Karen Petrou Remarks – Insurance In The Systemic Cross-Hairs: Workable, Realistic Regulation and Resolution for the New Age
It is an honor to speak again to NOLHGA’s annual meeting and to such an influential group of insurance-industry regulators and executives. You gather at a critical time when your expertise – orderly resolution of a faltering insurance company – is urgently needed by both U.S. and global policy-makers. Led by Peter Gallanis, I know you are at the forefront of this vital debate. I’d like this afternoon to put it in the broad framework of how policy-makers are assessing systemic risk both in insurance and other financial institutions to bridge a bit of the communications gap that has so far stymied efforts at a forward-looking, systemic risk-ready resolution solution.
You’re right – the current systemic framework is bank-centric, meaning that much of it doesn’t fit easily – if at all – with how insurance companies operate. However, policy-makers are also right – if there’s no clear, reliable way to ensure recovery and resolution for complex, cross-border, and non-traditional insurance companies, then all the insurance-suitable regulations are for naught. Indeed, they may even be counter-productive, because risk will simply move from big banks to big insurance companies. If only to cut this off, regulators believe they must govern insurance companies and, if they can’t quickly find an insurance-ready solution, then they’ll use the rules close at hand. Bank-centric rules could actually make some insurance companies riskier, but you will understand why policy-makers go to the rules they know to combat the risks they fear.
What I’d like to do in my remaining time is lay out some of the challenges I see and further actions you all could take to bridge the insurance/bank divide. I’ll do so by first describing the risks I think insurance companies pose to financial stability, some of these resulting from growing longevity risk in the sector, some from non-traditional activities, some because of governmental edicts, and some from what happens to big financial companies when big financial markets go bad. I’ll then lay out why I think what you do – resolution – is the most important way to address each of these risks.
Prudential rules protect insurance companies from themselves and make them more resilient to external shock. But, at the end of the day, what we need for market discipline and financial stability are firms that, if they take undue risk, immolate without causing conflagrations for policy-holders and the broader financial system – an end to too-big-to-fail is just as critical for insurance companies as it is for the very biggest banks.
I think it’s important first to recognize what U.S. insurance regulation does well and what it yet has well to address. Regulation has many goals – in insurance, these include ensuring appropriate business conduct and claims-paying capacity under normal market conditions. Post-crisis rules are beginning to address stress scenarios, but divisions among the states, opposition from the industry, and genuine disagreement about what happens to whom how in severely-adverse scenarios has slowed much of this work to a scary crawl. Risk, though, is racing ahead.
What scares me? Time doesn’t permit a detailed description of each insurance-sector risk with potentially systemic ramifications, but let me describe a few that worry me and, more importantly, global policy-makers and U.S. regulators:
- Yield-Chasing: You all know that life-insurance companies must pay out claims over extended periods of time based on actuarial risk. As a result, claims-paying risk is not meaningfully affected by macroeconomic or financial-stability considerations – at least it isn’t when these stresses come in short spurts and the insurance company’s risk profile isn’t heightened by other exposures. But, when external conditions create prolonged periods of ultra-low or, now, even negative real rates, the ability of insurance companies to earn enough prudently to ensure long-term, claims-paying capacity is seriously undermined. Higher-yielding assets pose all sorts of higher risks, risks not well addressed by insurance regulations such as concentration, correlation, “cliff-effect,” or interest-rate risk. As a result, even if companies can weather prolonged low rates and over time meet claims, they could at any point between now and market normalization be subject to significant risks based on how long it takes them to rebalance their portfolios and what happens in the meantime.
- Intra-Group Risk: Although most U.S. life-and health-insurance companies are far less complex than global systemically-important banks (G-SIBs), some companies engage in a wide array of activities across product segments, financial sectors, and national jurisdictions. Insurance companies are not generally subject to inter-affiliate transaction limits that protect insurance claims-paying capacity, nor are there usually limits on upstreaming dividends from insurance companies to the top-tier parent. Under earnings stress or risk at an affiliate (e.g., a captive re), the insurance subsidiary could become severely strained.
- Operational risk. This is a major concern of bank regulators, and it’s just as critical for insurance companies in this day of cyber-security risk. Recovery and contingency-funding plans are critical to operational-risk resilience, but companies can’t always be counted upon to lay in redundant systems, back-up liquidity, and other buffers if rules don’t mandate this.
- Non-Traditional Activities: You are well aware of ongoing work at the International Association of Insurance Supervisors (IAIS) to categorize non-traditional activities at large insurance companies. Securities lending is one to which I draw your attention. It is becoming an increasingly significant activity to support yield, but it creates balance-sheet risk for which insurance-company capital regulation is ill-designed. G-SIBs are under tough new rules affecting capital and liquidity requirements constraining their secfin operations, leading more of it to migrate to large insurance companies. This is a low-margin, low-risk business under normal circumstances, but risks can change fast under stress and it’s not at all clear that large insurance companies are well insulated from it. The same holds true for insurance-company banking and asset-management activities.
