that got you to open an account in the first place.

Maybe you mean the marginal value of a trade is less than $0, but I'm not convinced that's true ether.

However, AFIAK the rest of the information in this post is accurate.

The fact that they don't list exchange fees seems to suggest that they may end up routing orders to a market maker and collect fees from that (the exchange fees are generally 30 cents per 100-share lots). If so, they should note on the website that they are making money by selling flow

According to their site, they claim that you'll "buy and sell at the best possible price"; I'm not sure how they are able to make this claim if they cannot route your orders directly to the exchanges.

On their mockup on their site, the orders screen has an "Estimated Price" which is a strange field as typical order types should be either "Limit" or "Market" and should show the bid/ask spread on the current security; so I'm guessing that the best "possible" price is the "market price" and they are doing a combination of

1) Internal matching; orders aren't being routed to exchanges but if are internally matchable, just reconcile the books internally and they collect the bid/ask spread on both side of the trade.

2) Payment for order flow: they route your orders first to Citadel and Timberhill where marketmakers love to take the other side of retail order flow.

3) Internal marketmaking: they'll take the other side of your trades themselves and try to immediately close their negative positions by routing the order to an exchange for rebates and a better price.

My guess: they already advertise that they are constantly getting market prices from multiple sources - so they simply run a Best Price Execution algorithm for every trade and use the current state of their (locally cached) total market state. So a trade may get matched up against multiple bids, where a single bid may not be enough to satisfy the full order amount. So they apply BPE (think of it as a greedy algorithm) and find the next best available price to complete the trade. Iterate until the entire trade is matched and that's it.

Disclosure: I work for a betting exchange. We do BPE.

Doesn't that violate national best bid? Also, a lot of retail dumps market orders, so how do you determine the price to match buy-sell pairs at?

Retail trader Bob wants to buy 100 shares of MSFT at the market price Retail trader Sue wants to sell 100 shares of MSFT at the market price

NBBO of MSFT is at 36.70/36.72

You fill Bob at 36.72; and you fill Sue at 36.70. So Sue's account is credited with 3670 cash while Bob's account is debited by 3672 cash for the privilege of owning 100 shares of MSFT; while you collect the $2 difference and get RegNMS compliance.

Now, on the other hand, if you're Getco, you get to see these orders from every broker in the country. If you can trade against that, you can make money even with a bid/ask spread of $.02.

So if you're Getco, (hypothetically, I have no direct knowledge of them, except that they're big in market making) you can go to the retail broker and offer them price improvement.

E.g. in return for the order flow from that broker, Getco fills them "inside" NBBO. Getco bids 36.7001, Getco asks 36.7199. (I get confirms like that constantly from my retail brokers). Getco also gets rebates from the exchanges for "adding liquidity". I.e. NYSE or NASDAQ pays Getco for certain types of orders. Getco can kick back a portion of those rebates to the brokers.

This type of market making (known as specialists on the NYSE) was wildly profitable up until decimalization about a decade ago. Now there are only a few firms left, because the spreads are tight, so the volume needs to be high in order to make any money.

Plus, one bad day because of a computer error can totally destroy your company. E.g. when Knight Capital Group blew up just one year ago and was subsequently acquired by Getco.

The character Robin Hood stole from the rich to give to the poor. In contrast robinhood.io encourages every-day folks to compete against the elites in the stock market. Which explains why elite capitalists are funding the robinhood.io venture.

If you're not paying for the product, you are the product.

People have been restating this tedious point for years. Do you really think if you pay a few bucks a month for something like Netflix, they just decide "well we've already got this guy's 7.99, there's really no need to make any use of his data for additional gains".

I didn't feel any better about that when I saw that Google was one of their major investors.

Still hoping for the best though.

I guess what I'm saying is that if you are a retail trader you are the product regardless of commissions.

But you don't have that leverage with a free service. To them, (in the long run) making some money at the cost of pissing off their customers is better than making no money at all.

Free retail brokerage is something that needs to happen, and I applaud the effort. Brokerages provide a real value add for some services. Offering trading technology, market data, margin, dealing with block trades/portfolio trades, access to OTC, dealing with regulations/back office -- those are real services. Charging me to route an unleveraged, vanilla equity buy order to an exchange and pass the exchange's execution report back to me just because, by convention, exchanges don't want to deal with retail clients directly -- that's just introducing inefficiency and being a middle man.

