Libor has come under huge scrutiny since Barclays was fined for manipulating it. While few bankers believe it can be replaced, even its staunchest supporters say it needs to be reformed. But it is not obvious how.
When, in the run up to late 2007, derivatives traders at Barclays Capital were sending e-mails to colleagues on the money market desk asking them to manipulate their London Interbank Offered Rate (Libor) submissions – and celebrating with bottles of Bollinger when their requests were met – they had little idea just how explosive the repercussions would eventually be.
Soon after Barclays revealed it had been hit with a $450m fine from UK and US regulators in late June for fixing Libor, its three top bankers – chairman Marcus Agius, chief executive Bob Diamond and chief operating officer Jerry del Missier – were forced to resign (although Mr Agius has since resumed his role to provide stability and find a new CEO).
Libor itself has come under huge scrutiny. Policy-makers across the globe have attacked the process by which it is set and even its very concept. Ben Bernanke, chairman of the US Federal Reserve, told US Congress that the Libor system was “structurally flawed” and “a major problem... for the confidence in the financial system”.
Libor, a daily measure of unsecured interbank lending rates across 10 currencies and 15 maturities, has long had its critics. Ever since it was introduced in 1986 under the auspices of the British Bankers’ Association (BBA), bankers have described it as an imperfect benchmark. Scepticism was especially rife amid the turmoil in 2008 and early 2009. Some investors, citing banks’ widening credit default swaps and bond yields, felt they were fudging their Libor quotes to try to give the market the impression they were still able to borrow at low rates.
Regulators also had their doubts. In May 2008, Tim Geithner, then head of the Federal Reserve Bank of New York, sent a recently leaked e-mail to Mervyn King, governor of the Bank of England, outlining ways for Libor to be improved.
Enormously significant
But despite its shortcomings, Libor’s importance has only grown with time. It now acts as a reference for hundreds of trillions of dollars (estimates vary from $300tn to as much as $800tn) of securities ranging from complex interest rate derivatives to corporate loans to consumer products such as mortgages. As a result, even a small change in its setting of one or two basis points can have far-reaching consequences.
“Libor matters not so much as an indicator of market sentiment, but because so many contracts are referenced to it,” says Richard Fernand, director of industry relations at the CFA Institute. “While there are alternatives, the bulk of interest rate-related contracts are still based on Libor. It is of enormous significance.”
Nonetheless, thanks to the sheer outrage – among the public as well as in high financial circles – over revelations of Libor-fixing, few believe that the system can continue in its current form. “Change will happen for the simple reason that the public and regulators are demanding it,” says Mr Fernand. “Everyone is saying that it needs to be reviewed.”
Some technical reforms are almost inevitable. One of the most popular suggestions is to expand the number of banks that make Libor submissions. Between six and 18 are on the panels for each of the currencies. Having more would likely increase Libor rates, given that the existing contributors typically include the world’s best-rated banks. New entrants would thus decrease their average credit quality (Credit Suisse says US dollar Libor rates could rise by 15 basis points [bps] if the group of submitters for the currency was expanded from 18 to 30-odd). But more importantly, say several commentators, including more banks would decrease the consequences of a single lender trying to manipulate the final Libor rate. And it should ensure a figure more representative of the overall interbank market. “Expanding the panels would give you extra data points to work with,” says a money markets analyst in New York. “The quality of the metric would be improved.”
While there are alternatives, the bulk of interest rate-related contracts are still based on Libor. It is of enormous significance
In his leaked e-mail, Mr Geithner called for more local banks to be Libor contributors. He noted that only three US institutions – Bank of America, Citi and JPMorgan – were among the 18 US dollar Libor submitters, even though several more were active in the London interbank market. The situation is starker in other currencies. There are no local banks on the Swedish krone, Danish krone or New Zealand dollar panels, and only one (Commonwealth Bank of Australia) on that for Australian dollars.
Several commentators say the lack of local banks is strange given the BBA’s criteria of picking them based on the scale of their market activity and perceived expertise in the respective currencies. “If you wanted to know interbank rates in Swedish krone, is an international bank that is on the panel but probably borrows in the currency once in a blue moon going to be able to tell you?” says another analyst. “You’re probably going to get a better answer from SEB or Nordea.”
