I would like to thank Tim Taylor and Rob Harris for their help in preparing these remarks.
1
## The Bank and Benchmark Reform
Speech given by
Chris Salmon, Executive Director, Markets, Bank of England
Roundtable on Sterling Risk-Free Reference Rates, London
6 July 2017
## Introduction
Thank you to NatWest Markets for hosting this event, and let me add my own welcome and thanks to you all for coming.
The Governor has set out the need for a substantial transition from Libor to near-risk free alternatives.
I want to expand on his remarks by explaining what the Bank has been doing to advance this objective, and why. And, importantly, looking ahead, I want to highlight the need for action from many of you.
## The financial system has a Libor dependency
But let me first paint a picture of where we are starting from. The financial system has a Libor dependency. I'm sure this won't be news to you, but it bears some examination.
Usage of term Libor - that is the 3-month and 6-month tenors - is widespread, especially in Sterling and US Dollar markets, but also in Japanese Yen and Swiss Franc.
The greatest concentration of financial contracts referencing Libor is in derivatives markets. For example, in sterling, more than £30 trillion of 'over-the-counter' derivatives and £4 trillion of exchange traded derivatives (on a gross notional basis) reference Libor. USD Libor derivatives probably run to the hundreds of trillions. The economic purposes of these contracts include hedging or gaining exposure to: outright interest rate risk, basis risk, cross currency risk, and interest rate volatility. Banks, pension funds, insurers and hedge funds are all heavy users of Libor derivatives.
Libor is also important for the real economy in the UK, with more than £200bn of SME and corporate loans, around £125bn of Floating Rate Notes and £200bn of structured debt referenced to sterling Libor. Libor's role here is typically to provide a transparent and widely accepted reference point for the floating interest payments on loans.
The dominance of Libor might be taken as evidence of the overwhelming need for Libor-linked financial contracts. However I do not believe this is the case.
A large part of the reason Libor is so widely used is due to what economists call network effects. Using Libor is attractive because Libor instruments are liquid. This is self-reinforcing - liquidity drives use of Libor, which drives liquidity.
In many cases the use of other benchmarks would be preferable for users. For example, most users of interest rate swaps are seeking to hedge the general level of interest rates, not bank credit risk. For this, near risk-free rates would be more appropriate. However, liquidity for long-dated sterling swaps is firmly rooted in contracts referencing six-month sterling Libor, and individual participants cannot change this.
2
In effect the benefits of deep and liquid markets - ease of price discovery, minimisation of basis risk, tight bid/offer spreads - have outweighed the fact that Libor may not be a perfect match to users' underlying needs.
Whilst users of Libor clearly benefit from this concentration of liquidity, the flipside is that it makes the adoption of alternatives to Libor a very challenging task.
These network effects are reinforced by market convention, accounting practices, and the frictional costs of changing firms' systems and risk models. And use of Libor is perpetuated by the need to hedge and manage legacy positions. Overall, there is substantial inertia in the use of Libor.
## The case for change
The financial system's Libor dependency matters for two reasons.
First, as I have argued, in many applications users would be better served by a near risk-free rate: the bank credit component of Libor is in these cases neither necessary nor appropriate. Derivatives markets in particular could be more effective if there were liquidity in alternative reference rates.
The prospect of improved market effectiveness might not, on its own, justify change.
A second factor adds powerful impetus to the case for change. As the Governor emphasised, the truth is that the term deposit markets which underpin Libor fixings are no longer a liquid source of bank funding. Even a reformed Libor will need to rely on expert judgement to supplement transactional data. That is not a desirable state of affairs. The system-wide dependence on term Libor fixings is an unnecessary vulnerability.
A situation in which robust transactions-based benchmarks were more widely used would improve the resilience of the financial system. This is why the FSB recommended the development and implementation of RFRs.
And that is why the Bank - given its interest in effective markets and responsibility for financial stability - is committed to supporting the adoption of near risk-free rates as alternatives to Libor. Put simply, we want to see a transition to a less Libor-centric world.
## What is the Bank doing?
We have taken two important steps to support the implementation of a sterling risk free rate.
We have taken on the administration of SONIA, the sterling unsecured overnight interest benchmark, and instituted a reform process to strengthen it.
3
And we have set up the Sterling Risk Free Rate Working Group to forge a market consensus on the appropriate near risk-free rate and to coordinate wider adoption.
## SONIA reform
The Bank's aim in reforming SONIA has been to strengthen a benchmark which is already critical for sterling financial markets, but which is currently based on a market for brokered deposits which has limited transaction volumes.
A key change is to broaden the input data to include all overnight unsecured deposits, not just brokered deposits.
The Bank was uniquely placed to undertake this reform because we have the necessary data, collected under statutory powers. Indeed that is why we took it upon ourselves to reform SONIA.