That said, I don't like the fact that they have a "How We Make Money" section without more extensive disclosure. In my mind, either don't have one (I challenge you to find a single large brokerage that does), or have a more detailed explanation of how the modern brokerage business works. The truth, given the value proposition of free trading, is one that I'm happy to embrace.

I can't say with any certainty what they're actually doing; I can only speak to the industry on the whole, but most retail brokerages make money from:

1. Retail market making 2. Netting across client order flow (probably not applicable here) 3. Asymmetric exchange fees/rebates

The rules on all three are highly country/exchange dependent, but here's an abridged version.

1. Retail market making involves selling order flow to third parties who are able to execute it at a price better than anything that's currently showing on a lit exchange. I've included more details on this below the fold since a) it helps explain their estimated cost graphic, and b) it's one of the most hyped and misunderstood practices in finance, so people should at least decide how they feel about the practice based on correct information.

2. Netting comes about when you're dealing with lots of order flow at a bank/brokerage with multiple lines of business. Your clients might be, on average, and across some time horizon, buying and selling roughly the same amount of a security. You can fill your client at market price, taking the inventory down on your own book, or cross it immediately against an existing position. Most countries/exchanges still require you to 1) pay taxes and exchange fees and 2) print the trades on a market venue for disclosure/price discovery purposes, but there's still some benefit to be had as you can avoid market impact (moving the market when transacting a large order), crossing the spread (paying the differential between the buy price and the sell price), and "long sell" short restricted securities (many countries have regulations on short selling, some banning it all together, so having natural long inventory to sell against is valuable).

3. Asymmetric fees are the most straight forward. Many U.S equity exchanges charge a fee for taking liquidity (crossing the spread) and offer a rebate for posting it (submitting limit orders that don't cross the spread). By charging people this fee when their order does cross the spread and not giving them the rebate when it doesn't there's an easy differential to capture. Also, as noted in their fee structure, they're passing along all regulatory fees to the customer.

It's important to note that no matter what a brokerage does, the net effect is always a price that's better than or equal to what's showing on any public exchange, and what you as a client could get otherwise. In my mind at least, arguing that "I could have gotten a better price on my own if I had access to the same unfair advantages (read: technology/scale)" makes about as much sense as begrudging Google/AWS for buying hardware in bulk, spending billions on data centers that make more efficient use of power and bandwidth, and subsequently undercutting you in a web services platform pricing war. Anyone who wants to come along and usurp the throne is free to spend the money and hire the right people to do so. For me, I'm happy to let my broker engage in these activities if it gets me a better price than I could get for myself otherwise, after fees. I pay Google/AWS/Linode/Heroku to do things cheaper than I could practically speaking do them for myself.

Taking the above points into account, the feasibility of a free or nearly free brokerage (again, note the reg fees) is very real. I'm excited to see how this plays out.

==

Details on retail market making

U.S equities exchanges are highly fragmented compared to those in most other countries. It's common to have a single name trade on several lit venues, and when you count dark pools/other forms of liquidity, that number can easily approach twenty or thirty. As an investor you have a regulatory right to specify how your order gets routed. However, most people just want the best price (this sounds like a truism, but sometimes other considerations outweigh saving a millionth of a cent per share), and access to private dark pools isn't a god given right. There are thousands of pages of regulations regarding order routing, right down to what type of client account it is (is this pension fund money? is this an IRA account?) but the redux is -- you can never fill a client at a price worse than what's being offered on any public exchange.

Enter the retail market maker. For certain types of orders/accounts (back to the thousands of pages of regs...), if the client doesn't explicitly specify an exchange, the order can be routed to a retail market maker. Said market maker can fill the order at a price better than what the market has to offer, or immediately pass it along to the exchange. Surprisingly, they'll actually pay the brokerage for the privilege of doing so. Why would they do that? The name of your game as a market maker is netting. If you have a large, unbiased stream of order flow, statistically speaking you hope to see it balance out with market indices/other correlated equities (hedging) or itself (crossing) over a short time frame. Until it does so you have risk exposure, so from an economics/efficient market standpoint your job as a market maker is to provide liquidity and price risk premium.

