Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?

rollingstone.com

Post reply on HN
Print view View original

31–40 of 89 posts

Re: Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?

#31

Post by tptacek »

Afforess wrote: post #4

This article contains a rather poor explanation of what LIBOR is and the history around it. Matt Levine does a much better job: https://www.bloomberg.com/view/articles/2017-07-27/the-end-o... Back in the 1960s, a Greek banker in London[1] wanted to find a way for banks to make syndicated floating-rate loans. He found a very simple answer: The banks would lend money to a company, charging their cost of funds plus a sp…

Matt Levine is so much better on these topics than Matt Taibbi that one might hope the site would switch URLs. But, of course, accuracy is not the reason people read Taibbi's financial reporting. They're both humorous writers (though in different ways; Taibbi is the better of the two as stylists), but Taibbi is also emotionally satisfying.

It would be helpful if people kept providing Levine pieces to accompany Taibbi pieces.

Re: Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?

#32

Post by JumpCrisscross »

throw2016 wrote: post #27
JumpCrisscross wrote: post #26

Earlier quoted context omitted.

> regulators...concluded there is no basis for LIBOR Regulators did not conclude this. They concluded (a) better metrics for banks' costs of capital exist ( e.g. the Fed funds rate [1]), (b) the market Libor is based on (wholesale unsecured interbank term lending) is too small and inactive to provide the sort of precision Libor implies and (c) transitioning from Libor will be messy [2]. [1] https://fred.stlouisfed.or…

In other words there is no basis for libor.

> In other words there is no basis for libor

No, there is a basis. The regulator Taibbi cites speaks competently about this [1]; the least actively-traded currency-tenor traded only about once a month. (Every other currency-tenor traded more often.)

That may be a fine frequency for 6-month wholesale interbank rates in Danish krona (which, until recent reforms, was one of the currencies Libor was quoted for [2]). But turning it into a daily rate with three decimal places of precision is silly.

[1] https://www.fca.org.uk/news/speeches/the-future-of-libor

[2] https://en.wikipedia.org/wiki/Libor#Currency

Re: Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?

#33

Post by res0nat0r »

tptacek wrote: post #31
Afforess wrote: post #4

This article contains a rather poor explanation of what LIBOR is and the history around it. Matt Levine does a much better job: https://www.bloomberg.com/view/articles/2017-07-27/the-end-o... Back in the 1960s, a Greek banker in London[1] wanted to find a way for banks to make syndicated floating-rate loans. He found a very simple answer: The banks would lend money to a company, charging their cost of funds plus a sp…

Matt Levine is so much better on these topics than Matt Taibbi that one might hope the site would switch URLs. But, of course, accuracy is not the reason people read Taibbi's financial reporting. They're both humorous writers (though in different ways; Taibbi is the better of the two as stylists), but Taibbi is also emotionally satisfying. It would be helpful if people kept providing Levine pieces to accompany Taibbi…

The Spider Network, about the banker in the center of the scandal from a ways back just came out earlier this year and was an entertaining read.

https://www.amazon.com/dp/0062452983/ref=cm_sw_su_dp

Re: Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?

#35

Post by tigershark »

jkaptur wrote: post #29
JumpCrisscross wrote: post #10

Earlier quoted context omitted.

> there is no market that LIBOR measures Except that's not true. Interbank lending is still a $70 billion market in the United States alone [1]. Small compared to banks' balance sheets and less than the $500 billion from as recently as February 2008, but material nonetheless. Good rule of thumb in finance is to ignore Matt Taibbi. [1] https://fred.stlouisfed.org/series/IBLACBM027NBOG

I'm not a domain expert, but both could be right - isn't the point that some currencies and tenors are very illiquid, not that the entire market doesn't exist?

Neither am I, but you can easily interpolate from Fx rates and from other tenors plus the liquidity informations and create synthetic instruments from others that you already own. Everyone that trades in Fixed Income does (or should do) this. And as I read in another comment for sure everyone has been migrating to alternative benchmark like OIS for a while to get their official risk numbers.

Re: Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?

#36

Post by zzalpha »

JumpCrisscross wrote: post #24
zzalpha wrote: post #16

Earlier quoted context omitted.

And Matt Levine, who said the exact same thing in an article posted in this comment thread? But a second problem is that the banks might not even know. Libor surveys asked banks each day what they would have to pay to borrow money unsecured from other big banks, but over time the banks sort of stopped doing that, particularly in some of the more obscure combinations of tenors and currencies that nonetheless reported…

> We're well into ad hominem I'm not saying it is wrong because Taibbi wrote it. It's wrong because he got basic facts about interbank lending wrong, i.e. that it exists. I'm then passing along my observation that, whenever I've fact checked Taibbi, his facts have tended to be wrong. > Matt Levine...said the exact same thing Taibbi said there is no interbank lending. Libor is totally made up. Levine said that there i…

Honestly this is hair splitting.

Ultimately the point remains: there isn't sufficient market activity to build a real value for Libor so it's basically made up from whole cloth

Your nuance, while interesting if you care to dig deeply, doesn't change the conclusion. It's a distinction without a material difference.

