Ageing and pensions

Pensions Outlook 2014Pablo Antolín-Nicolás, OECD Directorate for Financial and Enterprise Affairs

Pension systems of all types are facing crucial and far-reaching challenges because of demographic trends, the continuing impacts of the economic crisis, and the environment of low growth, low returns and low yields. As a result, meeting pension commitments and having adequate pensions could become quite a challenge. The OECD Pensions Outlook 2014 discusses ways in which countries are addressing all these challenges, including the demographic challenge.

Ageing is the result of lower fertility rates and, especially, higher life expectancy, which results in an increase in the average age of the population. The impact of population ageing on pensions can be separated into the “baby boom”, a temporary factor, and improvements in mortality and life expectancy, a more permanent factor. Once the “baby boom” generations pass away their impact is gone. However, improvements in mortality and life expectancy are here to stay.

Most people including myself would consider living longer a good thing, so let’s prepare for it. An increase in life expectancy while keeping the number of years saving for retirement constant means that the ratio of years contributing to finance retirement to years in retirement will fall. Therefore, the same amount of savings will have to finance more years, and unless someone assumes that cost (e.g. governments through defined benefit public pensions or employers through defined benefit funded pensions) people will have lower annual pensions, although the sum of all pension payments throughout retirement will remain constant.

In the case that governments and/or employers assume the extra cost of more years in retirement relative to years saving for retirement, they may suffer problems of solvency or fiscal sustainability. There will be a problem of adequacy when people accumulate assets to finance retirement in defined contribution pension plans and they bear the risk of living longer. If they buy a life annuity the risk is transferred to the annuity provider.

Therefore, as the Pensions Outlook shows, population ageing and, in particular, the continued improvements in mortality and life expectancy, create problems of adequacy in defined contribution pensions, solvency in defined benefit funded pensions, and financial sustainability problems in PAYG-financed public pensions.

The Outlook argues that contributing more and for longer, especially by postponing retirement as life expectancy increases, is the best approach to face these challenges. The way to address the problems posed by improvements in life expectancy is to maintain the ratio of years saving for retirement to years in retirement constant, increasing contribution periods as life expectancy increases; or to increase overall contributions. So what are countries doing?

Many countries have responded to population ageing by increasing the statutory age of retirement. Some have linked retirement age to life expectancy.

The fairness of this solution, however, can be questioned when we look beyond the average. Gains in life expectancy have not necessarily been distributed equally across society. A skilled executive, for example, can expect to enjoy nearly four additional years in retirement compared to a manual labourer; this assuming that “retirement” begins at age 65. The inequality becomes more apparent when considering the period before retirement. Not only can the manual labourer expect to receive his pension for fewer years, but he can also expect to have made contributions to the system from an earlier age, as the highly skilled worker likely spent a number of years in higher education and began working later. Given the same retirement age, the unskilled labourer pays relatively more into the system to receive his pension for a shorter amount of time.

Mechanically linking retirement age to increases in life expectancy across the board may therefore be regressive. Life expectancy, time of entry in the labour market and improvements in life expectancy are not homogenous across the population, they vary across different socio-economic groups (e.g., low skill, low income groups). Hence, the best approach may be to link the number of years contributing to life expectancy. Unfortunately, the data needed for this is not available across all countries and the application across different socio-economic groups may be far from straightforward.

To compound the problem, future improvements in mortality and life expectancy are uncertain. Gains may continue as in the past, they may accelerate or decelerate. Improvements vary across different population sub-groups; they may converge or diverge further.

In defined contribution pension plans individuals bear the risk of outliving their resources in old age. They can insure themselves against this longevity risk by transferring the longevity risk to annuity providers, e.g., life insurers, as we said above. The OECD Roadmap for the Good Design of DC Pension Plans recommends default partial annuitization to provide protection from longevity risk. In defined benefit pension plans (e.g., PAYG financed public pensions or funded pension) the government, pension funds or employers who assumes the longevity risk.

Pension funds and annuity providers need financial instruments to mitigate the longevity risk. The OECD work on Mortality Assumptions and Longevity Risk examines the longevity risk that pension funds and annuity providers may be exposed to by looking at the (regulatory) mortality tables used to provision for future improvements in mortality and life expectancy and, in this way, be able to fulfil their pension promises. This OECD work also discusses several approaches for pension funds and annuity providers to manage longevity risk.