New York’s sleeping
Another idea is to change when Libor is set. At present, panellists are asked at 11am London time to give their quotes. Some argue it would be better if they made submissions only when the main markets for each currency are open. US dollar Libor might be improved, for example, if banks were able to take into account the changes in liquidity that often occur when New York, five hours behind London, starts trading. “If you’re going to set the dollar rate, do it when the US is open,” says John Rathbone, founder of advisory firm JC Rathbone Associates. “Equally, the yen rate should be set when Japan is open.”
More radical suggestions include making the quotes of each bank private and only publishing the final Libor figures (which are calculated as the mean of the submissions, excluding the top and bottom quartiles). Proponents say it is precisely because of the transparency of Libor that banks might falsify their rates, particularly in distressed markets when investors will be watching closely for any signs of a liquidity crisis.
Yet Libor’s lack of opacity, at least in the sense that all submissions are public, is seen as one of its strengths. In any case, now would hardly be the right time to take away a layer of accountability. “It wouldn’t help with the public’s perception if you had less transparency,” says Michael Cloherty, head of US interest rate strategy at RBC Capital. “And public perception is a greater near-term issue [than technical fixes].”
Almost all Libor-watchers agree that there are too many rates. Some 150 are calculated every day. Yet few of the 15 tenors, which range from overnight to 12 months, are considered important benchmarks. The vast bulk of instruments based on Libor are referenced to the three-month and, to a lesser extent, the six- and 12-month rates. The New York Fed made this point to the UK central bank in 2008. “For tenors such as the three-month tenor, Libor quotes provide valuable information to the public because of the volume of activity occurring at that tenor, while quotes for tenors at which little or no trading occurs, such as the 11-month, are... less valuable,” Mr Geithner wrote. “The current practice of soliciting rate quotes across 15 tenors, when only a subset of those tenors reflect[s] meaningful market activity, likely leads to more subjective and formulaic responses across all tenors.”
Flip the question?
Since 1998, Libor has been calculated by the BBA asking panellists the rates at which they think they could borrow cash. The CFA Institute’s Mr Fernand says it might be possible for the question to be modified, with contributors instead being asked for the rates at which they would lend to their counterparts. The banks could submit either a single, average lending rate they would offer the rest of the panellists, or provide rates for each of them.
The advantage is that banks would no longer be revealing potentially harmful information about their own ability to fund. “The one institution you wouldn’t be commenting on would be yourself,” says a banker involved in setting Libor.
Yet it would complicate the Libor process. The submitters would have to make a judgement about the credit quality of several different institutions. And they might be reluctant to reveal details about their capacity or willingness to lend. “It would be quite revolutionary [to flip the question],” says Juan Carlos Martinez Oliva, a visiting fellow at the Peterson Institute for International Economics in Washington, DC. “From an academic [point of view] it’s perfect. In reality, I’m not sure.”
A major criticism of Libor is that it is, in most cases, based on hypothetical trades. Contributors are asked for estimates of where they think they can borrow, rather than the rates at which they actually borrow. But this is largely unavoidable, given that no bank, regardless of its size, funds across multiple tenors and currencies each day.
The issue of whether to use perceived or real borrowing rates strikes at the heart of how to improve Libor. It has become increasingly important since late 2008, when interbank rates seized up. While better today, they have yet to recover fully. “The interbank market is still fairly illiquid, even if it’s nothing like as bad as it was in late 2008 and early 2009,” says Mr Rathbone, pointing out that three-month sterling Libor rates are roughly 35bps higher than the Bank of England’s base rate. “In a fully liquid market without anyone expecting base rates to move in either direction in the foreseeable future – which is where we are today – you wouldn’t think it would be more than 10bps or 12bps.”
But calls for a market-based or auction setting, whereby banks are made to borrow to determine their rates, are unlikely to be heeded, say analysts. Submitters would resist proposals forcing them to fund on a regular basis when on any given day they might not need to, particularly across a range of currencies and maturities.
Elusive substitutes
As such, some have argued for Libor to be replaced with benchmarks that are derived from actual trades. One suggestion is to use banks’ issuance of commercial paper (CP), or short-term bonds. But while yields on these could be tracked, they would not directly reflect conditions in the interbank market. “CP is an imperfect benchmark because that’s where a money market would lend to an issuer, not where another bank would lend to it,” says Mr Cloherty. “It’s a different beast.”
