Libor, long the most important number in finance, dies at 52
nytimes.com
nytimes.com
[2012]: https://news.ycombinator.com/item?id=4224873 "Lies, Damn Lies and LIBOR"
[2015]: https://news.ycombinator.com/item?id=9426247 "Deutsche Bank to Pay Record $2.5B to Resolve Libor"
[2017]: https://news.ycombinator.com/item?id=13497578 "Libor: the bankers who fixed the world’s most important number"
[2017]: https://news.ycombinator.com/item?id=14075230 "Libor: Bank of England implicated in secret recording"
[2017]: https://news.ycombinator.com/item?id=14994122 "Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?"
[2018]: https://news.ycombinator.com/item?id=16418629 "The brazenness of the LIBOR scam"
But at that time I was young and had no finance experience, so though that the adults in the room knew best. Turns out not!
I don't think the base problem is that people shaded their numbers one way or another, it's that the system is designed wrong.
Those dumb bankers. We computer scientists would never design something that failed to scale through 5 decades.
Of course, the locks of those individually-keyed mailboxes are also probably rakeable or bumpable or some other two-second attack. But at least that deters people who never bothered to look into how locks work.
Also, while many parcels are just purchases from companies, a parcel could very well be a priceless heirloom sent by a relative. Or, say, a live reptile.
The postal system bakes old assumptions into it, where ordering goods for delivery didn't used to be nearly so common, so a much larger fraction of parcels used to be of this personal variety. Which makes it even more curious that they're generally handled in such a blasé manner.
My only guess is that shipping parcels is expensive by itself, so almost nobody bothers to pay double to have their parcel handled securely as registered mail.
Aren't those the most corrupt part of society?
That's pretty much everyone everywhere in every industry. LIBOR really was good enough which is why it lasted.
Kind of funny, then, that banking was also the origin of the CQRS/ES paradigm.
The systems thinking is there — just not evenly distributed.
Traditionally the banking community of London was super close and built on reputation. People do huge deals based on people's word and people were expected to be honorable. That might sound naieve in the 21st century when global trade is much bigger but it worked for hundreds of years.
Secondly LIBOR really was pretty accurate, people talk a lot about how it could have been manipulated but the evidence is isn't so solid. Yes in aggregate a few bps adds up to a lot of money but for individual parties it doesn't really make a difference.
Look at crime statistics versus how safe people feel. People say “crime is worse than 20 years ago” when the murder rate is 1/5th.
Perception =\= reality.
That is: people's perception can be anything from a few months to a few decades outdated to the current state.
Usually, every four years, a new party in power can change the trend.
So, it is overwhelmingly likely that something better can be found to reflect this reality.
There's usually something weird going on with some of these indexes if you compare statistics with opinion-survey-based stats.
The numbers that LIBOR is measuring ultimately represent a human’s opinion. Well, it’s the opinions of several humans then the highest and lowest opinions get discarded and the rest averaged.
It’s not measuring a value derived deterministically from some inputs, so how else could you capture it?
SONIA is just one rate - last nights GBP overnight rate. LIBOR benchmark currencies have tenors, the future borrowing rates at different maturities.
They’re different things - so different that a “synthetic LIBOR” rate is now produced in order to help transition instruments that can’t rely only on the historical overnight rate like SONIA.
It’s true that LIBOR is being replaced (although a level 1 submission to ICE LIBOR was already based on value weighted averages from eligible transactions - it’s just that a level 3 expert judgement could override it) - but it’s not being replaced by something comparable. It’s being Replaced by an entirely different benchmark, it’s impossible to change some contracts to use the new benchmark because the new benchmark measures something different and lacks sufficient information to do everything that libor did.
Source: i inherited ownership of the system used to submit libor at one of the libor panel banks when i previously worked there.
This seems stupid to some tech people, who try to disrupt it by creating trustless systems such as distributed ledgers. But trust keeps on creeping back into the system. There are crypto custodians, crypto brokers, crypto exchanges, all of whom you have to trust to some extent. There is crypto lending. I wouldn't be surprised if we eventually have the Bitcoin Interexchange Offered Rate decided by a handful of the biggest exchanges.
If you boil everything down - trust a critical function. Your day is filled with trust of functioning. Without it you would live in pure chaos. I find these arguments facile.
