- Long-dated gilts are currently expensive by between 150 and 200 bps, on a pure valuation basis.
- The BoE’s monetary policy is responsible for part of the overvaluation, but far from all of it.
- The greatest driver of the gilt overvaluation continues to be demand from pension funds.
- Given the latent demand from pension funds for long-duration assets for hedging purposes, it is difficult to see this overvaluation abating much in the near term.
- The limited nature of any BoE policy reversal means it will have little impact on the overvaluation.
- With pension funds the likely buyers of gilts on any move above a yield of 3.0%, we see yields capped over the foreseeable future at 3.0% to 3.5%.
With UK government gilt yields near historic lows, the prevailing wisdom is that gilts are now prohibitively expensive and wise investors should steer clear. For those of us with long memories, this view is reminiscent of the experience of the Japanese government bond market 20 years ago, when Japan’s 10-year government bond yields breached below 2%. The perception that bonds at those low yields were too expensive was comprehensively refuted then, but will it be now?
To answer this question it is necessary to ascertain what fair value is for long-dated gilt yields and what it is likely to be over the coming years. Then, we will identify what the main drivers of gilt demand are and what impact they have had on the current level of yields. Supply will not be an issue, with the ever-profligate UK government projected to run budget deficits well into the latter part of the next decade.
In this paper we answer these questions, concluding that nominal gilt yields are between 150 and 200 basis points (bps) too low. However, with demand likely to rise as gilt yields rise, it’s unlikely we’ll see yields above 3.0% to 3.5% in the foreseeable future.
Assessing Fair Value
Over the long run, the return on long-dated gilts should be related to the long-term return from investing in the UK economy, less a premium for the certainty of return guaranteed by investing in government debt. For index-linked gilts, this should equate to the long-term real rate of UK GDP growth; for nominal gilts, it is the same plus long-term inflation. In practical terms, annual real growth has tended to be around 2.0% to 2.5%. Since the onset of inflation targeting in 1992, the Bank of England (BoE) has targeted an inflation rate between 2.0% and 2.5%, depending on the measure used. Fair value for long-dated nominal gilts should thus be in the region of 4% to 5% and for index-linked gilts, real yields should be below 2%.
Valuation—the Nominal Perspective
The 60-year history of UK nominal GDP and gilt yields is shown in Exhibit 1, reinforcing our suggestion of a rough relationship between yields and UK GDP growth as detailed in the prior section.
Perhaps a more pertinent history to use is the period since the UK government adopted an inflation target for the operation of monetary policy, post the UK’s exit from the European Exchange Rate Mechanism (ERM) in September 1992 through the present. Initially, the government was responsible for setting interest-rate policy under the new regime before passing it to the newly independent BoE in May 1997.
It is clear from Exhibit 2 that the nominal GDP correlation to gilt yields does indeed loosely hold, particularly after the new policy regime achieved credibility in the mid-1990s. Volatility is a factor, with the market arguably looking through periods of strong and weak growth to focus on the long-term trend. However, the last three years have perhaps seen a more meaningful decoupling.
Plotting the difference between 30-year nominal gilt yields and GDP growth over the last 30 years provides even greater insight (Exhibit 3).
The current differential between long-dated gilt yields and UK GDP growth is stretched, just over one standard deviation (the lower orange line) away from the five-year moving average (dotted grey line in Exhibit 3). Historically, this has been the point at which we have seen some correction in the differential; currently the spread is more than 200 bps wide. Clearly, the differential could narrow through gilt yields rising, growth slowing or a combination of the two. However, our view is that growth risks are to the upside for the UK over the next two years. Indeed, our central forecast for UK growth in 2017 and 2018 is above that of the BoE and the market consensus. On this basis, we expect long-dated gilt yields to rise over the next several months as the relationship normalizes.
Valuation—the Real Yield Perspective
The real growth to real yield correlation is weak at best and has clearly broken down in recent years. The demand for inflation-linked gilts is more likely driven by the need for inflation protection rather than income. This is borne out by the relative stability of breakeven inflation rates (BEI) over the last two decades (Exhibit 4), with BEI averaging 3.1% since BoE independence in 1997, with a low standard deviation of 0.4%.
Given that Retail Price Index (RPI) inflation has roughly averaged 2.75% since 1997, it could be argued that investors have or are still overpaying for inflation protection. However, when compared with long-term inflation expectations (five years forward) as given by the BoE TNS/NOP survey, the current levels paid by investors for future inflation protection look far less expensive (Exhibit 5).
In sum, from a fair-value perspective, both nominal and index-linked long-term gilt yields look expensive, and we would expect some form of correction to current levels over the medium term. The timing and durability of any correction is more difficult to predict.
Next, we take a closer look at the demand-side factors influencing the price and yields of gilts. Three primary contributors are currently driving demand: BoE central bank purchases, overseas buyers and UK private pension funds.
The Impact of Bank of England Asset Purchase Facility (APF) on Yields
The BoE has no doubt had a substantial impact on the level and structure of gilt yields. We conjecture that most of this influence has arisen from its actions on the policy rate rather than quantitative easing (QE) per se. The reduction in the BoE bank rate—the interest rate that the BoE charges banks for secured overnight lending—to a historic low of 0.25% has had an effect similar to the actions of the European Central Bank (ECB) and the US Federal Reserve (Fed) in moving risk-free rates down across the maturity spectrum. Viewed in the round, these can be seen as a continuation of the trend begun in the mid-1980s with the secular decline in yields in response to ever-lower inflation. However, the impact of the BoE’s QE programs can be seen quite clearly in Exhibit 6, especially the second and third phases of the policy.
