Archegos was too busy for margin calls
bloomberg.comFormer fund manager here.
Wow, this is really interesting. So CS actually knew what risk it was taking, the hypothesis that Archegos had kept its positions private from several lenders is false. They knew what was going on and couldn't ask for more margin, and when they did without a response they couldn't get themselves to just close out the customer's positions.
The problem is actually one of internal incentives, from my reading. If you've ever done business with CS (I was a PB customer), you'll notice there's a bunch of different entities, far more than you'd expect. I suspect when you have a number of committees in charge, actually nobody is in charge.
You also have a sales guy / risk guy issue here. Some person is in charge of the relationship and gets paid for bringing in the business. Another person is supposed to say stop once the risk to CS gets too high. There's a natural tension there and if the ownership over the relationship is vague, there's going to be a lot of meetings and not much decided. I saw this first hand at a firm I was at: risk guy comes in, asks to reduce positions. Trader says meh and then if there's no process the issue just sort of sits there awkwardly, with no resolution. Normally there's no blowup, but sometimes...
Another PB I worked with had a similar issue. A high rolling Gulf guy came in, wanting to do big FX trades on little margin. Risk said no, boss overruled them. Dude blew up, PB lost a lot of capital, boss got fired.
I'm speaking about this at a conference next week. Here is the MONEY QUOTE from Paul Weiss investigative report:
>>> ...further, the various Risk Committees only had access to data that were four to six weeks old.
As a consequence, Risk was unaware of, and unable to fully appreciate in real time, the magnitude and pace of the exponential growth in Archegos’s positions... <<<
Ref: https://www.credit-suisse.com/about-us/en/reports-research/a...
I'd like to also point out that unlike the 2008 crisis with margin calls which were highly subjective based on difficult projections of housing prices and defaults and correlations (think AIG-FP), here, these were pretty simple equity swaps. Valuing these assets is not difficult
There are many interesting quotes, but for me this is the one that explains everything:
"Indeed, the 'Action/Decision' for Archegos was for CRM to 'notify of any changes with the counterparty and revisit the counterparty at afuture meeting.' CPOC did not set a deadline for remediating Archegos’s limit breaches, for moving Archegos to dynamic margining with add-ons, or even for reporting back or revisiting the status of Archegos at a futuremeeting."
The Archeos situation was escalated as far up as possible, up to this CPOC committee, where the Chief Risk Officer of the Investment Bank was a member. Instead of the CRO breathing down their neck until the situation was remediated, they got away with a bullet point in some minutes, without even a set deadline.
It appears that Credit Swiss replaced their CRO though.
https://www.bloomberg.com/news/articles/2021-07-29/credit-su...
>Another PB I worked with had a similar issue. A high rolling Gulf guy came in, wanting to do big FX trades on little margin. Risk said no, boss overruled them. Dude blew up, PB lost a lot of capital, boss got fired.
I used to work in finance but quit. The industry is still by far driven by mostly by sales/relationships/politics. It seems like in this case the right person got fired. I knew of a trader on a credit desk that blew up their desk (lost many times their pnl in a year). His boss (head of desk) got fired instead and he got promoted to his position. Yet it was a completely unsurprising outcome and spoke so much about the industry
pessimists don't get promoted, but more importantly the most important skill is convincing your boss and the boss's boss that your course of action is the best path forward.
Any idea how they got away with this?
I think the generally acknowledged approach if you work on a trading desk at a bank want to put on some trade with desk-blowup potential is you first send out an email to your boss/head-of-desk outlying the trade and targets and stop loss and whatever, even if he sits right next to you. Once he replies in agreement, he's on the hook and you have the evidence. This makes it perfectly reasonable to argue that its the desk head's responsibility to make sure these kind of blowups don't happen. If head of trading also feels that head of desk wasn't exactly a stellar performer anyways, then poof he's gone. It just so happens in this case that they didn't want to replace headcount anyways so the trader next in line, the guy who actually blew the desk up, got promoted. Maybe it was just the bank I worked at but I saw enough of these charades and couldn't take the job seriously anymore. It was too much of a farce for me so I quite Lol, but hey I guess there are people who enjoy playing that game, nothing wrong with that.
That industry is relationship and influence driven.
Can you expand some of these acronyms? CS, PB, etc.? I'm guessing most of us won't know them. CS=Credit Suisse, PB=Personal Banking, FX=Foreign Exchange?
PB is prime brokerage
https://en.wikipedia.org/wiki/Prime_brokerage
CS is Credit Suisse, and FX is foreign currency trading.
Apologies, PB is prime broking. It's where the bank offers trade facilitation services, especially lending, to a hedge fund or family office.
