Capital quality and the illiquidity premium - The current debate on the illiquidity premium

[this page | pdf | references | back links | custom searches]

Article by Malcolm Kemp – 27 May 2010

Copyright (c) Malcolm Kemp 2010


Return to Abstract and Contents

Next Section


1.            The current debate on the illiquidity premium


The background to the meteoric rise in the perceived importance (to actuaries) of the illiquidity premium may be summarised as follows:


-          Some insurers have material exposures to illiquid liabilities, particularly in annuity books. Annuity policies are typically considered to be highly illiquid and relatively long-term, because they cannot normally be cancelled. Many such insurers have ‘matched’ this illiquidity in their liabilities by investing in correspondingly illiquid assets, e.g. illiquid corporate bonds generating similar cash flow payment profiles to those required to meet the annuity liabilities as they fall due.


-          These insurers have often also adjusted downwards the values they have placed on these liabilities, to reflect the perceived illiquidity yield premium available from investing in these assets. When commentators refer to the ‘illiquidity premium’ they are generally referring to the part of the overall yield spread (versus liquid highly rated government bonds) that is not perceived by the commentator to be merely a ‘fair’ compensation for the extra credit (i.e. default) risk carried by corporate bonds (or for the uncertainty in the future magnitude of this risk) but reflects differences in the current (and future) liquidity exhibited by these instruments.


-          Insurers have adjusted valuations of illiquid liabilities in this type of fashion not just in solvency valuations but also when pricing annuity policies, including bulk annuity purchases that pension funds might wish to make.


-          Even small differences in annualised yield spreads can compound to large differences in liability valuations. At the height of the recent credit crisis some commentators estimated the magnitude of the liquidity premium as in excess of 2% pa. This might equate to, say, a c. 20 – 30% change in value for an illiquid liability with a c. 10 to 15 year average duration. The impact of the illiquidity premium can be substantial both for sellers of such policies (insurers) and for buyers (pension funds as well as individuals).


-          Although the recent financial crisis is most commonly referred to as a credit crisis, it could arguably be better described as a liquidity crisis. The banks that failed were disproportionately ones that relied on the wholesale markets for their funding. It was when these funding sources dried up that they ran into problems, because they were then unable to source the liquidity they needed to continue as going concerns.


-          As a result, regulators have become much more focused on liquidity risk. This applies not just to banking regulators. Last autumn, EU insurance regulators (CEIOPS) preparing for Solvency II proposed that for regulatory capital purposes EU insurers should value annuity liabilities by discounting at a ‘risk-free’ discount rate derived from highly rated government bonds, i.e. without including an illiquidity yield premium.


-          Predictably, some insurers with large annuity books lobbied hard to have this stance reversed, claiming that it might require an extra £50bn of capital and lead to reductions in annuities to pensioners by between 10 and 20 per cent according to e.g. The Times (2009). The Association of British Insurers went on record as arguing that the proposed use of a yield curve derived from ‘AAA’ Government bonds would “cause massive disruption in the capital markets and a huge artificial inflation in the value of liabilities. A ‘liquidity premium’ should be recognised and taken into account in the risk free rate, in particular for long term non redeemable liabilities”. CEIOPS’s original stance has also been referred to by some as ‘reckless prudence’.


-          CEIOPS responded by commissioning a Task Force on the Illiquidity Premium, which reported in March 2010, see CEIOPS (2010). Its members were split about whether there was a sound theoretical basis for incorporating an illiquidity premium in the discount rates used to value illiquid long-term liabilities, and, to the extent that there was one, on how practical it might be to estimate its size reliably. This lack of consensus is perhaps not surprising. Kemp (2009) highlights that there are several different possible economic interpretations for liquidity risk. Some support the use of an illiquidity premium in this way whilst others do not. However, the Task Force did propose a pragmatic way forward, to the extent that it was felt appropriate to include an illiquidity premium at all, which involved deeming some proportion of the total yield spread available on corporate versus government bonds to correspond to an illiquidity premium.


-          The Task Force report was included in the Level 2 advice on Solvency II provided by CEIOPS to the EU Commission. The EU Commission subsequently indicated in the draft technical specifications for the Solvency II Quantitative Impact Study 5 that it was minded to allow inclusion of an illiquidity premium. The draft technical specifications included two sets of yield curves applicable as at 31 December 2009, one deemed to exclude the illiquidity premium and one deemed to include it. For liabilities with average durations of c. 10 to 15 years, the difference between them corresponded to a c. 5 to 10% difference in value, depending on the currency in question.


Contents | Prev | Next

Desktop view | Switch to Mobile