Game of Benchmarks: Season 1 - Episode 1: The king is dying.
Over the last years, I have reported different initiatives around the changes of benchmark interest rate indices.
With the announced death of Libor recently accelerated by FCA’s CEO, it is probably a good time to discuss what a good index would be and how the replacement of the current indices could take place.
I will present my opinion and suggestions regarding the interest rate benchmarks in a series of blog. Several episode are in the making, and certainly more than one season will be required before the drama unfold fully! The script of the next seasons still has to be written, by the current readers maybe.
The series is titled
LIBOR has been called “the most important number in the world”. It was widely known by this nickname only well passed its prime, when people started to question its significance in the wake of the Global Financial Crisis.
Why is it the most important number in the world?
Before explaining why it is important, I have to explain it is a number. It is a number in the literal sense, i.e. an arithmetical value, not a physical measure. For human kind, some measures are more important than the LIBOR number(s), like the temperature at the Earth surface. LIBOR is simply a number, an abstract quantity, from which some economical consequences, like the rate on loan or derivatives, are derived — “derived” and “derivatives” have the same root. By definition (1), LIBOR is at its core not a measurement or a economical reality, it is part of the imagination of a small number of “rate setters”. It is not a reality but a guess by someone about what someone else would offer him(2).
The number was at the start an internal convenience created by an association (British Banker Association - BBA) for the benefit of its members. Since, it had a dazzling career that brought it to royalty or maybe even to “imperium”, as it reigned not on a mere kingdom but on the world. How did this local man became a global force? With the help of men in power. Obviously, the bankers helped it on its ascent, this is expected and fair as it was their creature. But it was also helped, intentionally or not, by regulators, lawmakers and central bankers. Each time one of those public servants referred to it in regulation, law or analysis, it gave it more power. Soon LIBOR took over the world, not only for derivatives and interbank agreements, but also for almost everything linked to interest rates: retail mortgages, corporate lending, economic assessment, central bank policies, etc. Not that the LIBOR was the best number for any of those tasks, only that it was the only one available and nobody beyond its creators questioned it. It ascended to the thrones not on its own strength but from the support and ignorance of men in power.
The interest rate market is, by far, the most active asset class, well above equities or commodities. The traded notional of interest rate products is well above the amount of equities. Even if equity indices (like S&P 500, Nasdaq 100 and Eurostoxx) are attracting more news coverage, it is not where most of the activity of the financial markets take place. As a number, USD-LIBOR-3M is more important than SPX or SX5E.
The “L” in LIBOR stands for London and the first “B” in BBA for British, nevertheless the numbers have a worldwide impact. The reason is that the benchmarks are publish in many currencies — originally 10 and only 5 since 2012 (3)— and in particular in the dominant currency in the financial markets, the USD. The British Empire still survives, just not where you expected it. It has always been a surprise for me that the US market has accepted to trade most of its interest rate market based on an number coming from abroad and decided while they sleep (LIBOR is set at 11:00 am London time, which is most of the time 6:00am New York).
It is important to keep that history in mind, not to put the blame on the actors of the past, but to design a brighter future. To create a new benchmark, you have to be clear why you need it, how it will be used and what features you want from him.
The quick answer is: to save time. There are many circumstances where financial processes need to be repeated frequently by many people. For example you want to borrow money on a daily basis to manage the inventory of your shop. You borrow short term (overnight) and many shops like yours do the same. You have the choice, either every day you go to the bank and discuss with your account manager, who himself has to check with its treasurer, to agree on a rate, or you agree a reference figure that will be used for all your borrowings. The reference should not be set by one of the parties, or someone who has a conflict of interest with the parties. The best is to have as a reference a number which is a publicly available information proposed by a trusted third party source.
