The effect of LIBOR on banks
06 August 2019
RMA’s Global Markets Risk and Â鶹´«Ă˝ Lending councils have identified LIBOR replacement as a significant issue for banks over the next several years
Image: Shutterstock
What is happening with Libor at this time and where do you see it going?
The Financial Conduct Authority (FCA) would like to see the demise of LIBOR sooner rather than later. It has put out a very firm message to the industry that LIBOR will not be available beyond the end of 2021, and it would be a “black swan” event were that not to happen. They are pushing the industry extremely hard to work toward a safe conversion to the use of risk-free rates (RFRs). They’ve used their regulatory powers already in the UK, through a “Dear CEO” letter, to make sure banks are planning for the cessation of LIBOR and, ultimately, the adoption of the alternative risk-free rates, such as the Sterling Overnight Index Average (SONIA) in the UK and the Secured Overnight Financing Rate (SOFR) in the US.
However, there is a recognition that a significant problem lies in the existing LIBOR-referencing contracts that may not be migrated to RFRs by the end of 2021. The total notional amount of instruments—both cash and derivative—estimated back in 2017 was on the order of $350 trillion. Therefore, all market participants, banks, in particular, are obviously looking at their contracts to understand what will happen to them in the event LIBOR is no longer published—the so-called fallback language.
It would be fair to say that the derivative market is significantly ahead of the curve with respect to developing a fallback protocol. There has been an extensive consultation by market participants who would prefer an industry-agreed protocol as a fallback, and the preliminary results of that consultation have already been published.
The presumed LIBOR alternative for the US is secured. What about the proposed rate in other countries?
For the five LIBOR currencies—US dollar, pound sterling, euro, Swiss franc, and Japanese yen—the central banks have gone a long way toward identifying replacement interest rate benchmarks that could be successfully adopted according to the IOSCO principles, which were published in the aftermath of the benchmark-fixing scandals, based on transactions in liquid markets by arm’s-length participants.
Different central banks have come back with different solutions based on the unique characteristics of each market. For the US dollar, SOFR was published in April 2018, based on the daily repo rate for US government securities. For sterling, the Bank of England chose the already existing SONIA. This is an unsecured rate for overnight loan/depo transactions.
Although some changes were made to the underlying methodology at the same time, SONIA had been around for several years prior. The other currencies have headed down the unsecured route, likely because their repo markets are much less deep than in the US. Daily trade activity in transactions that would form the SOFR index is well over $400 billion per day.
This issue of secured versus unsecured rates will matter to those who trade derivatives, especially cross-currency swaps, as they will need to access different types of markets to hedge the resulting exposures.
Why should panel banks continue to contribute to the calculation of LIBOR?
The panel banks will continue to contribute until the end of 2021 because they have already agreed with the FCA to do so. The FCA recognised there was a very big problem: under EU benchmark regulations, there was a risk that LIBOR would be considered ineligible and could die within a two-year time frame. So the FCA asked all panel banks to continue contributing until the end of 2021 to provide a longer transition period.
I expect there was some arm-twisting involved, but the FCA got pretty much all the panel banks to agree to contribute in order to reduce the systemic risk that a poorly planned end of LIBOR could create.
Why wouldn’t panel banks contribute after 2021?
There are some good reasons why a panel bank would choose not to contribute beyond 2021. First, there is an ever-present small risk that a bank could be found failing in its governance of the controlled environments around its LIBOR submissions and thus be subject to fines.
Second, there is the direct cost of maintaining the infrastructure and staff time involved in providing those submissions. Many LIBOR submissions still rely on “expert judgment” due to the paucity of eligible transaction data, so there are systems and control processes that have to be maintained. And there have to be independent second-line-of-defence review processes in place.
Then there will be internal and external audit costs that have to be borne. This all comes at a time of increasing focus on costs. The FCA estimated the annual costs of being a panel bank at £2.4 million. That’s a lot of money at a time of other significant regulatory changes, such as the implementation of Fundamental Review of the Trading Book and Standardised Approach Counterparty Credit Risk.
Would you describe the process for submissions?
The first stage is to identify transactions that are eligible across all the branches for which the firm has indicated it takes deposits in that currency. It has to collect all the transactional data from the last 24 hours. Then this data has to be sorted into those that are potentially eligible as Level 1 transactions. Counterparties have to be classified by type to identify whether they’re eligible because these have to be wholesale transactions.
