Meeting worlds? Insurance and banking

Speech by Malcolm D Knight, General Manager of the BIS, at The Geneva Association 32nd General Assembly, 2 June 2005.


Mr Knight discusses how the worlds of banking and insurance are moving closer together, a process triggered by financial liberalisation and financial innovations. He explores the challenges that practitioners and policymakers face in three increasingly interrelated areas: risk management, prudential regulation and accounting reforms.

Full speech:

It is a great pleasure for me to have the opportunity to address such a distinguished audience of insurance industry representatives and insurance supervisors. The fact that a banker and former central banker such as myself has been invited to this event is a tangible sign of how far the banking and insurance worlds have gravitated closer together over the last couple of decades. Another such sign is that, since 1998, at the BIS - once the unique preserve of central bankers and banking supervisors - we have had the honour and privilege of hosting the International Association of Insurance Supervisors (IAIS). This has helped us to observe more closely a world that we could hitherto only glimpse from afar.

In my remarks today, I would like to share with you some personal reflections on this stronger gravitational attraction between the two worlds of insurance and banking. And I would like to do so by focusing on the unprecedented coming together of three areas, namely risk management, prudential regulation, and accounting standards. What are the forces behind this convergence? What are the implications for us all?

I will structure my remarks in four parts. First, I will briefly recall the long-term forces that have brought banking and insurance closer together - admittedly at a different pace across countries - as well as how these long-term forces manifest themselves in the evolving financial landscape. Then I will consider their implications for each of the three areas I have identified. While I will be drawing on the past, I will also be looking to the future. What are the challenges ahead? How best can we meet them?

My overarching message is that the increasingly common challenges faced by the banking and insurance industries in risk management, in prudential regulation and in accounting, put a premium on cross-fertilisation among these different perspectives and traditions. There is a unique opportunity to develop new and richer solutions based on a clear recognition of what the various parties - be they practitioners in banking or insurance, their supervisors and regulators, or their accountants - can bring to the table.

I - The evolving financial landscape: meeting worlds

By now, we are all very familiar with the fundamental forces that have brought the banking and insurance industries closer together. Even so, as a backdrop, it is worth recalling them briefly. Four such forces stand out.

First, there has been a major desegmentation of finance, as regulatory and technological barriers have come down. Regulatory and legal restrictions on balance sheets, products and prices, which were so common in the old days, have been substantially reduced. Quantum leaps in information technology have spawned a much greater variety of insurance and banking products and made it easier for institutions to contest each other's markets. As a result, a wave of heightened competition has swept through the financial industry.

Second, there has been an unprecedented marketisation of finance. By this I mean that financial markets have come to play a growing role as vehicles for the provision of financial services and the allocation of risks. Securitising funding and risks through securities and derivatives has become increasingly attractive. Crucially, private sector financial institutions have come to rely more and more on markets for the management of risk, notably through derivatives.

Third, there has been a globalisation of finance.This has been reflected in growing cross-border activity, be this through new establishments in foreign countries, acquisitions or - admittedly more rarely - the direct cross-border provision of financial services.

Finally, there has been an institutional repositioning of finance.By this I mean two different types of growing polarisation. One is between large, sometimes global, players, on the one hand, and smaller, often local or national ones, on the other. The second is between financial supermarkets - of which the growth of bancassurance is the clearest example at one end, and specialised players at the other.

Admittedly, these developments have not affected the banking and insurance worlds at the same time or to the same extent. The banking industry experienced these forces first and more deeply. Deregulation spread to insurance in earnest only somewhat later. The securitisation of insurance products has proved harder, not least because of greater difficulties in standardising contracts and in defining the triggering "events" clearly. Except for reinsurance, arguably the insurance industry is, on balance, less international or global than banking. Many argue that it is also more heterogeneous, with the life and non-life segments involving quite distinct types of business.

Even so, the similarities and growing areas of overlap are much more significant than the differences. Competition between banking and insurance has greatly intensified, especially between the asset management and life insurance segments, as the two industries have strived to gain the allegiance of a richer, ageing retail customer, left by a shrinking state to look more after his or her own retirement needs. Thus, the purely "financial" component of insurance products has been increasing; think, for instance, of single premium unit-linked insurance policies, which compete directly with other financial products as resting places for household assets. The establishment of banking-insurance conglomerates has been specifically aimed at exploiting synergies in the battle for the retail asset holder. It has promised to combine insurance companies' competitive edge in the "production" of insurance products with banks' edge in their distribution, through their vast retail networks. And it has spoken to retail customers' preference for "one-stop shopping". Marketisation has begun to make inroads in insurance too; the development of catastrophe bonds and the emerging securitisation of vehicle insurance are just the beginning of the process. And as investors in and suppliers of guarantees, insurers have become increasingly active in credit derivatives and structured finance products. Moreover, the broadening range of activities of The Geneva Association is testimony to the growing recognition in parts of the insurance industry of its increasingly international and global character. In other words, there is no doubt that the two worlds of insurance and banking have been converging, and will continue to do so.