The LIBOR started in that way to simplify the management of syndicated loans. Banks submitted their rates to their own association which was redistributing it to its members. There was no conflict of interest, or more exactly all the participants had the same conflict of interest and the same power. Then the indices were used in other circumstances that are somehow financially similar. The benchmarks were used in interest rate swaps — interbank or not —, for corporate bonds, commercial mortgages, retail mortgages, the remuneration of collateral, etc.
We need benchmark indices for convenience. But do we need one with imperium, like LIBOR is? Not really, we could continue to have LIBOR for interbank lending, some repo based benchmark for secure lending, some government bond benchmark and probably others. But I have to warn you already, if you have not read yet my book on the multi-curve framework — but I cannot imagine that to be the case —, each time you add a benchmark rate somewhere, your valuation framework is becoming more complex. Maybe you don’t care about valuation framework as you are not a bank, but that would be a mistake. More complex valuation framework means more difficult accounting for banks and corporate, disputes and trials to resolve them, difficulty to rescind contracts even amicably, more regulation, etc. It means unnecessary costs that in fine are paid by the end consumers, i.e. you, my dear reader. Human kind need to produce more food and shelter so that some of the members of the species (me in particular) can deal with the minutiae of multi-curve valuation and disputes around it.
More elements regarding the requirements of a good benchmark will be discussed in Episode 3: The Great Pretender.
We need benchmarks, probably several of them, but not too many of them either.
There are certainly a number of other benchmarks that play a role similar to the role LIBOR plays in USD, GBP, JPY and CHF (and EUR). Some of the other benchmarks, also based on interbank lending are: EURIBOR (EUR), CIBOR (DKK), HIBOR (HKD), TIBOR (JPY). In EUR, the EUR-LIBOR exists and replaced the pre-euro FRF, DEM, BEF and other LIBORs. But most (or perhaps even all) of the new instruments in EUR are linked to EUR-EURIBOR, and only the legacy trades to EUR-LIBOR. In JPY, a large portion of the market is in JPY-LIBOR, but an active market exists in JPY-TIBOR. There is even two versions of TIBOR, a domestic one and a euro-JPY one.
Other benchmarks play a similar role but are not based on interbank lending. In AUS, the leading Benchmark is BBSW, based on bank bills and in CAD the benchmark is CDOR based on bankers' acceptances.
There are other types of benchmarks that are very important but play a slightly different role. They are the overnight indices. They are usually computed as the weighted average of the rates on actual trades. They are more than “numbers”, they are the measurement of an actual economic reality. The best known are Fed Fund effective (USD), EONIA (EUR) and SONIA (GBP)
But none of those indices are real pretender to replace the LIBOR. In the next episode, I will review why it is so.
Next episode:
(1) The exact question ask to the LIBOR rate setters is: “At what rate could you borrow funds, were you to do so by asking for and then accepting interbank offers in a reasonable market size just prior to 11 am London time?” See the ICE LIBOR page.
(2) EURIBOR has even one extra level of guessing, as it is the rate at "which euro interbank term deposits are offered by one prime bank to another”. Each panelist has to guess what the other panelists’ banks will ask each other.
(3) See the Interest Rate Instruments and market conventions guide
With the announced death of Libor recently accelerated by FCA’s CEO, it is probably a good time to discuss what a good index would be and how the replacement of the current indices could take place.
I will present my opinion and suggestions regarding the interest rate benchmarks in a series of blog. Several episode are in the making, and certainly more than one season will be required before the drama unfold fully! The script of the next seasons still has to be written, by the current readers maybe.
The series is titled
Game of Benchmarks.
Game of Benchmarks: a no-fantasy series with no blood and no sex but plenty of greed, manipulation and money.
Season 1 - Episode 1: The king is dying.
The most important number in the world that nobody knew existed — The king is dying — Where is the successor?
The most important number in the world!
LIBOR has been called “the most important number in the world”. It was widely known by this nickname only well passed its prime, when people started to question its significance in the wake of the Global Financial Crisis.
Why is it the most important number in the world?
