Courtesy of Nicholas Burgess
The financial crisis of 2008 fundamentally altered the interbank market and the way banks fund themselves. Banks and regulators alike recognised the risks inherent in unsecured borrowing and lending between banks. Consequently, the London Interbank Offered Rate (LIBOR) derived from the London interbank unsecured lending market, was revealed to suffer from low liquidity, excessive leverage, and a lack of transparency because the rate was based on market participants’’ judgement rather than market transactions. Moreover, several high-profile rate manipulation scandals have undermined confidence in LIBOR’s reliability, robustness, and suitability as a global benchmark for such a large magnitude of transactions.
Unsurprisingly, central banks and regulators are reforming interest rate markets and encouraging financial institutions to migrate to alternative reference rates, which are benchmark rates compliant with the International Organizations of Securities Commissions (IOSCO) standards. These standards promote transparency, integrity, and accuracy. LIBOR rates are not IOSCO benchmark compliant because they are not representative of actual market funding levels.
The Impact on Financial Services
Many established financial institutions have legacy systems infrastructure and don’t fully understand the structural impact of the benchmark change. While this is not necessarily true for fintech businesses, the cost of upgrading infrastructure across an entire company with tens of thousands of employees can be costly and resource-intensive.
Every financial contract is impacted because every cash–flow requires discounting to evaluate its present value. This process relies upon yield curve construction to imply the discount factors required to evaluate the current or present value of every cash–flow. Furthermore, variable cash–flows also require yield curves to imply future market interest rate levels (forward rates).
LIBOR rates are embedded in every system and report in the banking industry, be it for balance sheet reporting, client reporting, or the assessment and management of trading positions. All systems, reports, and analytics need to be carefully reassessed and, for the large part, rewritten. The systems infrastructure impacted by Libor reform is staggeringly extensive and exceptionally costly to reconfigure and/or replace. Furthermore, the litigation risks for the industry of getting this wrong, with $350 trillion of contracts referencing LIBOR, are unprecedented.
Structural Interest Rate Changes
There are notable differences between LIBOR and alternative reference rates (ARRs). LIBOR rates are set in advance and are described as forward–looking, i.e. a three–month borrowing rate is set today. However, these rates no longer reflect actual market borrowing levels and have a built-in credit component to reflect the riskiness of lending in the interbank market.
ARRs are based on daily transactions and are consequently daily averaged rates. They are backward-looking, i.e. today’s three-months borrowing rate is not known until three months from today. Rates are set every day and aggregated into an average rate for the period, which is only known at the end of the period. The ARR term rate, or forward-looking rate, is not known, making it difficult for some investors to hedge. Finally, ARR rates have no interbank credit component, which is why some investors refer to them as risk-free rates.
ARR rates are also prone to market shocks. However, the impact of a one-day shock, while not negligible, is quite muted because the rates are averaged over many days.
Litigation risks, ISDA protocols & Fall-backs
The above structural changes have a broad and profound impact on product behaviour and pricing. For example, plain derivative products with bullet payments previously considered ‘European’ style are now more similar to ‘Asian’ style contracts with averaged payments, lower volatility, and lower prices. More complex transactions, in some cases, are unrecognisable. Contract renegotiation is a complex process and could subject the financial industry to extensive litigation when migrating existing contracts to incorporate market reforms.
For cleared transactions, this task is simplified and managed by the central clearinghouses such as LCH or CME, where transactions will be automatically migrated. Furthermore, for non-cleared transactions, the International Swaps and Derivatives Associations (ISDA) have extended their industry standards and contract protocols to define best practice for existing contract renegotiation. However, clients and investors are required to sign-up and agree to such protocols.
In the event of a contract not being renegotiated, existing contracts rely on fall-back provisions that usually do not support discontinuing the LIBOR rate. In some cases, the language suggests that variable LIBOR rates should default to the last available rate published. In a discontinuation scenario, this is tantamount to migrating from a floating rate contract to a fixed–rate contract, which could be subject to much litigation if not renegotiated in advance, due to the profit and loss implications.
