Daniel K Tarullo: Insurance companies and the role of the Federal Reserve

Speech by Mr Daniel K Tarullo, Member of the Board of Governors of the Federal Reserve System, at the National Association of Insurance Commissioner s International Insurance Forum, Washington DC, 20 May 2016.

The views expressed in this speech are those of the speaker and not the view of the BIS.

Central bank speech  | 
24 May 2016

It is a pleasure to be here with you this morning to discuss the statutory role of the Federal Reserve in supervising and regulating insurance firms. Since we inherited from the Office of Thrift Supervision the oversight of savings and loan holding companies (SLHCs) that contain insurance companies, we have looked at our role as complementing the role you - as insurance regulators - play. We value the interaction we have had with state insurance commissioners on policy matters and are pleased with the growth of ongoing, regularized cooperation in supervising the firms over which we both have oversight authority. Personally, I have profited enormously from the assistance provided by the staff of the National Association of Insurance Commissioners (NAIC) as I delved into the history and practice of capital regulation of insurance companies.

Today I would like to begin by describing how I view the Federal Reserve's role in insurance regulation and supervision - both what it is and what it is not. As I think you will find, we have tried to occupy that role in a way that is closely grounded in the duties that Congress has given us. Then I will present current thinking on a topic that I know has occasioned great interest - the Federal Reserve's plans for adopting capital requirements for the insurance firms over which we now have statutory responsibilities.

The Federal Reserve's role in insurance regulation and supervision

The U.S. insurance industry has historically been, and remains, a significant part of the U.S. economy. For the industry across all sectors, written premiums last year totaled nearly $2 trillion and earned premiums totaled nearly $1.8 trillion, representing increases from 2014 of approximately 6 percent and 4 percent, respectively.1 This represents over 7 percent of U.S. gross domestic product, as it has in recent years.2 The two largest sectors in the U.S. insurance industry, property/casualty and life/annuities, together wrote approximately $1 trillion of premiums in 2015, up nearly 3 percent from the year prior.3 Moreover, for the U.S. insurance industry as a whole, written and earned premium last year were the highest in five years.4

The Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act) gave the Federal Reserve regulatory responsibilities both for insurance holding companies that own a federally insured bank or thrift and for insurance companies designated as systemically important by the U.S. Financial Stability Oversight Council (FSOC). The insurance holding companies for which the Federal Reserve is the consolidated supervisor hold about $2 trillion in total assets, representing about one-quarter of U.S. insurance industry assets. These firms are highly diverse: they range in size from firms with total assets of approximately $3 billion to firms with total assets of over $700 billion. They engage in a wide variety of insurance and non-insurance activities. Some are fully domestic, while others have material international operations. Some are organized as mutual companies, while others are owned by public shareholders.

The approach of the Federal Reserve in regulating insurance holding companies is derived from its overall statutory responsibilities for financial regulation as those have evolved over the years, most recently through the changes made by the Dodd-Frank Act. First, as was the case prior to the enactment of the Dodd-Frank Act, we are responsible for protecting the safety and soundness of federally insured depository institutions affiliated with any kind of holding company. With the Dodd-Frank Act's transfer of SLHC oversight to the Federal Reserve, that statutory responsibility now includes insurance holding companies. While the particulars are somewhat different than for bank holding companies, the aim and many of the tools to achieve this aim are the same.

Second, the Dodd-Frank Act sharpened our statutory mandate to make clear that the Federal Reserve is to regulate and supervise holding companies with a view to the safety and soundness of not only the holding company itself, but also its functionally regulated subsidiaries, including affiliated insured depository institutions. Again, while the particulars differ, this mandate applies to SLHCs as well as to bank holding companies.5 The Gramm-Leach-Bliley Act of 1999 had, at the least, rendered ambiguous the question of whether the Federal Reserve could take supervisory measures that involved a functionally regulated subsidiary of a holding company unless a specific threat to an affiliated depository institution could be established. The financial crisis underscored the importance for safety and soundness of not waiting until problems in the holding company were manifestly affecting the depository institution.