Before explaining why it is important, I have to explain it is a number. It is a number in the literal sense, i.e. an arithmetical value, not a physical measure. For human kind, some measures are more important than the LIBOR number(s), like the temperature at the Earth surface. LIBOR is simply a number, an abstract quantity, from which some economical consequences, like the rate on loan or derivatives, are derived — “derived” and “derivatives” have the same root. By definition (1), LIBOR is at its core not a measurement or a economical reality, it is part of the imagination of a small number of “rate setters”. It is not a reality but a guess by someone about what someone else would offer him(2).
The number was at the start an internal convenience created by an association (British Banker Association - BBA) for the benefit of its members. Since, it had a dazzling career that brought it to royalty or maybe even to “imperium”, as it reigned not on a mere kingdom but on the world. How did this local man became a global force? With the help of men in power. Obviously, the bankers helped it on its ascent, this is expected and fair as it was their creature. But it was also helped, intentionally or not, by regulators, lawmakers and central bankers. Each time one of those public servants referred to it in regulation, law or analysis, it gave it more power. Soon LIBOR took over the world, not only for derivatives and interbank agreements, but also for almost everything linked to interest rates: retail mortgages, corporate lending, economic assessment, central bank policies, etc. Not that the LIBOR was the best number for any of those tasks, only that it was the only one available and nobody beyond its creators questioned it. It ascended to the thrones not on its own strength but from the support and ignorance of men in power.
The interest rate market is, by far, the most active asset class, well above equities or commodities. The traded notional of interest rate products is well above the amount of equities. Even if equity indices (like S&P 500, Nasdaq 100 and Eurostoxx) are attracting more news coverage, it is not where most of the activity of the financial markets take place. As a number, USD-LIBOR-3M is more important than SPX or SX5E.
The “L” in LIBOR stands for London and the first “B” in BBA for British, nevertheless the numbers have a worldwide impact. The reason is that the benchmarks are publish in many currencies — originally 10 and only 5 since 2012 (3)— and in particular in the dominant currency in the financial markets, the USD. The British Empire still survives, just not where you expected it. It has always been a surprise for me that the US market has accepted to trade most of its interest rate market based on an number coming from abroad and decided while they sleep (LIBOR is set at 11:00 am London time, which is most of the time 6:00am New York).
It is important to keep that history in mind, not to put the blame on the actors of the past, but to design a brighter future. To create a new benchmark, you have to be clear why you need it, how it will be used and what features you want from him.
Why do we need benchmark indices?
The quick answer is: to save time. There are many circumstances where financial processes need to be repeated frequently by many people. For example you want to borrow money on a daily basis to manage the inventory of your shop. You borrow short term (overnight) and many shops like yours do the same. You have the choice, either every day you go to the bank and discuss with your account manager, who himself has to check with its treasurer, to agree on a rate, or you agree a reference figure that will be used for all your borrowings. The reference should not be set by one of the parties, or someone who has a conflict of interest with the parties. The best is to have as a reference a number which is a publicly available information proposed by a trusted third party source.
The LIBOR started in that way to simplify the management of syndicated loans. Banks submitted their rates to their own association which was redistributing it to its members. There was no conflict of interest, or more exactly all the participants had the same conflict of interest and the same power. Then the indices were used in other circumstances that are somehow financially similar. The benchmarks were used in interest rate swaps — interbank or not —, for corporate bonds, commercial mortgages, retail mortgages, the remuneration of collateral, etc.
We need benchmark indices for convenience. But do we need one with imperium, like LIBOR is? Not really, we could continue to have LIBOR for interbank lending, some repo based benchmark for secure lending, some government bond benchmark and probably others. But I have to warn you already, if you have not read yet my book on the multi-curve framework — but I cannot imagine that to be the case —, each time you add a benchmark rate somewhere, your valuation framework is becoming more complex. Maybe you don’t care about valuation framework as you are not a bank, but that would be a mistake. More complex valuation framework means more difficult accounting for banks and corporate, disputes and trials to resolve them, difficulty to rescind contracts even amicably, more regulation, etc. It means unnecessary costs that in fine are paid by the end consumers, i.e. you, my dear reader. Human kind need to produce more food and shelter so that some of the members of the species (me in particular) can deal with the minutiae of multi-curve valuation and disputes around it.