It's probably best that the LIBOR system goes in favor of something more transparent. But I have a friend who is involved in the transition at a bank, and it's incredibly complex to move over such a huge mass of contracts to this new thing. Definitely a lot of work for a lot of people.
ISDAfix - swaps
Platts - oil prices
WM/Reuters - FX
High-Frequency Trading - equities
Commodities - Gold, Silver Stock indices
Are all rigged. With that history, how can we give that new system the benefit of a doubt. It will be gamed as well.
> Commodities - Gold, Silver Stock indices
?
It doesn't help for FX that it is effectively unregulated.
[1] https://www.cmegroup.com/market-data/cme-group-benchmark-adm...
There's a limit to how often the rate changes, etc.
Just looking at one now, and it's set up as follows:
1. Rate is allowed to change only on the 1st day of every 6th month (after the initial flat-rate period ends).
2. On that day, the new rate is set to a spread + the max of 0 and the value of the 30-day Average SOFR Index (as published by the NY Fed) as of 45 days before the day when the rate is changing.
Nothing being applied on the daily basis here.
What you're talking about is the credit spread adjustments from the ARRC formulation of the LIBOR replacement language (11.448 bps for 1 month, 26.161 bps for 3 months and 42.826 bps for 6 months) [0]. In reality, the credit spread adjustments for Term SOFR are still moving and the market has been all over the place - I've seen 10/15/25 bps at 1mo/3mo/6mo tenors (the most common formulation in the leveraged space right now) or a 10 bps flat adjustment (generally viewed as aggressive).
[0] https://www.newyorkfed.org/medialibrary/Microsites/arrc/file...
The authors must be very pleased with themselves about this one...
"Libor is survived by several successors, each making a claim to its crown. The Secured Overnight Financing Rate, or SOFR — a rate produced by the Federal Reserve Bank of New York that is based on transaction data, not estimates — has already been embraced by many banks in the United States and has the endorsement of the Fed. Others, like the American Interbank Offered Rate, or Ameribor, and the Bloomberg Short-Term Bank Yield Index, or BSBY, have their adherents. In Britain, the Sterling Overnight Index Average, or SONIA, seeks to inherit Libor’s place as the do-it-all benchmark."
My only remaining question is, since we used to have a single reference rate, and now we have multiple reference rates - how does this impact existing contracts?
This has kept lawyers busy for past the few years. The "LIBOR fallback language" was created and then inserted into existing contracts, to address this question. It's like a pull request for a patch. Most of the legacy contracts out there have been "patched" with the "fallback language" at this point, and as of Jan 1, 2022, no new LIBOR contracts are allowed.
It’s a pretty painless substitution though
> The transition to a post-Libor world would not be painless. Remember those $190 trillion of Libor-linked derivatives? Hardly any of those instruments — essentially contracts between two parties — provide a workable option for what to do if Libor were to vanish.
> In a worst-case scenario, banks and their customers would effectively have to negotiate how to end Libor-based contracts over the phone, said Darrell Duffie, a Stanford University finance professor. For a sense of what is at stake, Lehman Brothers was a party to more than 900,000 derivatives contracts when it went bankrupt in 2008, according to research published by the Federal Reserve Bank of New York.
> “It’ll be really nasty in terms of costly, difficult workouts,” he said.
The Most Important Number in Finance Is Going Away. Wall St. Isn’t Prepared.
LIBID is the rate at which you'd be willing to borrow. So the difference between the two is like bid vs ask.
After the LIBOR scandal, the EU brought the benchmarks regulation (BMR) which says that interest rate indexes have to be based on actual transactions, just as you say. Euribor, the equivalent of LIBOR for lending in euros, was reformed to be based on transactions:
https://www.emmi-benchmarks.eu/benchmarks/euribor/reforms/
The administrator of LIBOR proposed doing the same:
https://www.clarusft.com/rfrs-libor-is-changing/
But in the end, US and UK regulators decided just to abolish it, in favour of overnight indexes based on real transactions (SONIA for pounds, which already existed, and SOFR for dollars, which was created for this purpose).
I believe this divergence happened because of differences in the lending markets. In the euro area, there is still a lot of unsecured term lending, which is what Euribor measures. But in the UK and US, this kind of lending has largely dried up, but there is a lot of overnight lending, so they chose rates which measure that. I don't know why the euro area is different to the US and UK here. It's possible that the euro market will evolve to be more like the US and UK, in which case Euribor will stop being credible, and the euro will also move over to its overnight rate, ESTR.