- Policy Risk: Let me finally quickly mention one risk I think particularly acute for U.S. insurance companies. You all know this all too well – health insurance is now dictated in form, content, and price by federal and state regulation. This may well be warranted for fairness and even humanitarian reasons, but it changes the underlying business proposition. A wave of consolidation has swept the sector to squeeze out inefficiencies and sustain earnings. History teaches us that not all of these deals will go well and that resulting companies could pose even greater prudential and resolution challenges. The merged company might not be systemic in the near term because health-company risks are less significant in the areas I briefly noted. They could, though, grow in tandem with earnings pressure, especially if ultra-low rates continue. Health insurance is an urgent national service and its absence poses another serious systemic risk akin to the loss of critical market infrastructure like payment, settlement, and clearing.
Resolving Systemic Risk with Orderly Resolution
I hope recognition of these all-too-real risks leads to real prudential solutions, but effective orderly resolution is our safety valve. If rules don’t work, then markets and policy-holders will still be protected if, when an insurance company fails, no one beyond management, creditors, or shareholders gets badly hurt.
As with insured bank depositors, policy-holders protected by your guaranty associations are the top of the claims-priority order – hence the critical role of your guaranty associations. However, you can only do your job – protecting policy-holders – if prudential regulators do theirs, and here’s why:
- claims from a non-traditional activity or non-traditional insurance affiliate could overwhelm the life or health entity;
- claims-paying capacity within or without the entities you guarantee proves not to be resilient because of prudential risk within it or risks not captured by stress testing or other safeguards;
- resolution of an affiliated entity expropriates assets needed for claims-paying capacity; and/or
- risk within the guaranteed entity is correlated with financial-market risk (e.g., capital is in bonds subject to cliff-effect rating downgrades, short-term liquidity is insufficient for claims payment). FRB Gov. Tarullo has suggested that the new capital framework for U.S. insurers subject to the central bank should govern short-term funding risk just like the G-SIB surcharge. The “just like” idea won’t work, but the fact remains that insurance companies are indeed subject to liquidity risk.
With this in mind, what major issues confront you? I think they include:
- handling resolutions when an acquirer is unlikely;
- prioritizing policy-holder claims in the absence of an acquirer or in the presence of conflicting state thresholds for an interstate insurance company;
- anticipating the ability of reinsurance to back primary insurance under stress;
- anticipating problems if parent resolution is predicated on a single-point-of-entry approach that creates policy-holder risk. Multiple-point-of-entry resolutions might work better for diversified insurance companies, but how they would and who would run them is uncertain, especially in cross-border cases;
- knowing the extent to which an insurance company’s ability to exercise early-termination rights creates broader systemic risk;
- assessing conflicting policy objectives. The IAIS protocols subject policy-holder claims to systemic-risk considerations, but the U.S. framework prefers policy-holders in ways that arguably could increase systemic risk by undermining an insurance company’s solvency;
- recognizing the benefits of stress testing and the challenges of requiring this given discontinuities in U.S. insurance regulation;
- obtaining the ability to mandate robust recovery-and-resolution plans; and
- dealing with potential contagion risk if insurance failures occur outside the ambit of the guaranty associations. When several private mortgage insurance companies and bond insurers failed during the crisis, the risk to life- and health-insurance companies was minimal, although one MI was subsumed by the rescue of its parent (AIG). The paroxysm of the banking crisis otherwise distracted investors from the insurance sector. We still don’t know if this is because insurance then was so sound – although I think it generally was – or if insurance companies bought time because banks sucked up the market’s attention.
Banks – especially large ones – are now under a very stringent prudential and resolution framework that bolsters them. The next crisis could thus be far more focused on a non-bank sector – yours?
As I said, you all sit at this critical intersection: the road to making insurance resolution work under non-traditional or stress scenarios is the best way to prevent bank-centric prudential rules from redefining insurance in ways that inhibit customer service and policy-holder protection. Insurance companies could be more resilient than banks, but we need to think through how much of this comes from the differences between credit risk and longevity risk and how much instead comes from insulation from systemic buffers like automatic stays. Insurance companies could also, depending on the circumstances, be riskier than banks because they don’t have access to the Federal Reserve’s discount window, the prime source of stress liquidity support, nor do they have access to safe-haven repositories such as FRB excess reserves. How to solve for this and the other concerns I’ve sketched out today?
First and foremost, I think state guaranty associations should take two cues from the Federal Reserve and FDIC playbook for the biggest banks:
- stress-testing; and
- living wills.
Stress testing would demonstrate a firm’s resilience under an array of stress conditions across the group and throughout its assets, investments, and claims-paying obligations. I know that it is difficult for the guaranty associations to conduct stress tests. Therefore, I recommend robust resolution plans. If you do this within your state and take into account all the relationships between the entity you guarantee and its true exposures, you will better protect policy-holders in your state and, thereby, improve the industry’s overall systemic resilience.
Does this solve for interstate risk, let alone cross-border risk? No. Could it make insurance even riskier if assets and liquidity are ring-fenced within each state? Maybe. For sure this is a piecemeal solution. However, I think it’s a great place to start within the U.S. as you continue your vital work with your global counterparts crafting a more insurance-focused, forward-looking recovery and resolution framework.