II - Implications for risk management

The major changes in the financial landscape that have brought these two activities closer together have also left a major imprint in risk management. Risk management techniques have tended to evolve in similar ways and to draw closer, based on the slow but steady development of a common approach, or common "language", to address what have appeared as increasingly common problems. This common language, in turn, is deeply rooted in market concepts. And it has the potential to reinforce convergence between banking and insurance. Let me elaborate.

Risk management techniques have been profoundly shaped by two trends.

The first is the formalisation of risk measurement. By this I mean that the analytical quantification of risks has gradually gained ground at the expense of the use of less formal and more qualitative judgmental methods. To be sure, analytical models cannot, and never should, substitute for judgment. The point here is simply that breakthroughs in know-how in finance, coupled with the immense improvements in computer power, have led to a true revolution in the use of advanced statistical techniques as the basis of risk measurement and pricing.

The second trend is the homogenisation of risk measurement. A common analytical framework has been emerging to measure all types of risk, by deconstructing them into their granular components. This framework is based on building blocks such as underlying risk factors; stochastic processes for cash flows or asset values; cross-sectional aggregation through interdependence, as measured, for instance, by correlations; and time aggregation through discounting and risk premia. I think most of us are now familiar with the resulting constructs, including value-at-risk, dynamic stochastic asset-liability management, stress testing, and so on. This is the same framework that we use to try to deconstruct the behaviour of financial market prices.

To be sure, differences in some of the risk characteristics in banking and insurance remain and will persist. They relate, in particular, to the time pattern and degree of uncertainty in the cash flows. For example, the maturities of insurance liabilities in the life segment extend well beyond those in banking, and the option characteristics of bank deposits imply that liquidity risks in banking are quantitatively very different from those in insurance. Similarly, the payment for insurance policies is upfront (through premia); while in quintessential banking products, such as loans and deposits, it is predominantly back-loaded (through interest). Even so, the underlying structure of the risks is fundamentally the same.

No example illustrates more clearly how superficial diversity disguises fundamental similarities as a comparison of value-at-risk and stochastic asset liability management techniques - the two tools claimed to epitomise the irreconcilable differences between the worlds of banking and insurance. These tools do differ in terms of the breadth of instruments involved - individual asset portfolios vs a combination of assets and liabilities - and the length of the horizons employed - a few days vs typically a year. But, conceptually they are otherwise identical.

Moreover, differences within insurance can be at least as great as those between insurance and banking. As a result, where best to draw the line for risk measurement purposes can be somewhat arbitrary. Think, for example, of differences in the degree of uncertainty surrounding wholesale legal liability and standard retail insurance, reminiscent of, but perhaps even larger than the differences between wholesale and retail risks in banking.

The homogenisation of risk measurement has gone hand in hand with efforts to edge closer to a more integrated firm-wide approach to risk management. Firms are striving to manage the same risk in the same way, irrespective of business lines. And they are seeking to develop a common metric for aggregating different types of risk so as to define a single, bottom line measure of insolvency risk and calibrate the needed economic capital. To be sure, even cutting-edge enterprises are some way from achieving either objective. Across risk types, it has been easier to integrate traditional market risks and credit risks. By contrast, the integration of counterparty, liquidity, insurance and, quintessentially, operational risk, is proving much harder. Within organisations, harmonisation of risk practices has made more progress in comparatively more homogeneous firms, such as in investment banks. It is, for instance, very much in its infancy in conglomerates that combine banking and insurance. Even so, the broad direction is unmistakable.

In this process, the opportunities for cross-fertilisation between the banking and insurance perspectives abound. For instance, if banks have shown a comparative advantage in addressing and modelling risks over short horizons and in developing derivative instruments, insurance companies have a longer tradition in developing asset-liability management techniques, in stress-testing and in linking risk measurement and pricing policies in their mainstream products.

III - Implications for prudential regulation and supervision

The similar and increasingly convergent evolution of banking and insurance has also meant common challenges and responses for the prudential authorities that oversee the two industries. Let me briefly elaborate on four challenges and the responses to them.