Third, the Dodd-Frank Act also changed our statutory mandate to require that we regulate holding companies so as to promote the stability of the financial system as a whole. This means paying attention to interconnections among financial firms, correlations of asset holdings, and other characteristics that could endanger the nation's financial stability during periods of stress. One of the most notable differences between the Dodd-Frank Act and other financial legislation over the preceding half century is the frequency with which the phrases "financial stability" and "systemic risk" occur, either as the specific goal of certain new provisions or as a consideration to be taken into account in administering general regulatory authorities. As I have explained elsewhere,6 this systemic or "macroprudential" regulatory aim does not apply equally to all holding companies or banks. Indeed, with respect to community banks, for example, it does not really apply at all. It is important that financial regulation be tiered so as to regulate for the kinds of risks various groups of financial institutions actually pose, rather than to regulate in a monolithic fashion.

Fourth, for all the broadening of our statutory mandate, one feature of the financial regulatory system that the Dodd-Frank Act preserved was the functional regulation of holding company affiliates based on the kind of financial intermediation in which they are engaged. Thus, even though we do apply consolidated capital and liquidity requirements at the holding company level, we do not specifically apply them to broker-dealer, commodities merchant, or investment company affiliates. Nor do we have any role in the investor protection and market functioning mandates in the regulatory regimes administered by the Securities and Exchange Commission (SEC) and the Commodity Futures Trading Commission. Similarly, we have no role in regulating the types of insurance offered by affiliates of the holding companies we supervise or the manner in which the insurance is provided. These matters are the province of state insurance regulators. In fact, because of the overall structure of insurance regulation as specified by Congress in the McCarran-Ferguson Act and elsewhere, the line here is arguably brighter than in other areas.

In pursuing what might be termed its dual regulatory mandate of protecting the safety and soundness of depository institutions and their holding companies and promoting financial stability, the Federal Reserve must develop regulatory and supervisory measures that are appropriate for a range of non-bank financial intermediaries. Consider, for example, the risks associated with bank holding companies containing large broker-dealer operations. Unlike traditional commercial banks whose activities are funded with retail deposits, these bank holding companies are typically heavily reliant on repurchase agreements (or repos) and other types of short-term wholesale funding. The events of 2007 to 2008 reminded us that, while this type of funding may be very stable during normal times, it can disappear overnight in periods of stress. Runs by providers of short-term wholesale funding may not only precipitate the failure of an individual firm by depleting that firm's capital and liquidity buffers, but also precipitate contagion that can imperil the rest of the financial system.

As a result of this very dynamic during the financial crisis, the five large, formerly freestanding broker-dealers have mostly been absorbed by, or converted into, bank holding companies. In our post-crisis development of liquidity requirements and stress testing, and in reforming capital requirements, we needed to pay particular heed to these critical funding differences between traditional commercial banks and broker-dealers. Similarly, in building out our regulation and supervision of insurance companies, we are paying particular attention to how their funding patterns differ from other forms of financial intermediaries.

Here, though, the major relevant difference is that the funding structures of traditional insurers are generally much more stable than the funding structures of commercial banks, much less broker-dealers. A traditional insurance company's liabilities are largely composed of contingent claims based on the occurrence of specified events, such as the death of an insured person or the destruction of property. Because these claims generally cannot be accelerated, companies engaged in traditional insurance activities are less vulnerable to runs and, accordingly, to short-term pressures to sell assets into declining markets. This insulation from creditor runs in turn suggests that capital and liquidity requirements for insurance companies should be calibrated differently than capital and liquidity requirements for dealer banks. Because Congress modified the Collins Amendment in late 2014, we can now tailor capital requirements for insurance companies.

