Site map

Help us make USS better for everyone; join our research community

22 September 2017

Investment challenges: the facts

Much has been written about USS’s investment performance recently, with some suggesting it is the reason the scheme is facing some difficult decisions – so I want to set out some facts to address this.

We have, in fact, outperformed the scheme’s strategic allocation benchmark over all multi-year periods, as illustrated in the chart below. Performance has been more uneven compared with a liability benchmark comprised mainly of long-dated index-linked gilts. The 10 year period straddles both the global financial crisis and the further sharp fall in gilt yields following the Brexit referendum – only over this period have the scheme’s assets underperformed the gilts liability proxy.

DB fund vs Benchmark - to 31 July 2017

The strategic allocation benchmark represents a notional ‘reference portfolio’ of largely passive asset classes (equities, UK property, government and non-government bonds) chosen to provide the returns required by the scheme within the risk budget agreed with sponsoring employers.

Taking on diversified investment risks is expected to generate higher long-term returns than, say, a gilts only portfolio. And we know that gilt returns will be very low over the coming decades with inflation-linked gilt yields now around negative 1.5%. The risk budget aims to ensure that, even in a possible adverse outcome for the scheme’s investment portfolio, the participating employers can reasonably afford to support the pensions members have earned to date and also that the pensions promised for future service can also be provided with a high degree of confidence. This is a key point.

We challenge our investment team to outperform the reference portfolio by more broadly diversifying the portfolio across public and private markets and by actively managing the scheme’s investments: about 70% of our assets have been managed in-house over the periods shown above. According to independent analysis, our investment costs are £34m a year lower than our large global pension fund peers.

“Our investment costs are £34m a year lower than our global peers”

The second benchmark shown here is the returns provided by a basket of gilts that closely match the scheme’s liabilities.

In this year’s Annual Report & Accounts, we reported a 2% underperformance against the reference portfolio benchmark and fractionally behind the liability proxy for the year to 31 March 2017. Had our one-year performance had been measured at 31 July 2017 we would have been a few hundredths of a percent adrift the reference portfolio benchmark and several percent ahead of the liabilities proxy. That shows the limits of 12-month reporting periods and helps to explain our focus on rolling five-year periods.

As a long-term pension provider and investor, we want our managers to make good decisions for the long term and not to reject opportunities that might not play out over a year. Our performance goals are measured over rolling five year periods and we have consistently beaten them after costs.

In the five years to March 31 2017, our diverse investment portfolio – circa half in equity-like investments, one-third in fixed income and the balance in infrastructure, property, private debt, commodities and absolute return – generated an average return of over 12% per annum. This added £1.1bn of extra value against the strategic allocation benchmark, after investment-related costs.

Over the five years to July 31 2017, the scheme’s assets generated c.£10bn more in returns than the gilts liability proxy (mainly long-dated index-linked gilts).

"£10bn more in returns than gilts over the last five years"

Looking at our 10-year performance in the chart given at the top of the page, we are below gilts as a consequence of the scheme’s substantial exposure to equities (c.80%) at the time of the global financial crisis in 2008.

Since then we have progressively diversified our investments to the portfolio we operate today, seeking to maximize returns within the risk budget agreed with sponsoring employers.

Though investment performance has been averaging in double digits over the past several years, this cannot be projected into expectations for future returns.

Indeed, the principal challenge the trustee and its stakeholders face in addressing the 2017 valuation is a change in expected future investment returns since 2014.

In that time, changes to the financial market valuations have significantly reduced expected investment returns.

The contributions backing pension accruals in previous years have benefitted from the strong investment returns. But, new contributions into the scheme are expected to grow at a slower rate. This means that new pension accruals are likely to require higher contributions than in previous periods.

In USS, our proposals see the cost of supporting future accrual of defined pension benefits increasing by 35%.

"Since 2014, lower expectations for future investment returns have increased the cost of accruing future defined pension benefits by 35%"

The clearest indicator of this fall in expected investment returns can be seen from the fall in index-linked gilt yields. These are loans to the UK Government that pay a fixed return linked to changes in inflation measured by the RPI. As such they play a key role in the liability-hedging needs of many pension schemes.

Yields on 20 year inflation-linked gilts have declined from already historically low levels in 2014 of around zero percent to minus 1.75% at end March 2017 and close to minus 1.5% today. These investments guarantee a return well below inflation.

Inflation chart

Similar yield declines have been seen in nominal gilts and greater yield falls have occurred in corporate bonds.

Investors seeking above-inflation returns have been forced to invest in other markets. That increases competition for other assets, which has driven up prices now across the board.

This, in turn, reduces the returns the trustee believes it can expect in future – in the trustee’s view, our expectations for future returns reduced by around 1.35% per annum between our March 2014 and 2017 valuation dates.

"USS’s expected future investment returns have reduced by around 1.35% per annum compared to 2014"

Prospects for future returns, even for the long-term, are necessarily subject to considerable uncertainty. Our own forecasts are built up from fundamental building blocks for returns (such expectations for growth, inflation and the valuation of future cashflows). The results are not extreme when compared across a range of forecasters, including the scheme’s independent advisor, Mercer, and other asset managers.

Lower expected returns from investing new pension contributions means that a higher level of contributions is needed to provide the same level of benefits for future accruals. Compared to 2014’s forecasts, the maths works out at a 35% increase in the cost of accruing the current level of defined pension benefits in 2017. This would require the current level of contributions to increase by between 6% and 7%.

The expected cost of future pension benefits has increased for everyone, not just USS: as an example, a benchmark annuity at retirement now costs about a quarter more than in March 2014.

It is vital that employers understand the level of risk, particularly investment risk, underlying proposals for the scheme going forward. Any prediction is unlikely to be precisely correct and so, if our predictions turn out to have been over-optimistic, we need to make sure that the impact on employers would sit within their ability and willingness to pay extra contributions in future to make good on promises being made today.

That is why it is vital to consult employers before the price of new pension accruals can be finalised.

Once we have the employers’ views through the consultation with Universities UK, we can confirm the price for future pension accrual so that the stakeholders can then consider what the future benefit and costs sharing arrangements should be.

USS, as trustee, sets the price for supporting pension promises independently; employer and union representatives, through our constitutional Joint Negotiating Committee, then decide on future benefit structure and cost-sharing arrangements.

You can find out more about the valuation process on a dedicated section of our website.