## Andrew Large: Convergence in insurance and banking -some financial stability issues
Speech by Sir Andrew Large, Deputy Governor of the Bank of England, at a conference in the Mandarin Oriental Hotel, London, 12 June 2003.
## Introduction
In getting together my remarks, | found myself wondering why you asked me, a central banker and former investment banker, to address this illustrious gathering of the insurance industry. What is it that makes central banks so interested in insurance?
Central banks typically have two major preoccupations. Monetary stability and financial stability. Clearly those of us with an eye on the macro economic drivers which impact monetary stability cannot fail to have taken an interest in the volatility of equity markets. And that volatility has real implications for the life assurance world. Secondly, the way catastrophe and other cover can be extended in a world where terrorism is a fact of life has its own significance for the workings of the real economy.
But it's from the point of view of financial stability that I'd particularly like to develop some thoughts this morning. After all financial stability is about identifying and countering risks and threats which could ultimately impact confidence in the banking or wider financial system. And that gets to both the nature of interactions between insurance and banking on the one hand and to the environment in which we all operate -the areas of accounting, regulation, payment and settlement and law, the broad infrastructure if you will -on the other.
In simpler days, it hadn't occurred to most people that the world of insurance could be relevant to the issue of overall financial stability. The worlds of insurance and banking were relatively distant from each other as were the worlds of securities and banking. Separate silos, if you will. But today insurance companies are an important factor in mitigating risks which earlier would have been contained within the banking world. I'm talking about risks like debtor default. And now, there is increasing overlap and interaction with banks and capital markets.
A personal anecdote might help to make that point; and also highlight my view that the breaking down of the silos is a pretty recent phenomenon. | well remember a time in the 1970s. As an investment banker, | used to handle capital market deals in the early days of the eurobond market. | began to observe that, even though | couldn't articulate what was going on with precision, there were areas of the insurance industry which seemed to be taking similar types of risk as banks, but charging a wholly different price. | remember talking to my friends in the insurance broking industry about it. One or two of them are here in the audience! First of all, was it true? Secondly, if it was, why? And thirdly, if indeed there was an arbitrage opportunity then how could we unlock its value? | remember feelings of frustration. Somehow we couldn't quite see how to reconcile some of the essential differences, including regulatory and accounting distinctions. To say nothing of the language or jargon used in the two sectors. Well, the rapid growth of the derivatives industry during the late 1980s changed all that! All sorts of new instruments for exploiting arbitrage opportunities like the one | described came into being. The different aspects of risk were unpicked and recombined. Those developments depended partly on new information technology. And, together with securitisation the functional divide was crossed. That evolution provided colossal commercial opportunities and enhanced general economic value.
| recall this because it's a good example of how technical advances, some rocket science and market forces have changed the fortunes and capabilities of two great industries. And today, even though we don't have full integration we certainly have greatly increased interdependence and linkages. Just take the involvement of the insurance industry in credit risk transfer as an example. And just as with a lot of interdependencies you get network effects with undoubted economic benefits on the one hand, but potential vulnerabilities on the other.
Maybe | can start by making a few observations on the vulnerabilities -increased risk to firms or the overall financial system -that were less significant until the interconnections appeared. The starting point here will be to try to understand the nature of these vulnerabilities better, and their potential impact on the financial system.
BIS Review 29/2003
## Vulnerabilities: comparing banking and insurance
My first observation relates to the different "customs and practice" of the two industries.
In the banking world, obligations have historically tended to be for amounts which are clearly specified -either in money terms or according to some agreed formula -in advance. The event triggering delivery of the obligation is unambiguous. Often it's simply expiry of a set time period; and the timetable for payment once the obligation becomes due is fixed (and typically short). In the non-life part of the insurance world, on the other hand the amount to be paid is often a matter of interpretation of the contract. There might be greater scope for debate about whether the triggering event has actually occurred or whether any associated conditions have been met -even whether payment is due. And any payment due may not be made, at least in full until negotiations on these points have been completed. The fact that well-established channels may exist to resolve any disputes -including if necessary the courts -doesn't get around the fact that uncertainty regarding the timing and quantum of payment is injected into the process.