However, it is important to add that the balance sheet of an insurance company can look quite different from the traditional picture I have just sketched. For example, a life insurer may offer investment and retirement products with account values that can be withdrawn at the discretion of policyholders, sometimes with little or no surrender penalty. The option to surrender creates the potential for increased claims that could strain the liquidity of the firm. Relatedly, the extensive involvement of an insurance firm in securities lending, repo, over-the-counter derivatives, and other capital markets activities can create a balance sheet with much tighter connections to the rest of the financial system and greater liquidity risk in times of financial market stress. As with other financial intermediaries, insurers then become subject to demands for posting additional collateral or closing out positions as unfavorable market conditions take hold. In addition, if the positions of the insurance company are large enough, it could become a potential vehicle for transmitting distress at the company to other parts of the financial system.

The foregoing considerations suggest that we should distinguish between insurance companies that we oversee solely because they own an insured depository institution and those that have been designated as systemically important by the FSOC. This is precisely the path we are taking with regard to our supervision of these firms. For the former group, our supervisory efforts to date have focused on ensuring that the companies have strong internal controls and effective corporate governance, as well as satisfactory risk identification, measurement, and management. For firms designated as systemically important, our supervision has additionally emphasized capital and liquidity planning and positions, management of core business lines, and recovery and resolution planning. Board staff meets with the primary state insurance departments that supervise the insurance companies for which we have consolidated supervisory responsibility and, together with lead state supervisors, we co-host supervisory colleges and crisis management groups for AIG (American International Group, Inc.) and Prudential. We continue to discuss supervisory plans and findings for the insurance firms, extend invitations for joint inspection activities, and discuss common areas of supervisory focus.

As you likely know, the Board is also working to develop a regulatory framework for its supervised insurance companies. The Dodd-Frank Act authorizes the Board to establish minimum capital requirements for both systemically important insurance companies and SLHCs and bank holding companies predominantly engaged in insurance activities. For systemically important insurance companies, the Dodd-Frank Act directs the Board to establish enhanced prudential standards in order to mitigate risks to U.S. financial stability that could arise from the material distress, failure, or ongoing activities of these companies. The enhanced prudential standards must include, among other things, enhanced capital requirements, liquidity requirements, corporate governance and risk-management standards, and resolution planning requirements. We have already established resolution planning requirements for the systemically important firms. At present, we have three major regulatory initiatives underway for our supervised insurance firms: first, reporting requirements for the systemically important firms; second, enhanced corporate governance, risk management, and liquidity standards for those same firms; and third, capital requirements for all supervised insurance firms. Let me say a few words about the first two and then spend the rest of my remarks on capital requirements, which I know are probably of the greatest interest.

About a month ago, the Federal Reserve proposed new reporting requirements specifically for the systemically important insurance companies.7 These reporting requirements are designed, among other things, to keep us informed as to the financial condition and risk profile of these firms, as well as the extent to which their activities and operations could pose a threat to the financial stability of the United States. The proposed reporting requirements differ substantially from reporting requirements for bank holding companies, and from available U.S. Generally Accepted Accounting Principles (U.S. GAAP) financial information filed with the SEC. They are highly tailored to the assets, liabilities, and risks of insurance companies. We look forward to receiving public comment on the proposal.

Regarding enhanced prudential standards, the Federal Reserve's enhanced corporate governance and risk-management standards for systemically important insurance companies will likely build on the core provisions of our consolidated supervisory framework for large domestic and foreign banking organizations, with appropriate adjustments to reflect these firms' predominantly insurance business model.8 The framework will mandate standards for the systemically important insurance companies to follow in managing risk and will identify the parties responsible for ensuring the integrity of a company's risk-management practices. Enhanced liquidity requirements will likely include internal control requirements, general comprehensive cash flow projections, contingency plans to manage liquidity stress events, and internal liquidity stress testing requirements. I expect that these proposed enhanced prudential standards for systemically important insurance companies will be considered by the Board in the coming weeks. Of course, we look forward to public comment on these standards, once proposed.