Empirical studies have found a positive impact from BoE QE. Joyce, Tong and Woods (2011) estimated a 100 bps decline in long-dated gilt yields due to the first QE program, while Meaning and Warren (2015) estimated that the first two rounds of QE reduced yields by 25 bps. Obviously, the conclusion reached will depend on the assumptions made in the underlying models, but what is clear with the ending of BoE asset purchases is that the future impact from the APF will be limited.
The Impact of Overseas Capital Flows
The UK, having run a current account deficit for the last 30 years, is reliant on capital inflows to fund itself. Gilts benefit from some of this inward investment. This demand is unlikely to diminish in the near term, as the current account deficit remains around 4.5% of GDP and whilst the UK benefits from its safe-haven status. While some correlation exists between overseas purchases and the direction of gilt yields, given that most overseas investors tend to favor shorter-dated maturities, the impact on long-dated securities is muted.
The Impact of Pension Funds
Pension Fund Solvency—The Real Driver of Gilt Demand
The current solvency situation in the UK private pension industry, as determined by the Pension Protection Fund (PPF), has worsened in recent years, increasing the demand for gilts. As of March 2016, funding levels under Section 179 of the Pensions Act 20041 and a benefit paid on a full buyout basis2 equate to a deficit of £220 billion and £780 billion, respectively. Funding levels as a percentage of liabilities are 85.8% and 63.2%, respectively. If these deficits were immediately plugged by pension plan sponsors and the monies invested into the gilt market, this would represent 11% and 39%, respectively, of the existing stock of nominal and index-linked gilts together.
Pension Fund Asset Reallocation—Already Supporting Yields
Clearly, moves on this scale are extremely unlikely to happen. What is more likely is that pension funds switch assets as solvency ratios improve, moving from riskier equities into bonds. This is exactly what has happened over the last 10 years, with plan sponsors appreciably lowering their exposure to equities from 61% to 30%, while increasing bond investments by 23% to 51% and hedge fund allocations to 7% from nothing.
Within the main asset classes, there have been significant moves. Since 2008, index-linked bonds have grown by 6% of the full pension fund bond allocation, explaining the robust nature of the index-linked market over recent years and the ‘expensiveness’ of inflation protection. Meanwhile, overseas equities now make up an additional 17% of the full pension fund equity allocation.
Impact of Nominal Gilts and Index-Linked (Linkers) on Pension Solvency—Just How Valuable Are Gilts as a Liability-Matching Asset?
To fully understand pension fund demand for gilts, one needs to understand their effectiveness in immunising funds from solvency risks. We first look at the efficiency of nominal gilts and linkers, then compare them to equities. This relative efficiency is key to determining whether or not pension funds will add to gilts if yields rise, thus preventing yields from moving back to fair value.
Exhibits 7 and 8 highlight the far greater sensitivity fund valuations have to real and nominal yields versus that of equities.
This makes intuitive sense and is consistent with findings elsewhere. The PPF and pension regulators found that in the UK:
- A 10 bps rise in gilt yields decreases the aggregate deficit by £22.7 billion.
- A 2.5% rise in equity prices would reduce the aggregate deficit by £9.9 billion (or 0.7% of liabilities).
- A 10 bps fall in gilt yields raises the aggregate liabilities by 2.0%, but increases asset values by only 0.6%.
- An increase in the assumed rate of inflation by 0.1% with nominal gilt yields unchanged raises liabilities by £10.4 billion (or 0.7%).
By further developing this analysis, using the PPF valuation at the end of March 2016, one can estimate the simultaneous change in bond yields and equity moves needed to move aggregate pension plans into surplus using s179 methodology. For example, if equities are assumed to remain unchanged, it would take a rise in 30-year gilt yields and linker yields of approximately 100 bps to eliminate the overall funds deficit. If equities rose 2.5%, the required move in gilts would be closer to 95 bps.
Supply—A Final Word
Since the UK election on 8 June, political uncertainty has clearly risen. While markets always hate uncertainty, and will increase volatility in the short term, we feel these events are unlikely to impact the long-term fundamentals for gilt yields. As we discussed in a recent paper, Hung by the Polls: Mayhem in June, a minority government administered by the Conservative Party is now in place, and mild fiscal slippage should be expected. Therefore the supply picture looks cloudy, but as discussed earlier, we feel it plays a minor role in gilt yields compared to the prospects for demand.
On a valuation basis, long-dated gilts are currently expensive by between 150 and 200 bps. We find BoE policy responsible for part of that overvaluation, but far from all of it. The greatest driver remains pension fund demand. Given the latent demand for long-duration assets for hedging purposes, it is difficult to see this overvaluation abating much in the near term even if the BoE reverses policy. With pension funds likely buyers on any move above 3.0%, we see yields capped over the foreseeable future at 3.0% to 3.5%.
- Section 179 of the Pensions Act 2004 (s179) stipulates the levels of benefits paid under the PPF capped to a maximum £33,700
- Full buyout basis implies benefits paid in accordance with what pensioners have accrued