Hacker News Reader

Comment by gruez | Hacker News Reader

>DTCC who increased their margin requirements on buyers for seemingly no reason other than [...]

...except credit risk

https://www.bloomberg.com/opinion/articles/2021-01-29/reddit...

>[T+2 settlement] means that the seller takes two days of credit risk to the buyer. I see a stock trading at $400 on Monday, I push the button to buy it, I buy it from you at $400. On Tuesday the stock drops to $20. On Wednesday you show up with the stock that I bought on Monday, and you ask me for my $400. I am no longer super jazzed to give it to you. I might find a reason not to pay you. The reason might be that I’m bankrupt, from buying all that stock for $400 on Monday.

View on HN

The future price of the stock has nothing to do with anything. If I want a $400 stock, and you agree to sell it to me for $400, it is a done deal, is it not? The order is filled, regardless of settlement time or future price changes.

>The future price of the stock has nothing to do with anything

The quote in my previous comment covers that, specifically:

>I am no longer super jazzed to give it to you. I might find a reason not to pay you. The reason might be that I’m bankrupt, from buying all that stock for $400 on Monday.

Yes, I read it. You can’t cancel an order in settlement because you’re “no longer super jazzed”, else every order where the stock price increased would be cancelled by the seller so they could sell again at the higher price. Do you have any evidence that orders can be cancelled in settlement?

>Do you have any evidence that orders can be cancelled in settlement

It's not that they'll show up and say "on second thought I don't want that stock anymore, please cancel my order". It's that they'll go bankrupt in the meantime. This was already mentioned in the original comment.

>The reason might be that I’m bankrupt

The broker doesn’t have some way to credibly lock up the money so the buyer can’t back out, regardless of jazzedness or bankruptcy status?

>The broker doesn’t have some way to credibly lock up the money

They do, that's what the collateral is for. From the linked article:

>The way that stock markets mostly deal with this risk is a system of clearinghouses. The stock trades are processed through a clearinghouse. The members of the clearinghouse are big brokerage firms—“clearing brokers”—who send trades to the clearinghouses and guarantee them. The clearing brokers post collateral with the clearinghouses: They put up some money to guarantee that they’ll show up to pay off all their settlement obligations. The clearing brokers have customers—institutional investors, smaller brokers—who post collateral with the clearing brokers to guarantee their obligations. The smaller brokers, in turn, have customers of their own—retail traders, etc.—and also have to make sure that, if a customer buys stock on a Monday, she’ll have the cash to pay for it on Wednesday.

The catch here seems to be that the broker can't use customer funds for collateral (see my other comments in this thread), so the broker has to come up with the money themselves by drawing on lines of credit. If those lines of credit run dry, then they can't take any more orders.

Right, that’s what’s confusing. To rephrase, I’m hearing that brokers can execute some primitive operation <lock up> which correctly and credibly persuades the financial market that $X is available to settle a transaction.

But <lock up> is only a “valid operation” when executed with “the broker’s own funds”. Why? To whom would “locking up $X from broker funds” be a valid operation, but “locking up $X from client funds” wouldn’t be?

Your comments are saying “that’s just how it is”, but that’s not a reason. Is it a wholly arbitrary operation handed down from on high that accomplishes nothing? Is it a matter of the financial system not being able to trust client-originating money?

>But <lock up> is only a “valid operation” when executed with “the broker’s own funds”. Why?

regulations, presumably to protect the retail trader. It was part of dodd-frank. https://twitter.com/kralctrebor/status/1354952686165225478

>Is it a matter of the financial system not being able to trust client-originating money?

Yeah pretty much. If every trader had their funds sitting in one place (with the clearinghouse or the fed), then the clearinghouse wouldn't need a deposit system since they can easily validate whether everyone has the funds.

I still don’t get the model whereby <lock up> is unsafe with client funds but not broker funds. Were you presenting one? Did I miss it?