Just a general suggestion for comments: if you’re going to use abbreviations, it’s probably a good idea to spell them out on the first mention.
PB is probably prime brokerage, it is a managed service offered by large institutions to other large institutions.
> The problem is actually one of internal incentives, from my reading.
It looks a lot more like a feature of those incentives. Functionally the risk persons job is to tell gambling addicts "I dunno guys, looks too juicy. High rollers only."
To paraphrase Sinclair: It's difficult to get someone to employ second-order thinking when his salary increases dramatically and there's no real downside if he doesn't.
couldn't get themselves to just close out the customer's positions.
Exactly. CS let Archegos get into a position where, if they pulled the funding plug on Archegos, Archegos would go bust. At which point everybody hates everybody.
Sunk cost fallacy. If you get to the point where pulling the plug means someone has gone bust, it's already over and you need to pull the plug sooner rather than later.
Well the folks at Credit Suisse aren’t perfectly rational beings either, it’s totally understandable why they behaved like that, most folks would engage in wishful thinking when staring at a multi billion dollar loss that hasn’t been realized yet (several hundred times their compensation).
We changed the URL from https://www.credit-suisse.com/media/assets/corporate/docs/ab..., which is one of those annoying links that downloads a file, and the title from "Postmortem in finance: How Credit Suisse lost $6B [pdf]", which was editorialized. (Please read the site guidelines! https://news.ycombinator.com/newsguidelines.html)
A user suggested the Levine article as a better alternative, so we'll use that for now. If anyone knows a better URL, we can change it again.
A little hopefully constructive feedback:
1. The original link was to the source, which I would think would take precedence over a second-party review of it, even a skilled one like Levine's.
2. The source's actual title is "Credit Suisse Group Special Committee Of The Board of Directors Report On Archegos Capital Management". The editorizialized title ""Postmortem in finance: How Credit Suisse lost $6B [pdf]" isn't bad, includes a [pdf] warning, and is more descriptive of the content for anyone not aware that Archegos lost $6b of CS's capital. And how they lost it is certainly interesting.
3. Maybe if you change a post link from the original source, also link the source back in the comments. Levine includes it in his article below the fold [1][2], at least, but that might not always be the case. Lots of web writers these days like to write about reports and studies without actually linking to it.
[1]:https://www.credit-suisse.com/about-us-news/en/articles/medi...
[2]:https://www.credit-suisse.com/media/assets/corporate/docs/ab...
> 1. The original link was to the source, which I would think would take precedence over a second-party review of it, even a skilled one like Levine's.
A 170 page report isn't a really useful thing to drop on anyone especially if they don't already know what's happening. This article has a good summary of what happened and includes a link to the report if you want to see it.
Yeah, I hate to be pedantic, but there seems to be some disconnect with the actual HN guidelines here:
> Please submit the original source. If a post reports on something found on another site, submit the latter.
https://news.ycombinator.com/newsguidelines.html
Perhaps this rule needs a refreshment?
I agree with dang that the Bloomberg article is more useful than the report itself, but the moderation action does not seem to be in line with the actual guidelines.
> Maybe if you change a post link from the original source, also link the source back in the comments
The originally posted link is in my GP comment, no?
Why do you change URLs instead of submitting the one you like instead as a regular user and let upvotes decide? I appreciate and am thankful to the effort you put into moderating HN, but the user did not submit this URL and changing it from under them is super weird, even with a comment explaining it. Not only that you changed it to a paywalled link which is simply bad form.
Once a story is on the front page, it's in a completely different category than the /newest page, so we can't just "let upvotes decide". Even when there's a strictly better URL, by far the most likely outcome is that it would languish unnoticed on /newest. If it did get noticed and lead to a front page discussion, we'd then have the problem of two separate front page discussions (i.e. dupes). The upshot is that improving the URL (hopefully it's an improvement!) on an existing front page post is one of the important things that mods do here.
"Changing [the url] from under them" is standard moderation practice on HN - we do that every day. Ditto for titles. So I'm not sure it counts as "super weird"?
The paywall issue is orthogonal to this one. If there is a workaround, a paywalled URL is ok. If there is no workaround, it's not ok, no matter how "good" the URL is.
It was weird to find out that it's done in the first place! I didn't know moderation took a post that hits front page to become HN's more than the author but can see now how it makes sense. Thanks for answering and for keeping a great community going for so long!
HTML version of the original PDF: https://www.sec.gov/Archives/edgar/data/0001159510/000137036...
The new URL (at bloomberg) seems to be paywalled. Is there an alternative link?
you can usually get one read out of it through incognito/private
Sorry - it's not always easy to predict what is paywalled where.