Capital requirements for insurance companies supervised by the Federal Reserve

In formulating capital standards for the insurance companies we supervise, our goals should track our overall goals in financial regulation, as I described them a few moments ago. Thus we should aim, first, to protect depository institutions and, second, to promote financial stability in a manner that takes account of both the unique form of financial intermediation reflected in the traditional insurance business and the activities by a few firms that are much more closely connected to short-term financial markets and the rest of the financial system. As with some of the other measures I have mentioned, these factors argue for different approaches to capital requirements for the systemically important insurance companies, on the one hand, and for the rest of the insurance companies we supervise, on the other. That is, indeed, the direction in which I believe we are headed.

There are, as all of you know, a lot of ideas out there as to how we should construct the capital requirements we will apply to insurance companies. Some, such as variations on the Solvency II approach used in the European Union, strike us as unpromising. The valuation frameworks for insurance liabilities adopted in Solvency II differ starkly from U.S. GAAP and may introduce excessive volatility. Such an approach would also be inconsistent with our strong preference for building a predominantly standardized risk-based capital rule that enables comparisons across firms without excessive reliance on internal models. Finally, it appears that Solvency II could be quite pro-cyclical.

Other ideas have some conceptual appeal, but strike us as impractical for the foreseeable future. Most prominent among these is an approach based on internal cash flow stress testing. The appeal of this idea lies in its melding of capital and liquidity considerations. The relationship between a firm's capital levels and liquidity profile is something that needs to be kept in mind when formulating regulatory regimes for all forms of financial intermediaries. However, as I noted earlier, we are reluctant to rely on internal models for basic regulatory capital requirements, which makes this idea less appealing to us. The novelty and untested nature of this idea as a basis for regulation would, in any case, necessitate extensive development, likely over a period of years. While we have consciously taken a good bit of time in considering a capital regime for our supervised insurance firms in order to make sure we get it right, we cannot wait years more, especially because there is no guarantee that even this investment of effort would produce a feasible internal models-based capital rule. However, I think it is worth continuing to explore internal cash flow stress testing as we build the supervisory stress testing program for systemically important insurance companies.

Another set of possibilities is raised by the work of the International Association of Insurance Supervisors (IAIS). As part of its effort to create a common framework for the supervision of internationally active insurance groups, the IAIS began work on a comprehensive insurance capital standard (ICS) in 2013, issued an initial consultative proposal on the ICS in late 2014, and will continue work on the ICS for at least the next few years. The ICS would be the first international, group-wide capital standard broadly applicable to internationally active insurance groups and would incorporate internationally compatible valuation principles, definitions of capital resources, and risk-based capital requirements for assets and insurance liabilities.

Progress in developing such a global capital standard for internationally active insurance firms has been slow. I will mention some of the difficulties that have been encountered. First, there is considerable heterogeneity among the insurance products offered in different jurisdictions. As one example, annuities and other savings and retirement products tend to be offered less frequently in countries with relatively expansive government pension programs. In the context of developing capital requirements, a second factor has been jurisdictional variations in accounting and valuation standards on which global capital requirements for insurance companies can be built. A third factor has been significant disagreement on the extent to which capital standards for internationally active insurance companies should be built on internal models. These challenges and the protracted process they have occasioned have, as a practical matter, rendered the ICS insufficiently developed to be an option as the Federal Reserve moves forward with capital requirements applicable to the insurance companies we supervise.

