Alex Waite
The gilt yield is generally accepted as the benchmark for risk-free returns. It’s what Government debt over various time-frames will reward you with and it is accepted as a benchmark because the chance of the UK Government defaulting on the debt is negligible.
It is also accepted as a benchmark because it is not subject to the vagaries of supply and demand because its major stakeholders – the Government and pension schemes – are in equilibrium – supply meets demand.
This equilibrium is being challenged according to LCP’s Alex Waite, who sees a potential for a large scale sell-off of gilts by pension schemes as they prepare for and enter into deals with insurers which will see pension liabilities switch from being matched by gilts to being matched by corporate bonds.
“The anticipated sell-off would naturally drive down gilt prices and hence elevate yields, potentially accounting for the observed circa 0.5 per cent pa discrepancy relative to other markets.”
High yields = high discount rates = happy days on solvency. But is the anticipated sell-off as likely to happen as the market imagines? If it isn’t , then current solvency may be measured against a benchmark which may be a bit skew-whiff.
This may all sound a little academic and pension-nerdy, but there’s a bigger problem associated with gilt pricing than just vagaries in pension fund valuations.
One of the Chancellor’s golden rules, mentioned in his Mansion House Reform speech was that the reforms supported the integrity of the gilt-market.
It is hard to see how a wholesale sell off of gilts as DB schemes move to buy-out could do anything but destabilise the gilt market as Alex Waite suggests,
“For investors and pension schemes with a strategy to hold their gilt portfolio to maturity, the gilt yield arguably remains a reliable indicator of the ‘risk-free’ rate that they can expect to receive.
However, the billion-pound question remains: Will pension schemes stay the course to redemption? If market sentiment indicates a shift, the gilt yield’s reputation in the pensions industry as the de facto benchmark for the long-term ‘risk-free’ rate could well be due for a reassessment
“As we step into the new year, it’s clear that, while the UK government’s commitment to honouring gilt coupons is not in question, the market participants’ commitment to holding these instruments might well be.”
LCP are not just respected pension actuaries but seen as pivotal to the “de-risking” of DB pensions through buy-outs and buy-ins. Their questioning the reputation of the gilt-rate might have been considered heresy until recently, right now it looks like common sense.
Conventional Gilts will remain the benchmark risk free interest rate for different durations because they are risk free and always widely available. Both more so than swaps. They are a market set rate. If a lot of schemes sell gilts, then the market price will fall and the market interest rate will rise until the market reaches a point where fresh buyers turn up to mop up the sellers stock.
Not a free market rate, taking into account all risks and rewards – but a rate influenced by regulatory interference. Such intervention always distorts, and ultimately burst the bubble.
Its not the job of the Pension Scheme trustees to ensure the integrity of the gilt-market, and their advisers should not be puppets of DMO / Govt monetary policy.
The blog response of xxx below are in point. It s a completely asinine and intellectual myopic approach to value all other assets off of a risk free gilt rate, and one that has led to the wrong valuation of liabilities and a huge mis-allocation of free capital. Its an actuarial simplification to help them explain the black-box (and to sell services), but it does not cover all of the risk, and certainly not the systemic impact and risk from the DB universe’s excess holdings of single issuer risk. Its a poor indictment of the Actuarial profession to ignore the risk associated with single issue debt.
UK Govt debt is now AA- rated. That is more than ‘negligible’, and similar to the rating Lehmans expected back in the day…. If we are to use the Gilt+ model, then apply the risk characteristics and loss potential of the relevant market rating.
“it is not subject to the vagaries of supply and demand because its major stakeholders – the Government and pension schemes – are in equilibrium – supply meets demand.”
This is a tautology, of course supply (from the government) equals demand (from pension funds, and other investors). The question is at what price, and what is the variability in the supply, demand, prices and yield.
Simply plotting the proportion of DB scheme’s assets held not in gilts makes it very obvious that the demand for gilts from DB schemes is likely to fall in future. They do not have many more assets to divest from productive investments (equities), into unproductive investments (gilts). Hence, absent some other source of demand for gilts, the yield is likely to rise.
Thus supply=demand is absolutely not an argument in favour of using gilts as the basis for determining the likely returns on all other assets (i.e. gilts+ approach).