We started collecting daily data on money market transactions in February 2016 from over 50 banks, building societies and investment firms. These data give us a much clearer picture than we have previously had of what is going on in sterling money markets. A subset of the data will be used for reformed SONIA.
The reformed rate will be much more robust, with daily transaction volumes three to four times greater than current SONIA. Any change in SONIA's properties will be minimal - so far the reformed rate has been around one basis point lower than on the current methodology.
Another key reform is that the Bank has consulted on and will implement a changed definition for SONIA. The new definition will allow the methodology and input data of SONIA to evolve if material changes to market structure or functioning threaten to undermine the integrity of the rate. This unique feature will ensure the robustness of SONIA over the long term, giving users confidence that their contracts will not be disrupted.
Let me emphasise that, from the perspective of a user of SONIA, implementing these reforms should be seamless. However, one operational consequence that users should be aware of and prepare for is that the publication time for SONIA will move from the evening of the day in question to the morning of the following day. The Bank is working with ISDA, LCH and others to ensure this change is implemented smoothly.
These reforms, including to the timing of SONIA publication, will take effect from March/April next year.
Working Group on Sterling Risk-Free Reference Rates
Let me now turn to the issue of forging a market consensus on the appropriate risk free rate.
4
Here the Bank set up the Risk Free Rate Working Group with the objectives of identifying an RFR and promoting its use as an alternative to sterling Libor.
We used our convening power to do this because we recognised that individual market participants, acting alone, cannot overcome network effects. Only coordination and cooperation across a broad spectrum of market participants can create the conditions for change.
We also believe that the solution must come from the market - that is why the Group is market-led.
As you know, and having considered the merits of two secured benchmarks alongside SONIA, the Group has now reached a decision to recommend SONIA as the RFR.
This is welcome progress: the main priority from the Bank's perspective was for the market to coalesce. We were committed to supporting whatever decision the Group made. That is what having a market-led process means.
As it happens we were not surprised by the outcome because the transition path to a secured rate would clearly be much more difficult than it will be for SONIA.
A consequence of the Group's decision is that SONIA will grow in importance. Administering a systemically important benchmark is a new departure for the Bank - but it is a role we accept given our statutory ability to collect the underlying data. We will administer SONIA as a utility, for the benefit of the market, on a cost-recovery basis.
I know François will say more about how the Sterling Working Group made its decision. For my part, I would like to take this opportunity to thank François, members of the Working Group, and their institutions, for their commitment and hard work in getting to this point.
However the task is not yet finished. Attention now turns to the adoption of SONIA as an alternative to sterling Libor.
The Working Group has an important ongoing coordination role to play.
The main focus so far has been on the derivatives markets. Indeed we invited dealers to form the Working Group in the first place because our experience of derivatives reform suggests they can be most influential in catalysing a transition.
But it is stating the obvious to say that demand for derivatives stems partly from the desire to hedge underlying cash market instruments. That reinforces the need, ultimately, for a broadly-based transition.
5
That can be to the benefit of any user of Libor who bears exposure to bank credit risk not through preference but through lack of established alternatives. I personally suspect that many users would be in this category.
In fact the scope for SONIA adoption in markets beyond derivatives is a key issue we want to explore - and will be discussed in the panel sessions later today.
We do not underestimate the complexity of reducing the financial system's Libor dependency. We are at the beginning of the process, and at this early stage the challenges are not all clearly in focus.
That is why the engagement, help and support of the wider community of users of sterling interest rate benchmarks - issuers, investors, banks, as well as dealers - will be essential.
The Working Group needs your help to identify the impediments to transition as you see them - and, equally, to identify where there are opportunities. That is why you are here today.
## The way ahead
So what happens now?
I have argued that, over time, we need to see a broad transition away from sterling Libor towards SONIA.
That will require changes in the behaviour of firms that carry some cost.
But, where SONIA better matches the risks firms are hedging, there will be long-term benefits. Moreover the financial system will be safer.
The Bank has already laid the groundwork for transition by reforming SONIA and creating the risk-free rate working group.
We will continue to use our convening power to support the Working Group in driving adoption of SONIA.
We will help market participants to understand changes in the environment and how they might be affected.
And, where there are coordination issues we will stand ready to assist the market in solving them.
What we are asking from you is to engage with the Working Group.
As I've already said, we are under no illusions that transition will be easy. That is why it is so important that you help identify the challenges and opportunities of adopting SONIA more broadly.
The Working Group has produced a paper which seeks your input on a number of specific questions. Please do respond.
6
It also invites participation in the work of the Group over coming months. Please do get involved in shaping the future of sterling markets - starting from now.
I am looking forward to hearing your thoughts this morning.
7