Views from the LDI desk – UK Edition – April 2022

Inflation was already a hot topic in 2022 with the re-opening of economies post-covid, but the conflict in Ukraine has only served to amplify the challenge and push it even further up the domestic political and economic agendas. The cost-of-living crisis from the inflationary supply shock is starting to impact both monetary and fiscal policy decisions.

In second half of this piece, we explore what this means for pension scheme funding levels and inflation hedging programmes, but first we recap some of the key events over the past few weeks, what this means for gilt supply, yields, asset swap levels and the repo market.

Beware the Ides of March

The start of the second half of March kicked off a busy period on both the monetary and fiscal policy fronts. In both cases, a fine line is being walked between considering controlling inflation and maintaining underlying economic growth and avoiding stagflation.

The Bank of England (BoE) increased base rates by 0.25% on the 17 March, taking the base rate back to its pre-pandemic level of 0.75%. This hike was expected by the market, however the decision for one voter (member Cunliffe) to vote for no hike was a surprise with some risk of a 0.5% hike previously being priced by markets. The decision was viewed as more dovish as wording around future rate hikes was softened, for example a further modest tightening in policy “might be appropriate” versus “is likely to be appropriate” in February.

After the meeting, markets scaled back expectations around the speed of hikes, however since then much of this ground has been regained and markets continue to price a relatively rapid pace of hikes in the attempt to control inflation.

Dovish comments from the BoE in the March meeting have tempered expectations on the speed of rate hikes but the market is still currently pricing the same end point over the next 12 months

Next up was the Chancellor with his spring statement on the 23 March. While various policies aimed at combatting the cost-of-living crisis were unveiled, such as a 5 pence fuel duty cut and removal of VAT on solar panels and heat pumps, the overall cost of these policies in the grand scheme of annual gilt issuance was modest. Tinkering with national insurance thresholds to spread the impact of the planned national insurance increase served to reduce the total value of this tax hike to the Treasury, but again did not fundamentally change the fiscal picture.

The Office of Budgetary Responsibility (OBR) published its semi-annual Economic and Fiscal outlook immediately post the spring statement, downgrading 2022 growth from 6% to 3.8% as forecasts were updated to capture the impact of higher energy prices and the instability created by the Russian invasion of Ukraine. In normal circumstances a 2.2% reduction in forecast growth for a year might be recessionary, but 3.8% feels still a healthy growth figure. A large part of this is the impact of base effects from a heavily COVID impacted 2021.

A spring statement that cut some taxes and duties (and promised further tax cuts to come), combined with a sharp cut in growth forecasts might have been feared to lead to an oncoming deluge of gilt issuance. However, as we will explore the twists and turns of economic forecasting by the OBR in an economy swinging between rapid growth and contraction with alarming regulatory has offered the Treasury and therefore the Debt Management Office (DMO) some breathing room. Thanks to a significant surplus from funding during the current 2021/2022 fiscal year to the tune of £46.6bn, total gilt issuance was predicted to be a lower than expected £124.7bn for 2022/2023. This surprised markets, where many economists were expecting a figure of £150bn, while just last October the OBR forecast was £200bn. Immediately following this news, gilt yields fell both on an outright basis and relative to swaps.

Smaller than expected DMO remit still leads to an above average net remit vs. the last few years due to a lack of purchases under Quantitative Easing programmes going forward

Within this headline gilt issuance number a lot of additional detail is hidden in terms of the split of issuance both by tenor and type of gilt. We explore three dimensions to this, the tenor of conventional gilts, green gilt issuance, and the level of index-linked gilt issuance.

Comparing year on year gilt issuance splits

Conventional Gilt Issuance – Barbell Approach

The DMO has announced conventional gilt issuance skewed more towards 3-7yr gilt issuance (shorts) and 15+ year issuance (longs).

While the headline gilt issuance figure was lower than expected and lower than some other net remits over the past decade, the amount of longer dated issuance in the absence of any BoE buying creates the potential for this to be a record year for net gilt supply in PV01 terms. As we wrote in the February edition of views from the LDI desk, record gilt supply combined with potentially attractive conditions for buy-out activity have the potential to push longer dated gilt yields higher. This may be on both an outright basis and relative to swaps and this record net PV01 of longer dated gilts for the market to digest continues to support this view.

