Views from the LDI Desk – 2021 UK LDI Outlook

19-Nov-2020
  • BlackRock

RPI reform - the final countdown?

After much anticipation, on Monday 9 November the Chancellor announced in a letter that a response to the Government and UK Statistics Authority’s (UKSA) joint consultation on Retail Prices Index (“RPI”) reform will be delivered alongside the Spending Review on 25 November.

To recap, in September 2019 the Chancellor issued a response to the UKSA’s proposed changes to RPI.

The Chancellor rejected the notion of ceasing to publish RPI due to the potential for significant disruption to end users but recognised the UKSA’s suggestion for aligning RPI with the Consumer Prices Index including owner occupied housing (“CPIH”), stating that such a change was to be implemented no sooner than February 2025. The letter also highlighted that the UKSA continued to need consent from the Chancellor to make such changes to the RPI before 2030.

This prompted a joint consultation being launched by the Government and the UKSA on 11 March focusing mainly on the timings of the change and other technical points which closed on 21 August.

Potential Consultation Outcomes

There are a broad range of potential outcomes possible as a result of the RPI Reform Consultation. We have explored some of these in more detail in our previous piece “RPI Reform: The thin end of the wedge”. However, for simplicity, we can broadly categorise these into two groups:

Potenital consultation outcomes

From BlackRock polling of broker views and market movements, market consensus appears to be converging strongly around the assumption that RPI is redefined to equal CPIH at some point between 2025 and 2030. As such, if the outcome is broadly in line with this, the market reaction should be relatively muted in comparison to if one of the more unexpected outcomes prevails (i.e. one that would suggest that any new methodology does not have a meaningful impact on realised RPI compared to the current methodology used). Our previous estimates suggested that CPIH is approximately 0.8%-0.9% below RPI over the long run, which would suggest, very bluntly, a range in market levels between these two outcomes to be somewhere between this order of magnitude for the affected forwards.

We can show this in the stylised diagram below for RPI forwards:

RPI swap forwards

Source: BlackRock. Illustrative only and based on estimates from observed market moves.

How much of a change has the market priced in?

There have been a multitude of attempts to quantify exactly how much of the expected reform outcome has been ‘priced in’ to the market, including our own attempt back in August.

As a simplistic estimate on more clearly attributable impacts, we can look at the market moves on the day immediately after significant moments in the RPI reform debate.

Change on day after significant RPI Reform events

Change on day after significant RPI Reform events

Source: BlackRock. Data as at 11 November 2020.

This should give a significant underestimate of what is priced into the market, with the logic that further market moves ‘pricing in’ the impacts were typically spread out over a longer period rather than simply taking place on the next trading day alone.

As another approach, we can look at the move in the forwards across the whole period since the House of Lords report was first released in January 2019.

Move in Inflation forwards between 16-Jan-19 and 11-Nov-20

Move in Inflation forwards between 16-Jan-19 and 11-Nov-20

Source: BlackRock. Data as at 11 November 2020.

This shows significantly greater moves, and, at first glance, would suggest that a change is nearly fully priced in. The average YoY forward from 2031-2068 has dropped by 69bp over the period 12-Nov-18 to 12-Nov-20.

However, it is worth also noting the global trend for lower inflation (with EUR HICP and US CPI moves over the same timeframe also shown in the chart above), caused by a plethora of factors, not least the ongoing Covid pandemic. As such, it would be bold to conclude that the drop off in RPI forwards over this long time-horizon can be attributed to RPI reform alone.

Another approach we can use to gain insight is to look at the evolution in RPI-CPI spread levels.

Historical RPI - CPI spread

Historical RPI - CPI spread

Source: BlackRock, Lloyds. Data as at 10 November 2020.

Observing the spread between Zero Coupon RPI swaps and the CPI equivalents shows a significant fall in the spread from after then-Chancellor Sajid Javid’s letter last September. Whilst the headline Zero Coupon swaps give us some guide, perhaps more meaningfully we can look at the 10y10y forward spread. The logic behind this is that it would start around 2030, the latest date where reform could begin, and would not extend past the 20y maturity, which has typically been around the final maturity of much CPI supply, originating from Electricity Market Reform renewable energy projects. This 10y10y forward has dropped from 76bp in early September 2019 to 13bp on 13 November 2020. This would suggest that the vast majority of a convergence to CPIH beyond 2030 has been priced in. However, we would note two significant caveats to this conclusion. Firstly, these levels are based on CPI and not CPIH, so there is another basis to consider (which we examine more later). Secondly, and more importantly, the CPI market is illiquid, with volumes falling even more sharply throughout the uncertainty associated with RPI reform. As such, drawing a conclusion based on market data underpinned by a very limited number of transactions is not appropriate.

