Robinhood maxed out credit line last month amid market tumult
bloomberg.com
bloomberg.com
We hired the Global Head of Clearing [0] from a big bank and on their first day they were giving a presentation on their background and how financial clearing works.
Someone asked "What is your nightmare scenario?"
The response:
We have a large overnight position in a security. Our Prime Broker (PB) [1] comes back and says "We disagree with you on the position so you can't trade." Even if it turns out that the PB is wrong, by the time it all gets sorted out, the market has moved so much in that security that it bankrupts us. I've seen it happen to other firms and it's not pretty.
I also started at Knight right after their big outage. [2] People like to talk about the big tech outages sinking financial firms but it can just as easily be plain old issues with bookkeeping that can blow up a firm.
0 - https://www.investopedia.com/terms/c/clearing.asp
1 - https://www.investopedia.com/articles/professionals/110415/r...
2 - https://www.thestreet.com/investing/stocks/knight-capital-sh...
One, that could result in a fire sale. Freezing is safer than forcing liquidation. If your liabilities outweigh your assets, the latter are the PB's; getting pennies on the dollar instead of dimes is against their interest.
Two, the client may be able to post collateral. You don't want to lose an institutional relationship because you were trigger happy.
Three, you might have made a mistake. Freezing the position until you get your facts straight is prudent in most markets.
If they are right, making you put on a trade to close the position they don't think you have will actually put you into a position of equal size in the other direction.
eg:
You are my pb and I think I'm long 20 S&P e-mini futures. You don't know that trade so think I'm flat. If you make me "close" my position and you're right then I'm actually now short 20 futures.
This type of situation can happen a lot more easily than you might think especially in markets where there are multiple trading channels (eg voice and electronic) and settlement instructions etc can go wrong.
One place I worked we had a trade be DKd because the instructions were sent by fax and needed to be confirmed by voice. The pos Solaris printer driver used to leak memory so one day it leaked so much memory that it just fell over and died and one of the trade confirms wasn't printed out, therefore the human who had the shitty job of picking them up off the printer and calling the other side to confirm the trade didn't confirm. 5m dollars right there.
That aside, I generally trust the discount brokerages a bit more than people like BofA/Merril and Wells Fargo and JP Morgan, just because I feel like the discount brokerages are a bit more content to just sit on deposits and earn interest (interest accounted for > 50% of schwab's revenue last year) rather than get up to highly leveraged hi-jinks or open fraudulent advisor accounts.
Money chases the sun. At the end of trading in New York you make a swap with a company in Japan that at the end of trading in Japan does the same with London then back to New York. Any money not swapped doesn't make money overnight.
This is called the overnight repo market. But if you don't get your money back in the morning, you've got a problem.
http://www.fintools.com/docs/Warren%20Buffet%20on%20Derivati... is nearly 20 years old, but Warren Buffet's criticism of how we account for the value of derivatives is as true today as it was then.
This would most likely be hitting smaller to medium sizes firms (e.g. Hedge Funds) that are holding a larger position in a security. The small/med firms don't "self clear" so they tend to depend on a larger firm to do that for them.
I should also point out: I used the example of one big position but it could also be a bunch of medium sized positions and then the PB says "Hmmm, we don't agree with 1 or more of those positions so we are stopping trading with you on all of them."
Headline should read "Robinhood used credit that was available to them and paid it back"
But that wouldn't generate the desired panic or clicks now would it
It's yet another sign that this company is teetering on the edge of collapse. That isn't irrational panic. They were completely offline for an entire day during the biggest market rally in over a decade, and have had smaller outages during other recent volatile trading days.
This company deserves to crash and burn. Anyone who still keeps a balance there is insane. People have lost millions and are going to continue to lose because Robinhood is run by amateur clowns.
Robinhood accounts are protected by the SIPC. Although with other discount brokers introducing free trading, I would tend to agree with you that continuing to use RH with its simplistic interface and appalling execution is pretty silly.
This is correct. But SIPC reimbursement can take months. For non-trivial balances, one may need a lawyer to prove ownership.
After that, an accountant would likely be needed to reconcile records, including for tax purposes. (Tax forms are not automatically generated for brokerages in receivership.)
But, I'm sure that Robinhood must be omitting those values on purpose, and the only reasoning I can come to is that they want their users to be less educated because it is more profitable for them.