Net issuance of long dated conventional gilts in PV01 terms looks set to be record with no QE purchases to offset issuance while index-linked gilt issuance continues to be restrained

The increase in issuance in shorts (and also treasury bills) was desired by the market where short dated gilts had become difficult to purchase or borrow and increased in value relative to other instruments as shown in the chart below. In part this is due to the effects of quantitative easing, with the BoE holding a high proportion of the free float of many of these bonds. It is also being driven by a technical change to a facility offered by the DMO whereby market participants can borrow bonds from the DMO at a fixed spread below the base rate. Where certain bonds are tightly held and liquidity is poor, it is common to see it become very expensive to borrow the bonds in the repo market (often referred to as trading special) and this DMO facility acted as a floor on how expensive this can become.

As base rates have been increased the margin below the SONIA rate that is used as the floor has been increased, in line with historical precedents. This has served to further increase the cost of borrowing such bonds, impacting on the ability of market markers to go short, the ability to undertake IOTA trades (long linkers and short conventional gilt) and the attractiveness of other financing trades such as netted repo transactions.

As base rates increase and the DMO borrowing window is widened repo rates on some bonds have diverged from average levels also helping to drive asset swap spreads lower

As base rates look set to continue to increase the risk of the floor this facility provides continuing to fall away causing short dated gilts to further outperform swaps is present, however as supply returns to the market it is not unreasonable to expect to see short dated gilt yields increase relative to swaps.

Key Takeaway: For schemes holding sub 10 year gilts, particularly if these are funded via repo, this could be an opportunity to consider switching some of this hedging into swaps and this is an approach we are undertaking for discretionary LDI accounts.

Green gilt issuance – No new maturities

The remit update also provided further details on the plans for further green bond issuance. With a minimum of £10bn to be issued in 2022/2023 the supply was lower than hoped, with many expectations of similar amounts to previous year which would have been more like £15bn.

As a percentage of total issuance this represents 8% and is broadly in line with the previous year on a proportional basis. However, we feel it doesn’t make sense for green issuance to scale in line with the total amount of gilt issuance and if there are sufficient green projects available for the bonds to fund, given the greeniums previous issues have come with it would be cost efficient for the DMO to issue as many green bonds as possible. With shifting dynamics around energy policy driven by the conflict in Ukraine there may be greater uncertainty on projects that will qualify for green bond funding in the years to come, which may have led to caution in the amount to be issued.

It was also confirmed that this issuance will be through further taps of the existing 10yr and 30yr gilts, rather than through the issuance of a new green gilt as many had expected. This might have been a 20yr or 40yr point to continue to build out the green curve. We had also suggested the DMO consider issuing a green index-linked gilt, which would increase the opportunities for schemes keen to maximise green bond holdings to do so more easily, as much of their liability is inflation linked.

Key Takeaway: The news that the two existing bonds would each be tapped again rather than issuance driving a new bond has led to greeniums falling and the yields on green bonds increasing relative to surrounding non-green bonds as the market prepares to digest this additional supply, particularly in the 2033 bond.

Greeniums on 2033 and 2053 green gilts have fallen on the spectre of more supply in these bonds

Index-linked Gilt Issuance – Post financial crisis lows

While the percentage in index-linked gilts has increased from last year, the lower outright issuance level for gilts means the total value of index-linked gilts is likely to be below £20bn for the first time since 2007/2008. While its highly likely that around a quarter of this issuance will be further supply of the ultra long duration 2073 index-linked gilt, this may still leave the market short of overall PV01 if pension de-risking continues, which we give further consideration to in the next section on inflation.

Overall the upcoming quarter is expected to see a significant total gilt issuance with both the 2073 index-linked gilt and potentially also the 2073 conventional gilt being issued. This leads to a particularly big contrast with Q1 2022 where issuance has been very constrained. Q2 could see in excess of £60m of PV01 issued, compared to around half that amount in Q1, while QE buying still also continued and bought up much of this issuance.

Gilt issuance continues to follow the feast and famine pattern with Q2 2022 potentially providing record PV01 with two ultra-long syndications

Accelerating Inflation – what does it really mean for pension schemes?