Moreover, many lenders are cutting their CP issuance because of Basel III. Among its regulations, banks have to hold liquid assets against debt maturing in less than 30 days. Since they would also be required to hold Tier 1 capital of 3% on top of these assets, the effect is to make it more expensive than before to borrow for under a month or even three months (as the rules would apply for as much as one-third of the life of the debt). “It’s extraordinarily difficult to head towards a transaction-orientated rate when Basel III has the potential to knock out all your benchmarks,” says RBC Capital’s Mr Cloherty.
It wouldn’t help with the public’s perception if [Libor had] less transparency. And public perception is a greater near-term issue [than technical fixes]
Interest rates that track interbank trades do exist. Among them are Sonia (Sterling Overnight Interbank Average Rate), which is a weighted average of borrowing rates between banks, and its equivalent for secured lending, Ronia (Repurchase Overnight Index Average). For the US market, the Federal Reserve funds effective rate, which is based on the trading of balances held at the Fed, could be used as an unsecured benchmark. And a newly launched repo futures rate might work as an equivalent for secured transactions.
But few of these benchmarks are liquid beyond an overnight tenor, especially when it comes to currencies other then the dollar, euro, sterling and yen. “Do we want rates across the maturity spectrum?” asks Mr Fernand. “If so, these [benchmarks] won’t fit the bill. If you want transaction-based data, you’re going to have to accept that there’ll be very little of it.”
Others argue that Libor was never intended to be a benchmark based on actual trades, and that changing it would increase its volatility. They say that using estimates smooths the rate in the medium term, which is what the derivatives market and corporate borrowers of Libor-based loans require. “Using a poll, rather than a traded rate, has its advantages,” says another interest rates analyst. “You are essentially asking professionals who are in these money markets every day to give their best estimates of where they could borrow cash. On any given day, those markets might not be that active. But in that situation, a traded rate might end up being based on one or two transactions that aren’t representative of the overall market, the result being volatility.”
Libor’s future
Most bankers and investors agree that there are no obvious substitutes for Libor and that replacing or fundamentally changing it would not help financial markets. “It’s very difficult to think of something better than the basic system,” says Mr Rathbone. “People have been looking at it for a long time, but without success.”
Nonetheless, Libor will change. The BBA began a review before Barclays was fined and is expected to announce proposals later this year. Along with technical reforms, the rules are likely to be tightened to prevent more manipulation. Bankers say accusations of panellists “lowballing” their quotes in distressed times are almost inevitable, given that the markets can be so illiquid an accurate rate might be impossible to determine. They are more worried about the type of fixing that occurred at Barclays, and doubtless many other banks, between 2005 and 2007, when traders enhanced their positions by asking for their institutions’ Libor submissions to be made artificially high or low. Clear punishments for this type of behaviour are needed for Libor to regain the credibility it has lost in the past two months, say commentators. And the UK, as the home of Libor, needs to be able to punish individual bankers specifically for falsifying rates or colluding with others to do so. “The market abuse regime does not apply directly to Libor fixing,” says Mr Fernand. “There’s now a pretty compelling argument that it should do.”
For this to work, the BBA might have to cede some of its oversight of Libor to another entity. That could be the Bank of England or the central banks for each of the currencies. “They’re possibly the best candidates should new actors enter the system,” says Mr Martinez Oliva. “They have the competence, experience and reliability necessary to control the process. They know the local banking market.”
If Libor contributors feel that any new regulations or the fines they face are too onerous, they could decide to leave the panels altogether. They are, after all, on them voluntarily. And bankers say that while there is seemingly little financial benefit to being a submitter, there are big risks, both material and reputational, as Barclays’ fine and the subsequent backlash demonstrated.
Nonetheless, having clear rules about what counts as manipulation and what does not should allay banks’ concerns. “If the mechanism is transparent and works properly, it makes sense for the banks to contribute,” says Mr Martinez Oliva. “Being on a panel is a good advertisement for them, suggesting they are considered reliable and safe.
“Libor has done its job for decades. Reforms would help. But I don’t see any reason to replace it. Rather, the rules need to be improved. After that, Libor can continue to play its crucial role as a benchmark.”
