Loading ...
Global Do...
News & Politics
2
0
Try Now
Log In
Pricing
S16 INSURANCE RISK ● RISK JUNE 2002 ● WWW.RISK.NET Regulatory capital A s financial services firms converge, the range of products they offer nat- urally increases. The next task is to choose the best methods to most effec- tively manage risk and economic capital. This article, prepared under the auspices of the UK actuarial profession’s Action Group for Banking, looks at the various developments that occur as such conver- gence progresses.1 First, we outline how significant regu- latory capital arbitrage opportunities may exist due to the way the existing regula- tions have evolved in an arbitrary fash- ion, with little or no regard given to similar risks written by different types of finan- cial services entities. We also seek to show that if regula- tory capital is set equal to economic cap- ital (possibly plus a margin), with economic capital determined relative to both aggregate assets and liabilities (rather than relative to just the assets or the liabilities of a single business line), it may be possible to significantly reduce capital arbitrage opportunities. Finally, we demonstrate how ‘natural hedges’ may be exploitable for certain risks, thereby reducing hedge costs and increasing capital efficiency. For the purposes of this article, we as- sume a financial services company com- prises at least a life assurance company and a retail/wholesale bank operating under UK regulation. Regulatory changes The regulatory landscape is set to change for UK financial firms, with the intro- duction of principal UK financial regu- lator the Financial Services Authority’s (FSA) consultation paper CP97, which sets out new rules governing the regu- lation of financial services firms, the new Basel Accord2, fair-value accounting and other potential new legislation/regula- tions – such as that eliminating the use of capital to cover the solvency of both a subsidiary and a parent. It is to be expected that the new reg- ulatory environment will attempt to har- monise the regulatory capital required to back a particular risk, irrespective of where that risk is written in a financial services company. This is reasonable, provided the amount of capital required to be held against a risk is ‘fair’, bearing in mind the size and type of risk. Risk collection, hedging and management Examples of the types of risks financial services firms collect as a result of their product offerings and distribution chan- nels, together with typical types of col- lectors, are shown in table A (page 15). Once these risks have been collect- ed, firms essentially have three possible approaches: ■ Keep the risk on their own balance sheet, without fully hedging it – for example, life assurance with profits funds investing in equities; retail banks promoting mortgages; and reversion companies purchasing reversions. ■ Keep the risk on their own balance sheet, but hedge it internally – for ex- ample, life assurance unit-linked funds investing in equities and passing the market risk onto their unit-linked pol- icyholders; or general insurer/whole- sale bank groups offering catastrophe insurance and passing the catastrophe risk onto bond investors. ■ Remove the risk from their balance sheet, at least in part, by hedging the risk externally – for example, reinsur- ance of life assurance annuitant longevity risk, ie, the risk that annui- tants live longer than has been priced into a life assurance company’s annu- ity rates; or asset securitisations of re- tail bank residential mortgages. This article is concerned mainly with the second of these approaches. In particu- lar, as financial services firms evolve into financial services conglomerates, risks collected by one component of a con- glomerate may ‘sit better’ – or be better matched – on the balance sheet of a different component within the group. For example, it may be more appro- priate for the mortality/morbidity risks collected by a reversion company to be held by a life assurance company. Moreover, a risk collected by one com- ponent company within the group may form a ‘natural hedge’ (see later) for a risk on another component company’s bal- ance sheet. As financial services conglomerates move risks between their component company balance sheets, the amount of regulatory capital required to back the risk will generally vary across balance sheets, opening up the possibility of regulatory capital arbitrage. The position we take in this article is that: ■ the economic capital required to back a particular risk should not depend on the balance sheet on which the risk is held; and ■ regulatory capital should broadly equal economic capital – although often regulators may require an addi- tional margin over and above this. The only exception to this position is where a natural hedge exists on a partic- ular component company balance sheet and the risk requires less economic cap- ital there as a result. Capital links As mentioned earlier, the regulatory en- vironment is moving towards one where regulatory capital is more closely linked with the amount and type of risks a fi- nancial services firm takes on. In short, regulatory capital is moving closer to eco- Managing post-convergence risks in financial conglomerates Bruce Porteous looks at how financial conglomerates are rising to the challenge of managing the potential capital requirements associated with the diversity of products on their books 1 Anyone interested in participating in further/related work should contact Mark Symons at marks@actuaries.org.uk 2The new Basel Accord – Basel II – stipulates new risk-based rules that set out the amount of capital that banks must hold to cover their banking risks Regulatory capital nomic capital. The new Basel Accord, Basle II, which will eventually apply to banks worldwide, is perhaps the best cur- rent example of this. However, since