Another fun quirk is that SONIA and ESTR measure unsecured overnight lending, whereas SOFR measures "repo", which is essentially lending secured with government bonds as collateral. There is a sterling overnight repo rate, RONIA, but i don't think it's used much. I think repo volumes are higher than unsecured lending volumes; if that difference gets stark enough, perhaps sterling and euro regulators will force another switch, to the repo indexes.
The market - interbank rates - is quite small in terms of participants. I think the problem is the number was incredibly useful even if it wasn't accurate.
LIBOR + bank's minimum interest rate + a rate based on creditworthiness = your offered interest rate
Planet Money recently had a really good episode on how some banks are deciding on a replacement rate: https://www.npr.org/2021/10/08/1044598674/libor-pains
If anyone else was more accurate at this measurement, then they had arbitrage against those using the measure, giving those doing the initial measurement incentive to get it as right as is humanly possible, since they usually worked at places that use LIBOR to price things.
Since LIBOR underlied hundreds of trillions in assets, there are ample papers on all aspects of LIBOR, including those trying to see how well it was computed versus post outcomes.
It holds up well. https://scholar.google.com/scholar?hl=en&as_sdt=0%2C14&q=LIB...
It's rare to see shade thrown so overtly in the Times, because it's so rare it can be done this deniably, and always makes me chuckle when it does.
For derivatives, it wasn't so bad because ISDA (the industry association for derivatives users) published an IBOR fallback protocol which counterparties could adhere to. All contracts between two adherents to that protocol were deemed amended so as to include market-standard fallback language.
There was no such neat solution for bonds and loans, so banks had to look at them pretty much one-by-one. The economic and legal terms of the amendments required to replace LIBOR were mostly standardised across the market, so they typically didn't involve any tough negotiation - the issue was more the operational burden of amending many thousands of contracts.
In a simple bilateral loan the process is straightforward: bank reaches out to borrower, borrower and bank sign amendment agreement, done. But bonds which are widely held through clearing systems posed a much bigger problem, because material amendments typically need the consent of at least half (or sometimes two thirds or three quarters) of bondholders.
A single bond issuance can be held by thousands of (ultimate) investors, and ownership can be heavily intermediated: an investor might hold her bonds in an account with her broker, that broker might hold the bonds in an account with a custodian, that custodian holds them in an account with a securities depositary, etc. The issuer does not know who the ultimate holders are; it can only send out a consent solicitation through the clearing systems. Even if that solicitation manages to work its way through the ownership chain to the end investors, most of them will probably just ignore it.
So a lot of consent solicitations fail even for routine, unobjectionable amendments. When this happens (or is likely to happen), issuers need to look at other ways to push the amendments through, like asking the security agent (who basically represents the bondholders as a class) to consent to the amendment without first receiving the consent of the underlying bondholders. Most deal documents allow security agents to do this where the proposed amendments are not materially prejudicial to bondholders, but security agents are very reluctant to make that determination.
What are the chances that this "club of gentlemen bankers" always intended to manipulate Libor to some extent?
I know finance and banking is very complicated; maybe someone will come along who happens to have been a banker in London in 1986 and will set me straight. Otherwise I find it hard to limit my cynicism when enormous amounts of money are involved.
Well, then it grew. US banks got involved in EuroDollars and foreign banks got US subsidiaries and everything kind of ballooned.
But the big change was the invention of interest rate swaps. Interest rate swaps create a linkage between the Money Market (terms < 12 months) and the Capital Market (terms > 12 months). There are a bunch of economic explanations as to why interest rate swaps exist and some of them have to be true, but they're irrelevant to the LIBOR story. A vanilla fixed-floating swap needs a floating rate, and that's LIBOR. A couple of trillion dollars (notional) worth of derivatives later, instead of simply being a pragmatic way to quote rates in the money market, it then drove P&L of derivatives desks.
I think everyone knew the potential for manipulation was always possible, but for a long time I think, until the tail started wagging the dog, it worked. But there's no going back now.
You should always be cynical when large amounts of money are involved. It helps you avoid large losses.
Side note - one of the best questions to ask in any deal is “how are you making money on this”. If the other party doesn’t tell you, then they probably know something that they don’t want you to know. If you doesn’t get a straight answer, walk away [0].
