Core infrastructure: through its first cycle
Our young asset class is going through its first cycle in terms of interest rates, and proving itself. Nominal1 expected returns2 have increased recently, but nominal asset values are holding up. In this write-up, focusing on core infrastructure,3 I provide my view on the historical relationship between the cost of debt and the expected returns, and claim that inflation’s positive impact is greater on asset values than the negative impact of rising discount rates for core infrastructure. Going forward, looking at two of the most likely macroeconomic scenarios, I also claim that core infrastructure will continue to outperform – only if it is done right, with a disciplined core approach.
The fluctuating returns: a brief history of infrastructure
Expected returns are now in the 9 – 12% range in core infrastructure. It used to be 6 – 9% very recently, and 10 – 12% in the early years, up until 2011-12. For a long time, since the emergence of infrastructure as an asset class two decades ago, expected returns were only going one-way. One reason was, and still is, the increasing investor demand for infrastructure. Allocations to infrastructure in institutional investor portfolios have been increasing steadily. That created a small compression in the equity risk premia, but “too many dollars chasing too few deals” type arguments never materialised except for a few trophy assets here and there. Increasing allocations played a limited role in lowering the expected returns in our view, mostly in some specific sectors such as renewables. The elephant in the room, the overwhelming reason behind the declining expected returns, not just infrastructure but for all asset classes, was the interest rate super-cycle. As the central banks of the world kept lowering the interest rates, discount rates kept decreasing, and so asset prices kept increasing. There was asset price inflation in all asset classes, not just in infrastructure.
Figure 1: Core infrastructure equity expected return range in Europe versus euro area 10-yr government bond yields
Sources: Eurostat for euro area 10-yr government bond yields, which is the Maastricht Treaty EMU convergence criterion series as reported in April 2023. Core infrastructure return range estimate is based on Arjun Research views and observations. It is based on anecdotal evidence and does not rely on a
robust data set.
In hindsight, or simply applying the basics of asset pricing theory, it is obvious: Expected return is the sum of the risk premium and the risk-free rate. The equity risk premium in core infrastructure has been going down gradually, but the strong continuous decline in the risk-free rate dominated. In some periods in which the risk-free rate declined very rapidly, such as the central bank responses to Covid-19 in 2020-21, infrastructure’s perceived equity risk premium actually increased – since infrastructure expected returns are very sticky due to the high illiquidity of the asset class, expected return cannot respond very quickly to market volatility. But the sum, the expected return, kept decreasing.
Then, the 40-yr interest rate declining trend ended abruptly. The 10-yr government bond yield, the widely accepted risk-free rate proxy in infrastructure, increased by about 250 bps in almost all high-income OECD economies since January 2022.4 The expected return range for core infrastructure, on assets with similar risk profiles, increased from 6 – 9% to 9 – 12% in our view.
Valuation impact: inflation dominates discount rate
While most independent asset valuators are careful in adjusting the discount rates gradually, especially if the interest rates stay where they are today, a roughly 200 bps increase in valuation discount rates is inevitable. The asset values, though, are not going down in core, even increasing in many cases in our observation. The main reason is inflation’s significantly positive and perpetual impact on the future cash flow forecasts.
Let us consider a simplified example to highlight the fact that the positive impact of inflation on valuations can easily dominate the increase in the discount rates.
Consider a hypothetical infrastructure asset that was acquired in 2019:
EBITDA is constant at €100 in inflation-adjusted terms (e.g. an asset with both revenues and O&M costs are inflation indexed)
15-yr economic life with zero terminal value5
flat 2% base case inflation forecast at acquisition
50% loan-to-value ratio
15-yr fixed rate amortizing debt at 4% coupon6
base case expected IRR and 10-yr average yield at 8.0% and 5.6%
For the acquisition in 2019, and the valuation in 2023, let us assume the inflation forecasts in Figure 2, below. The cash flow forecast, which is a straight inflation-indexed line in 2019 gets a perpetual step increase due to the higher-than-expected inflation in the 2021-2024 period.