There are also minimum-size criteria for the transactions. And there are related-party criteria. If customers transacted three times with different branches or three times during the same day, these transactions would have to be aggregated to ensure that no attempt is being made to manipulate the market.
Once Level 1 submissions have been determined, eligible transactions from previous days can be adjusted by market movements to compile Level 2 submissions, according to a very rigorous specification including time and volume weighting. Absent the ability to compute a Level 1 or 2 submissions, then we’re back to a Level 3 submission: “expert judgment.” This requires the submitter to estimate the rate at which it could raise funds in the wholesale market for the requisite term. Given that this market barely exists for many tenor/currency pairs, this is a very challenging task—and ultimately the reason why the regulators want to kill off LIBOR. It’s too subjective.
What should the second and third lines of risk management be looking at with regard to the transition between now and the end of 2021?
ICE Benchmark Administration (IBA) took over the administration and publication of LIBOR following the financial crisis. It has worked with the regulators and the panel banks to make the submission process much more robust and thus de-risk the process as much as possible. The latest innovation is the waterfall methodology of Level 1 to 3 submissions. This ensures money market transactions are used in the determination of the submission wherever possible.
It is only when recent transactions are not available that the submitter has to use expert judgment. Panel banks have been migrating to the waterfall methodology since the summer of 2018, and all were expected to be using it by April 2019. The second and third lines of defence have to ensure that the data capture of potentially eligible transactions is accurate and robust. Then the algorithm to compute the ultimate submission must be validated against IBA’s specification.
Finally, of course, all Level 3 submissions where expert judgment is applied need daily testing to ensure that the bases for final submission can be substantiated by reviewing multiple markets data sources.
At a recent RMA round table, attendees discussed the issue of people trying to determine a credit spread and term rates. Would you explain the issues?
LIBOR has two major differences to the new RFR benchmarks. LIBOR comes in seven different tenors—overnight, one week, and one, two, three, six, and 12 months—and aims to measure the risk of unsecured lending between banks: a bank credit spread component.
But the new RFRs come in one flavor—overnight—so there’s no term, and essentially no credit component, especially for the US dollar SOFR index, which is a secured rate. The industry’s challenge is to transition all the existing instruments from LIBOR to the new RFRs. But they are not “plug and play.” So the industry has to do two things: first, wean itself off its dependency on trading LIBOR-linked products and begin transitioning to the use of RFRs; and second, know that it won’t achieve that for the entirety of its legacy stock of LIBOR products and recognise the need for a conversion process.
FCA chief executive Andrew Bailey made it very clear in his July 2018 speech that he wants industry participants to remove the stock of LIBOR instruments as much as possible before the end of 2021: “The wise driver steers a course to avoid a crash rather than relying on a seatbelt. That means moving to contracts which do not rely on LIBOR and will not switch reference rates at an unpredictable time.”
So the banks are now working on significant outreach programmes with their clients to ensure they are aware of LIBOR’s pending demise, the need for new products to be based on one of the RFRs, and—most challenging—to work out what to do with their existing stock of LIBOR-referencing bilateral derivatives with maturity dates after December 2021. That could be to renegotiate them to reference RFRs.
What are the risks in being an early or a late adopter?
The challenge is that these markets in RFRs are not yet liquid and not well accepted or understood by clients. It’s a chicken-and-egg situation: traders won’t trade something until there’s a narrow bid-offer spread. They will continue doing what they like to do, which is to trade LIBOR-based products because those markets still remain very deep and very liquid. So if they’re trading swaps today, they are more likely to trade a fixed versus floating swap, where the floating leg is LIBOR. Few swap traders are trading RFRs, especially for longer-dated swaps where the market barely exists.
While there is an increasing number of floating legs that reference an RFR, it will be some time before the RFR market is the size of the LIBOR market. The Bank of England recently published volume data, showing that a traded volume of SONIA futures was just 3 percent of LIBOR futures. The SONIA swaps market is more developed with LCH showing that SONIA swaps were at 67 percent of the LIBOR swaps (notional traded YTD). And while swap traders’ books only contain LIBOR-based risks, they won’t use an RFR-based swap to hedge a LIBOR risk because that is simply opening up a significant basis risk.