[0] - I work in finance and never do business with someone who is not transparent about this. Been doing it a long time and it serves me well. I learned it from an old hand, and cringes the first few times he asked it. Then, I got the nerve to ask why he asked such a cringeworthy question. Glad I did!
There's a ton of good info in there on what Libor is, why it needed to live, why it needed to die, and some attempts to replace it.
“Libor was the interest rate that banks themselves had to pay, so it offered a convenient base line for the rates they charged customers who wanted to borrow cash to buy a home or issue a security to finance a business expansion.”
“Because Libor relied on self-reported estimates, it was possible for a bank to submit a rate that was artificially high or low, thus making certain financial holdings more profitable.”
“Libor could no longer be used to calculate new deals as of Dec. 31 — more than six years after a former UBS trader was jailed for his efforts to manipulate it and others were fired, charged or acquitted.”
Interesting history on LIBOR, was created related to an Iran loan in the 1960s. Doesnt seem that long ago but that is pushing 50 years now. Imagine apple told you that mouse will no longer be supported on OSX, please everyone switch to touch screen -- the adoption would be slow at best.
And it was mostly fine before the world started drowning in derivatives and derivatives of derivatives.
It was a traumatic and tragic moment for me at the time. But in hindsight it was the event that lead to my eyes being opened to the world of Silicon Valley tech companies. Until then, I had naively thought that working as a developer in the IT departments of Wall Street investment banks and hedge funds* was the pinnacle of a SWE career.
* Not referring to places like Citadel or Jane Street or Two Sigma, etc.
I found that being aware of whether you will be part of the cost center or profit center in a company is very useful when deciding where you should work.
Now, 25+ years later, I am the head of technology (C-Level) for a large financial services firm (Fortune 200). I report to the CEO, I lead thousands, I am handsomely compensated, but I am professionally lonely.
Over the years, I have become very, very good at explaining technology concepts to non-tech peers (I think it was an intrinsic skill that got me here), but honestly, I am exhausted. I don't think I have it in me to explain technical debt, or the importance of investing in our platform, or how to run a build/buy process or why having an engineering culture is so important. I long to work at a company where my work is intrinsically respected. My peers are polite, but treat the work my team does like magic. It felt deferential at first, but now it feels condescending. I think I've done a great job of creating a real technology culture, but in the last year I realized I am never going to turn us into a technology company, no matter how hard I try.
The lesson is - if you want to work at a technology company (revenue is directly generated through licensing or SaaS fees), then don't compromise. You won't be able to change the nature of your employer no matter how high up the ladder you climb.
My litmus test is this: If you couldn't imagine a company installing a former engineer as their CEO, don't consider it a tech company no matter what the leadership claims.
If the executives from the CEO on down fail to understand technology as the source of a serious competitive advantage, then you will be seen merely as a glorified janitor, or maybe plumber. They do absolutely essential work, but nobody respects them.
And one guarantee, if your company (or one you are considering) looks at technology as a cost center, then I can guarantee that they do NOT and WILL NOT see technology as the source of any competitive advantage. You'll be nothing more than a plumber on a team of plumbers who will be ignored, until a pipe breaks, then you'll be blamed for it happening even if they congratulate you for fixing it to your face. Good luck with that.
I got the same advice from my father, but it meant something different. I was told if I went into computer science or any engineering, I'd always be a servant to management and my job would be outsourced to India. I would be easily replaceable. Best to be "close to the money" instead... that was management. Also, to make "real money", I'd have to move up from engineering to management and wouldn't be programming anyway.
So I went to business school. Might as well optimize and skip the engineering step and go straight into managment. And to be even closer to the money: finance degree.
15+ years later, while finance has been fine, I just really like programming. I have a real aptitude for it. Had to teach myself to code, started side hustle online business (finance is still day job). I might get the same salary as a FAANG software engineer (without the skyrocketing stock), but I always wonder what if I did comp sci instead.
Then, on HN I see comments like yours. Many here hate management, or in your case, moved up to the top of IT management and still seem unsatisfied.
Now, I figure grass is greener on there side... Management says they are treated like a cost center and engineering is "closer to the money" as in profit center. Engineers, even in the profit center, gripe at those MBAs who are "closer to the money" as in directing the business plan, budget and timelines.
In my next life, I'll just do what I enjoy and am good at.
