A new report from Towers Watson, which analyses the pension disclosures of FTSE350 companies with financial years ending 31 December 2011, has found that:

  • On average, companies expect the gap between RPI and CPI inflation to be 0.9% per year. This means they expect recent legislative changes to have a significantly smaller impact on the pension increases that will be awarded in future than the Government has assumed.
  • If two companies have similar schemes, the liabilities disclosed at 31 December 2011 could vary by about 5 per cent depending on how they selected a discount rate to calculate the present value of future payments. In 2010, this technical decision would have made little difference.
  • Forthcoming changes to accounting standards may have knocked around £4 billion off reported corporate profits if they had been in force in 2011.
  • The average life expectancy assumed for pensioners aged 65 in 2011 was 87.3 years for men and 89.3 for women. Higher life expectancy (an extra 2.0 years for men and 1.8 years for women) is assumed for scheme members turning 65 in 2031, reflecting an assumption that life expectancy will continue to improve significantly for many years to come.

The consultancy’s modelling also highlights how volatile pension numbers calculated for accounting purposes can be. The total pension deficit for all FTSE350 companies is estimated to have been £49bn both at the end of 2010 and at the end of 2011. In between, however, it rose to £81bn and fell to just £4bn. And this volatility has continued: deficits have nearly doubled since the end of 2011, to £90bn in May 2012.

Inflation assumptions
On average, companies assumed that CPI inflation would be 0.9% a year lower than RPI inflation over the long term. Assumptions about the size of this gap varied widely from 0.5% to 1.2%. The most common assumption was 1.0%, which was adopted by 37% of companies. In July 2011, the Department for Work and Pensions assumed a gap of 1.2% for its regulatory impact assessment; this reflected an estimate from the Office for Budget Responsibility, which the OBR has since revised to 1.4%.

Neil Crombie, a senior consultant at Towers Watson, said: “Market prices provide few signposts to the future gap between RPI and CPI inflation. Instead, companies have to think about whether the goods in one basket will increase in price more quickly than those in another, and how much impact using different formulae to calculate the two measures might have. This is also a moving target: the statisticians are looking at how both measures of inflation are calculated and the Government has not said whether a parallel CPI which includes owner occupiers’ housing costs would be used to set pension increases. There is a lot of uncertainty but all companies in this report and most others say their best estimate is that the gap between CPI and RPI will be smaller than suggested by official projections.”

Two-thirds of companies assume that RPI inflation itself will average either 3.0% or 3.1% a year, but there is a wide range with assumptions varying from 2.5% to 3.5%. Neil Crombie said: “Gilt prices imply that markets expect inflation to be higher further into the future, so companies whose pension payments are due sooner will make lower inflation assumptions. Strong demand from pension funds and insurers for inflation-linked assets may also mean that gilt prices overstate expectations of future inflation and not all companies will have the same view of how big a factor this is.”

Discount rates and pension liabilities
Companies disclose the present value of pension payments that will fall due over several decades. The lower the interest rate used to discount future payments, the bigger pension liabilities look.

Neil Crombie said: “Accounting standards require this discount rate to reflect the interest rates on high quality corporate bonds of equivalent duration to the company’s pension liabilities. In 2011, how this instruction was interpreted could make a significant difference to the published liabilities – around £50 million for a £1 billion scheme.”

Some companies simply take this interest rate from an index of long-dated AA corporate bonds. Others favour more precision and weight the yields on different bonds to reflect when their pension payments fall due.

Neil Crombie said: “Even with an index of long-dated bonds, the average duration will be shorter than a typical company’s pension liabilities. For December 2010 accounts, that did not matter much because bond yields flattened out once maturities reached about 14 years. However, in 2011, the highest yields were over longer periods so the more precise approach allowed smaller liabilities to be disclosed to investors. Some companies may want to review what approach they use going forward. However, auditors will not let them pick and choose a methodology from year to year, so companies cannot use whichever happens to produce the smallest number at the time.”

Changes to accounting standards and corporate profits
Revised accounting standards which apply for financial years beginning on or after 1 January 2013 will change the way that expected returns on pension scheme assets feed into companies’ reported profits. At the moment, income statements include an expected return on the assets that pension schemes hold. In future, this will instead reflect returns on the corporate bonds used to calculate pension liabilities, which will usually be lower.

Neil Crombie said: “On average, companies banked a 5.5% expected return on pension assets in their income statements. Using the average discount rate instead would have reduced this to 4.8%. Applying this change across the board would have reduced reported profits by around £4bn in 2011, though the impact will be much bigger for some companies than others. New accounting rules mean investing in riskier assets will no longer provide an immediate boost to recorded profits. However, for most companies the main motivation for doing this remains the hope that it can reduce the cash contributions needed to pay off deficits”

Life expectancy
Company accounts assume average life expectancies for pension scheme members that are higher than the life expectancies for the population as a whole used in national population projections – for example, an extra 1.2 years for a man aged 65 in 2011.

Neil Crombie said: “There is evidence that pension scheme members live longer than non-members. This has long been reflected in the assumptions that companies make when budgeting for pension liabilities. More companies are now drilling deeper into the data to see how factors such as where their members live can help to explain how long they might live.”