These groups have access to good technology and are well integrated with all of the lit venues/dark pools. Their volumes are huge so they get exchange discounts and dark pool fees (as any individual trader who dealt in those volumes would). They also have good credit and large account balances, so their clearing/margin/and funding costs are lower. As such, the brokerage makes money (risk free), the retail market maker might or might not make money (depending on how good they are at their job, and the space is competitive enough that only the good ones are left), and the client gets a better fill price than they would have on the exchange. Ironically the only people hurt by this are the HFT guys who now have highly refined (read, directionally correct over a few second time span) orders hitting the exchange.

> Robinhood is venture-funded by Google, Andreessen Horowitz and many others, which affords us the freedom to focus on building a wonderful brokerage experience rather than short-term profits.

...just smells of lock-in. Is "free trades" for perpetuity?

I guess that's why there will be a Premium service and an API.

With zero brokerage, you can write bots that make money and abuse the market to no end with very small capital.

http://screen.yahoo.com/first-citywide-change-bank-2-0000005...

"All the time our customers ask us, how do we make money doing this? The answer is simple. Volume."

Their website is ugly and old-school, and their online interface - at least the basic one - feels like it's from ten years ago, but it's cheap (and has different pricing plans depending on how you want to trade).

   $0 trade Commissions are currently only available
   for Robinhood Financial self-directed brokerage
   accounts via mobile devices

More fine print: only available in CALIFORNIA

https://www.robinhood.io/faq/

* charge for API access * provide a margin facility (charge interest on a loan provided to facilitate trading).

You must pay at least $10 in commissions each month or they charge you the difference. Depending on how much you trade this might be a deal breaker. There is also a minimum initial deposit.

Virtualbrokers is another option with more choice of fee structure. I don't think they have a minimum monthly fee but they charge you for data.

Its spot on that High Frequency traders pay nothing beyond regulatory fees (or get money back) to trade as they provide (theoretically) market liquidity. Not sure you can extend that argument to individual investors unless they scale massively, but the technology for trading isn't terribly complex.

One wrinkle is in less liquid names, where their algorithms will likely lag those of larger brokers for a while.

(You can't execute a trade at any given time during the day. You put in "Buy $100 worth of AAPL" and they execute it at the end of the day. Same with sell orders.)

1. http://www.renaissancecapital.com/ipohome/pricings/priced.as...

There was a company called Zecco (zecco.com) which also did something similar with zero commissions (hence the name). Not sure if they're still active.

Update: looks like they (zecco) were bought by tradeking.

So from the FAQ Page it looks like they will make money from margin accounts (presumably swap/interest) and also by charging for API access.

It was developed and designed in Denmark by an agency i co-founded and then incorporated in the US a year later.

They merged with TradeKing a couple of years ago.

The intent with ZEro Cost COmmission was to be free and making money on margin accounts. But as I think these guys will find its going to be tough to build a business on that.

As circumstances will have it I am having lunch with one of the original zecco people today :)

http://www.russellbeattie.com/blog/artisanal-websites-the-ri...

(Beattie credits evocative short name, stable British administration, lots of available short and English word domains.)

Another clever new-ish one is .ac, but it's more expensive than the already expensive .io.

.io domains are very popular in a ton of different areas, though!

One could take trading systems previously designed to operate on large minimum balances of $50,000+ (so as to render the impact of transaction fees negligible) and run that same system on comparatively small account balances, perhaps even as low as $500.

In theory, the percentage returns in both cases should be the nearly same. Obviously a 25% annual return on $500 is a lot less exciting than a 25% return on $50,000. However, the fact it's now possible to even do this (again, as far as I can tell) is exciting.

As other comments have pointed out, trade execution quality may end up being merely average on this platform. But, if your trading system is operating on a sufficiently long time frame, then executions become far less important.

Concerning other comments regarding classification as a Pattern Day Trader and being required to maintain a minimum $25k balance, I believe it only applies to margin accounts. Obviously this means no short selling, but small loss, considering.