UK life assurance and banking regulations, for example, have evolved and developed in arbitrary and unilateral ways, the same risk written in different types of financial services firms may require different regulatory capital charges. To create a clearer picture, consider the following examples: ■ For UK life assurance companies, reg- ulatory capital is measured mainly relative to liabilities, whereas, for banks, it is measured mainly relative to assets. So a mortgage written as an investment of a life fund may require a different amount of regulatory cap- ital compared to a retail bank writing the same mortgage. ■ Interest rate risk written by a UK life assurance company requires regula- tory capital via the ‘resilience test’ – an asset/liability mismatch test pre- cribed by the FSA from time to time WWW.RISK.NET ● JUNE 2002 RISK ● INSURANCE RISK S17 Here we consider two specific examples of financial ser- vices products that require different amounts of regulatory capital, depending on where they are written or how they are structured within a UK financial services firm. A property reversion Consider a financial services firm offering a reversion equity release plan (ERP) product to asset-rich, income- poor pensioners who own their home. The company receives a share in the capital value of the home, valued at a discount to its market value and realis- able only when the property is vacated by the pensioners. The pensioners, in return, receive an income for life and/or a cash lump sum, together with the right to live in their home for life – or until they are unable to look after themselves and must go into care. For the purposes of this article, we will consider the pure income variant of the product where, for a retail bank, the income is provided by the purchase of a life annuity. In this instance, the product asset is the rever- sion and the product liability is, broadly, the funding required to purchase the annuity. The reversion is, in effect, a house price inflation (HPI)- linked zero-coupon bond of uncertain term, and the fund- ing might take the form of a retail price index (RPI)-linked zero-coupon bond with the excess (HPI minus RPI) expo- sure hedged. In this example, therefore, the retail bank’s assets and liabilities are reasonably well matched. There is, however, a cashflow mismatch, as the reversion does not generate any income until the property is vacated and sold, but this is of secondary interest here. When the bank writes this product on its balance sheet, the amount of regulatory capital that will be required is around 10% of the market value of the booked reversion – direct investments in residential properties are 100% risk- weighted, and we assume that the bank has a risk asset ratio of around 10%. We will assume no additional capital is required for interest rate risk as the assets and liabilities are broadly matched, and this is consistent with the treatment of banking-book interest rate risk under Basel II, the new Basel Accord. To this initial 10% we now add the capital that must be held by a life assurance company to back the income annu- ity – that is, 4% of mathematical reserves. So, assuming the annuity reserve is about equal to the initial booked market value of the reversion, and that it is well matched in the life fund of the assurance company, the total initial capital required to write a reversion ERP product is around 14% of the initial market value of the reversion. If a life assurance company writes a reversion ERP, it is more likely to treat the reversion as an asset of the life fund, with the income, or annuity, as a corresponding liability. So, an effective HPI-linked zero-coupon bond may be used to back an annuity, resulting in a potential asset/liability mis- match, ie, the volatility of the liability may be greater than that of the asset under bond-yield movements – assuming that bond yields net of HPI, and so reversion market values, are relatively stable under pure bond-yield movements. Nevertheless, specimen calculations show that a life assurance company will hold capital of less than 14% of the market value of the reversion, provided the associated resilience reserve – the additional reserve that life assurance companies are required to hold to cover the resilience test, an asset/liability mismatch test prescribed by the FSA from time to time – is less than around 250% of the annuity mathematical reserve, and this will certainly be the case. This example shows that if a life assurance company writes a reversion ERP in its life fund, it will generally require less regulatory capital than a bank/life assurance company combination product would – assuming a product structure that is more appropriate for a bank/life assur- ance company product. The products and risks are identical and, at least in principle, the regulatory capital ought to be independent of the product structure. Gilt/mortgage swap Under Basel I, the current Basel Accord, retail mortgages are risk-weighted at 50%. For example, a bank with a risk- asset ratio of 10% has to hold 0.1 x 0.5 = 0.05 of capital per unit of mortgage issued. Under Basel II, the amount of capital required to back a mortgage is likely to fall significantly (Porteous, 2001), but capital will still be required. Gilts, on the other hand, can be risk weighted at 20% under Basel I, but will require no capital at all under Basel II. Provided assets and liabilities are duration-matched, the amount of regulatory capital a life assurance company must hold to back a particular class of business will be insensitive to whether the assets are gilts or mortgages. It seems likely, therefore, that under Basel II, bank/life assurance firms could exchange ownership of the mortgages written by the bank for gilts held by the life assurance com- pany, thereby significantly reducing the capital requirements of the