Explanatory power of SARON term and actual rates
Hedging of SARON exposure, Part I
On 5 March 2021, the Financial Conduct Authority (FCA) announced the official dates of the cessation and loss of representativeness of the LIBOR rates. As a result, 31 December 2021 will be the last day on which the CHF LIBOR will be published. During the last months, a multitude of information were provided by the Alternative Reference Rates Committee (ARRC) and the International Swaps and Derivatives Association (ISDA). This article aims to provide an analysis of the prediction power of term and in advance rates.
Refresh on the compounded methodologies
There is an essential difference between contracts linked to CHF LIBOR and compounded SARON. When using a term reference rate (e.g. 3-month LIBOR) as a benchmark in a financial contract, the interest payments are known at the beginning of the interest period. This type of forward-looking term rate reflects the expected level of interest in three months, equivalent to a sequence of expected overnight rates. On the other hand, using compounded SARON, interest payments are the result of a daily compounded interest rate. These rates are defined as backward-looking rates and they reflect the realized level of interest over the past three months, equivalent to a sequence of realized overnight rates.
Moreover, there are generally two ways to compound a reference rate:
- Compounded in arrears: backward-looking starting at the end of the fixing period.
Figure 1 – Ex-post / Backward-looking at the end of the Fixing Period / Compounded SARON in arrears.
The blue arrow represents the payments period, while the green arrow the rate calculation period.
- Compounded in advance: backward-looking starting at the beginning of the fixing period. In other words, the rate that will be paid at the “End Date” 01.03.2021 is the previous realized 1-month rate compounded over the period 01.01.2021 – 01.02.2021.
Figure 2 – Ex-post / Backward-looking at the beginning of the Fixing Period / Compounded SARON in advance.
The blue arrow represents the payment period, while the green arrow the rate calculation period.
On the other hand, a 1-month forward-looking rate is derived from market expectations, e.g. from the swap SARON curve. For more information on the difference between backward- and forward-looking rates, and on the different compounding methodologies please refer to the article IBOR Reform in Switzerland, Part II: Calculation of compounded SARON.
Term rates vs Actual rates
The world is discussing the new reference rates and how to implement them. However, little information on the prediction power of those rates is present in the market.
David Bowman, Senior Associate Director at the Board of Governors of the Federal Reserve, made a speech during ARRC’s first SOFR symposium, highlighting that market participants do not have any guarantee on the availability of a SOFR term rate (forward-looking alternative to LIBOR) in 2021. Moreover, according to the different National Working Groups (NWGs), there will not be any recommendation to use forward-looking rate by the middle of this year, encouraging market participants to continue to transition to the new reference rates using the tools available now such as SARON averages and index data that can be applied in advance or in arrears. This concept was also stressed during the ISDA Benchmark Strategies Forum on 13 April 2021. With respect to the Swiss market and as mentioned in the article IBOR Reform in Switzerland – Part I (February 2020), the Swiss NWG assessed the feasibility of a forward-looking term rate based on SARON derivatives (e.g. SARON swaps and futures) and argued that a robust fixing is unlikely to be feasible. The main reason behind this choice is that the poor liquidity in the derivatives market would lead to an incorrect expectation of future rates.
David Bowman highlighted that most of the time there is a difference between what the market expects (e.g. 3-month Effective Fed Funds (EFF) term rate) and what actually happens (e.g. 3-month EFF compounded rate plain in arrears), particularly when the rates move down sharply. From this point, it is possible to start thinking about which strategy is better to use to explain the actual rates.
In Figure 3 and Figure 4 below, it is possible to see how the fixed-leg SARON swap with maturity 3-month (i.e. the term rate) is not better than the SARON 1-month or 3-month in advance (Last Reset methodology) in terms of predictability. The analysis starts by looking at a certain date, e.g. 01.07.2020, and calculates the SARON 3-month Plain in arrears over the previous period 01.04.2020 – 31.06.2020. The goal is to find the tenor that best predicts the future rates. This is done by comparing the following SARON rates:
- SARON 3-month term rate fixed on 01.04.2020. It is the current price paid today for the expected compounded 3-month SARON (not yet realized), based on the 3-month SARON swap.