Re: Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?

#37

Post by JumpCrisscross »

jkaptur wrote: post #29
JumpCrisscross wrote: post #10

Earlier quoted context omitted.

> there is no market that LIBOR measures Except that's not true. Interbank lending is still a $70 billion market in the United States alone [1]. Small compared to banks' balance sheets and less than the $500 billion from as recently as February 2008, but material nonetheless. Good rule of thumb in finance is to ignore Matt Taibbi. [1] https://fred.stlouisfed.org/series/IBLACBM027NBOG

I'm not a domain expert, but both could be right - isn't the point that some currencies and tenors are very illiquid, not that the entire market doesn't exist?

> isn't the point that some currencies and tenors are very illiquid

Yes. It's a subtlety bulldozed over in this article because nuance doesn't sell clicks like outrage. Recapitulating an earlier comment, the regulator Taibbi cites speaks competently about this [1]; the least actively-traded currency-tenor traded only about once a month. (Every other currency-tenor traded more often.)

That may be a fine frequency for 6-month wholesale interbank rates in Danish krona (which, until recent reforms, was one of the currencies Libor was quoted for [2]). But turning it into a daily rate with three decimal places of precision is silly.

[1] https://www.fca.org.uk/news/speeches/the-future-of-libor

[2] https://en.wikipedia.org/wiki/Libor#Currency

Re: Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?

#38

Post by lordnacho »

JumpCrisscross wrote: post #12
lordnacho wrote: post #2

Pretty much everything that has derivatives tied to it is manipulated. Option expiries, FX fixes as well. Or it was when I was looking at it. The thing is there are derivatives that are sometimes non-linear, things with triggers and barriers. When some large enough fish has one of these (eg by taking the other side vs a customer) they have an incentive to move the rate in whatever way they can. Whether it's getting s…

> This would happen even in exchange traded options. If you knew the specifics of the settlement window, you would know when the price would go wonky. Exchange-traded options settle on a predictable window; there's no "if you knew" to it [1]. Also, the price goes wonky near expiries because that's when delta approaches one. Market makers switch from hedging with derivatives to hedging with spot; that means jitters. N…

Yes, I used to work in this. "If you knew" should say "Because you knew"...

Might actually be a reason for the exchange to fuzz the expiry (eg random time) but that has its own problems. People do have to look at the options as they expire for legitimate hedging reasons.

Re: Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?

#39

Post by lordnacho »

ThrustVectoring wrote: post #20
lordnacho wrote: post #2

Pretty much everything that has derivatives tied to it is manipulated. Option expiries, FX fixes as well. Or it was when I was looking at it. The thing is there are derivatives that are sometimes non-linear, things with triggers and barriers. When some large enough fish has one of these (eg by taking the other side vs a customer) they have an incentive to move the rate in whatever way they can. Whether it's getting s…

Doesn't this work as an argument against cash-settling options and futures? Owning a giant pile of call options and driving the price up during the settlement window doesn't help you any if you have to unload a massive quantity of the underlying stock.

Forgot to mention it was mainly indexes I was talking about, and they are cash settled.

Single stock options tend to be settled with actual stock, and at least in Europe tend to be a lot less liquid. And there's other problems with trading them, like other people knowing a lot more than the market maker about what's happening. You also have pin risk which seems to be like a magnet.

Re: Is LIBOR, Benchmark for Trillions of Dollars in Transactions, a Lie?

#40

Post by JumpCrisscross »

zzalpha wrote: post #36
JumpCrisscross wrote: post #24

Earlier quoted context omitted.

> We're well into ad hominem I'm not saying it is wrong because Taibbi wrote it. It's wrong because he got basic facts about interbank lending wrong, i.e. that it exists. I'm then passing along my observation that, whenever I've fact checked Taibbi, his facts have tended to be wrong. > Matt Levine...said the exact same thing Taibbi said there is no interbank lending. Libor is totally made up. Levine said that there i…

Honestly this is hair splitting. Ultimately the point remains: there isn't sufficient market activity to build a real value for Libor so it's basically made up from whole cloth Your nuance, while interesting if you care to dig deeply, doesn't change the conclusion. It's a distinction without a material difference.

> there isn't sufficient market activity to build a real value for Libor so it's basically made up from whole cloth

The least active currency-tenor, since deprecated, traded once a month. Most currency-tenors trade many, many, many times a day. There's plenty of market activity to build Libor-esque metrics.

> It's a distinction without a material difference

It's a world of material difference. The Fed Funds rate in the United States is based on the same kind of wholesale unsecured interbank lending as Libor is supposed to be. The metric, and the market it's based on, work.

We can have something like Libor based on market activity. It just won't be published every day for every tenor and currency.

If you just read Taibbi, the answer would seem to be to scrap any attempt at measuring the market because you cannot measure something that does not exist. If you understand the nuance, you walk away better appreciating what (a) went wrong, (b) we should do to improve future metrics and (c) one should look for when evaluating other metrics purporting to do similar things. You also gain an understanding for the kinds of scaling problems financial markets run into, which are quite unlike scaling problems in other contexts.

Post reply on HN