If you weren’t, you’re not really resolving my confusion here.

>I still don’t get the model whereby <lock up> is unsafe with client funds but not broker funds.

"unsafe" is the wrong word here. That would imply the clearinghouse doesn't think the collateral is as good if it came from the customer rather than the broker. This isn't the case. The requirement to use broker funds is not to protect the clearinghouse, it's to protect the customer in case the clearinghouse seizes the collateral (as they're allowed to do) when things go south. By using broker (or their creditor's funds), the customer's funds aren't at risk.

So ... the clearinghouse is perfectly capable of seizing collateral and running off without delivering the promised shares, but client funds "aren't at risk" when the broker puts up separate collateral, even though the clearinghouse could seize that just the same?

This still doesn't make sense.

>but client funds "aren't at risk" when the broker puts up separate collateral, even though the clearinghouse could seize that just the same?

The collateral is lost either way, but when the broker puts up the collateral the broker is on the hook for the loss, not the customer.

Why wouldn't they be? You said the clearinghouse could seize the client purchase funds just as well as the collateral. Why couldn't they seize both? At the very least, it's a bizarre threat model that the clearinghouse can seize both, but it will only ever seize collateral (which is what your claim requires to make sense).

That is, it's some kind of capricious being capable of seizing and willing to seize any money trusted to them, with no recourse, but somehow the presence of collateral makes it all better, even though that could be seized too.

>Why wouldn't they be? You said the clearinghouse could seize the client purchase funds just as well as the collateral. Why couldn't they seize both?

Regulations? Presumably customer funds are segregated from company funds, so the company and the clearinghouse can't raid it if they need money.

>At the very least, it's a bizarre threat model that the clearinghouse can seize both, but it will only ever seize collateral (which is what your claim requires to make sense).

>That is, it's some kind of capricious being capable of seizing and willing to seize any money trusted to them, with no recourse, but somehow the presence of collateral makes it all better, even though that could be seized too.

by "seize" I don't mean the clearinghouse can walk into the offices (or bank accounts) of any of their member and grab whatever they want. They're seizing (or more accurately, refusing to return) the deposit that the member sent on the day of trade.

>Regulations? Presumably customer funds are segregated from company funds, so the company and the clearinghouse can't raid it if they need money.

Okay, and again, if "regulations" are enough for one case, why not the other? If a "regulation" can prevent the clearinghouse from holding on to asset they're not entitled to, why not use that instead of requiring the broker to put up money that will compensate the customer when the CH holds on to an asset they're not entitled to?

>by "seize" I don't mean the clearinghouse can walk into the offices (or bank accounts) of any of their member and grab whatever they want. They're seizing (or more accurately, refusing to return) the deposit that the member sent on the day of trade.

Great, that's how I was using it too.

>Okay, and again, if "regulations" are enough for one case, why not the other? If a "regulation" can prevent the clearinghouse from holding on to asset they're not entitled to, why not use that instead of requiring the broker to put up money that will compensate the customer when the CH holds on to an asset they're not entitled to?

But the clearinghouse is entitled to it. The clearinghouse member's collateral is used to make up the difference (the credit risk) should the member fail to pay the required amount on the day of settlement. If you prohibit the clearinghouse from seizing it, then that kills the point of the collateral.

Yes, obviously the CH should seize assets they’re entitled to. I think that you forgot we were talking about the case where they seize money they’re not entitled to. (Which is how we got to talking about the need to protect the client from that, and why the broker has to put up the extra collateral, and how “regulations” stop that money from being misappropriated but somehow not the client funds.)

I think we’re going in circles. And if I may give some unsolicited feedback, if I understood this topic as well as you’re claiming to, I probably would have given answers that avoided reaching that point. If I didn’t, I would have confessed as much, earlier in the thread, once I started giving your answers.

It's a done deal until your $400 doesn't show up two days later, and then what?

Reply on news.ycombinator.com