The reason you want to be trading RFR products today is that you want to be out there demonstrating your commitment to reform to the regulators by promoting the adoption of these new RFR indices. There is a significant amount of pressure and influence being exerted by regulators, the US Federal Reserve, the FCA, and the Prudential Regulation Authority to do just that. Banks, US agencies, and supranational are responding by issuing RFR-linked debt, so SOFR and Sonia-linked bonds are now being issued by institutions that want to be seen as participating in those markets and promoting the use of the new benchmarks.
Some markets are ahead in the UK. An early adopter is the liability-driven investment community—pension funds. They are happy to transact in swaps referencing RFRs. They never needed bank credit spread risk built into their risk profile, so having a pure interest rate reference is a much cleaner hedge for their risk.
If you were tasked with writing a letter from the FCA to chief risk officers at banks, what would it say?
The FCA has essentially done that but the letter went to CEOs. The FCA has a history of using “Dear CEO” letters as a way of delivering a clear message to the industry about potential systemic failings in business practice. In this case, while the letter was only sent to a limited number of the largest London banks, the underlying message is that all banks must take heed. That letter is available on the FCA’s website.
Clearly, the FCA is concerned that banks are not putting sufficient resources into planning for the end of LIBOR. The purpose of the letter was to seek assurance that firms’ senior managers and boards understand the risks associated with this transition and that they are taking appropriate action now so that they can transition to alternative rates before the end of 2021.
The FCA has also used its senior manager regime very effectively by requiring recipient banks to specify the individual who will oversee the implementation of the LIBOR transition. That is going to focus attention.
Which financial institutions do you see being impacted the most?
The banks that likely will be most impacted are those that are the biggest derivatives dealers and the biggest lenders in the leveraged loan space, such as BAML, J.P. Morgan, and Morgan Stanley.
If a financial institution gets this transition to LIBOR wrong, how bad could it be?
It could be awful. If it were a smaller bank, the risk is a massive operational risk. If it’s a small commercial lender in the US, the risk is that an awful lot of manual workarounds will be required for a long time while the systems get straightened out. These banks are probably very dependent on vendors who actually deliver the changes that small- and medium-sized institutions need and that in and of itself will be a challenge.
Small regional banks with loans referencing three-month dollar LIBOR may not even have picked up on this, or they may just be starting to pick up on it and don’t really understand its seriousness. The other risk is that they don’t deal with their clients early enough and their clients insist on staying on LIBOR-based contracts after the market has moved on. The big risk here is ending up with major litigation disputes.
What should business people or traders be talking about now?
It’s all about educating your own staff and then your clients. There are hundreds of documents available on LIBOR transitions—from the big banks and the major consulting firms. Smaller institutions can educate themselves and then explain to their clients that they need to consider how they will end their dependence on LIBOR-based products, whether that’s swaps, bonds, mortgages, or loans, and ultimately adopt the new SOFR index for US dollars. The industry is working on finding a fair, equitable, and operationally simple solution.
A major goal here for the industry has to be finding a solution that doesn’t create winners or losers. This is going to be very difficult, but we all have to work together. It’s not just about derivatives or valuations. It’s about operational capability. It’s about legal robustness of contracts—about sitting down and understanding what the exposure is between a bank and a client. What are the fallbacks that exist? Are we happy with those fallbacks and, if not, how will we renegotiate them using RFRs? What would be sensible and fair for both?
Could a single universal formula be used to make the process less challenging?
The clear regulator messaging is that we need to tackle this problem now, not leave it to the last minute. Between now and the end of 2021, there’s time to renegotiate contracts with clients and counterparties.
A clear second-best option is to use the fallback protocols being constructed by the industry associations for each asset class. It’s likely that most asset classes will follow an International Swaps and Derivatives Association-endorsed protocol. At least that way, cash instruments hedged with derivatives could remain hedged in lockstep if they both “convert” to an RFR index in the same way at the same time.
One area that will be greatly impacted is model development and, more importantly, model validation?
Yes. This is certainly a competing priority problem, alongside the fact that the standards around model vetting that originated from the Fed have rippled through the industry in such a way that the standards for vetting in 2018 and 2020 are much higher than a decade ago.
I hesitate to think about the number of model-vetting quants that banks now, have relative to a decade ago, but it would be an interesting survey. There are probably double the number of people doing this compared to a decade ago. If you’re a US institution, it’s probably a lot more than double.
