De-risking has been a pervasive mantra in the world of UK pension professionals for more than 20 years. Yet it is arguably the most treacherous word in the investment lexicon.
Consider, first, what it means in practice. For people in defined contribution pension funds where the size of the pot reflects investment returns and contributions, de-risking implies shifting the balance of the pension as retirement approaches from risky assets such as equities to supposedly safer assets such as gilts.
In defined benefit schemes, where pensions are related to pay and length of service, it means adopting so-called “liability-driven investment”. This involves holding assets, mainly nominal or index-linked gilts, that produce cash flows timed to match pension outgoings, thereby minimising interest rate and inflation risk.
There follows a so-called endgame in which risk is transferred to an insurer either via a buy-in or bulk annuity purchase, or a buyout, where the scheme transfers all its liabilities to the insurer.
To be clear, the actuarial consultants who promoted the de-risking idea were not entirely misguided. Seeking to minimise risk as people approach retirement makes eminent sense. To be exposed to a collapsing market and thus a shrunken pension pot at the point of retirement is a dire predicament. Reducing exposure to market volatility therefore makes absolute sense.
In defined benefit schemes that are becoming mature, with a high proportion of cash flow being paid out in pensions, liquidity afforded by supposedly safe assets — that is, the ability to realise investments without moving the market against you — becomes more important. Liability matching also reduces a fund’s vulnerability to deficits. The difficulty arises with the definition of safe assets. In actuarial lore and economists’ conventional wisdom the safest of safe assets are government bonds.
These are assumed to be less volatile than equities and to provide valuable diversification benefits — what Harry Markowitz, the great pioneer of portfolio theory, called the only free lunch in investing because diversification allows investors to reduce risk without sacrificing return.
Yet from the financial crisis of 2007-09 to the upturn in interest rates prompted by the recent resurgence of inflation, government bond markets were hostage to an unprecedented bubble.
In the low growth post-crisis world a secular decline in real interest rates was intensified by the Asian savings glut and more specifically China’s huge surplus of savings over investment. Central banks’ policy rates turned negative, notably at the European Central Bank and the central banks of Denmark, Japan, Sweden and Switzerland.
The idea was that by forcing banks to pay a penalty for parking their excess cash at the central bank, commercial banks would be encouraged to lend out those funds instead, so countering weak growth after the financial crisis. And across the developed world yields on much government debt turned negative so that investors were paying for the privilege of lending to governments while borrowers were paid to borrow money.
At the worst point in 2021 the yield on 10-year index-linked gilts was a minus number of more than 3 per cent — a swingeing penalty for investing in what had been regarded as the ultimately safe asset. The decline in yields and rise in prices in the UK was conspicuously worse than in the US. This, as my colleague Martin Wolf has argued, is because of greater regulatory pressure on UK pension funds to match their liabilities with gilts — a pressure that was welcomed by sponsoring employers.
As Iain Clacher of Leeds University and Con Keating of consultants Brighton Rock recorded in evidence to the work and pensions committee of the House of Commons last year, finance directors had been traumatised by two decades in which ultra-low interest rates had caused pension liabilities to balloon and companies with pension deficits to make big calls on their cash reserves. When rates are low, companies must typically set aside more money to meet their liabilities.
The resulting weight of money pouring into gilts distorted the market. Meantime, extremely low nominal interest rates on retail bank deposits had the perverse effect on retail investors of causing what economists have dubbed a “savings reversal”. When rates were approaching their all-time lows household savings actually increased. As in the 1970s, when real interest rates after adjusting for inflation were negative, people felt obliged to save more to achieve retirement income targets.
Note, in passing, that a malign consequence of this supposed de-risking was a pension fund exodus from equities. There lies an important part of the explanation for the relative undervaluation of the UK equity market against the US and some continental European markets. The decline in risk appetite in a vital part of the UK financial system inevitably has a deleterious impact on investment in the real economy.
The worst of it is that the notion of bonds as safe assets is pure myth. Elroy Dimson, Paul Marsh and Mike Staunton in the latest UBS Global Investment Returns Yearbook point out that between October 1946 and December 1974 UK government bonds lost 74 per cent of their value. In fact, UK bond investors lost half their real wealth during the inflationary period between 1972 and 1974.
Then in the bond bear market that began in May 2020 real bond returns fell by 52 per cent. Their conclusion: sovereign bonds are not “safe” assets and their real value can be destroyed by inflation. They add that historically such bond market retreats have been larger or longer than for equities. Nothing could more devastatingly highlight how mainstream economists, actuaries and regulators have lost the risk and return plot.