The first step is to recognise the existence of longevity risk and provision accordingly. For this, regulators and policy makers should ensure that pension funds and annuity providers use regularly updated mortality tables that incorporate future improvements in mortality and life expectancy. In addition, these mortality tables should be based on the mortality experience of the relevant population.

The regulatory framework could also help to ensure that capital markets offer additional capacity to mitigate longevity risk, for example recognising the reduction in risk exposure of using index-based financial instruments to hedge longevity risk, and by publishing a longevity index to serve as a benchmark for the pricing and risk assessment of longevity hedges, improving the standardisation, transparency and liquidity of these markets.

The issuance of a longevity indexed bond could be considered, though with care. While it may be helpful in kick-starting the market for longevity hedging instruments by providing standardisation, a benchmark for pricing and liquidity, it would also significantly increase the exposure of the government to longevity risk, while many governments already have significant exposure on their balance sheets

Demand for protection against longevity risk will only increase as individuals are expected to live longer, and the sustainability of pension funds and annuities providing this protection for individuals has to be ensured. Sufficient provisioning for longevity is essential to guarantee that future payments will be met, and the ability for providers to manage and mitigate this risk will allow them to continue offering protection in the future.

Useful links

OECD Pensions Outlook 2014

OECD Roadmap for the Good Design of DC Pension Plans

Mortality Assumptions and Longevity Risk


Hey you, stop wurfing and read about the 26 billion buck haircut!

Haircuts can be dangerous

OK, wurfing (surfing the web at work) didn’t make it into the new edition of the Oxford Dictionary of English published today, but toxic debt and quantitative easing did.

Speaking of which, haircut was already there. Eh? Haircut, you know: “US informal: a reduction in the stated value of an asset”.

That’s one of the terms you need to understand to follow the argument in a new OECD Working Paper on the stress tests 84 European banks passed so brilliantly in July.

Adrian Blundell-Wignall, Special Adviser to the OECD Secretary-General on financial markets, and his colleague Patrick Slovik point out that the tests only considered “trading book” exposures to sovereign debt, while over 80% of exposure is on the “banking book”.

The trading book consists of the securities a bank buys and sells regularly, even daily, while the banking book contains the products the bank would normally hold on to until they matured, including the bonds used to finance sovereign debt.

For the trading book, the haircut is around €26 bn in the stress tests. No haircut was applied to the banking book, on the grounds a default would be virtually impossible over the two-year period considered. The tests also assumed there would be no bank failures.

Blundell-Wignall and Slovik argue that these two assumptions help to explain why despite the encouraging test results (only 7 banks failed) equity markets are still performing poorly, bond spreads remain high, and banks are still reluctant to lend.

If a bank fails, it cannot hold on to the longer-term assets on its banking book, which would have to be sold for whatever they are worth on the day, even at a loss, and in fact there would be no difference between the trading and banking books. In other words, shifts in the market value of sovereign debt do matter, unless you assume that the stress-tested banks never fail. 

That’s a brave, or foolish, assumption in light of what we’ve seen since the crisis broke, but the assumption of no sovereign default over the next two years seems reasonable, given the €720 bn European Financial Stability Facility (EFSF) agreed earlier this year.

The EFSF could more than cover all the funding needs of the most exposed countries, even in the highly unlikely case that no securities could be sold on the open market.

So why are ratings agencies like Moody’s worried about the sovereign debt of even the US, Germany, France and the UK, countries they consider “well-positioned at AAA” in their latest figures?

They’re not worried about the next couple of years, but many analysts foresee problems in reforming labour and pension markets to ensure sustainable growth before the stimulus packages run out. In the medium-term, budget restraints will make these reforms more difficult.

In the longer term, Moody’s is afraid of a situation where states delay pension reform for political reasons, leading to a downward spiral as they try to borrow more to finance deficits, while at the same time, conflict between younger and older generations destroys the social cohesion needed to stabilise debt.

Countries in this situation would lose their triple-A rating.  

Useful links

OECD Insights: From Crisis to Recovery

OECD work on financial markets

This OECD paper warns that retirement income may become a “lottery” unless default strategies are carefully designed