Can you really lose GBP1.5Bn in a couple of quarters when investing in a water utility?

Investments in infrastructure like water utilities are very often presented as ‘boring’, slow-moving, low-risk assets, that also happen to be cash cows. That latter point is true, as my colleagues have shown in previous research. The rest is wishful thinking at best, but mostly it is lazy thinking that reveals a lack of understanding of infrastructure and of finance.  

In March 2022, some investors in Kemble Water Holdings (the HoldCo for Thames Water Utilities Limited) were still increasing their marks by 15%, reflecting the belief that all was well with their investment in the company supplying water and wastewater services to London and the Thames valley. By December 2022, the largest shareholder had to recognise a loss of 28% of the fair market value (FVPL) of its investment in Thames, via entities incorporated in Singapore and the Netherlands.  

What changed in the space of nine months? Thames Water was not hit by a missile or partly renationalised. In practice, nothing changed. Reality simply caught up with investors who clearly had been in the dark until then.  

The Thames Water Utility Company is tracked in infraMetrics, our data platform, like hundreds of other private infrastructure companies. Its fair market value is measured monthly, going back twenty years and this is the opportunity to see an interesting pattern.

The volatility of Thames was always high when compared with various benchmarks, and over time it grew higher. Meanwhile, logically, its fair market value declined, as shown in table 1. From the time Macquarie sold its stake until now, we estimate the value slowly dropped by c.30% as risk increased. In other words, the loss that was booked in December 2022 didn’t happen overnight.  

These practices make a mockery of the notion of fair market value, which has solid theoretical groundings in academic finance and clear rules set by IFRS. They leave investors clueless about the risks they take when investing in private assets, which is not supportive of their prudential and fiduciary duties. They also end up hurting the infrastructure asset class and the business of investing in infrastructure, by creating confusion in the public eye and among regulators as to what private investors in infrastructure are up to.  

Private infrastructure investments are not risk-free. Some are highly exposed to certain risks, starting with real rates. They are all exposed to systematic risk factors that yield a return (a risk premia) in the market for these assets, and taking these risks is what justifies the returns investors expect to receive.  

We have shown that it is possible to measure the price of risk for private infrastructure equity using market transactions and a good asset pricing model. This is what infraMetrics is all about: produce robust estimates of the risk premia and discount rates of infrastructure assets that match actual market prices and therefore measure the changes in market value (the risk!) of these investments. Indeed, we have shown in multiple papers that the pricing of key risk factors that infrastructure companies have in common like size, capex or leverage, largely explains the changes in the fair market value of private infrastructure companies, along with changes in the yield curve.  

Looking at the volatility of the reported appraisals, the shareholders of Thames Water must have thought that it was indeed, low risk. However, as anyone who has looked at a dataset relying on appraisals should know, appraisal do not reflect market values and do not capture the risks to which investors are exposed. How can the idea of Thames being a low risk asset be reconciled with the fact that you can loose 30% of the value of the assets? 

Table 1 also shows the volatility of total returns (in local currency) for Thames water, a benchmark of UK Water Companies and a global benchmark (the infra300 index). We see that water utilities as a whole have volatile returns: this is due to risk factors that drive their expected returns, their large size, variable profits and capex, increasing leverage over time, as well as their exposure to interest rate risk (they are perpetuities after all). Looking at Thames alone, we see that it is even riskier than the benchmark and that its valuation decreased over time when the sector average has increased. It should be clear from this data that Thames has become a higher risk investment overtime and continues to be so looking forward. 

Similar conclusions could have been drawn by investors looking at the actual (marked-to-market) cost of capital data of Thames Water and its benchmark in infraMetrics. Incidentally, this would have helped the firm make its case to Oftwat as well, which uses an extremely low (and artificial) WACC to regulate these companies. Adequately measuring fair market value, using the right yield curve and properly measured risk premia, shows that the writing was always on the wall for Thames shareholders. Were they looking at the wrong numbers? What were their auditors doing? Everyone involved seems to have indulged in the old narrative: “infrastructure is boring”, “there is no need to update the discount rate, as we don’t want to sell”, etcetera, etcetera.   

So is Thames Water just a case of bad management? Looking at the complex structure, the whole-of-business securisation, the second layer of debt on top of that, the management fees extracted by some shareholders via the HoldCo, it is true that investing in Thames does not look like your plain vanilla infrastructure project. But none of this is a secret, especially not for its shareholders.  

Had the same shareholders measured the risks they were taking with respect to a comparable benchmark, like UK Water Utilities, or even the global market as proxied by the infra300 index, they may have behaved differently, asked questions, required preventive action. Instead, they failed to recognise the risks they were exposed to because they did not measure them and therefore could not manage them.  

The irony is that the numbers were available.  

Table 1: Risk & Valuation Metrics for Thames Water Utilities, UK Water, Global Utilities and the infra300® market index  

Source: infraMetrics®