This has been terrible news for older members of defined contribution pension schemes. The great majority take a default option whereby they leave it to the trustees to do the job of asset allocation. In a typical scheme the default option, often referred to as “lifestyling”, involves a progressive shift, starting up to 15 years from retirement, out of equities and into bonds, credit and a modicum of cash.
What that meant in the period between 2008 and 2021 was that the default option pushed people into the most overvalued bond market in history. So they now nurse big losses on supposedly safe investments after the recent normalisation of interest rates, as the FT has reported. That is because rising rates depress bond prices. Among the more lethal investments, when inflation became rampant, were index-linked gilts, which plunged further than nominal gilts.
As I have written before in FT Money, index-linked gilts only provide inflation proofing if held to maturity. Their prices are driven by relative real yields, not inflation. So, if nominal gilt yields rise, index-linked gilt yields have to rise to offer a competitive return, which destroys capital value regardless of what is happening to the general price level, since rising yields mean falling prices.
Also bad news for defined contribution scheme members is that since the era of ultra-low interest rates came to an end bonds and equities are no longer negatively correlated. That is, they have tended to move in lockstep, so bonds have failed to provide a risk-reducing hedge against equities.
For defined benefit scheme members the perils of de-risking are rather different. Adopting liability driven investment is a low risk/low return strategy. The return-seeking proportion of the portfolio shrinks, thereby restricting a pension fund’s ability to pay discretionary increases to pensioners, reduce company contributions and recycle surpluses to threadbare defined contribution schemes.
If a scheme is passed to an insurer, the chances of a pensioner getting a discretionary increase are very much reduced. Yet the case for transferring risk to an insurer is less compelling now that 80 per cent of defined benefit pension schemes are in surplus. In effect, such transactions swap the company sponsor’s guarantee for the insurer’s guarantee. They also entail substituting the backing of the well-funded Pension Protection Fund for that of the industry-funded but more generous Financial Services Compensation Scheme.
The insurer’s promise is supported by collateral in the shape of the assets that have been transferred by the pension fund; likewise by regulatory capital requirements. Yet the robustness of this capital regime is open to question since insurers may transfer much of the risk to reinsurers in offshore tax havens where solvency requirements can be less stringent.
This is a very opaque market. What is clear is that it offers ample room for insurers to economise on capital, or look for ways to meet legal requirements more cheaply, which cannot be good for pensioners. Buyouts and buy-ins are also costly and have been hugely profitable for a narrowly concentrated group of insurers operating in an exceptionally capital-intensive business. This means high barriers to entry for would-be competitors.
William McGrath of C-suite Pension Strategies argues that the benefit to pension scheme members of having a life insurer paying their pension as opposed to the corporate sponsor is modest and falling. There is real upside, he says, in a corporate sponsor running its scheme: pensions are paid out of the fund’s current income so that surpluses can be used to improve benefits and reduce or eliminate its pension costs, including those for current defined contribution employees.
The news is not all bad for defined contribution folk. The Labour government’s proposal for expanding collective defined contribution schemes to allow multiple employers to take part in larger pooled funds would increase scheme members’ exposure to higher return investments for longer than is currently the case. This has the potential to enhance retirement incomes, and in theory would be good for Britain’s capital markets. Meanwhile, for those approaching the point of de-risking, gilts are no longer wildly overvalued.
That said, there are good grounds for questioning asset allocation in the de-risking phase in many default portfolios, whether those portfolios are invested in different fund choices shaped for scheme members to move either to an annuity, drawing down from an existing fund, or turning the pension pot into cash. Given the historic evidence that bond market retreats have been larger or longer than for equities; that gilts do not provide fail-safe diversification; and bonds underperform equities over the long run, there is a case for suggesting that de-risking funds are overexposed to gilts.
And while it sounds counterintuitive, many are underexposed to cash as well as equities. Cash, of course, underperforms most asset categories over time. Under, say, 2 per cent inflation — the Bank of England’s target — a 10-year investment in cash will reduce an investor’s capital in real terms by 18 per cent over the period. But as value destruction goes, that is much less savage than the bear markets that afflict gilts and equities. Moreover, cash offers genuine de-risking because it is a true diversifier against other major asset classes.
But even if we accept that there is scope for pension professionals to work towards more effective de-risking, there is no escaping an important underlying reality. Defined contribution schemes, collective or otherwise, are not really pension schemes. As Clacher and Keating point out, they are no more than tax privileged savings funds — ultimately, a second best answer to the challenge of obtaining a secure income in retirement. We forget that at our peril.
