Freek Zandbergen: The introduction of the alternative methodology is seeking a compromise to include more market information beyond the last-liquid point (LLP) in the extrapolation of the curve and to limit the immediate impact on the valuation of long-term liabilities.
Initially, the European regulator was looking at shifting the curve entirely, because it felt that starting from an LLP of 20 years is not in line with what it has observed as market practice.
I think this is fair, because in normal times you see some trading for example at the 25- and 30-year point and even at the 40-year and 50-year point in the interest rate swap market.
"Regulators are now stepping back a little bit from their original idea of shifting
the whole curve and that’s why this alternative curve comes in."
There are two good reasons to reconsider the change to longer LLP, first is the massive hit in terms of solvency ratio the market needs to absorb, second is the limited amount of bond trading on the long-end of the curve.
Consequently insurers would need to increase their synthetic hedge programs and this introduces more types of risk, for example liquidity risk and a mismatch in spread duration matching. In addition, the amount of trading during market stress is fairly limited – signalled by a inverted curve on the long-end.
So, I have the feeling that the regulators are now stepping back a little bit from their original idea of shifting the whole curve and that’s why this alternative curve comes in. The old method fully ignores market information beyond the last liquid point and in the alternative it is gradually phased in, so at least there’s some exposure to long-term rates.
Freek: It solves at least a couple of the strange implications from the old Smith-Wilson method. But it is still complex and a little bit artificial. That’s because it’s still not a closed-formula approach, so you need some numerical work procedures just to end up with your curve.
The question is: how big is the impact? For day-to-day management for companies who have a proper solvency ratio, the impact may be limited because they might think: “Okay, let’s just accept this and observe some volatility in the ratio, just like we have experienced in the past”.
On the other hand, companies with a low solvency ratio will need to model this very closely. Because if your Solvency ratio is below your commercial capital target, or maybe even close to the level of where the regulator starts to intervene, then you really need to avoid dropping it any further.
"Companies with a low solvency ratio will need to model this very closely."
One of the key things here is that a new parameter alpha, which is managing the speed of convergence to the UFR, is brought into the extrapolation.
This new parameter might be a new instrument by how regulators can start to steer and start to slow things down in the extrapolation. Previously the only room to have impact was in gradually lowering the UFR over time.
And then, of course, you come to the ethical discussion: what is good and what do you want to see? Do you want to have stable solvency ratios across the industry, knowing that some uncertainty is not fully captured on the long end of the curve? Or do you want to have (very) volatile solvency ratios, which might reflect economic reality immediately?
I prefer the first; not saying that we should act like ostriches and ignore the facts, but the life insurance industry is a long-term business.
"Monitoring is required, and insurance companies should analyse
and understand the scenarios of prolonged low-rate scenarios."
Important to note, what is the economic reality for a 50-year liability? That’s a discussion without an end, because how should you discount that liability when there is no liquid market (or it is not manageable for an insurance company to fully hedge it)?
There’s not a single answer to that one and it just comes to personal opinion. In my opinion it is a strange incentive to enforce insurance company to trade on a 50-year interest rate swaps with a negative fixed rate, in particular when this is caused by a lack of balance between supply and demand.
So, monitoring is required, and insurance companies should analyse and understand the scenarios of prolonged low-rate scenarios, but these scenarios should allow an insurance company to generate profits on their assets and on other businesses.
Translating this uncertainty into either a technicality in how to extrapolate your risk-free rates or to other things like the volatility adjustment or matching adjustment concepts, (which are relevant in the UK) to reduce volatility is then a good way to go.
Freek: If you look at the process so far, a year ago the review was announced, and they had the first wave consultation papers shortly afterwards. In October, a second wave was published, which contained the details on the quantitative side.
EIOPA, has suggested five different options (technically four as one included pillar 2 and 3 measures only) to change the extrapolation and three of them are used in the first impact analysis in December 2019.
Then in March this year, the regulator launched the holistic impact assessment, where the it opted for one set of changes on the whole range of topics which it is evaluating.
"I think there’s a strong indication that the alternative method is
EIOPA’s preferred way forward."
In the holistic impact assessment, the alternative method to extrapolate the curve was chosen, and I think there’s a strong indication that this is EIOPA’s preferred way forward.
It’s also been clear from the first impact study that moving the LLP is a solution which comes at a great price, as it will jeopardize the Solvency ratios of a couple of solid insurance industries (at least in the Netherlands and in Germany).
I can imagine that this is an undesired outcome from such a change, so that’s why I think it’s a logical solution to go for this middle route.