How is a browser downloading an html file any different than requesting a PDF other than size? If anything it's useful to be able to pop it over into the Documents app and read later.
I guess it's not responsive but to be honest that's a feature not a bug lol. I guess the editorial concern is the bigger issue.
"Downloading" here means that it sends something to your downloads folder and doesn't participate in the normal browser UI. When I use a browser, I generally expect to use its UI. If it transferred a PDF but showed it in the normal UI (colloquially not referred to as "downloading," although yes, it still is an inbound transfer of data), it'd be less annoying.$ curl -SsfI https://www.credit-suisse.com/media/assets/corporate/docs/about-us/investor-relations/financial-disclosures/results/csg-special-committee-bod-report-archegos.pdf | grep -i disposition content-disposition: attachmentgeofft has it - I wasn't saying pdfs were annoying, but rather that it's annoying to click on a pdf (or html or anything else) and have it save something to my hard drive rather than displaying the damn thing. First, I want to see it, not save it; and second, I want to decide for myself what to save locally. Anyhow, no big deal! just a pet peeve
> Notably, this is not a situation where the business and risk personnel engaged in fraudulent or illegal conduct or acted with ill intent. Nor is it one where the architecture of risk controls and processes was lacking or the existing risk systems failed to operate sufficiently to identify critical risks and related concerns. The Archegos risks were identified and were conspicuous. The persistent failure of he business and risk to manage and remediate the risks, and pervasive issues of business competence and resourcing adequacy, described in detail in this Report, require CS’s urgent attention.
wow. this reads like 'culprits have been found and let go with their bonus packages'.
This is common practice in post mortem analysis to try to determine the root cause in a blameless manner.
On the next iteration of this issue the eventual culprits of the first probably won't be there to remember what went wrong.
What might survive until then though is the set of procedures and guidelines put in place following the investigation recommendations.
So it's more effective to determine what process were insufficient than which person had the capacity to avoid something.
In cases like this, I doubt you can determine which process were insufficient without noticing which people had the capacity to avoid the problem, yet did not exercise it.
That's correct, but unless the investigation determined they acted with malice, they should not be explicitly mentioned.
Good point.
They’ve clawed back nearly $100m of bonuses…
My question is, how profitable are these services that they can keep losing money on an implosion here and there and still want to be in that business? I guess if the other 99 PB clients are netting you $10B you can lose some to Malachite Capital Management and Archegos now and then.
It's tough to quantify. It's not just about the PB revenue itself. It's the fact that it's an entry point for a lot of institutional client activity that's necessary for a lot of other lines of business.
For example, investment banking is a very lucrative business. Let's say you're trying to win a big-ticket IPO. To run a credible IPO you need to be able to fill an order book of deep-pocketed, credible public equity investors. That requires having a relationship with these institutions. The primary way those relationships are built is through a prime brokerage business.
Maybe the motivation is not a constant profit for your bank, but a big bonus at the end of the year. With some luck you can collect a few of those bonuses before the next implosion.
PB is a loss leader for everything else.
If you’ve ever traded with anything better than Reg-T margin, you know this whole system is ripe for implosion. The leverage is insane!
Reg-T (like the kind Robinhood has) doesn’t allow long options to be bought on margin. Other margining systems allow as low as 6% down even on options.
And when you’re talking real money, your personal risk/compliance team understands their employment is contingent on looking the other way! They’re at-will employees too.
Can individuals access that sort of margin?
The best I've seen is portfolio margin at some brokers.
What's the name of the "margining systems" you are referring to?
The closest for retail customers is IBKR, needs >110k they have lowest margin rates. They approve pretty easily.
On other hand Fidelity will reject portfolio margin even if you have significantly more than 100k.
Caveat emptor: IBKR's TOS makes it clear they reserve the right to liquidate your positions at any time for basically any reason, including margin requirement changes.
Every brokerage has those terms. Even Credit Suisse in this article had this right over Archegos's positions - they just were too scared to exercise it.
Of course. IBKR has lower rates and approves more easily, therefore they will be much more trigger happy. CS timidly emailed about potentially discussing margin rate adjustments, IBKR can liquidate your positions for instantaneous margin violations without warning or a margin call. The assumption is your leveraged position is hedged well enough to protect against untimely margin calls, and the cost of hedging is roughly equivalent to the nominal margin rate savings they offer.
Of course they have to put that in TOS, this whole discussion is about how CS didn't enforce changes in margin on time, and they ended up bag holding.