What are Yield Curves?
Yield curves have two purposes. First, they tell us what a cash-flow in the future is worth today and imply the appropriate scaling or discount factor to apply. For example, a cash-flow of $100 to be received in 10 years is perhaps only worth $90 today, i.e. it has a discount factor of 0.9 or 90%. It is more intuitive to think of the reverse scenario, where a cash-flow of $90 will grow in a savings account to $100 in 10 years.
Second, yield curves are also used to calculate forward rates: the borrowing and lending rates for different maturities. For example, calculating what lending rate to charge a client to borrow $1,000,000 for ten years.
Without a yield curve, it is not possible to evaluate the time value of money. Every financial transaction is based on the evaluation of cash-flows. Yield curves are required for the pricing of all financial products, and there are $350 trillion financial contracts directly referencing LIBOR forward rates.
Consequently, as yield curves depend upon, and are constructed from, LIBOR rates, clearly all cash-flows, all financial systems, and all reports must be enhanced to be compatible with LIBOR reforms and the new ARR rates. This gives an idea of the scale and cost involved in managing the benchmark rate transition from LIBOR to ARRs. It also highlights the central importance of the yield curve calculation to the overall process.
Why is Yield Curve Calibration so Hard?
Building or calibrating yield curves is a complex process. We take a collection of market instruments, typically, futures and swaps, and solve for the forward rates and discount factors such that these instruments reprice entirely with high precision.
Interest rate swaps typically quote with a bid-offer spread of 1/10th a basis point, which implies that the accuracy must be at least to five decimal places. To achieve such precision, we require eight decimal places of accuracy in the corresponding discount factor calculation.
For a typical USD yield curve for Overnight Indexed Swaps (the discount curve), we must price over 100 instruments simultaneously to imply a full curve of 10,000 daily forecast rates and 10,000 daily discount factors out to 40-50 years.
Furthermore, this has to be done with high accuracy within a fraction of a second, as fast as 5-10 milliseconds for moderate performance in electronic markets. Finally, risk must often be calculated at the same time.
Because of these stringent requirements, it can take highly skilled professionals several years to set up and test a yield curve model. Clearly, there is much work to be done to reconfigure yield curves to satisfy such high expectations.
As a result of LIBOR benchmark reforms, financial contracts referencing LIBOR will be replaced with ARR equivalent contracts. At present, $1 trillion of contracts reference new ARR interest rates. But this is not sufficient to support the migration and active management of the existing $350 trillion of financial contracts currently referencing LIBOR. There is much liquidity for short-term futures contracts refencing ARRs in the 1-5-year region, however, longer-dated liquidity in swaps contracts is almost negligible and very thinly traded.
Clearinghouses are migrating to ARR or risk-free discounting in June 2020, which will significantly alleviate this problem. This is because cleared swap transactions will automatically migrate to risk-free discounting, which will involve distributing many LIBOR-ARR basis swap contracts to investors to manage their risk portfolios over the migration period.
The process will involve compensating investors and adjusting P&L balances to account for any trade discrepancy so that there will be no profit and loss impact whatsoever. More importantly, investor risk positions will not be impacted, and investor portfolios will be credited with LIBOR-ARR basis swaps in such a way to preserve the existing risk profile of the investor despite referencing the new ARR discount factor.
LIBOR benchmark reforms will have a broad–reaching impact on financial services and potentially society at large. Financial institutions and investors have a lot to do to prepare for the scale of change. Systems infrastructure, financial reporting, and transactional changes will all require costly upgrades.
Yield curve analytics and system calculations central to trading, pricing, and risk management are complex, resource–intensive, and costly to enhance. They provide a vital service that needs to be skilfully managed to meet market deadlines. Finally, financial contract migration and renegotiation need to be delicately managed to prevent costly litigation.
LIBOR is the most significant number within the financial services industry, impacting every cash-flow, every transaction, every report, and every system. The scale of leaving LIBOR is indeed a landmark transition.