A second relevant initiative of the IAIS involves the design of an enhanced regulatory and supervisory framework for global systemically important insurers (G-SIIs). In 2014, the IAIS released the Basic Capital Requirement (BCR) for G-SIIs. It is the first international consolidated capital standard developed for the insurance industry. The IAIS developed the BCR in an effort to ensure the soundness of global insurance firms with the largest systemic footprints. In 2015, the IAIS supplemented the BCR with a Higher Loss Absorbency (HLA) capital standard for G-SIIs. But the IAIS will need to revisit the BCR after the first version of ICS is complete and will also need to revisit HLA after further refinement is made on its G-SII identification methodology. This somewhat provisional character of the BCR, along with its reliance on a method of valuation not in use by U.S. companies and regulators, mean it does not really fit our need for an approach that can - as a practical matter - be developed and implemented in the relatively near term.

These are some of the paths we do not intend to take in formulating capital requirements for insurance firms. The approach we do intend to follow will be elaborated in what I anticipate will be an advance notice of proposed rulemaking (ANPR) to be issued in the coming weeks. As I suggested earlier, the ANPR will likely put forward two different methodologies - one for the insurance firms that we supervise solely because they own a bank or thrift and the other for the firms designated as systemically important.

A. Insurance companies owning a bank or thrift: the building block approach

The Federal Reserve has traditionally set capital requirements for holding companies on a consolidated basis. Among other things, a consolidated capital standard deters firms from moving assets around its affiliates in order to take advantage of lower requirements applied to different affiliates for particular assets. Insofar as publicly traded banking organizations are required by securities laws to file consolidated, U.S. GAAP-consistent financial statements, there is both a parallelism between accounting and capital regulation and a measure of economy in the compliance costs of a consolidated capital rule.

In contrast, most insurance firms subject to our regulation only because they own depository institutions do not produce consolidated financial statements. Furthermore, Congress has prohibited the Federal Reserve from requiring insurance firms that file only Statutory Accounting Principles (SAP) financial statements to produce U.S. GAAP consolidated financial statements.9 Given that these firms have not been designated systemically important and that the depository institutions they own tend to be relatively small parts of the total firm, the compliance costs of requiring a move to some non-U.S. GAAP form of consolidated approach may well not be worth the incremental safety and soundness benefits of doing so. Together, these considerations are leading us to what we refer to as the building block approach (BBA), which is likely the approach that we will put forth for comment in the ANPR for this group of firms.

The BBA would aggregate capital resources and capital requirements across the different legal entities in the group to calculate combined, group-level capital resources and requirements. A firm's aggregate capital requirements generally would be the sum of the capital requirements at each subsidiary. The capital requirement for each regulated insurance or depository institution subsidiary generally would be based on the regulatory capital rules of that subsidiary's lead regulator - whether a state or foreign insurance regulator or a federal banking regulator for depository institutions. The regulatory capital requirement for any non-insurance, non-banking subsidiaries - which at present are relatively minor parts of these firms - would probably be determined under the standardized risk-based capital rules applicable to affiliates of bank holding companies.

Should we adopt this approach, we will need to develop a formula or scalar to put the different regulatory regimes on a comparable basis. With an eye towards the kind of issues that are usually addressed by a consolidated capital approach, we will also need to fashion standards for adjustments to address inter-company transactions and possibly some other exposures. An example of the latter would be potential adjustments to harmonize permitted accounting practices that vary across states.

The BBA would efficiently leverage existing legal-entity-level regulatory capital frameworks that already apply to the various units of the supervised insurance group. As such, it would involve relatively low regulatory burden for these entities, even as it produces regulatory capital requirements that are reasonably well tailored to the insurance-related risks for each distinct jurisdiction and business line of the group.

B. Systemically important insurance companies: the consolidated approach

By definition, the insurance companies designated by the FSOC are systemically important. For these firms, application of an aggregated approach like that of the BBA could pose significant risks to the Federal Reserve's statutory aims of safety and soundness and financial stability. The BBA, recall, would simply aggregate capital requirements that may not be founded upon financial stability considerations. A consolidated form of capital requirements is key to ensuring that risks to the financial system as a whole (as compared, say, to investor or policyholder protection) are accounted for.