Much has been made in the mainstream media of recent high inflation prints and increases in energy bills, with eye catching 50% increases in bills due to hit home in April. But what does this spike in short-dated inflation really mean for pension schemes and how much is this feeding through to longer dated inflation expectations?

Firstly, it is worth considering the impact of very high inflation prints in the short run. Many 1 year RPI inflation projections in the coming months (e.g. May 2021 to May 2022) have at times been pricing RPI in excess of 10%. Many schemes are likely to have a proportion of their liabilities linked to LPI(0,5) - RPI inflation capped at 5% and floored at 0%.

As inflation expectations increase, the sensitivity of these LPI(0,5) liability streams to inflation fall and for the majority of schemes hedging these liabilities with RPI linked instruments, the proportion of cashflows that are real (linked to inflation) should theoretically be reduced to match this lower “delta” to inflation. With inflation at 10%, many of these deltas would be close to zero. However, most schemes do not regularly refresh these characteristics for a variety of practical reasons. In the current situation where inflation has passed so far through the cap, should inflation start to realise at these levels this can have material positive impacts on the funding level and implications for interest rate and inflation hedges.

Taking the example of a scheme with a set of LPI(0,5) liabilities that were perfectly hedged in May 2021 with RPI instruments with an inflation base date in May 2021 and considering current market implied pricing from inflation swaps to realise, this is estimated to have a 5.8% impact on the value of this liability stream relative to the hedging assets. No scheme is likely to be 100% LPI(0,5) linked, but many have a reasonable proportion, so this may translate to a few percentage points improvement in the funding level of the scheme. Unfortunately, this benefit comes at a cost to scheme members, who experience a real reduction in the value of their future pension and living standards as inflation prints above the cap, as many workers also are, where wages are not keeping pace with inflation.

The market is currently pricing 5% inflation caps to be hit over the next one to two years

So schemes with a lot of LPI(0,5) and an inflation hedge that hasn’t been regularly recalibrated by their actuary or consultant may find themselves experiencing this unexpected funding gain as high inflation realises, but what about their hedging? Should they be recalibrating and does inflation look expensive for those with unhedged exposures?

The effect of the year-on-year inflation caps ripple through to future years, as the capped forward inflation feeds through to future inflation uplifts. This impacts not only the inflation sensitivity and the delta to inflation but the size of the future cashflow itself and therefore its interest rate sensitivity. Schemes may also find this results in the interest rate sensitivity of their assets being higher than their liabilities – potentially not ideal in an environment of rising rates, so considering a recalibration is important. Should inflation continue to be volatile in the coming months, this may need to be re-visited several times.

The impact of a 5% inflation cap being hit in year 1 is felt on inflation rates in all years to come

On inflation levels themselves, while 30 year RPI inflation rates look and feel historically high, much of this is driven not by high expectations for inflation in the long run but by the extreme valuations on shorter dated inflation.

While 30yr inflation is close to highs over the past 10 years despite planned RPI reform in 2030, levels are dragged higher by extreme inflation in the first few years

If longer dated inflation expectations are isolated from the extreme levels at shorter tenors then levels of forwards look, while perhaps not cheap, more reasonable given the BoEs long term target for inflation at 2%. Allowing for some inflation risk premium and term premium, not to mention the supply and demand imbalance inherent in the index-linked gilt market when the size of the pensions and insurance market is compared to the size of the index-linked gilt market, levels can still be justified and have actually shown very little change since the start of the Russian invasion of Ukraine and subsequent spike in short dated inflation.

While since the start of the Russian invasion of Ukraine shorter dated forwards have increased, longer dated forwards are flat or have fallen

Long dated inflation has appeared expensive relative to a 2% BoE target for a while and will likely continue to be expensive while supply and demand trends continue to play out, as they look set to with a limited amount of inflation supply in the 2022/2023 DMO remit. Schemes adverse to hedging shorter dated inflation as they believe the current spike will be transitory (or with very little exposure to this anyway due to inflation caps) could consider the purchase of forward starting inflation. However, there is growing risk of inflation expectations de-anchoring as the BlackRock Investment Institute recently wrote, so any schemes planning to avoid short dated inflation on the basis it looks historically expensive should carefully test their views and level of conviction. It is a hedge after all.