Insurance and systemic risk: No easy conclusions

Christian Thimann

31 May 2016



Three years have passed since the Financial Stability Board (FSB) designated the first set of nine insurance companies as ‘systemically important’ (G-SIIs): five companies from the EU, three companies from the US and one company from China.1 Since then, supervisory colleges among the local insurance oversight authorities most relevant for these companies – so-called Crisis Management Groups (CMGs) – have been formed and detailed plans of financial linkages and management have been provided by the designated insurers to these colleges. These reports have been closely reviewed by the respective home supervisor. So far so good.

The core of the current debate is whether these nine insurance companies should hold more capital than their competitors; if so, how this is assessed on a comparable basis in a world where the regulatory regimes for insurance differ profoundly between Europe, the US and other constituencies such as China; and whether one needs and can establish a global capital standard in insurance.

While this debate is going on, two things happened recently. Most importantly, a US court rejected the designation of MetLife by the US government, which is based on a framework not fundamentally dissimilar to that of the FSB, as ‘arbitrary and capricious’. The reasoning of this decision is broadly that the designation was based on non-transparent method, with a lack of proper risk analysis and a lack of consideration of costs entailed by designation. The US government will appeal, but the very process raises questions on the designation of systemically important insurance groups as it has been done to date.

Second, the IMF came out for the first time in its Global Financial Stability Report with a dedicated chapter to the ‘insurance sector – trends and systemic risk implications’ (IMF 2016). This report is a very positive signal of growing interest in the business model of the insurance sector, as it remains generally insufficiently understood by many financial actors and too quickly assumed to be similar to the banking sector. How can the IMF chapter on insurance be assessed?

Useful clarification and trends detailing the insurance business model

Regarding systemic risk, the IMF report clarifies a number of points and brings valuable insights on the systemic role of insurance in the current context of ultra-accommodative monetary policies and lasting low interest rates (Bean et al. 2015).

  • A first set of insights are about the sources of systemic risk: the IMF confirms that size, direct exposure to banks and concentration are not relevant indicators for the insurance sector. The importance of this finding cannot be overstated. It puts into question the designation of the largest companies in insurance, as through diversification, size can be a source of stability and capacity of risk absorption, not a source of risk.
  • A second set of insights regards potential liquidity risks linked to the new environment of low interest rates and customers’ response to this new reality: liquidity risk stemming from lapse is actually declining in the current environment of extremely low interest rates. This is a recurrent theme; the main question is how to deal with the fact that some insurance contracts can be liquidated at short notice with the fact that during times of crisis, savers usually leave their long-term savings untouched.

Figure 1 Liquidity and runs
(Lapse rate experience, 2004-2014)

Source: IMF (2016).

  • A third set of insights regards the right regulatory focus in such an environment: the IMF suggests that supervisory attention should be more focused on the sector as a whole, compared with an approach targeting a few individual entities. Indeed, the debate in the FSB as well in the International Association of Insurance Supervisors (IAIS) is shifting in this direction. This, too, is very welcome: many challenges – low long-term interest rates, population ageing and the retirement savings gap, innovation, data protection – affect the sector as a whole and not just nine companies.

Still some confusion resulting from a remaining lack of appropriate differentiation between bank and insurance functioning

In turn, a few points in the report are worth further discussion, as they implicitly refer to bank-centric concepts and approaches that are not suitable for analysing the risks potentially present in the insurance sector.