For these firms, then, the ANPR is likely to seek comment on what we have internally been calling the consolidated approach (CA). Let me note at the outset that this is not the consolidated capital framework we apply to bank holding companies. As with our capital requirements for bank holding companies, the CA would categorize all of the consolidated insurance group's assets and insurance liabilities into risk segments, apply risk factors to the amounts in each segment, and then set a minimum ratio of required capital comparing the consolidated capital requirements to the group's consolidated capital resources. However, the CA would use risk weights or risk factors that are more appropriate for the longer-term nature of most insurance liabilities.

Like the BCR of the IAIS, the CA would take a consolidated approach to capital requirements. However, unlike the BCR, the foundation of the CA would be consolidated financial information based on U.S. GAAP, with appropriate adjustments for regulatory purposes. Our current thought is that the CA would initially be relatively simple in design, with a relatively small number of risk categories. Thus there would be a broad risk segmentation of asset classes and insurance liabilities. However, as we gained experience with the CA, we would have the option of making it increasingly granular in order to achieve greater risk sensitivity.

The CA, as a consolidated capital framework, would help prevent intra-group regulatory arbitrage opportunities and the potential for double leverage.10 It would also be compatible with supervisory stress testing, which - as you all know - is now a central component of our supervisory program for making systemically important firms highly resilient, as well as for giving us a more complete picture of risks in the financial system as a whole. Obviously, the compliance costs associated with the CA will be higher than those of the BBA. But this enhanced regulation is justified by the increased risks to the financial system associated with these firms. Indeed, it is our obligation under the Dodd-Frank Act to impose stricter regulation on these firms. Even so, by constructing a separate consolidated approach to capital for systemically designated insurance firms, compliance costs for these firms should be considerably lower than if they had to conform to the bank holding company capital regime. Finally, I would note that adoption by the Federal Reserve of a fully consolidated capital requirement for systemically important insurance companies might also help advance the international development of appropriate and effective capital requirements for insurance firms of global systemic importance.


As noted earlier, the Board's prudential regulatory objectives pertaining to supervised insurance companies are to protect the safety and soundness of the consolidated company, to protect any subsidiary depository institutions of the company, and to mitigate any threats to financial stability that might be posed by the activities, material financial distress, or failure of the company. Just as we have done with other non-bank financial intermediaries for which we have been given regulatory responsibilities, we will fashion a regime that takes account of the particular characteristics and risks of those intermediaries. We have taken a good bit of time in arriving at some potential insurance regulatory capital frameworks precisely because we wanted to consider these issues thoroughly. Our tentative conclusion is that a bifurcated approach to a capital regime for insurance companies makes sense in light of these considerations. But we anticipate using the vehicle of an ANPR in order to give all interested parties an opportunity to comment on this approach before we turn to developing and proposing its details. In this process, we of course especially welcome the views of those of you on the front lines of insurance supervision.

Thank you very much.

2 Ibid.; Bureau of Economic Analysis (2016), "Gross Domestic Product: First Quarter 2016 (Advance Estimate) (TXT)," April 28, Table 9; Federal Insurance Office (2015), "Annual Report on the Insurance Industry (PDF)," September, p. 12.

4 Ibid.

5 See 12 USC § 1467a(b)(4)(A).

6 See Daniel K. Tarullo (2014), "Rethinking the Aims of Prudential Regulation," speech delivered at the Federal Reserve Bank of Chicago Bank Structure Conference, Chicago, Illinois, May 8.

7 See 81 Fed. Reg. 24097 (April 25, 2016).

8 See Board of Governors of the Federal Reserve System, Division of Banking Supervision and Regulation (2012), "Consolidated Supervision Framework for Large Financial Institutions," Supervision and Regulation Letter SR 12-17 (December 17).

9 The SAP are a set of accounting rules for insurance companies established by the NAIC and used to prepare those companies' financial statements.

10 "Double leverage" refers to a parent company's use of debt to fund an equity investment in a subsidiary.