Timing of Reform: 2025 - 2030

Given the increasing market consensus of a convergence to CPIH without a spread, much of the market debate has focused on the timing of any such reform measures. The letter from then-Chancellor Sajid Javid indicated that a change would not be acceptable before February 2025, whilst the consultation specifically asked around impacts from changes in 2025 and 2030. To get a sense of the relative market assessments of these timings, we can look at the evolution of the 2025-26 market implied inflation level vs the 2030-31 level.

Comparing 2025/26 and 2030/31 RPI swap rates

Comparing 2025/26 and 2030/31 RPI swap rates

BlackRock, J.P.Morgan. Data as at 10 November 2020.

This shows that, at present, the 2030/31 forward is approximately 30bp lower than the 2025/26 level. As such, if an announcement was made of a 2025 reform date, we would expect a significant drop in the forwards from 5-10y, and a more general steepening of the curve (as these lower forwards in 5-10y would have a more diminished impact in basis point terms for longer tenors once this change is averaged out for longer). Given the previously estimated saving to Government finances of ~£6bn p.a alongside the current challenging fiscal environment, we believe that the likelihood of an early reform is underestimated by the market.

The likely market reaction for any announcement is hard to gauge, given the difficulties in assessing what is priced in. However, it seems less contentious to state that the 5-10y sector of the forward curve could see some of the most enhanced volatility under any outcome.

Inflation supply outlook

Whilst the impending reform announcement is likely to give some much-anticipated clarity to the market, there is still the (very topical to all things 2020) question of what else is needed to return closer to ‘normality’. There is a reasonable expectation that the announcement on the 25-November will likely unlock both latent demand and supply that has been sitting on the sidelines waiting for the reform response. However, what is less clear is the quantum and timing of the latent flows that may emerge.

One piece of the puzzle to look at is Inflation Linked Gilt supply. If we look at the evolution of the Debt Management Office (“DMO”) supply in the past years, a vast majority of the increase has arisen in nominal bonds. Indeed this is not an unintended phenomenon, with the DMO stating in the 2020-21 Debt management report that “as a result of the government’s responsible approach to fiscal risk management [the proportion of annual debt issuance that is in index linked Gilts] has been reduced”.

Gilt Supply and QE purchases by fiscal year

Gilt Supply and QE purchases by fiscal year

Source: UK Debt Management Office (DMO), Bank of England, BlackRock. Data as at 10 November 2020.

However, it is worth noting that Quantitative Easing (“QE”) has been targeted in nominal Gilts, so if we look at supply net of QE purchases, the decline in index-linked gilt issuance is still evident, but certainly less pronounced.

Linker Issuance as % of total Gilt Issuance (exc bills)

Linker Issuance as % of total Gilt Issuance (exc bills)

Source: UK Debt Management Office (DMO), Bank of England, BlackRock. Data as at 10 November 2020.

However, what may be most pertinent to driving market liquidity and levels is the risk sensitivity (measured by present value of a basis point, “PV01”) of inflation linked bonds issued. The chart below comparing the last three years shows that despite the often-quoted narrative that there is much less inflation linked Gilt supply coming to the market, this is only true when looking as a proportion. Whilst inflation linked syndications have paused, the increased frequency of auctions means that average PV01 is not meaningfully lower for 2020 than vs previous years.

PV01 in Index-Linked Gilt Supply - last three years

PV01 in Index-Linked Gilt Supply - last three years

Source: UK Debt Management Office (DMO), BlackRock. Data as at November 2020.

The DMO has recently split their December supply from the January-March schedule, which will be published on Friday 4 December, after the results of the RPI reform consultation are public. The market reaction to the consultation results may help to shape the upcoming supply schedule, and whether we may see an Inflation Linked Gilt syndication at the beginning of 2021.