What do you want them to do? bag hold a loosing or risky trade and pass the effects to other clients?
portfolio margining does what I describe and has a regulatory minimum of $125,000 or so net liquidation value (portfolio size), SPAN margining can as well which is what the futures and futures options market uses at any portfolio size
so yes individuals can access both
the primary benefits are cross margining, using other assets to fulfill or calculate margin requirements for a new position
and portfolio margin requirements are easy to calculate, just take the loss at a 15% move and whatever that loss is becomes your current margin requirement, it is a 6% move for indexes (like S&P)
the way to get in trouble is by confusing the margining system and hiding exposure behind synthetic positions (example, a combination of derivatives to make it look like you are long stock)
> Can individuals access that sort of margin?
Yep, I can get lots and lots of margin from IBKR. It helps that I'm not an American.
Can someone provide color as to why this was such a discretionary process? If I get a call from my broker, and I don’t post collateral, I’m liquidated immediately. I’d at least expect a fixed date which triggers liquidation if you catch someone picking their nose.
Also, I find it hilarious that this whole thing was bound to blow up after the recent Chinese regulatory crackdown anyway.
There's the saying of "if you owe the bank $100k you have a problem but if you owe the bank $100 million they have a problem." You're small enough fish you don't have a nearly dedicated person who's bonus and compensation is tied to keeping you from taking your pile of cash and going somewhere else. Individually your business and risk mean basically nothing to your broker most likely.
Do you have multiple billions of dollars at stake with the broker?
If yes: he might annoy you enough with his impertinent margin demands that you go elsewhere, which makes his boss Very Unhappy.
If no: his boss has never heard of you.
Also it sounds like liquidating many of these positions would be no easy feat, when they represented 3-5 days mean trading volume in the underlying asset.
In a quote from the report it says it would have taken two weeks to a month to liquidate their positions just at Credit Suisse.
If I have multiple billions at stake, I would assume my broker would be even more demanding of responding within a time frame. Note, I'm not calling for automatic liquidation, but rather some timeframe wherein discussions must be held or else the posted collateral is seized.
The problem there is at that level you're also a significant source of money for both the company and whoever gets bonus/commission on the money you bring in so suddenly there's a lot of reason to not make you annoyed enough to take your business elsewhere.
They had a contract for static margin and wanted to change to dynamic margin. If your broker changes their terms of service, they likely won't immediately liquidate you pending approval.
I think from my reading of the article that the dynamic margin was a compromise that would require less margin from them than what the standard formulas would require. Business balked at requiring the actual $1B margin that the original contracted margin method would require so they were trying to switch to a new one.
Matt Levine did a pretty good writeup on this: https://www.bloomberg.com/opinion/articles/2021-07-29/archeg...
See dang's comment. HN article link previously pointed to a pdf, now it points to this link.
I did enjoy reading through his updates about the situation as it was developing, pretty interesting stuff.
I don't remember the exact line Matt Levine wrote, but he made a point of stating normally you couldn't get away with being this over leveraged, but since Archegos was a family firm it avoided some scrutiny. Interesting to think there will probably always be a way to avoid scrutiny when you're moving that much money around.
I find this accident to be too suspicious in this context. I know that I am looking backwards and finding patterns where maybe there aren't any, but it's just too tempting to connect this dot to the other dots.
https://snbchf.com/2020/02/durden-credit-suisse-md-dies-frea...
Nothing new under the sun. Read Arthur Hailey's The Moneychangers.
I've recently stumbled upon Matt Levine. His writing is wonderful.
I now read his column most days despite starting with zero interest in finance. Mostly because he's funny as hell
Who wrote this?
The authors are listed on page 1.
Some of the formatting of this document is remarkably poor.
I am surprised there is no conspiracy theory about some third party forcing archegos to blow up on purpose. it would be a convenient way for a secret police or other extra legal group to punish or blackmail a wealthy person while making a little money on the side at the expense of a big bank. I mean billy only put away a few hundred million protected in trusts and charities… do deca billionaires blow up this spectacularly? Doesn’t seem regular to me.
Hanlon's razor: Never attribute to malice that which is adequately explained by stupidity.
And never attribute to stupidity what can be explained adequately by self interest.
I cannot believe that Archegos was "too busy" to read a request for additional collateral. They knew exactly what was going on, and they were hoping to buy more time and pray that the market would change in the direction they needed.
This sounds like a desperation play that didn't succeed, and Credit Suisse was left with the loss.
Now, I'm not much educated in finance... but risk is risk, and whether you have models or you have tarot cards, sometimes things just do not go the way you thought they would. If you're betting 1x, you can lose what you have. If you're betting with leverage, you can win bigger, or lose 1x (and potentially the creditor can lose the remaining Xx). Unless the creditor can automatically liquidate and close out a client's position if they can't make a margin call, it seems you would have to be crazy to be a creditor for leveraged clients. Eventually you will lose big because of a client.
I think you missed Levine's dripping sarcasm.