The Financial Conduct Authority (FCA) would like to see the demise of LIBOR sooner rather than later. It has put out a very firm message to the industry that LIBOR will not be available beyond the end of 2021, and it would be a “black swan” event were that not to happen. They are pushing the industry extremely hard to work toward a safe conversion to the use of risk-free rates (RFRs). They’ve used their regulatory powers already in the UK, through a “Dear CEO” letter, to make sure banks are planning for the cessation of LIBOR and, ultimately, the adoption of the alternative risk-free rates, such as the Sterling Overnight Index Average (SONIA) in the UK and the Secured Overnight Financing Rate (SOFR) in the US.
However, there is a recognition that a significant problem lies in the existing LIBOR-referencing contracts that may not be migrated to RFRs by the end of 2021. The total notional amount of instruments—both cash and derivative—estimated back in 2017 was on the order of $350 trillion. Therefore, all market participants, banks, in particular, are obviously looking at their contracts to understand what will happen to them in the event LIBOR is no longer published—the so-called fallback language.
It would be fair to say that the derivative market is significantly ahead of the curve with respect to developing a fallback protocol. There has been an extensive consultation by market participants who would prefer an industry-agreed protocol as a fallback, and the preliminary results of that consultation have already been published.
The presumed LIBOR alternative for the US is secured. What about the proposed rate in other countries?
For the five LIBOR currencies—US dollar, pound sterling, euro, Swiss franc, and Japanese yen—the central banks have gone a long way toward identifying replacement interest rate benchmarks that could be successfully adopted according to the IOSCO principles, which were published in the aftermath of the benchmark-fixing scandals, based on transactions in liquid markets by arm’s-length participants.
Different central banks have come back with different solutions based on the unique characteristics of each market. For the US dollar, SOFR was published in April 2018, based on the daily repo rate for US government securities. For sterling, the Bank of England chose the already existing SONIA. This is an unsecured rate for overnight loan/depo transactions.
Although some changes were made to the underlying methodology at the same time, SONIA had been around for several years prior. The other currencies have headed down the unsecured route, likely because their repo markets are much less deep than in the US. Daily trade activity in transactions that would form the SOFR index is well over $400 billion per day.
This issue of secured versus unsecured rates will matter to those who trade derivatives, especially cross-currency swaps, as they will need to access different types of markets to hedge the resulting exposures.
Why should panel banks continue to contribute to the calculation of LIBOR?
The panel banks will continue to contribute until the end of 2021 because they have already agreed with the FCA to do so. The FCA recognised there was a very big problem: under EU benchmark regulations, there was a risk that LIBOR would be considered ineligible and could die within a two-year time frame. So the FCA asked all panel banks to continue contributing until the end of 2021 to provide a longer transition period.
I expect there was some arm-twisting involved, but the FCA got pretty much all the panel banks to agree to contribute in order to reduce the systemic risk that a poorly planned end of LIBOR could create.
Why wouldn’t panel banks contribute after 2021?
There are some good reasons why a panel bank would choose not to contribute beyond 2021. First, there is an ever-present small risk that a bank could be found failing in its governance of the controlled environments around its LIBOR submissions and thus be subject to fines.
Second, there is the direct cost of maintaining the infrastructure and staff time involved in providing those submissions. Many LIBOR submissions still rely on “expert judgment” due to the paucity of eligible transaction data, so there are systems and control processes that have to be maintained. And there have to be independent second-line-of-defence review processes in place.
Then there will be internal and external audit costs that have to be borne. This all comes at a time of increasing focus on costs. The FCA estimated the annual costs of being a panel bank at £2.4 million. That’s a lot of money at a time of other significant regulatory changes, such as the implementation of Fundamental Review of the Trading Book and Standardised Approach Counterparty Credit Risk.
Would you describe the process for submissions?
The first stage is to identify transactions that are eligible across all the branches for which the firm has indicated it takes deposits in that currency. It has to collect all the transactional data from the last 24 hours. Then this data has to be sorted into those that are potentially eligible as Level 1 transactions. Counterparties have to be classified by type to identify whether they’re eligible because these have to be wholesale transactions.
There are also minimum-size criteria for the transactions. And there are related-party criteria. If customers transacted three times with different branches or three times during the same day, these transactions would have to be aggregated to ensure that no attempt is being made to manipulate the market.