Also of interest, is the prospect of corporate and infrastructure supply post-reform. Whilst a lot of this supply takes place in the derivative markets, thus complicating any attempt at quantification, we can see from the drop-off in the volume of corporate UK-inflation linked bonds issued in the past couple of years that this is also a market that has been heavily impacted by the ongoing uncertainty around RPI reform as well as utility regulator consultations around future inflation linkages.

Corporate UK Inflation Linked Bond Issuance by Year

Corporate UK Inflation Linked Bond Issuance by Year

Source: BlackRock, Bloomberg. Data as at 11 November 2020.

The corporate issuance market can typically take longer to react to changes in market conditions, with ongoing pressure on some utilities to deleverage likely to provide an additional headwind to a quick rebound in inflation supply. Other key factors, such as the potential for more announcements of renewables projects, are often cited as determinants of a return of more supply, particularly in CPI, and will be closely watched.

The potential lag before the re-emergence of supply can be thought of as a reason why even if we come to the conclusion that a reform of RPI to CPIH without a spread occurs, and the market has not fully priced this in, it is far from certain that we would see RPI market levels fall into year end.

Additional nuances to consider

Aside from the potential headline moves in implied inflation rates in both assets and liabilities, it is also worth pension schemes considering some of the nuances a little under the surface.

Data from The Pension Regulator suggests that 22.4% of memberships post-1997 are linked to CPI rather than RPI. For such linkages, whilst a reform of RPI to align with CPIH would reduce the basis between RPI-based hedges and CPI liabilities, there is still some residual volatility. After the growth of the CPI market in the past years, this brings into question how a two-tier linkage with more minimal basis between them will evolve. It seems likely that many with contractual CPI linkages will be comfortable to run the basis to CPIH, which, in turn, challenges the longer-term viability of an active CPI market.

Spread statistics: last 15y

CPIH-CPI RPI-CPI
Average Spread -0.11 0.63
Standard Deviation 0.32 1.05
Range -1 to 0.4 -3.4 to 2.3
Interquartile range -0.3 to 0.1 0.5 to 1.1

Source: ONS, BlackRock Oct 2020

CPIH - CPI spread vs RPI - CPI spread

CPIH - CPI spread vs RPI - CPI spread

Source: Office for National Statistics (ONS). Data as at October 2020.

It has been well covered that RPI reform will have a significant impact for schemes which have been hedging CPI liabilities with RPI instruments. However, even for those hedging RPI-based exposure with RPI assets there are some nuances worthy of consideration.

For schemes that have hedged inflation and rates through separate derivative instruments, it is worth being aware of the ‘cross gamma’ impact – as moves in implied inflation will affect the nominal value of projected future payments and hence will change the interest rate sensitivity from the discounting of these cashflows.

Given the high proportion of schemes that have limited price indexation or ‘LPI’, with caps and floors inherent in inflation liabilities, it is worth noting the non-linearity of such linkages. As an example, for a liability linked to RPI floored at 0% and capped at 5% (as is typical in many schemes), a fall in RPI market levels will increase the value of the floor and decrease the value of the cap. This means that, in the event of a fall in RPI levels, RPI linked assets in place as hedges are likely to fall in value to a greater degree than the LPI-linked liabilities they are hedging.

We also note that the small print in many corporate inflation-linked bonds, including Network Rail issuance, will be closely examined, where reference is made to possible compensation to bondholders in the event of changes to indexation.

Conclusion

After a long journey, that we can trace back at least to the January 2019 House of Lords Report, and arguably many years previously to discussions around the suitability of RPI, it appears that the Reform Consultation Response on 25 November will likely begin to bring much needed clarity to the market. While the consensus for RPI becoming CPIH now appears strong, the timing remains uncertain and we think a 2025 reform date is under-priced still. Of course, a range of surprise conclusions or further delay remain on the table as potential outcomes but feel unlikely.

With the potential for follow-up questions arising from the consultation response itself, as well as the need for a market to find a ‘new normal’ whilst many pension schemes tackle the consequences of any announced changes, much of the work to adapt to the changing UK inflation market may only just be beginning.

The opinions expressed are as of November 2020 and are subject to change at any time due to changes in market or economic conditions. The above descriptions are meant to be illustrative. There is no guarantee that any forecasts made will come to pass.