- SARON 3-month compounded over the period 01.01.2020 – 31.03.2020. The name for this compounding strategy would be SARON 3-month in advance Last Reset. In other words, the current price paid today is based on the past realized 3-month SARON rates.
- SARON 1-month compounded over the period 01.03.2020 – 31.03.2020. The name for this compounding strategy would be SARON 1-month in advance Last Reset. In other words, the current price paid today is based on the past realized 1-month SARON rates.
Figure 3 shows how the market seems to take into account new information that dropped the rate. It also highlights that the SARON 3-month in advance has a similar miming capacity of the term rate. However, Figure 4 shows that the term rates result in a better predictor for a sharp decrease in actual rates. A similar analysis is done with respect to SARON 6-month in arrears, which you can find in Figure 12 and Figure 13 in the Appendix.
Figure 3 – Compare SARON 3-month in arrears (Plain methodology) to SARON 3-month term rate, SARON 3- and SARON 1- month in advance (Last Reset methodology).
Moreover, backward-looking reference rates reflect the past interest rate movements and not just the market expectations as the forward-looking rate does (e.g. CHF LIBOR or SARON term rate). For this reason, backward-looking rates should be the baseline for explaining rate movements. From Figure 4, it is possible to assert that the appropriate rate to predict the SARON 3-month in arrears is the SARON 3-month in advance (Last Rest methodology), which has the smallest average basis by considering the rates from January 2010 to January 2021. It is possible to argue that in case of raising interest rates, the lender will not be paid with the raised rate but with the 3-month compounded rate calculated in advance over the previous three months. However, the lender will be compensated with the next quarter payment, which will be compounded in advance (Last Reset methodology) taking into account the raised rates. Similar outcomes result from the 6-month analysis (see Appendix).
Figure 4 – Basis between the SARON 3-month term rate, SARON 3- and 1-month in advance (Last Reset methodology) to SARON 3-month in arrears (Plain methodology).
Finally, the in advance rates create a basis risk1 , but any basis created by an in advance rate will tend to average out over time. Taking into account the basis existing over each single month in an unstressed scenario (e.g. SARON 3-month Plain in arrears vs 3-month in advance Last Reset over each month in 2020), a loan lasting one year with quarterly reset (e.g. a loan over 2020 with the same characteristics) will cut the basis by 86% if the worst month in terms of basis risk is considered (e.g. from 0.043% to 0.006%). Over a stressed period, e.g. between July 2014 and July 2015, the basis is cut by circa 73% (e.g. from 0.662% to 0.178%).
The analysis assesses the explanatory power of the SARON term rates (not yet realized), against the SARON in advanced rate with Last Reset methodology (realized in the past) with respect to the actual in arrears compounded SARON rates. The current term rates do not seem to be a better predictor than the in advance calculated rates for the actual in arrears rates. It is possible to assert that the appropriate rate to predict the SARON 3-month in arrears is the SARON 3-month in advance (Last Rest methodology), which has the smallest average basis. The Part II SARON hedging strategies will explain the different compounding methodologies with respect to the hedging opportunities for SARON exposure.
Please find the appendix here.
Notes and references
1) Imperfect hedging due to the mismatch between futures contract and underlying “cash” position. This imperfect correlation between the two investments creates the potential for excess gains or losses in a hedging strategy, thus adding risk to the position.
2) ARRC. Insights from National Working Group Chairs. March 2021.
3) National Working Group on Swiss Franc Reference Rates. Discussion paper on SARON Floating Rate Notes. July 2019.
4) National Working Group on Swiss Franc Reference Rates. IBOR to RFR transition: Effects on financial reporting. July 2019.
5) SIX, Swiss Average Rate Overnight. The new Swiss franc benchmark. 2018.