Once Level 1 submissions have been determined, eligible transactions from previous days can be adjusted by market movements to compile Level 2 submissions, according to a very rigorous specification including time and volume weighting. Absent the ability to compute a Level 1 or 2 submissions, then we’re back to a Level 3 submission: “expert judgment.” This requires the submitter to estimate the rate at which it could raise funds in the wholesale market for the requisite term. Given that this market barely exists for many tenor/currency pairs, this is a very challenging task—and ultimately the reason why the regulators want to kill off LIBOR. It’s too subjective.
What should the second and third lines of risk management be looking at with regard to the transition between now and the end of 2021?
ICE Benchmark Administration (IBA) took over the administration and publication of LIBOR following the financial crisis. It has worked with the regulators and the panel banks to make the submission process much more robust and thus de-risk the process as much as possible. The latest innovation is the waterfall methodology of Level 1 to 3 submissions. This ensures money market transactions are used in the determination of the submission wherever possible.
It is only when recent transactions are not available that the submitter has to use expert judgment. Panel banks have been migrating to the waterfall methodology since the summer of 2018, and all were expected to be using it by April 2019. The second and third lines of defence have to ensure that the data capture of potentially eligible transactions is accurate and robust. Then the algorithm to compute the ultimate submission must be validated against IBA’s specification.
Finally, of course, all Level 3 submissions where expert judgment is applied need daily testing to ensure that the bases for final submission can be substantiated by reviewing multiple markets data sources.
At a recent RMA round table, attendees discussed the issue of people trying to determine a credit spread and term rates. Would you explain the issues?
LIBOR has two major differences to the new RFR benchmarks. LIBOR comes in seven different tenors—overnight, one week, and one, two, three, six, and 12 months—and aims to measure the risk of unsecured lending between banks: a bank credit spread component.
But the new RFRs come in one flavor—overnight—so there’s no term, and essentially no credit component, especially for the US dollar SOFR index, which is a secured rate. The industry’s challenge is to transition all the existing instruments from LIBOR to the new RFRs. But they are not “plug and play.” So the industry has to do two things: first, wean itself off its dependency on trading LIBOR-linked products and begin transitioning to the use of RFRs; and second, know that it won’t achieve that for the entirety of its legacy stock of LIBOR products and recognise the need for a conversion process.
FCA chief executive Andrew Bailey made it very clear in his July 2018 speech that he wants industry participants to remove the stock of LIBOR instruments as much as possible before the end of 2021: “The wise driver steers a course to avoid a crash rather than relying on a seatbelt. That means moving to contracts which do not rely on LIBOR and will not switch reference rates at an unpredictable time.”
So the banks are now working on significant outreach programmes with their clients to ensure they are aware of LIBOR’s pending demise, the need for new products to be based on one of the RFRs, and—most challenging—to work out what to do with their existing stock of LIBOR-referencing bilateral derivatives with maturity dates after December 2021. That could be to renegotiate them to reference RFRs.
What are the risks in being an early or a late adopter?
The challenge is that these markets in RFRs are not yet liquid and not well accepted or understood by clients. It’s a chicken-and-egg situation: traders won’t trade something until there’s a narrow bid-offer spread. They will continue doing what they like to do, which is to trade LIBOR-based products because those markets still remain very deep and very liquid. So if they’re trading swaps today, they are more likely to trade a fixed versus floating swap, where the floating leg is LIBOR. Few swap traders are trading RFRs, especially for longer-dated swaps where the market barely exists.
While there is an increasing number of floating legs that reference an RFR, it will be some time before the RFR market is the size of the LIBOR market. The Bank of England recently published volume data, showing that a traded volume of SONIA futures was just 3 percent of LIBOR futures. The SONIA swaps market is more developed with LCH showing that SONIA swaps were at 67 percent of the LIBOR swaps (notional traded YTD). And while swap traders’ books only contain LIBOR-based risks, they won’t use an RFR-based swap to hedge a LIBOR risk because that is simply opening up a significant basis risk.