And their haters. There's a very vocal group of traditional investors who hate Robinhood (and its users) for a variety of (mostly trivial) reasons. Robinhood's service problems over the past couple weeks are finally a legitimate reason for them to air their hate and they would love to see Robinhood fail.
It's true that many other brokers from bit players to top-tier firms occasionally have trouble during intense market days. I've worked in finance on and off for decades, for brokerages, hedge funds, and market makers. I've written systems that process real-time market data and trade directly on exchanges. I'm well aware of the challenges in trade processing systems, and Robinhood simply hasn't built enough testing, redundancy, or scalability into their systems, as we're now seeing.
It also happened on leap day, and there's speculation that their systems were simply unable to handle that. Robinhood denies it, but they also had problems on leap day 2016.
There's no reason that a company like Robinhood can't succeed, and someone in that space will succeed. It may very well end up being Robinhood. But are you willing to bet your own money on that right now?
The real question to come is if a drying up of liquidity is going to torpedo lots of ships. It is simply not possible for a loss making entity to continue losing money forever (unless you are the US Government)
Once free money dries up, a lot of darlings will go down.
> “Companies don’t tap their credit line unless they need to,” said David Ritter, an analyst at Bloomberg Intelligence, who spoke generally about the issue without commenting directly on Robinhood. When companies do, it’s “perhaps not a good signal with regard to their cash burn, which could make creditors nervous.”
“Companies don’t tap their credit line unless they need to,” said David Ritter, an analyst at Bloomberg Intelligence, who spoke generally about the issue without commenting directly on Robinhood. When companies do, it’s “perhaps not a good signal with regard to their cash burn, which could make creditors nervous.”
Should I always stick to big firms like Vanguard or Fidelity? Thing I don't like about firms like Vanguard and Fidelity is that you cannot buy fraction of a unit which Robinhood/Stash allow you to do.
Any advice?
Fidelity has started to offer fractional shares, but it has some limitations: Only market and limit orders, only trades through the basic trade ticket in the mobile app, limited to NYSE/NASDAQ stocks, etc. [1] Another thing to watch out for is that fractional shares can't be transferred; they must be liquidated if you want to switch brokers.
Schwab has said they'll be launching fractional shares over the summer.
Note that most brokers already support fractional shares indirectly through DRIP [2], but that's unrelated to trading.
[1] https://www.fidelity.com/trading/fractional-shares
[2] https://www.investopedia.com/terms/d/dividendreinvestmentpla...
What happens if I buy .5 shares of AAPL every week for 53 weeks? Do I have 26 undivided shares that can be transferred and one half-share that has to be liquidated, or do they treat it as 53 half-shares that all must be liquidated?
If the latter, seems like a great way to keep people locked-in. But the former could be a headache for basis reporting, especially if you could transfer to a brokerage that doesn't natively support fractional shares.
Vanguard, Fidelity, Schwab, TD Ameritrade are all legitimate operations as well, but IB is what professionals use.
Pretty much anything else is just gambling.
When the stock market plunges, some people will see that as an opportunity to buy more stock at a reduced price. If they don't have any liquid funds in their account, borrowing money from the broker is instantaneous (assuming you've previously enabled it). I don't know what happened, but I could see their line of credit getting eaten up very quickly if enough investors decided to leverage their available margin.
If investors sold those shares within two weeks, they could pay back the credit line.
Does anyone know what happens to any open positions in the event that they do become non-operational?
Not that it really matters for the, what, $2 I put into Dogecoin, but still.
Most will talk big about leaving and then end up staying put anyway because it's easier than jumping ship.
I would suppose drawing on a credit line like that would have something to do with maintaining adequate funds in various accounts for settlement of funds transfers and stock transaction as well as the various margin products that brokerages have during a usage spike – without touching corporate accounts for raised funds and various other capital sources. I have no knowledge of the situation but it seems like they pulled in cash from a credit line created for this exact purpose to act as a buffer.
Under the hood of all of these banks and brokerages are a series of settlement periods and bank accounts shared at different firms. You end up needing buffer cash because nothing is instant, you want to keep that to a minimum, and during a usage spike you have to add to the buffer to make sure nothing goes negative. Just the underlying mechanics of banking which has nothing at all to do with the firms financial health.
If covered by SIPC, you get up to $500,000 back.
Personally, I'd get my funds off the meme brokerage and on to a real trading platform.
EDIT: changed frontrunning -> tailgating.
That's a pretty serious criminal allegation, do you have any more info?
> and selling your trade history to hedge funds
I have not heard about this either, do you have more info?
