RegulationNov 10 2020

Designing a business risk model

twitter-iconfacebook-iconlinkedin-iconmail-iconprint-icon
Search supported by
Designing a business risk model
Pexels/Fauxels

At the same time, the decrease in swap yields will have increased the value of their liabilities on their Solvency II balance sheets.

However, we do not believe all is grim for insurers.

Instead, it is worth focusing on certain high-level considerations in relation to the development of stress and scenario testing for multi-asset insurance investment portfolios.

Immediate impact of Covid-19

The global pandemic and its immediate impact on the drawdown in equity prices (and now subsequent partial retracing) and widening credit spreads had adversely affected investment performance over the first five months of 2020.

However, significantly lower interest rates and higher volatility associated with equities and credit spreads have, together, reduced expectations for forward-looking returns and increased the risk around them for the short to medium term.

Does this imply that we should increase the time devoted to global pandemics? I think not.

There is a common behavioural bias encountered in investing – the so-called recency bias.

Key points

  • The current situation has been challenging for insurers
  • Recency bias is a common behavioural bias in investing
  • Some are concerned about credit-related events

This describes our tendency to focus on events that have recently occurred in lieu of more distant experiences and use the recent events to predict and prepare for the future. 

While it is true that the global economic slowdown and potential recessions in various parts of the world are a direct result of the Covid-19 pandemic, it may not be the case that the event we should be modelling and considering is the same one that is occurring now.

Lightning does not strike in the same place twice – or so we are told. We noted that after the 2008 global financial crisis there was a reactionary focus on credit risk and systemic financial risk.

It was a natural response to the crisis of 2008. Generally, we need to think in broader terms to prepare for and increase our knowledge around potential future outcomes.

What could be coming next?

There is growing concern that the events that should be most under consideration are credit-related events that are currently partially hidden under the cover of the overwhelming central bank intervention that has placed a cap on credit spreads.

One of a number of possible medium-term events playing out over the next several months and into next year could be related to a deterioration of performance of credit and its implication on sovereign debt, municipal bonds, corporate debt and structured credit. 

The Prudential Regulation Authority’s supervisory statement on the prudent person principle, issued in May 2020, explicitly states the expectations from the PRA on companies to “pay particular attention to the measurement and control of credit spread and default risk, including credit transition downgrade/upgrade risk”.

Market commentary

We have begun to see some of the fault lines in the credit space. In emerging market debt, the spotlight is on Argentina, where a technical default has recently occurred. This is after Argentina issued a 100-year bond only three years ago. 

In the municipal bond market, we have recently seen the state of Illinois seeking $41bn (£31.5bn) of financial support, including $10bn to support its underfunded pension schemes.

Investment-grade credit spreads and high-yield credit spreads have suffered significant widening prior to the central bank intervention.

Design of stress tests and scenarios

Given these developments, it is important to have a carefully chosen set of stress tests and scenarios to bound and estimate the impact of potential events on the valuation of your investment portfolio and liabilities.

It is not always an easy task to create a small set of appropriate stress tests and scenarios with market-consistent assumptions that are relevant and tailored to both your investments and liabilities. 

Adding to the complexity and difficulty, the PRA expects companies to demonstrate that:

• Their solvency appetite will not be breached by a moderate stress scenario.

• Their solvency will not be threatened in a severe stress scenario.

• The company is able to recover from a severe shock and restore compliance with all its regulatory requirements.

The design of the appropriate stress tests will be related to the areas of the market to which your portfolio is most exposed from a monetary perspective relative to potential risks with similar likelihoods. 

Stress tests may be designed to include shocks to individual factors – for example, large capitalisation equities or investment-grade credit spreads – or, as suggested by the regulator, using a combination of simultaneous stresses.

Stress tests can also include propagated risk. For example, a shock of wider investment-grade credit spreads will also likely be accompanied by (correlated to) wider high-yield credit spreads and a drop in equity prices.

These changes, taken together, may be useful to model as a single multi-factor ‘credit’ stress test. 

Designing scenarios that combine changes in many market variables consistently and model a particularly concerning situation or time is useful from a risk management perspective.

Scenarios can be drawn from situations that lead to concerns in relation to your risk appetite limits or strategic asset allocation. This may include floors on investment performance or limits on capital buffer erosion. 

Modelling scenarios

There are two ways to build and design scenarios. First, scenarios can be built based on statistical models. The changes in market levels of factors can be drawn from the same statistical models driving insurers’ internal capital models. 

Alternatively, the scenarios may be modelled based on the relative changes that have occurred over a particularly interesting period of time in history.

This could include, for example, the period around the default of Lehman Brothers in 2008. Clearly, in the future, a Covid-19 scenario will be created. Both methods are useful, and they can be used in combination. 

Though the level of economic uncertainty and market volatility will likely remain high until Covid-19 is contained, organisations can position themselves for future growth while quantitatively measuring their risks.

Ash Belur is a director of Willis Towers Watson’s insurance investment team