To see how asinine the gilts+ approach is consider the implications – it presumes that the yield on gilts affects the returns on all other possible assets (e.g. see the ludicrous modelling from the USS). However, is it plausible that the yield on gilts will affect the returns on Japanese or US equities, bonds or real estate? This is completely implausible and not supported by any empirical or theoretical evidence.
This approach leads to wildly volatile estimates of liabilities. But worse than this have been the consequences for DB schemes’ investment strategies and the terrible reduction in diversification across the entire DB sector. For example, see the horrific losses experienced by the Bank of England pension scheme in the last year, which had divested entirely into gilts. How can something be “low-risk”, if it leads to 34% loss in assets in a single year? That is 7pp higher than the losses USS experienced during the GFC.
Sigh. How often does this need to be explained? Gilts are risk free in the sense that the UK Government has always paid investors what was due when it was due. They are not risk free if you want the money when it isn’t due.
Yes, understood. Sovereign states never default, until they default.
Has always not same as will always.
Or the more ‘acceptable’ level of soft-default, re-categorise / disconnect the inflation uplift from the actual cost of living experienced by pensioners.
They had the same assumption for local authorities given their tax raising powers too. Local authorities can’t go bankrupt, until they do.
I always thought Wilkie models started from inflation (RPI to begin with, and later CPI), not gilt yields. In fact, when asset classes were introduced later in the model
it was the yield on undated Consols rather than fixed or index-linked gilts which was used at first.
In a 2019 interview, Professor Wilkie said “all of my research was directed toward the practical problems of life offices, getting new usable mortality tables, getting usable fixed interest indices, and reserving allowing for the stochastic nature of investments”.
But I believe many scheme actuaries have been using “modified Wilkie models” for years to advise on DB pensions funding.
Gilts are risk-free in the sense that the payments of interest and repayment of principal are fixed and guaranteed by government. It does not follow that the gilt yield is a risk-free rate because the yield depends on gilt prices on the market – and it varies subject to the market forces of supply and demand. Calling the gilt yield “risk free” is therefore an undergraduate error.
Why is such an obvious error not corrected – especially when it has such serious consequences? In the case of the USS for example, where the whole valuation is based on discount rates expressed relative to gilt yields as a “benchmark”, it has led to wild instability. Changes to discount rates (ie gilt yields) since 2020 have contributed £44 BILLION towards the transformation of the valuation from deficit to surplus.
QE has distorted gilt yields so that not only are they not risk-free (that is a very narrowly defined construct for gilts in the sense the Government can print money to repay in theory) but they are no longer the lowest risk asset necessarily either, in terms of market pricing. QE deliberately rigged gilt yields (it artificially drove down gilt yields by buying huge swathes of the market with newly created money). Even worse than this, because pension regulators and industry advisers insisted that the liabilities must be measured by gilt discount rates, and because annuities were priced on a gilts-related basis, the impact of QE led to pension schemes competing with the central bank for the supply of gilts. That means the demand side was significantly distorted – and again this drove up the price of other assets too but it is still not clear what QE has done to investment ‘risk’ and the models on which capital asset pricing is based.
We saw in 2022 that supposedly lowest risk assets are no longer necessarily low risk at all. The assumptions of the capitalist system are based on asset pricing models that measure investment risk relative to the base asset, which is gilts in the UK. The assumption is that investors will be better rewarded for taking risks in assets that are expected to deliver more volatile but overall higher expected long-term returns. However, once you have prevented the free market from setting gilt prices, you cannot know what has happened to the investment risk of other assets relative to this base because the base itself has been unstable. It became a self reinforcing vicious circle that the more money the Bank of England printed to fund its buying of gilts and the more gilts it bought as existing holdings matured, the more downward pressure it put on gilt yields and the more it artifically inflated pension liabilities as measured using the mark to market gilt yields.
In addtion, as we move into QT, the central bank has been distorting gilt yields in the opposite direction. We are now heading for a situation in which the Bank of England is selling gilts and pension schemes that have been buying gilts in anticipation of buyout will hand their holdings over and the insurers will then also sell the gilts, again in competition with the Bank of England.
I believe we have made terrible mistakes by relying on bonds instead of higher expected return assets and have done significant damage to economic performance. QE was a big experiment, kept in place for far too long, but it suited politicians and financial markets, as well as the wealthiest groups, to have a ready buyer for gilts and increase other asset prices.