That said, even companies where SWEs and tech are the profit center are certainly capable of laying you off, so profit vs cost center isn't really any insurance to avoid that sort of fate. Even at the banks I've worked at the traders (profit center) would face the axe before us lowly peasants in the tech departments.
Rather, it's more an issue of respect...and relative compensation.
The problem with many tech-as-a-cost-center companies is that you will quickly run into a person on that hierarchy who doesn't (often at or near the intersection between tech departments and the profit center business departments).
I also have a problem with ex-developers who work their way up the structure with time. Generally they drift away from the tech and what tech takes and end up serving their higher masters. So you end up with someone who thinks they know what it takes but hasn't actually done it for many years. Literally had this again the other day when I gave an estimate for a piece of work one of the devs had done a decent bit of investigation on. Bluntly told that was too much time from someone who had really no much more info than the subject line of the bug report. Of course who had the weight to get their estimate across.....(not me)
The problem with most tech companies these days is they dont make profits.
Not sure why you would believe this. It is much harder for regulators to interview employees about their activities when they are no longer centrally located for convenient discussions.
Its the same tactic they use in bury investigators with paperwork but for people.
What was the reasoning behind this thinking? While finance often does have some pretty advanced tech behind it, you can find more cutting edge and complex work at purely technical companies. Or are you speaking of compensation?
In the days before the internet, technology was niche (as was the knowledge to develop and operate it) and super expensive, so only mega corps had decent tech to work with.
This reputation persisted for sometime into the new millennium until we started to see more of these scruffy younguns starting to make noise in the business, and techno, spheres.
1. Just plain old not knowing better.
2. Being in NYC, my social circle had a lot of Wall Street finance types. In fact, this is still true - I know more finance people amongst my friends than fellow SWEs.
3. Hacker tech culture wasn't as widespread or widely known back then, especially in NYC. Or maybe it was just me. Going back to #1, I thought things couldn't get any better than dressing up in business casual monkey suits every morning and being ordered to dance like a monkey by a hotshot trader in hopes of picking up after his glorious scraps.
4. I thought the money wasn't half bad, making like $80-90k back then. I thought no one could pay better than the Masters of the Universe on Wall Street.
I don't know what a company like Google was paying their engineers back then, but I think it's a somewhat recent phenomenon where tech SWE compensation grossly and utterly outpaced tech-in-finance SWE compensation. Again, I'm not referring to ultra-elite finance firms like Jane Street or Citadel where I understand SWEs do get paid on par (or better) than FAANG.
I distinctly recall even ignoring a Google recruiter's solicitations back then (2007? 2008?) because I naively thought working at my investment bank was utterly superior to anything Google could offer.
Now I know better, and knowing is half the battle.
They are already behind some big tech in total comp, so brain drain is inevitable.
I personally find finance pretty interesting, but mortgage contracts are pretty much the definition of boring for many people.
1. Its LIBOR, not Libor
2. LIBOR is not a number, its a rate. Thats literally what the "R" stands for
3. LIBOR cant die, as its not a person
Its just an awful title because it calls LIBOR a number, which its not, then goes on to treat it as a human, which its also not.
As for "treating it as a human", the financial industry has long used the term "mr market". Ie it's sort of an inside joke.
To the target audience it is not an awful title, nor an awful article.
As for the semantics of numbers v rates - the rate is specified in numbers. And LIBOR v Libor - the number was used so extensively for so long many people didn't even know what the acronym meant. It's like correcting someone for calling a tissue a kleenex.
Anyways, it seems like there is a need for a system that is transparent, immutable, and accessible to everyone... Crypto? Matt Damon is calling.
Cheers
Interest rate swaps, credit, loans, fixed income, exotic derivatives, CDO, whatever.
You have to hedge the interest rate risk somewhere or get stuck with huge collateral requirements/XVA.
Cryptocurrencies don't solve the problem since they're largely traded on opaque exchanges and other l2 solutions even less trustworthy than the libor cartel.
While there are rumors of wash tradings on some exchanges, the price at which such trades would be going will be constrained by the larger network, and the risk of triangular trade will limit the possible divergences to a larger spread (instead of going one direction only)
Add enough data, and you may get something that would be almost impossible to trick, simply due to the sheer number of exchanges, and bots that would gladly take the money of those who would try to rig the game.
Maybe the big players could agree to record a decent portion of the transactions transparently across a clearinghouse or blockchain that would be audited manually or cryptographically. Then you have real-time approximation of rates.