Bankers find it hard to reconcile these realities of the insurance world with the way they manage risk and undertake their business. Clearly, this fundamental difference in approach can raise issues for those of us with an eye on financial stability. Specifically, these different characteristics can give rise to cash flow and liquidity implications which, for bankers and others in the financial markets, are potentially highly disruptive. It's fair to say that confidence in the banking world, which enables the system to function relies on obligations being met as they fall due. As the volume and complexity of inter-linked contracts in the system increases the need for certainty on this point also increases. Higher volumes and greater complexity also increase the interest of financial authorities wno have responsibility for stability of the financial system.
To be more specific, a potential problem can arise if, at any point in the chain expected payments do not materialise. Despite the progress which has been made there is still some caution on the part of banking regulators towards recognising the risk mitigation provided by insurance contracts. A recent example is the debate over recognition of insurance cover for operational risk under the new Basel II capital adequacy regime.
That general point is illustrated by the post-Enron case involving JP Morgan Chase and various insurance companies. Whatever the rights and wrongs of the case the very fact that it became necessary illustrates the potential for uncertainty in relation to contracts at the boundary between banking and insurance. Furthermore the JP Morgan Chase case was ultimately settled out of court. So it might not have taken us much further forward in terms of clear guidance or greater legal certainty.
I'm sure you will be revisiting these issues tomorrow afternoon, in your session covering legal systems. So, | won't labour my point much further. But there is a vulnerability here which | think we would do well to try to understand better. We've got to enhance confidence that contracts straddling the two sectors can be relied upon to produce expected outcomes in a timely way and even in adverse conditions.
## Risk transfer
My second observation relates to risk transfer more specifically.
Alan Greenspan made the observation that the successes of the banking system in diversifying risk have increased its ability to withstand significant shocks in recent years, including the Asia crisis, Russia, LTCM, 9/11 to say nothing of Enron, the IT boom and bust, and the telecom write offs. A lot of people agree with him. So the development of markets for the transfer of credit risk is of real value. It allows institutions that are best placed to originate a loan to do so, without necessarily requiring them to continue to bear the risk. Having a market to transfer risk allows institutions to diversify their exposures across different sectors and regions. And of course it generates price information that would not otherwise be available. That means credit exposures can be marked to market in a way that had not previously been possible.
But it raises other questions. The transferors may have proper insight into the nature of the risks. But to whom have the risks been transferred? Are the transferees actually aware of the risks they have taken on? And are they in the best position to monitor these risks as they evolve? How do they acquire day to day knowledge about the quality of those risks? Going further, what might they do when they find out?
BIS Review 29/2003
Some of the concerns being expressed by private and public sector commentators in these areas are, in my view, important and worthy of further examination. Those concerns I'm talking about include: a lack of aggregate data on derivative positions; lack of transparency in accounting for them; and the potential for unexpected concentrations of risk to build up. Although in principle this redistribution of risk should be benign we simply do not have the information at present to judge what the actual impact has been.
But why is all this relevant to insurance? It's certainly true that in a lot of cases originating banks have laid risks off to other banks. There are many possible reasons for this: for diversification or to avoid concentration; opportunities to earn fees; regulatory arbitrage; and different views on pricing and spreads. But it is also well known too that risks have been transferred from banks to various areas of the insurance world. That includes reinsurance, where the degree to which such risk has now been concentrated is difficult to assess. Moreover, these transfers have taken place, not only through OTC contracts, but also through securitised debt instruments, of various credit quality. In my view, this opacity is not conducive to long-term confidence in the strength of the industry. And, it certainly doesn't contribute to overall systemic robustness.
So this is the second area where gaining better understanding of the potential vulnerabilities before we face adverse outcomes could be of real value.
## Supervision and regulation
Obviously, it's pretty easy to make lists of risks and potential vulnerabilities. Central bankers are meant to be especially good at it! But I've already referred to the significant benefits of risk transfer. They include wider distribution of risk, and potentially at lower cost to buyers of protection from those risks. That's got to be of real value. The question is: can these gains be achieved efficiently whilst at the same time mitigating the risks?
