Those charged with managing the investment portfolios of the world’s insurers are a cautious breed. During the post global financial crisis years of negligible or zero interest rates few were seriously tempted by the blandishments of asset managers queuing up to ply them with all manner of alternative assets.
This innate caution might be expected to be reassuring to financial regulators. But, in a volatile age, where markets are constantly buffeted by greater geo-political uncertainty than we have experienced for several generations, these are being overlaid with new risks from climate change. Regulators have been scarred by failures that have happened on their watch and are more determined than ever to ensure that the risks of corporate failure are minimised and, in particular, that the risk of contagion across a sector are minimised and correlation risks are better understood and contained.
The latest tool in this new armoury of prevention is stress-testing. It is being deployed across insurer operations, from product design and pricing, risk identification and mitigation, modelling and right through to investment strategies. It is an exercise that asks difficult questions, testing the ability of many insurers to respond to them, especially to demonstrate the ability of key teams to proactively collaborate to face the new challenges.
“The scenario being tested is also very extreme – way beyond the
one-in-200-year regulatory requirements under Solvency II."
Are the questions asked always realistic? Insurance Europe (IE), the trade body that represents European insurers, does not think the European Insurance and Occupational Pensions Authority (EIOPA) has always struck the right balance when it comes to creating realistic scenarios.
“EIOPA’s latest stress test exercise assesses the impact of exacerbated geopolitical and economic tensions, leading to a market stress environment coupled with high inflation. This scenario is similar to a more extreme version of the market environment experienced during the Covid pandemic”, said Angus Scorgie, Head of Prudential Regulation & International Affairs/Reinsurance at IE.
Scorgie said that while the results of this exercise may be helpful to identify potential macroprudential issues, insurers will already understand their own risk profile and sensitivities to the shocks being tested and as such, the exercise does not provide significant benefits from a micro-prudential perspective.
“The scenario being tested is also very extreme – some of the resulting asset stresses are assessed by the European Systemic Risk Board (ESRB) to be one-in-3000-year events, way beyond the one-in-200-year regulatory requirements under Solvency II," he said.
The questions this sort of extreme scenario posed to chief investment officers were not always easy to answer.
“We were asked to predict the responses of markets and central banks to unprecedented volatility, beyond anything we have experience of, and justify those responses”, said one CIO of a mid-sized European insurer.
“Has it prompted us to look at our asset mix? Our strategies are always cautious and heavily weighted to high grade assets, whether bonds or equities," they said. "I’m not sure what else we could do to protect against some of the severe downsides implicit in the stress-testing scenarios. However, we are looking at some of the issues around correlation and our ability to access liquidity”.
"What I hear from industry is that their management actions
are very tailored to the type of portfolio that they have."
The market is still trying to absorb the lessons of the initial focus on long-term insurers, especially those European insurers offering products with guaranteed returns, said William Gibbons, Senior Investment Consultant at Mercer. While not fashionable in the UK, these are still a feature of many markets:
“Maybe some products with guarantees are coming back as there is more scope for them with higher rates”, he said, although Gibbons added that there were concerns elsewhere in the world, especially China where guarantees are often linked to property investments which have crashed.
There is also a danger that the intense scrutiny of stress-testing could create greater uniformity in product design and portfolio construction, just at a time when many are concerned about enhancing competitiveness in the financial services sector.
“What I hear from industry is that their management actions are very tailored to the type of portfolio that they have,” said Kareline Daguer, a Director at Deloitte. “If you are very concentrated on, let's say, equity release, or if you have a lot of infrastructure, or a lot of real estate, why would you choose in the event of an incident to only sell 25% of everything across the board, and then swap it and replace it for investment grade investments,”
Daguer added that it “makes no sense to do such a thing in reality”.
“That's something that firms would have to work on to explain those differences, explain that to the regulator, explain to investors,” she said. “That will happen before we see that mass movement towards everybody having a portfolio that is similar, and that reacts in a similar way under the stresses that the regulator is imposing.”
In short, insurers will fight hard to remain in control of their own destinies, responding to market challenges and new opportunities while all the time reassuring regulators they have the skills, models and internal controls in place to understand and manage the risks.
With stress-testing here to stay we can expect regulators to explore a wider range of risks, said Linda Hedqvist, Senior Manager at Deloitte. “Stress-testing, we think is fundamentally a good thing and helps insurers and regulators to assess exposures, especially to these new types of risks that we're seeing.
“And there are quite a few of them. Geopolitical risk, which is something that we've heard clients speak about quite a lot in the last year, both banks and insurers. Then there are quite a few different elements of climate risk. Hedqvist added that with physical risk, the general insurance industry at least argued that is well understood because they're used to dealing with catastrophes and weather-related risks.