bank, but with broadly neutral effects otherwise. This means that a structuring may be achievable where the mortgages/gilts switch from the bank/life assurance company balance sheet to the life assurance/bank balance sheet, in such a way that the investment return earned by the bank/life assurance company is broadly unaltered. As the risks taken on by the firm in aggregate are unaf- fected by this arrangement, the aggregate capital require- ment should also be unaffected by it. Specific product examples Regulatory capital – whereas interest rate risk written in a banking book may not require any additional capital at all, depending on the view of the regulator. These two examples show that reg- ulatory capital should be based on eco- nomic capital, where economic capital is measured relative to both sides of the balance sheet. What is important is how robust the balance sheet is as a whole in terms of high-risk events – that is, events that may have a low probability of occurrence but a big impact on the balance sheet. Determining economic capital It will be useful at this point to outline how a company might determine its aggregate economic capital requirement – this approach may also assist financial services firms in allocating economic capital across their separate businesses. The following method is one that UK life assurance companies often use for managing their capital: ■ All assets and liabilities are consoli- dated into one balance sheet, possibly sub-aggregated by business line. ■ Assets and liabilities are valued at mar- ket value, or at a value approximating to market value, so any conservatism or margins built into the asset/liabili- ty valuations should be removed. ■ Associated assets and liabilities should be valued consistently relative to each other. For example, the discount rate used to value a liability/asset should be consistent with the associated asset/liability’s market value. Assets and liabilities will, as a result, have valuations that are dynamic and con- sistent relative to each other. The aggregate economic capital re- quirement of the firm is then deter- mined as the amount of capital required by the firm to achieve a target level of solvency following a series of prescribed shocks to the aggregate balance sheet. These shocks can be deterministic, stochastic, or a mixture of the two. If we consider the equity release plan (ERP) reversion product discussed in the box (previous page), we can see that, under the approach outlined here, the economic capital requirements of the product will be broadly the same under the two proposed product structures. This is because under the bank’s product structure, its liability in respect of the loan used to purchase the annu- ity will be broadly equivalent to the life company assets backing the annu- ity, and these two items will then broad- ly cancel each other out. Similarly, the effect of the gilt/mortgage asset exchange will also be neutralised. Natural hedges Finally, as we have mentioned the term ‘natural hedges’ in this article, it will be useful to consider an example of such a hedge. Reversion companies, banks and building societies may become very large collectors of housing price index (HPI) risk through equity release re- version products (see the Report on Eq- uity Release Mechanisms (2001) prepared by the UK’s Faculty and In- stitute of Actuaries). However, such firms may be reluctant to keep this risk on their own balance sheets, but at the same time may find it difficult to hedge the risk externally in sufficient volume at the right price. Possible natural hedges might include the following: ■ A life assurance company that is a part of the same firm may be able to use the HPI risk to back HPI-linked prod- uct offerings. ■ As HPI is normally highly correlated with average earnings, the HPI risk might be usable by pension funds, with the residual risk hedged outside the firm. It may therefore be possible for banks and building societies to hedge the HPI risks present in their reversion assets with those present in liabilities held elsewhere in the group. The economic/regulatory capital required to back this risk will, as a result, be greatly reduced. Firms have the option of hedging risks via arrangements with third parties, but if a natural hedge exists: ■ firms will be able to avoid the cost of the third-party hedge; and ■ less aggregate capital will be needed across the financial services system, thereby increasing the capi- tal efficiency of the financial services markets. Bruce Porteous is financial risk manager at Standard Life Bank in Edinburgh. e-mail: bruce_porteous@standardlife.com S18 INSURANCE RISK ● RISK JUNE 2002 ● WWW.RISK.NET Type of risk Collector Mortality/morbidity Life assurance, health insurance, pension funds, reversion companies Business retention All Expense All Market risk (for example, equity investment) Life assurance, asset management, pension funds, investment banks House price inflation Life assurance, retail banks, reversion companies Credit Life assurance, asset management, pension funds, retail/wholesale/investment banks Interest rate Life assurance, retail/wholesale/investment banks Currency Life assurance, pension funds, wholesale/investment banks Retail price or earnings inflation Life assurance, pension funds Liquidity All Claims experience risk General insurance, health insurance companies Operational All A. Types of financial risk The Actuarial Profession, 2001 Report on equity release mechanisms Basel Committee on Banking Supervision, 2001 Risk management practices and regulatory capital: cross-sectoral comparison Bank for International Settlements Porteous B, 2001 The Basel capital Accord The Actuary magazine, September Rule D, 2001 Risk transfer between banks, insurance companies and capital markets: an overview Financial Stability Review, Bank of England, December References