The main way that I'm aware of that Robinhood makes money is payment for order flow. The idea is that HFT firms want make risk-free money by providing liquidity without taking a position (via the bid/ask spread). The problem in the larger market is that every once in awhile the price will move against the HFT firm because some other market participant knows something they don't, and so they get caught holding equities that they didn't want to hold. Instead, they pay Robinhood to send them order flow from retail customers who are unlikely to actually know anything and should have approximately even/random orders on either side of the bid/ask spread.
Orders still have to be executed at the best possible price though, they can't send your order to Citadel and have Citadel charge you $0.02/share more than what you could get from someone else.
It is common to pay for order flow, but it appears that Robinhood turned out to be exactly the sort of crooks that their customers assumed all the rest of the brokers were.
It's instructive how you can be cynical, and believe that equates to sophistication, and end up conning yourself.
Front-running: Broker gets order, broker buys for self, broker executes order. (illegal)
Tailgating: Broker gets order, broker executes order, broker buys for self. (legal)
Broker gets order, sends it to their best buddy instead of normal market, buddy takes other side of order (at the same price as market would have given).
The assumption being that someone who crosses the spread on the general market may have some good info they are trading on. But a robinhood trader who crosses the spread is basically just trading randomly, and is giving away the spread for nothing.
I am not sure if you meant it this way but "best buddy" implies that something shady is going on. If you want to pay Robinhood more for their order flow than Virtu & Citadel I'm sure they would be happen to send you all the orders you could fill.
It arguably undermines the existence of price ticks. I think you know this but for the lurkers, the purpose of price ticks is to make the market first-come-first-serve. Without them people can raise the bid by 0.0000000001 to get first in line, making the market last-come-first serve.
But as far as shadiness goes, it's a very light shade of gray.
The canonical case of market-making is that you think "you know, I think people are going to ask me to buy 100 shares this afternoon, so I'd better buy 50 this morning to hedge". (Or "you know, I've just bought 100 shares for a customer, I'd better buy 50 for myself".) There's a core similarity here - you're buying shares based on your customers' trades - but I think it's also clear that you're not really taking advantage of anyone.
This smells like an unexpected collateral call. Their maxing out the line is material for estimating operational continuity. But given the market's volatility, I wouldn't read it as a solvency problem per se.
> Does anyone know what happens to any open positions in the event that they do become non-operational?
The SIPC [1] insures deposits up to $500,000.
Brokerages, particularly discount brokerages, are leveraged beasts.
When prices move, they have to post collateral. The brokerage has to then make up that deficit by calling customers' capital and/or liquidating securities. Between those two, the brokerage has to draw on reserves and, lacking that, borrow. If it can't, it goes into receivership.
> and is now backed by venture capital firms including Index Ventures, Andreessen Horowitz and Sequoia
So these big players (aka "Smart Money") have granular access to everyone's personal & financial details, have access to trading information, are gaining deep knowledge on how Retail (aka "dumb money") operate/trade, their patterns, per age/gender/region/etc. That information is priceless, and then this happens: Sardine Feeding Frenzy: Whale, Shark, Dolphin and Sea Lions | The Hunt | BBC Earth https://www.youtube.com/watch?v=6zOarcL1BSc
I'm sure it's happened before, but I have to imagine such a thing is incredibly rare and illegal if violating the stated privacy policy (due to GLB law).
I recognise that's reductio ad absurdum, and I'm not saying it's impossible that data is misused. But the idea that equity holders of regulated financial institutions can just get whatever data they want out of that institution is... not true.
I have seen this happening so many times.
And there are myriad ways to get the data. The easy ones is through their DWH where suddenly (with the correct mapping) all reports etc work on day1 with the new data.
If you know this from experience (ran due diligence pre M&A, worked with the DWH post M&A) or have some other experience on the matter, feel free to share names. I will only share two SocGen and BofA and leave it at that.
> That heavy load caused the so-called Domain Name System, or DNS, to fail.
Not a developer. Is it not customary just to hard-code IP addresses instead of domain names when the app is communicating with your own servers? I would assume that would make every interaction faster and is something that would be done anyway.
If you're managing your own server farm though, yes, you can statically allocate IP's and cut out DNS in it's entirety. A practice to this day I do not understand why more places don't. Then again, I was always comfortable with memorizing phone numbers and the like.
Much quicker to remap an elastic IP than wait for DNS changes to propagate.