  • With historical hindsight, one could fairly say that nobody would have started a global regulatory debate on insurance and potential systemic risk had AIG not collapsed. However, one must be clear that it was AIG’s non-insurance, banking-like activities that lead to this disaster. AIG is no longer engaged in such activities and no major insurance company other than AIG has been operating such business offering financial protection and de facto rating guarantees – business that would have required banking supervisory tools and not insurance ones. There is also a consensus by now that an inappropriate supervisory organisation let this happen, and that measures taken since 2008, notably to install ‘group-wide consolidated supervision’, would probably have prevented such operations at that time.
  • Many conclusions, in this IMF report and in many others, about systemic risk in insurance are based on models of market-based capital shortfalls models (as developed initially by Viral Acharya and the NYU Stern team in 2009). This piece of research is however conceptually bank-based – and not yet accurately interpreted for insurance companies. The reason is that it assumes that capital has the same purpose in insurance companies as in banks, which is not quite the case. Capital in both cases helps to absorb losses, but in banking it is also vital to secure liquidity funding in the interbank market, to control leverage essential for capital raising, and to ensure refinancing with the central bank. Therefore, if a bank has a capital shortfall, its entire funding structure is immediately in danger and systemic risk can unfold. In turn, capital in insurance is exclusively used to absorb exceptional losses that technical provisions (i.e. reserves) cannot cover, and ensure that the long-term commitments to policyholders are met if reserves are not enough to do so. As a consequence, a temporary capital shortfall would not necessarily lead to the unravelling of risk in insurance. Examples of insurance companies living periods of temporary capital shortfalls while staying liquid are well known from insurance supervisors, while totally unthinkable for bank ones. Interestingly, Acharya and his co-authors have started to distinguish more clearly the differences in the banking and insurance business models, and provided a framework for considering how systemic risk of an insurance company needs a comprehensive review of the nature of liabilities and their potential to generate ‘runs’ and risks of fire sales, insolvency and capital shortfalls.
  • Continuing on the different purpose of capital in banking and insurance, policy proposals to implement countercyclical capital buffers are not suited for the insurance sector. In particular, as insurance does not play a role in money creation and maturity transformation, capital buffers will hardly have the expected macro-prudential countercyclical effects. In particular, capital buffers do not help if risk stems from sudden liquidity shortfalls resulting from market movements or unexpected major change in customers’ behaviours. Additionally, local insurance supervisors already have numerous tools and powers to play a counter-cyclical role by adapting their supervisory behaviour to the position of the insurance sector in the financial sector cycle, and are used to do so discretionarily when engaging in their continuous dialogue internal risk managers when the situation requires it.

A way forward for the global regulatory framework for systemic risk in insurance

These points confirm that there is no easy conclusion to be drawn on systemic risk in the insurance sector and the potential policy tools to be proposed. The context of durably low long-term interest rates also raises new questions regarding the best supervisory approach in a situation where the risk mostly comes from the macro and real economy impact, due to the fundamental and largely policy-induced drop in long-term interest rates in capital markets.

So far, the IAIS framework of regulation and supervision has focused very much on nine so-called global systemically important insurers (G-SIIs), which are essentially multi-line diversified global insurance companies. This today appears at odds with the attention to be given, as stated by the IMF, to smaller and less diversified life insurers and more generally the sector as a whole. Size is indeed a factor of stabilisation through diversification in insurance, and particularly when size is coupled with diversified geographical footprint allowing diversification of investment allocation and exposure.

Figure 2 US life insurers’ higher-risk assets
(% of total assets)

Source: IMF (2016).

One way forward for the FSB and IAIS framework of systemic risk in insurance would be the following two-pronged approach.

  • Entrusting the continued close oversight of activities and risk management of the large global insurance groups to the supervisory colleges, with annual updates provided to the IAIS and FSB. Supervisors acting in coordination will be able to see whether there are potential residual risks to the broader financial markets after application of the respective regulatory frameworks in place. This approach should make full use of the existing reports shared between the large companies and the supervisory colleges.
  • Focusing the attention of the FSB and IAIS on the insurance sector as a whole, and more specifically at the disruption to the sector brought by durably low long-term interest rates. This would also make it possible to study new pockets and allow to truly focus on relevant activities.


Bean, C, C Broda, T Ito and R Koszner (2015), Low for Long? Causes and Consequences of Persistently Low Interest Rates, Geneva Reports on the World Economy 17, ICMB and CEPR Press. 

IMF (2016) “Chapter 3: The Insurance Sector—Trends and Systemic Risk Implications”, in Potent Policies for a Successful Normalization, IMF Global Financial Stability Report.

Thimann, C (2014a), “Regulating the Global Insurance Industry: A Compendium of Motivations and Challenges”, CESifo Forum 3/2014, September

Thimann, C (2014b), “How Insurers Differ from Banks: A Primer in Systemic Regulation”, LSE Systemic Risk Centre Special Paper 3, July.


[1] These nine companies, as updated in November 2015, are Aegon, Allianz, American International Group (AIG), Aviva, AXA, MetLife, Ping An Insurance (Group) Company of China, Prudential Financial and Prudential plc. (



Topics:  Financial regulation and banking Global economy Macroeconomic policy

Tags:  IMF, insurance, systemic risk, global systemically important insurers

Head of Regulation, Sustainability & Insurance Foresight, AXA; Professor, Paris School of Economics