Hacker News Reader

Comment by SilasX | Hacker News Reader

But that wouldn’t make sense as being a constraint that the Robinhood’s upstream counterparties would demand, since the collateral is purportedly to protect them. No one has ever argued this point in terms of “oh we just wanted to make sure no one took your money without giving shares”, or, if they did, they are really bad at communication.

Edit: note that this other authoritative explanation claims the failure mode is RH holding the bag for a client not depositing funds as promised.

View on HN

Here's a concrete example why that rule might be wanted:

1. Clearinghouse member A buys 1000 shares at $200 each, costing a total of $200k. On the day of trade (not settlement) they put up $20k in collateral.

2. Clearinghouse member B buys 2000 shares at $200 each, costing a total of $400k. On the day of trade (not settlement) they put up $40k in collateral

3. Something bad happens. The price of the stock drops to $150, and clearinghouse member A goes bankrupt. They're supposed to pay $200k for the shares they bought, but they can't and those shares are now worth $150k, so they owe the clearinghouse $50k. They only put up $20k in collateral so there's a $30k shortfall.

4. The clearinghouse somehow socializes the losses, presumably using some of the collateral from member B to make up the deficit. Now member B is short $30k.

For a small amount they might be able to cover it out of pocket, but if it was sufficiently large they won't be able to. In that case the net result is that the customer had their funds seized (because their funds were used for the clearinghouse collateral) but their brokerage can't pay for the stock. Forcing the brokerage to use their own funds prevents this problem. The brokerage and their creditors might still lose money, but their customers shouldn't be affected.

I don't know what that has to do with a customer who put up (100% of) $10k to buy $10k of stock and is prevented by their broker from backing out.

Because there are two distinct layers to the system, the clearinghouse to its members, and its members (brokerages) to their customers. All the clearinghouse cares about is that the member can put up the collateral on the day of trade, and on the day of settlement they show up with the rest of the money. Whether the clearinghouse member's customer paid for their stock in cash or on margin doesn't concern them. There isn't a mechanism for robinhood (or any other brokerage) to tell the clearinghouse "hey this trade is 100% good because the customer is buying it with settled cash in his account, so waive the collateral requirement".

I think what you're trying to ask is that if clearinghouse somehow knew that a given buy order would be good (ie. was paid with settled cash), then this whole fiasco wouldn't have occurred. That's true, and that's essentially what T+0 settlement is (ie. you have to come up with all the money on the day of trade).

This is a copy of the previous thread. As before, this latest reply just prompts the question: how can a broker promise that the collateral is good, but not the initial funds for the purchase? [1] Apparently, they have a way to say “oh no these funds are good, you can trust them” … but only for the collateral, not for customer funds that have lived for years in the same account.

And for that matter, how can bankruptcy even matter? These are all funds that live in a separate bucket from the rest of operations. Even in a bankruptcy, they wouldn’t disappear (like your coat at a dry cleaner’s.

[1] Recall the comment about how you’re implicitly supposing a “credible lockup of funds” primitive, which has a strange structure, on closer examination: https://news.ycombinator.com/item?id=27694540

>how can a broker promise that the collateral is good, but not the initial funds for the purchase? [1]

1. the collateral is good because it's transferred/wired on the day of purchase.

2. as mentioned in my previous comment, that could theoretically be done today (ie. "trust me, I can definitely come up with the money on the day of settlement"), but it's not done for various reasons.

>And for that matter, how can bankruptcy even matter? These are all funds that live in a separate bucket from the rest of operations. Even in a bankruptcy, they wouldn’t disappear (like your coat at a dry cleaner’s.

but in the scenario mentioned above, how are you going to recover the funds? specifically, from where are you going to get the money to make the seller whole?

It makes perfect sense for a regulation that intends to protect the retail customer's interests. I suspect Robinhood would put customer money up for collateral if they were allowed to do so.

What is that responding to? The problem is the lack of a consistent threat model for whom you're protecting and what you're protecting them from. Half the time it's "the consumer could reneg on the purchase" and half the time it's "the shares could be stoken from the consumer". And 100% of the time, the super-confident, I-get-this-and-you-don't explainer doesn't realize the inconsistency.

I'm not sure what you're driving at. The sibling clearly outlines a scenario where the collateral is lost. But it almost doesn't matter so long as you assume the clearinghouse isn't some vestigial intermediary.

I can't answer why customers with cash in hand couldn't directly settle the trade. Maybe most of the time it doesn't matter and it's easier for brokers to have one settlement process. Maybe the industry has come to depend on the delayed settlement in some unrelated way. Maybe it's something that hasn't fully been adopted due to change taking time. Maybe the referenced regulations get in the way of an uncollateralized process.

>I'm not sure what you're driving at. The sibling clearly outlines a scenario where the collateral is lost.

They outlined a (dubious) scenario where the collateral protected the clearinghouse or counterparties, not the retail consumer, which your original comment appealed to. So no, you can’t mix and match and equivocate between the two justifications, and if your comment is going to appeal the latter, you can’t equate it with the former.

>I can't answer why customers with cash in hand couldn't directly settle the trade. Maybe…

Okay, I hate to come down hard on you, since you’re far from the only (or worst) offender, but … if what you’re saying is true, you really shouldn’t be commenting on this thread.

If you’re coming in with the view that Gamestonkers are just misinformed about how things work, and you want to tell them so, it’s incumbent on you to actually understand “how things work” at a deeper level, which includes being able to answer follow up questions.

I linked an earlier thread where I, following the Hacker ethos, applied my curiosity to pin down a model that would explain the Robinhood failure and reconcile any deficiencies in my worldmodel. That exchange was an attempt to pin things down rigorously, and a rigorous answer requires that you be able to answer questions like that.

So this really isn’t the place for “fake it till you make it” or throwing around wild speculation. If you speak authoritatively while lacking the requisite understanding, then you’re adding noise, not signal.

It is not a strike against you that you lack that understanding. Heck, I’m the same way! But it is when you want to call others’ understanding deficient. And it does mean you should leave room for others who can (justifiably) offer understanding.

I never claimed to know anything about the subject and I don't see how you got from what I said to "check your tone". I read the thread you linked and shared my own conclusions.

What I can say is that the explanation given is consistent with motives that make sense for the relevant parties. And I feel that was perfectly clear in the linked thread as well. And guess what, the involved parties are publicly saying things that are consistent with all this too!

I never mentioned or in any way disparaged "gamestonkers".

As to the rest of your comment: Actually understanding how this part of the financial system works is going to require a good deal more work than being the obtuse end of a discussion. And I expect you'll be disappointed with what you learn.

Reply on news.ycombinator.com