This helped in the shorter term but has left us with massive overhang and instability in asset markets for the future. Pension schemes have lulled themselves into a false sense of security thinking today’s bond yield discount rates (regardless how distorted and unreasonably low) are the accurate measure of long-term liabilities, and market asset prices for other assets should not be predicted into the long term but taken at today’s level, giving rise to the violent swings in deficits and now potential surplus which are treated as gospel but are just estimates!
Time to diversify pension funds away from bonds and into higher expected return assets – the sooner the better. And way over time to ensure pension assets boost growth, rather than chase down gilt yields which do not match pension liabilities properly anyway!
Which is why I pointed out that the Wilkie actuarial model was based relative to inflation/purchasing power rather than a gilt yield.
Like jnamdoc, I believe the actuarial
profession owe us answers over this corruption of model-based advice which seems to have gone on far too long ….
That is all clear and makes perfect sense, Ros.
To become an actuary you have to be the sort of person that will study very hard and be good at following complex maths rules. Unfortunately once they’ve been told something is a rule, especially once they’ve made a model for it, they struggle to look beyond or outwith the narrow parameters of their model (in my experience).
The Morris reviewed raised concern about the prevalence of groupthink (not surprising in a profession necessitated by complex rule following) and the unsuitability of the profession to be providing advice on investments.
The Morris review also signalled concerns with the vested commercial interests in the profession. And we are all suffering the consequences now of the professions’s hijacking of the DB arena, and the unchallenging promotion of rules or guidance promulgated by TPR. The profession should more properly have (and to be a fair, a small few of the more enlightened did) provided an independent source of advice and challenge for Trustees against the regulatory induced investment cul de sac (hopefully not abyss) we find ourselves in. Too often I’ve heard “this is what the Regulator says” in place of advice.
Back to basics per the legislation and to invest (place money with a view to making a return) in a diversified portfolio of assets. Not to act as an agent of DMO nor of TPR’s shepherding of schemes to Insurers.
Spot on. Two further points:
1. There’s no theoretical or regulatory reason for DB pension scheme liabilities to be discounted on a gilts+ basis. This is an easy change; just switch to a CPI+ basis for determining discount rates. Base the distribution of investment returns on some public empirical evidence on expected asset returns (i.e. Jorda et al. https://academic.oup.com/qje/article/134/3/1225/5435538, or better). This will remove the correlation between discount rates, liabilities and gilt yields.
2. It’s also worth noting how the simplistic gilts+ approach favoured by actuaries, TPR, and USS subverts prudent person rule and monetary policy. Ordinarily, when gilt yields fall, this is a signal for investors that expected future returns on gilts will fall, and so they respond to this signal by *increasing* investment in other assets and the real economy. Thus, the behaviour of a prudent investor in the face of decreasing gilt yields is to increase their exposure to other assets with higher expected returns. In contrast, DB schemes following simplistic gilt+ approaches *decrease* their demand for other assets, increasing allocation to gilts in response to reduced gilt yields. Liabilities increase, and schemes attempt to buy *more* gilts, increasing demand and further suppressing yield. This fundamentally undermines the monetary policy, which works on the presumption that if interest rates fall, real investment and economic activity will increase.
If this thesis is correct, the abysmal modelling favoured by actuaries has resulted in hundreds of billions of pounds of losses to the real economy and largely explains the economic stagnation observed in the UK from 2008 to 2023. Go actuaries!
Ros, as ever, a well argued piece. Nothing highlights the market distortion more than the difference between the global credit rating of UK Gilts (‘AA-‘) and the European Investment Bank (‘AAA’) – need I say more!
First, who believes that HMG doesn’t default? As at a year ago, Con Kearing and I estimated that the upcoming ILG switch from RPI to CPI would cost investors £29 b (https://www.ukrpi.com/reforming_rpi.htm). It may well be lower a year later (we’ll check) but that is still a default against valid expectations, right?
Second, if one is looking for a realistic cost assessment, discounting at a gilt rate is insanely wrong. If one oretends that one is seeking prudence, how can one know that the gilts rate is prudent, imprudent or just recklessly prudent? No, one can’t. Actually using a discount rate at all instead of robustly based Monte Carlo emerging costs is an even bigger problem than the rate being used (discrate.com)!