I wish trading stocks was as easy and cheap as this. This is one of the reasons why I'm too nervous to buy invest in the stock market or just put some money in it for fun, the barriers to entry are bafflingly complex.

This is often cripplingly high. I mean, if you're just putting £100 on, you'll wave goodbye to £10 for each trade you make. Even if you just buy one stock and sell it one year later that's £20 gone plus any capital gains taxes.

I know the stock market is for people making long term investments with big lump sums, but for people at the bottom, the barrier to entry is too high

http://en.wikipedia.org/wiki/Dividend_reinvestment_plan#Cash...

Most brokers also have some range of mutual funds with no trading fee.

Building up funds over time is probably a better use of small amounts than trying to earn percentages trading them. For one thing, even an amazing annual return on £100 isn't going to cover a whole lot of your time.

Sharebuilder is ok too, but more expensive. I think bitcoins are actually more complicated to trade.

Sharebuilder is great when you're just starting out or want to automatically dollar cost average into positions. I started there, and then moved to TradeKing.

One cool thing about Bitcoin is you can use "colored coins" [1] to track shares of stock, then you can use Bitcoin's "script"[2]/contracts[3] to do trustless/atomic trades. It requires the company (or some other counterparty) to issue the stock as colored coins though, so for now it's limited mostly to Bitcoin related companies. You can even do trustless cross-chain trades, for example if a you wanted to trade Litecoin for a Bitcoin (or Bitcoin blockchain-based colored coin) [4].

[1] http://coloredcoins.org/

[2] https://en.bitcoin.it/wiki/Script

[3] https://en.bitcoin.it/wiki/Contracts

[4] https://en.bitcoin.it/wiki/Contracts#Example_5:_Trading_acro...

No minimum balance, $0 commissions. Trading for the normal person.

The name Robin Hood is perfect.

Commissions, fees, and minimum account sizes can be barriers that prevent participation at all.

1) How will they afford SIPC coverage? (That is the insurance for your money in case they fail.)

2) Will they do any securities lending with purchased stock?

3) How much will they charge for selling a stock short?

4) Taxable reporting? Do they handle that or is that up to you, the investor?

I would bet that a good portion of their outside investments would go toward this while they are growing. If they can make enough on their API and margin trading that would cover it long term. I would also bet their margin trades will have a bit higher lending costs than a typical trading platform.

As far as I know, it's a legal requirement for your brokerage to report your income and transactions to you and the IRS on the various 1099 forms.

(disclaimer, I work there)

And their going to make money on margin accounts and API access? Anyone sophisticated enough to be trading on margin or paying for API access has enough funds to not care about $10 trading fees. Seriously - on a $10,000 trade, that's a tenth of a percent. You've got to be a hedge fund with millisecond precision to care about that.

$10 each way fee can seriously eat into the EV of people looking to make small orders, and disadvantages this group against the wealthy (who already have a host of other edges!)

I'd like to spend a few hundred on several stocks, but the fee makes this not worth it. Leaves me with the option of don't, or put it all in one stock which isn't a good idea.

So basically, they will attract lots of small, active trading accounts. I don't see the model.

I'll be interested to see if they can find a sustainable business model. I'm on the fence about whether I'm going to open an account. Even if they go under, there's basically zero risk to customer funds, but it is a hassle to do paperwork on another account.

People seem to forget this often.

Taxation in the 12th century was also pretty much limited to the landowning few.

King Richard (the Lionheart) gets kidnapped while fighting in the Crusades. Prince John burdens his brother's people with an excruciating tax which John says is for the ransom to set their beloved king free.

To me it looked like each trade is an electronic entry in market and should cost near-zero dollars givens the volume of all trades.

My current broker offers me great execution, trading on a number of international markets, allows me to settle trades and hold a number of different currencies, and good research tools.

At the moment this service seems to be worth exactly what they charge for commissions...

A 10-25bp difference in execution on a $10k+ trade easily covers any commission.

- their tagline and what they do "$0 commission stock brokerage"

- why I should sign up "stop paying $10 for every trade"

- clear call to action "get early access"

- giant iPhone graphic gets cutoff and leads you to scroll down

Wish you Super Duper Success :)