More elements regarding the requirements of a good benchmark will be discussed in Episode 3: The Great Pretender.
We need benchmarks, probably several of them, but not too many of them either.
What are the other indices similar to LIBOR?
There are certainly a number of other benchmarks that play a role similar to the role LIBOR plays in USD, GBP, JPY and CHF (and EUR). Some of the other benchmarks, also based on interbank lending are: EURIBOR (EUR), CIBOR (DKK), HIBOR (HKD), TIBOR (JPY). In EUR, the EUR-LIBOR exists and replaced the pre-euro FRF, DEM, BEF and other LIBORs. But most (or perhaps even all) of the new instruments in EUR are linked to EUR-EURIBOR, and only the legacy trades to EUR-LIBOR. In JPY, a large portion of the market is in JPY-LIBOR, but an active market exists in JPY-TIBOR. There is even two versions of TIBOR, a domestic one and a euro-JPY one.
Other benchmarks play a similar role but are not based on interbank lending. In AUS, the leading Benchmark is BBSW, based on bank bills and in CAD the benchmark is CDOR based on bankers' acceptances.
There are other types of benchmarks that are very important but play a slightly different role. They are the overnight indices. They are usually computed as the weighted average of the rates on actual trades. They are more than “numbers”, they are the measurement of an actual economic reality. The best known are Fed Fund effective (USD), EONIA (EUR) and SONIA (GBP)
But none of those indices are real pretender to replace the LIBOR. In the next episode, I will review why it is so.
Summary: The previous king is dying, from his own fault — manipulation — and from the same causes that brought it to the thrones — used where it shouldn’t have. For convenience, one or several successors would be very useful. No great pretender to the thrones as of yet.
Next episode:
Game of Benchmarks. Episode 2: Nobody is running for the throne.
(1) The exact question ask to the LIBOR rate setters is: “At what rate could you borrow funds, were you to do so by asking for and then accepting interbank offers in a reasonable market size just prior to 11 am London time?” See the ICE LIBOR page.
(2) EURIBOR has even one extra level of guessing, as it is the rate at "which euro interbank term deposits are offered by one prime bank to another”. Each panelist has to guess what the other panelists’ banks will ask each other.
(3) See the Interest Rate Instruments and market conventions guide
Hi Marc,
ReplyDeleteThanks for your post it's very interesting.
Would you consider it a good idea to rely on a transaction based index (on repos for instance) coming directly from global trade repositories like dtcc ?
Do you think that moving to a low credit risk index (like overnight or collateralized index) will de facto remove the multi-curve paradigm created by the risky libor ?
would it be a good idea to consider several publishing hour just like what is done for swap rate at ice or would it create a new basis (again) ?
Thanks,
AyoubZ
For me, one of the qualities of a benchmark should be transparency (more on that in a forthcoming episode). Global repositories are there for some form of transparency. That could be a good idea to link both. I don’t know enough about the details of the trade repositories to have a definitive answer. Who is publishing on them, what information is provided, at what time?
DeleteFor example the counterparties ID are not provided. It means that a benchmark based on Interbank market between banks of prime quality could not be computed from the repository; there is not enough information to discriminate between the relevant trades. Also, I think that large transactions notional is not provided/capped.
For the disappearance of the multi-curve, I don’t think it is imminent! Even if the main benchmark is changed, there will still be multiple benchmarks. I don’t know how the future will turn out, but I can see one alternative where there is not one benchmark but multiple benchmarks (ON, term + repo, interbank, government, etc.) and an even greater need of the multi-curve framework.
For you last question, my opinion will be part of “the great pretender” episode.