A beginner’s guide to the Solvency II interest rate
The 'risk-free' interest rate used to establish technical provisions has taken centre stage recently as the key obstacle delaying progress towards Solvency II implementation. In this article we take the opportunity to brush up on the key components of this theoretically straightforward topic.
Why are swaps contracts 'risk-free'?
The safest thing to do with one’s money is to hold on to it in a bank account with instant access. The problem with keeping your money in the bank is that it will earn the overnight deposit rate, which could change every day and would be expected to change frequently over the course of several years. The future rate, useful for calculating long-term liabilities, is therefore unknown.
The solution is to enter into a swap contract. In a swap contract, the investor hands over the interest received from the cash deposits at the bank. In return, the swap counterparty agrees to pay the investor interest at a fixed rate. The investor now has certainty over the interest he will receive without introducing any counterparty default risk; if the counterparty were to stop paying the fixed rate the investor would then stop paying the floating rate, losing nothing except the cost of setting up a new swap arrangement with a different counterparty.
Solvency II therefore uses swap rates when determining a risk-free interest rate.
Does this cause any issues?
Many insurers hold government bonds as their 'risk-free' asset, but the yields on certain Eurozone government bonds (e.g. Italy and Spain) are not moving in line with Euro swap rates, meaning that assets and liabilities move by different amounts, placing a strain on solvency.
Are swaps contracts completely risk-free?
The snag with the argument presented above is that most swaps require the investor to pay the 6 month LIBOR rate. To be 100% sure of being able to honour his commitment under the swap, the investor would need to deposit his cash for a 6 month period without the option to withdraw early. This introduces the risk that the bank could default during the 6 month period.
Solvency II therefore makes a deduction to the swap rate to allow for credit risk.
Does this cause any issues?
One way to derive the credit risk adjustment would be to compare the rate of the 6 month LIBOR swap with that available on a swap based on the overnight cash rate; but any differences may be due in part to liquidity preference or differences in the availability of the two instruments rather than the market’s understanding of the credit risk.
A particular concern is pro-cyclicality: the spread tends to increase during times of market turbulence. Using a greater credit risk deduction during these times would push up insurers’ capital requirements and so encourage them to de-risk and start selling assets, which adds to the turbulence.
Interpolation
The problem with selecting swaps as the risk-free instrument is that the market in them is only sufficiently liquid at a few select durations e.g. 5, 10 or 15 years, requiring some sort of interpolation between these points to construct the full yield curve.
There are various issues that need to be resolved such as whether the curve needs to go through every liquid point, whether any attention should be paid to those swaps that are available at the illiquid points and whether there is an argument for looking at gilt yields between the liquid points. These will be addressed by the implementation technical standards that EIOPA will publish.
Extrapolation
The harder question is how to construct the yield curve beyond the term of the swap market. The proposed approach is to assume that interest rates are mean-reverting and so will converge towards a long-term average 'ultimate forward rate' (UFR). This is a forward rate as opposed to a spot rate; it is the rate we expect to earn on our money at some point in the far future (by which time markets have reverted to their 'average' condition).
Does this cause any issues?
The issue is defining what is specifying how liquid the market must be to provide information. The proposed extrapolation methodology leads to the present value of long-term liabilities being sensitive to the choice of 'last liquid point'.
A particular issue raised by Evans et al1 is that it ignores all swap data beyond the last liquid point. Hence the present value of a liability due in 50 years’ time does not depend on the 50 year swap yield. Therefore the best way to match this liability is therefore with a combination of swaps with durations shorter than the last liquid point and not with a 50 year swap as might be expected.
The choice of UFR is also proving controversial after a proposal to use a different UFR for each currency. Insurers based in some countries would end up using a higher interest rate than those in others (since insurers tend to write a large proportion of their business in their home territory), making the issue a political one; politicians naturally wish to protect the interests of insurers in their own country.
Adjustments to the basic risk-free rate
There are two further adjustments proposed to the interest rate:
- the matching premium/adjustment
- the counter-cyclical premium
It is not really possible to justify including either of these in a 'risk-free' interest rate on theoretical grounds alone. The matching premium was discussed extensively in my blog of 8 November. We describe the counter-cyclical premium here.
The counter-cyclical premium is essentially an addition that EIOPA can allow insurers to make to the basic risk-free rate during times of market turbulence. Its purpose is to ease capital requirements during times of turmoil to prevent insurers selling their risky assets and so making the problem worse.
Does this cause any issues?
Insurers do not like the uncertainty over when the counter-cyclical premium will be turned on. The argument against this is that the counter-cyclical premium should be an emergency measure and that insurers should not be able to pre-empt when it will be switched on and allow for this in their pricing.
There are also arguments about whether the counter-cyclical premium should be switched on for a particular country or a particular currency and whether the decision to apply it should be taken by EIOPA or individual member states.
Conclusion
Given the difficulties that Solvency II poses in terms of determining 1 in 200 year events in order to establish the SCR, it is perhaps surprising that progress is being held up by the issue of a risk-free interest rate. This illustrates that overall capital requirements depend not only on the SCR but also upon the underlying technical provisions.
1Ross Evans, Angelina Lai, Eamonn Phelan, Derek McLean; “Hedging your bets”; The Actuary, 25 July 2012: http://www.theactuary.com/features/2012/07/hedgingyourbets/