The first challenge is the risk measurement challenge. This is the need to ensure that prudential standards keep up with developments in risk management that have tended to blunt their effectiveness, especially by encouraging regulatory arbitrage. And it has meant aligning prudential frameworks with risk management practices, while at the same time avoiding excessive interference in management decisions and taking into account differences in the degree of sophistication among firms.

The second is the level playing field challenge. In a world where prudential restraints have become potentially more important determinants of competitive advantage as other regulatory constraints have been eased, the authorities have faced the pressure to promote a level playing field along different dimensions, ie among firms within the same sector; among firms in different sectors; and, for those competing internationally, among different regulatory jurisdictions.

The third is the multiple jurisdictions challenge. This means ensuring the effective supervision of groups operating across jurisdictions, while at the same time avoiding the risk of an excessive cost burden - a burden that reflects the disconnect between the shift towards group-wide risk management, on the one hand, and segmented and overlapping regulatory jurisdictions, on the other.

The fourth is the systemic risk challenge. This is the need to monitor and control risks not just on an institution-by-institution basis, but on a system-wide basis. This perennial need has been highlighted further by the growing interdependencies and common exposures within and across sectors and by the ease with which risks can be transferred across them, not least through derivatives.

The common responses by prudential authorities have evolved along several lines. First, the authorities have strengthened their emphasis on improved risk management in several ways: they have increased the risk sensitivity of their standards (so as to make them more effective); placed a greater focus on the qualitative aspects of risk management (so as to allow flexibility for firms); relied more on transparency and disclosure (so as to enlist market discipline); and sought to avoid a one-size-fits-all approach (so as to tailor standards more closely to the risk management capabilities of individual institutions). Second, prudential authorities have increasingly sought to achieve more harmonised standards across countries and sectors. Third, they have strengthened the mechanisms for the exchange of information among themselves. Fourth, they have improved the monitoring of risk transfers across sectors.

Admittedly - and this reflects the timing and extent to which the forces shaping the financial industry have affected banking and insurance - in some respects the process has not yet gone as far in insurance as it has in banking. For example, reflecting the less international nature of much of the insurance industry, co-operation among national authorities is of more recent vintage; the IAIS was established in 1994, the Basel Committee has been in existence since 1974. And the naturally stronger focus of insurance authorities on policyholder protection, rather than the systemic stability concerns, which are paramount for banking supervisors, has meant that a more system-wide approach to risk monitoring among insurance supervisors has emerged only more recently. Examples include the work done by the IAIS to increase the transparency of the systemic risks in reinsurance and, together with other prudential authorities, on credit risk transfers across sectors.

Even so, the broad common direction is, again, unmistakable. For instance, both the IAIS and the Basel Committee have developed core principles for their respective sectors. Within Europe, the Solvency II Directive resembles closely Basel II in its three-pillar structure. Globally, the IAIS is developing risk based solvency and capital standards. Moreover, nationally, the establishment of integrated supervisors and, internationally, that of the Joint Forum and the Financial Stability Forum, have given impetus to the cross-fertilisation and more intensive cooperation between banking and insurance regulators. Through mutual learning, the two worlds are bound to draw even closer.

The fact that the BIS has been hosting the secretariats of both the Basel Committee and the IAIS has no doubt helped this cross-fertilisation. It has allowed the two secretariats to coordinate their activities more closely, to benefit from day-to-day contact and to cross-participate in selected working groups on the whole range of common issues that they are increasingly facing, including solvency and financial reporting challenges. In addition, I believe it has also permitted them to benefit from the analytical resources present at the BIS, ranging from expertise in risk measurement to purely statistical issues. For example, the long-standing experience of the BIS in the production of statistics was very valuable in the IAIS's effort to develop statistics to increase the transparency of the reinsurance sector. And it has allowed them to be more intimately exposed to the broader concerns of the central banking community, essential in a world in which systemic issues have become so important.

IV - The accounting challenge

What about the accounting challenge? Insurance and banking are both in the process of adjusting to major changes in International Financial Reporting Standards - changes that have in turn been spurred largely by the same forces which have been bringing the two industries closer together. Not surprisingly, this has raised similar issues and responses in the two industries.

In this final part of my remarks, I would like to highlight the nature of the challenge and suggest how best to address it in a constructive and common way. Here, as in other areas, risk measurement will play a key role.