My suggestion here relates to the environment in which the industries operate: the general infrastructure | talked about earlier. If we want to get the full benefits of risk transfer, but with the vulnerabilities contained at acceptable levels then it's worth reflecting on the value of co-ordinating some major elements of that environment: supervision, regulation and accounting, as they relate to the worlds of banking and insurance. | can't help feeling that a genuine effort and real engagement in this area will help develop a better understanding of where risk really resides. In turn, that would mean greater predictability, enhanced confidence and better business opportunities.
I'd like to give you a few examples. For a long time, the functional areas of banking and insurance were silos. So were the regulations and indeed the regulators! After all, they mirrored the different business realities | have just described. Here in the UK, the creation of the FSA to oversee both of the areas -and other parts of the financial system as well -is a huge step forward. It helps in developing a regulatory framework which meets business realities on the one hand and public policy requirements on the other. Whatever the complexities it may generate, | believe we are already seeing great benefit. And similar developments are taking place in other countries as well.
Now although the integrated structure of regulatory authorities is making significant progress in some countries, the nature of the regulations themselves are converging only slowly. There is as yet no real equivalent to the Basel Accord in the insurance area: no common prudential template for insurance supervisors exists yet internationally. That said, | fully recognise and applaud the work of the International Association of Insurance Supervisors in setting out best practice and promoting co-operation between supervisors. That's a very helpful base for further work in future. We have to wish them well. But, the current position is that approaches to insurance regulation can vary widely from country to country. On top of that, there is sometimes fragmentation of markets and regulatory responsibilities, notably in the US. Unfortunately that complicates progress towards international agreement. Maybe even more so than what the brave souls in banking supervision who negotiated Basel II experienced!
Of course, progress is being made. In the UK, work is concentrating on the development of a riskbased solvency regime as a counterpart to capital adequacy in the banking arena. This is part of the FSA's Tiner project to overhaul the regulatory environment for insurers. At the EU level the first stage of a review of solvency requirements for insurers has now been completed. Designers of the next stage are now looking to the approach taken for banks under the Basel Accord, focusing on the three pillars to develop further EU solvency standards.
BIS Review 29/2003
It's worth looking a bit at the three pillar approach itself. For those of you unfamiliar with the banking jargon this approach involves combining requirements for minimum capital (the first pillar) with a supervisory review of each institution's profile and internal processes (the second pillar). The third pillar involves market discipline where effective disclosure should allow a better understanding of the risks being taken on by different institutions. That allows the market itself rather than armies of regulators to do some of the work. The same set of principles should be equally relevant for insurance. But, we've got to recognise that the scale of change required is truly formidable!
That's especially true since disclosure leads us inexorably to accounting. After all meaningful disclosure requires a common approach to recording the value of insurance contracts. Otherwise you couldn't achieve standardised capital requirements; and greater transparency is of limited value unless comparisons can be made. So the work being undertaken by the International Accounting Standards Board on a disclosure standard for insurance will be absolutely crucial in this respect.
## Conclusion
My observations about the value of convergence in what | have called the 'infrastructure environment', come down to these areas of focus:
First, convergence will enable better understanding of the reality, or absence, of systemic risk. In particular I'm talking about the potential for transmission of shocks between the insurance industry and the banking system.
Second, it will discourage arbitrage that is based on technical accounting and regulatory factors. Instead, convergence will enable the market to focus on arbitrage which reflects the substantive evaluation of risks where real economic value may reside.
And third, the added transparency of interfaces between the two industries will enhance confidence in the risk transfer arena. That will help dissipate the sort of unease which | sense around the place at the moment.
The data deficiencies we confront today where the actual location of risk is so opaque is to my mind a major challenge. Better transparency will surely help to avoid adverse consequences of shocks which are inevitable from time to time. The infrastructure environment has to be robust enough to withstand these shocks.
Of course there are challenges. For example: in avoiding over-engineering and over-prescription in regulation and accounting standards, and sticking more to a principles-based approach. Those principles have got to be based on best practice in the markets on the one hand, and the public policy needs on the other. Here, I'm including adequate systemic security. My thesis would be that the very effort necessary to define those principles and the necessary engagement between leaders of the industry on the one hand and public authorities on the other will reward us with a system where all parties can operate and contribute with confidence.