Figure 2. Annual inflation and cash flow forecasts
Source: Arjun Research view.
For the valuation in 2023, any change to the cost of debt is irrelevant due to the fixed rate long term debt. In reality, as asset lives are usually longer than the debt maturities, there will be refinancing risk. But if the existing debt term is long enough, its impact will be small – as it should be for investments done with the core approach.7 The only necessary update is on inflation: historical values between 2019 and 2022, and a conservative forecast back to the flat 2% line by 2025.
Because of the higher-than-expected inflation, the actual cash flow to equity in 2021 and 2022 was higher than the underwriting case. More importantly, the cash flows to equity in the future years remain significantly higher as inflation index perpetually stepped up. The investment outperforms the 2019 acquisition base case significantly.
If we keep the discount rate constant at the expected return of the acquisition case, then the net asset value (NAV) increases 23% in 2023 compared to the underwriting case. As high as a 4% increase in the discount rate to 12% keeps the NAV constant. In practice, the discount rate increase is likely to be strictly between 0% and 4%, most likely around 2%, implying a NAV increase in this example: inflation’s positive impact to dominate discount rate’s negative impact.
Source: Arjun Research.
What is next?
There are two likely macroeconomic scenarios in my view: Either interest rates will go down quickly together with inflation, or they will stay at where they are at the moment and the decline in inflation will be gradual. I expect both scenarios to be beneficial to core infrastructure:
Figure 4. Euro area 10-yr government bond yields and two likely scenarios
Source: Eurostat for euro area 10-yr government bond yields, which is the Maastricht Treaty EMU convergence criterion series as reported in April 2023.
Going back to the “lower for longer” interest rate environment, say risk-free rate at around 1% levels, in the next 12-18 months implies that the core infrastructure expected returns will go down to the 6 – 9% level. There will be a significant discount rate compression for investments made in the current high rate environment. For investment already in the investor portfolios, as the impact of higher discount rate fades away, the perpetual impact of the higher inflation index will remain, crystallising the significant positive impact of inflation.
In the “no change” interest rate environment, for which the likely case is inflation to stay above central banks’ comfort levels, the perpetual positive impact of inflation compared to the underwriting cases will keep increasing, again leading to overperformance.
Core requires discipline
Applying the core investment strategy in infrastructure requires discipline. Since the asset lives are long and the level of illiquidity is the highest compared to other asset classes, the required level of discipline is also higher. Many asset managers, claiming to have a core investment approach but opting for short-term debt instead of the more appropriate but more expensive long term debt got away with it as the cost of debt kept decreasing – until now. This cycle with high inflation and high cost of debt is the first real test on asset managers, claiming to be implementing core strategies.
Core infrastructure, only If it is done right, is outperforming and it will continue to outperform in the near future, more or less regardless of the macroeconomic situation.8 The lesson we are learning from the ongoing crisis is the importance of the discipline on the core approach.
1 I have long argued that infrastructure should be analysed in real (inflation-adjusted) returns rather than nominal ones. Yet the world we live in is nominal, it takes an effort to convert nominal figures to real ones, and many investors still think in nominal terms.
2 I use the terms “expected return” and “discount rate” interchangeably.
3 Our preferred definition of core infrastructure is based on the forecastability of the cash flow to equity – necessarily including a prudent leverage strategy with long term fixed rate debt.
4 One notable exception is Japan, where the increase is around 50 bps.
5 Economic lives tend to be longer unless the investment is on a concession.
6 Normally the debt maturity is a few years before the end of life (or contract length), but assuming 13-yr debt does not change the main points.
7 See our previous research paper “Core infrastructure: keeping it real” for a discussion on the importance of a prudent approach to leverage in core infrastructure; available at https://www.arjuninfrastructure.com/post/core-infrastructure-keeping-it-real
8 IRR and yield based on actual and future cash flows between 2019 and 2034.