The reason you want to be trading RFR products today is that you want to be out there demonstrating your commitment to reform to the regulators by promoting the adoption of these new RFR indices. There is a significant amount of pressure and influence being exerted by regulators, the US Federal Reserve, the FCA, and the Prudential Regulation Authority to do just that. Banks, US agencies, and supranational are responding by issuing RFR-linked debt, so SOFR and Sonia-linked bonds are now being issued by institutions that want to be seen as participating in those markets and promoting the use of the new benchmarks.
Some markets are ahead in the UK. An early adopter is the liability-driven investment community—pension funds. They are happy to transact in swaps referencing RFRs. They never needed bank credit spread risk built into their risk profile, so having a pure interest rate reference is a much cleaner hedge for their risk.
If you were tasked with writing a letter from the FCA to chief risk officers at banks, what would it say?
The FCA has essentially done that but the letter went to CEOs. The FCA has a history of using “Dear CEO” letters as a way of delivering a clear message to the industry about potential systemic failings in business practice. In this case, while the letter was only sent to a limited number of the largest London banks, the underlying message is that all banks must take heed. That letter is available on the FCA’s website.
Clearly, the FCA is concerned that banks are not putting sufficient resources into planning for the end of LIBOR. The purpose of the letter was to seek assurance that firms’ senior managers and boards understand the risks associated with this transition and that they are taking appropriate action now so that they can transition to alternative rates before the end of 2021.
The FCA has also used its senior manager regime very effectively by requiring recipient banks to specify the individual who will oversee the implementation of the LIBOR transition. That is going to focus attention.
Which financial institutions do you see being impacted the most?
The banks that likely will be most impacted are those that are the biggest derivatives dealers and the biggest lenders in the leveraged loan space, such as BAML, J.P. Morgan, and Morgan Stanley.
If a financial institution gets this transition to LIBOR wrong, how bad could it be?
It could be awful. If it were a smaller bank, the risk is a massive operational risk. If it’s a small commercial lender in the US, the risk is that an awful lot of manual workarounds will be required for a long time while the systems get straightened out. These banks are probably very dependent on vendors who actually deliver the changes that small- and medium-sized institutions need and that in and of itself will be a challenge.
Small regional banks with loans referencing three-month dollar LIBOR may not even have picked up on this, or they may just be starting to pick up on it and don’t really understand its seriousness. The other risk is that they don’t deal with their clients early enough and their clients insist on staying on LIBOR-based contracts after the market has moved on. The big risk here is ending up with major litigation disputes.
What should business people or traders be talking about now?
It’s all about educating your own staff and then your clients. There are hundreds of documents available on LIBOR transitions—from the big banks and the major consulting firms. Smaller institutions can educate themselves and then explain to their clients that they need to consider how they will end their dependence on LIBOR-based products, whether that’s swaps, bonds, mortgages, or loans, and ultimately adopt the new SOFR index for US dollars. The industry is working on finding a fair, equitable, and operationally simple solution.
A major goal here for the industry has to be finding a solution that doesn’t create winners or losers. This is going to be very difficult, but we all have to work together. It’s not just about derivatives or valuations. It’s about operational capability. It’s about legal robustness of contracts—about sitting down and understanding what the exposure is between a bank and a client. What are the fallbacks that exist? Are we happy with those fallbacks and, if not, how will we renegotiate them using RFRs? What would be sensible and fair for both?
Could a single universal formula be used to make the process less challenging?
The clear regulator messaging is that we need to tackle this problem now, not leave it to the last minute. Between now and the end of 2021, there’s time to renegotiate contracts with clients and counterparties.
A clear second-best option is to use the fallback protocols being constructed by the industry associations for each asset class. It’s likely that most asset classes will follow an International Swaps and Derivatives Association-endorsed protocol. At least that way, cash instruments hedged with derivatives could remain hedged in lockstep if they both “convert” to an RFR index in the same way at the same time.
One area that will be greatly impacted is model development and, more importantly, model validation?
Yes. This is certainly a competing priority problem, alongside the fact that the standards around model vetting that originated from the Fed have rippled through the industry in such a way that the standards for vetting in 2018 and 2020 are much higher than a decade ago.
I hesitate to think about the number of model-vetting quants that banks now, have relative to a decade ago, but it would be an interesting survey. There are probably double the number of people doing this compared to a decade ago. If you’re a US institution, it’s probably a lot more than double.
NO FEE, NO RISK
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Â鶹´«Ă˝ Finance Times
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Â鶹´«Ă˝ Finance Times