The policy challenge in creating internationally accepted accounting standards is twofold. First, it is to manage the tension between business practices that have adjusted to differing national accounting conventions and the need for common and reliable accounting standards in internationally-active industries. This is especially true in insurance, where differences in national accounting standards are even greater than in banking. Second, it is to manage the tension between the approach of prudential supervisors, more naturally inclined to favour accounting principles that are seen as conducive to prudent financial risk management and financial stability, and that of accountants, more naturally inclined towards what they regard as an "unbiased" (or "true and fair") portrayal of the financial condition of a firm.

This difference in perspective is evident in the current debate over accounting reforms. Prudential authorities have generally been more inclined to accept certain accounting methodologies that are consistent with risk management practices that they see as reinforcing prudent behaviour on the part of firms. One example is that banking supervisors would like accounting standards to accommodate more forward-looking provisioning; they often see current rules as too backward-looking. In insurance, similar issues have arisen concerning the valuation of catastrophic reserves. Another example is that banking supervisors would like greater flexibility for firms to macro-hedge their risks, and, in the process, to recognise the economic behaviour of core deposits. In insurance, analogous issues pertain to the valuation of renewal policies.

To my mind, three basic principles could help reconcile the perspectives of accounting standard setters and prudential authorities.

First, as a long-term goal, it would be useful to keep distinct the pursuit of two objectives: that of portraying the best approximation to an "unbiased" picture of the financial condition of the firm, on the one hand, and that of ensuring prudent behaviour, on the other. This "decoupling" would mean using different tools for each. Specifically in this ideal world, one would instill discipline by relying on prudential instruments that applied to that unbiased portrayal; would start with a common set of valuations rather than from one for accounting and another for prudential purposes; and would avoid creating hiddensafety cushions through deliberately conservative measures of value and income.

Second, an "unbiased" portrayal of the financial condition of the firm would ideally go beyond point estimates of value, income and cash flows - the information typically included in the primary accounts. It would also encompass information about the risk profile of the firm, as captured by point estimates of items like VaRs and stress tests - information sometimes already included in supplementary disclosures. And it would, in addition, give a sense of the degree of uncertainty surrounding the point estimates in primary accounts and in supplementary risk disclosures, at least for the most important pieces of information. This would reflect the uncertainty surrounding the key assumptions made in the calculation of those estimates, such as in the case of the risk premia used in arriving at the value of insurance liabilities. Sensitivity analysis, for example, might be helpful in this context. Full transparency also requires acknowledging the limits of our knowledge.

Third, it is essential to ensure a smooth transition to this long-run "unbiased" accounting valuation. Critically, the prudential authorities should always be able to offset any adverse implications that the adoption of new accounting standards may have on the safety and soundness of supervised institutions. Insurance supervisors, for instance, are justifiably concerned about artificial volatility in the accounts, as the shift to marked-to-market valuation of assets will likely begin well before an international financial reporting standard for valuing liabilities is agreed and implemented. Banking supervisors have faced similar issues, especially in relation to hedging practices. So-called "prudential filters" can be useful as a general offsetting device.

All of this calls for an intense dialogue among prudential supervisors, accounting standard setters and you, as market participants. You and prudential supervisors have a lot to bring to the table, based on a long-standing expertise in risk measurement, validation and control. This expertise is extremely valuable not only in developing risk information, but also in articulating standards for the primary accounts. For instance, valuations fundamentally depend on risk premia, whose measurement remains a major challenge. Moreover, recognising sound risk management practices and hence also valuations on an aggregate portfolio basis - rather than individual instrument basis - is essential to achieve alignment of underlying economic realities with financial statements. I am pleased that the dialogue between prudential authorities and the IASB has intensified in recent years. In fact, insurance supervisors have been drawn more closely in the development of standards than their banking counterparts, as an international standard for valuing insurance liabilities is not yet in place. This dialogue should intensify further in the future.

To conclude, let me very briefly underline my main message. Financial liberalisation and financial innovations have drawn the worlds of banking and insurance closer together, by desegmenting the financial industry and by spurring competition. As a result, practitioners and policymakers have faced similar challenges, and tended to respond to them in increasingly similar ways. This has been reflected in risk management, which has been developing a common analytical basis and a common language to help discipline the essential role of human judgment. It has been mirrored in prudential regulation, which has sought to ensure safety and soundness in ever more market-friendly ways. And it has been evident in the response to accounting reforms, which is forcing a rethink of hitherto well established practices. Finding adequate answers to the challenges in these three increasingly interrelated areas puts a premium on the cross-fertilisation of different perspectives - on the ultimate recognition that what we have in common and brings us together is stronger than what sets us apart. At the BIS, by bringing closer together the Basel Committee and the IAIS, we are fully committed to supporting this process.