“But there are other elements of climate risk too, like litigation, for example, that are also becoming big,” said Hedqvist and added that cyber risk is a major example of this that the UK regulator has been focused on.
"Some insurers will be in a good place and have probably done
cyber insurance stress-testing but for others it might be new."
“What we are seeing is that the types of stress test exercises are becoming more exploratory in nature,” she said. “That's something we've seen on the banking side too, where regulators are trying to work with the industry to understand these new types of risks.”
With the scope of stress-testing constantly expanding, how well-placed are insurers to respond? The technical demands of regulatory reform stretch very wide, with reforms of Solvency rules in Europe and the UK underway and extensive changes to accounting rules imposed by the International Accounting Standards Board also imposing new burdens.
While most insurers are capable of managing these demands there could be some issues, said Hedqvist. “General insurers are trying to enhance their modelling capabilities overall but not knowing exactly what [the regulators] will focus on means you do not know exactly what expertise they need," she said. "Some insurers will be in a good place and have probably done cyber insurance stress-testing but for some others it might be new. The same goes for climate related risks for example so there will be a wide range, but I would expect there to be some need to beef up capabilities over the next year.”
In the UK, general insurers are nervously awaiting the next round of stress-testing from the Prudential Regulation Authority (PRA) in the first half of next year. The scenarios they will be presented with are being kept under wraps until nearer the time, a move by regulators to introduce what they claim is more realism into the testing.
“The dynamic nature of the 2025 exercise represents a significant change from previous exercises and will involve simulating a sequential set of adverse events over a short period of time”, is the PRA’s description of its new approach.
There are few clues as to what the PRA might be focussing on although potential over-reliance on reinsurance is an area of concern already flagged up by the PRA.
Another is liquidity risk, especially if institutional investors all jump the same way at the same time in response to a market shock as Dave Ramsden, Deputy Governor, Markets and Banking at the Bank, told a recent ESRB event:
“Recent events have revealed that shocks can uncover vulnerabilities, exacerbating rapid or sharp moves in market interest rates – independent of the impact of monetary policy – leading to additional liquidity stress in the financial system, harming the functioning of core markets … there is ongoing work at central banks to address the vulnerabilities that these events revealed.”
"It is fundamental that we maintain strict discipline
when it comes to cost-benefit analysis."
Getting models fit to face the challenge of the next round of stress-testing will be key, said Gibbons. “The tough thing with stress-testing is you do not always know what you are facing, and the models can fluctuate wildly if they are not carefully calibrated”.
He expected climate change related hazards to figure prominently in the next round of stress-testing, especially for insurers with significant exposures to property catastrophe risks in the United States.
Hanging over all this regulatory activity is the question of how the results will be used, something insurers are very sensitive about. “While it is understandable that regulators want to assess the potential impact of these complex scenarios, it is fundamental that we maintain strict discipline when it comes to cost-benefit analysis and correctly identifying which stress test data can and should be used for which purposes,” said Scorgie.
“For example, multi-decade climate scenarios heavily rely on assumptions and often do not reflect the ability of insurance companies to adapt to new situations. These are therefore useful when identifying and assessing potential long-term impacts of climate change but not for calculating capital requirements.”
Alongside this new enthusiasm of regulators to present demanding stress-testing scenarios has been a series of major changes to accounting standards driven by the International Accounting Standards Board (IASB).
This collection of International Financial Reporting Standards (IFRS) has been in formulation for a long time, but the changes are now beginning to bite, imposing more demands on the financial and technical resources of insurers. After nearly a decade of arguments between insurers and the IASB, from January 2023 IFRS17, and its cousin IFRS9, imposed new rules requiring the profit on long-term policies to be accounted for evenly across their lifetime.
“We can expect some vigorous lobbying by insurers of the IASB
later this year to get some movement on these points."
For many insurers used to accruing the profit on a policy immediately it is taken out this has required a huge trawl through legacy data and a major adjustment to their balance sheets, not always favourably. It has also absorbed a huge amount of technical resource just as regulators were launching stress tests that similarly require the extraction, manipulation and presentation of vast amounts of data.
Insurers are still not happy with the new IFRS regime, especially as it imposes a more relaxed regime on banks, said Phillipe Angelis, Manager, Corporate Reporting & Sustainable Finance at Insurance Europe.
He says there are still issues around how alternative assets are treated and the failure to deliver some of the promises around dynamic risk management, especially the restrictions on linking hedging instruments to liabilities. “We can expect some vigorous lobbying by insurers of the IASB later this year to get some movement on these points,” he said.