Not Your Father’s Infrastructure (Part 4)
Implications for investors and lenders
This is the fourth and last in Infrastructure Ideas’ series on the impacts of technology on the infrastructure business (Not Your Father’s Infrastructure-overview, Implications for Policy-Makers, Implications for Infrastructure Companies). This post will explore some of major implications of this change for providers of capital.
The headline for lenders and investors is similar to the headline for infrastructure companies: more opportunities, and more risks.
Before deciding this is a completely boring message, let’s step back to where we’ve been for… decades. The starting point is that infrastructure has been a set of sectors where the big issue – for decades — has been opportunities. Infrastructure investors and lenders, more than anything, have been short of assets. All have looked these sectors as ones with attractive fundamentals: long-term stable returns for institutional investors, often taking quasi-sovereign risks for good yield spreads over sovereigns; relatively strong risk-adjusted returns, and as a bonus high societal impact — making this a priority for all IFIs.
The issue for investors and lenders has been never enough deal flow. Given the societal or developmental importance of most infrastructure, public groups have looked for years at how to increase infrastructure investment. G-20 and G-8 reviews of infrastructure in developing countries have consistently come to the same conclusion, pointing to the lengthy and complex process of bringing infrastructure assets to market as a target for policy interventions. One can add to this that much of greenfield infrastructure is historically funded by governments, and in some big economies – especially in Asia – dominated by State-Owned Enterprises.
With the advent of large-scale wind and solar generation, this picture is changing. From a tiny base, continued large cost declines for wind and solar power have driven growth to the point where these two sub-sectors may account for up to 1/3 of global infrastructure investment. In a space where investors and lenders have struggled to find opportunities, wind and solar have added on average some $20-30B in new infrastructure assets. Sure, to an extent, these are substituting for other types of generation investments – thermal generation, especially coal, has been cannibalized to a point by wind and solar. But there are two big elements of this picture that have made wind and solar additive, and not substitutive:
1. Faster growth rates. Governments, and other buyers, have been turning to wind and solar for a number of reasons. One is to meet electricity capacity requirements, as has been the case historically for other sources of power. But in other cases governments have done so for emissions-related policy reasons – whether for climate considerations, or in the case of China in particular, local health concerns – and now many are turning to wind and solar because of their cost advantages relative to alternative sources. Capacity additions for these reasons have not substituted for other options, and so have pushed the rate of new generation investments above what it would have been for simply meeting capacity requirements. Another way to observe this development is to note the advent of oversupply of electricity in several markets. Bottom line, this means more assets for investors and lenders to consider than would otherwise have been the case.
2. More private sector. A notable aspect of the rise of wind and solar (notable and insufficiently discussed) has been that it has been all private sector. Infrastructure generally is owned and operated more by the public sector than the private sector. Power generation has increasingly been more a private sector business, but SOEs still dominate many large markets in thermal or hydropower generation. With the exception of China, the public sector is out of the picture. Wind and solar have been exclusively a matter for private players – who have a much greater ability to understand rapid technology change and its implications, and take advantage of it. PLN in Indonesia is a great example of an SOE dominating generation – but essentially unable to capture the benefits of low cost wind and solar. This shifts assets from the public sector – where investors and lenders cannot finance them – to the private sector – where they can.
Through different dynamics, and to a lesser extent than with wind & solar, the application of technology to natural gas has had a similar effect. Assets for LNG transport and storage are now far more plentiful and available for private financing. Technology change promises to continue delivering new deal flow in other areas in the coming years: energy storage, distributed generation, smart grids, new mobility technologies. In some ways, these technology-driven opportunities represent the fastest increase in available infrastructure deal flow in decades. The big traditional problem for infrastructure investors and lenders is on its way to being solved! Great! Celebrate!
Now for a time-out from breathless enthusiasm.
At the outset of this post, we noted not one big impact from technology coming to infrastructure, but two. Yes, the historical BIG problem, deal flow, is getting unexpected solutions. But… the historically big advantage of these sectors, their long-term stability and low risk, is getting undermined. Let’s look at how this is happening.
The first thought of many risk managers, with new technology, is to worry about technology risk. Completely logical. But not the big problem. Yes, in the early days of wind and solar some turbine gearboxes were prone to problems, and some of the early solar panels delivered less than expected. But most technological problems were sorted out in fairly short order, and non-performing assets related to technology risk have been minimal. We should expect more of the same for the next set of technologies – some early teething issues, but which dissipate fairly quickly.
The two main sources of risk, and loss of stability, for lenders and investors, come instead from compression of returns, and from value obsolescence. Neither of these have been present previously in infrastructure.
The compression of returns on wind and solar generation is observable from the plummeting auction prices worldwide. As bids at solar auctions drop from $0.10 a kilowatt hour only 5-6 years ago to five cents three years ago, to three cents in multiple countries in 2017, to two cents now in Mexico, the news is increasingly good for governments and consumers – cheap clean power. Making money is another story: the total price of power now offered is less than the margins for solar or wind only five years ago. At 20% cost of capital, which characterized some of the early wind and solar transactions, capital charges would run about two cents a kilowatt hour. One therefore has to wonder, if the power is being paid for a two cents, and it costs something to deliver it, what’s left for covering financing costs? The answer: not much. Hence some of the very creative balance sheet and financing approaches being taken by a number of players. The bottom line here: not much left for return on capital, and not much buffer to cover fixed borrowing costs.
Value obsolescence is new to infrastructure. It is a common concept in almost any business exposed to changing technologies – someone comes along who builds a better mousetrap, or a much cheaper one, and does so quickly. Entertainment, music, communications, are fields littered with examples of value which disappeared completely – think of cassette tapes, walkmans, palm pilots and more. This movie is coming to an infrastructure portfolio near you. Coal plants have been seeing this for years now – the number of coal plants being retired annually in the US and China eclipsing the number of new plants ever built annually. Increasingly, early 2000s vintage wind and solar plants are becoming obsolescent – it’s much cheaper and more efficient to replace them with new versions than to keep running them.
Note that this risk is – unfortunately — not limited to technology-intensive assets. While in some ways it is intuitive that technology-intensive assets, say a solar farm, would be exposed to the risk of how fast the technology keeps improving, it is less intuitive – but just as true – that non-technology intensive assets are as exposed to this risk. Technology change has major impacts on non-technology intensive segments through substitution effects – wind and solar displace thermal power, and in cases hydropower plants; new mobility models and technologies affect the use of toll roads, or public transit.
What this means is that portfolio managers for infrastructure assets are in for a lot more headaches than usual. An asset’s position on a merit order dispatching curve will deteriorate faster, and less predictably, over the life of a long maturity loan that before. Contractual protections, notably Power Purchasing Agreements (PPAs) in electricity, will come under much more stress. We have already seen in countries such as Spain, Bulgaria and others that governments will rethink subsidy schemes aimed at promoting adoption of early-stage wind and solar assets. In those cases, the governments were willing to bear the implicit costs of policy discontinuity, and its negative signals to markets, in exchange for stepping away from fiscal arrangements which were no longer comparatively economic. Expect in the future to see other offtakers, government or industrial, question the merits of staying in long-term PPAs when cheaper and cheaper alternatives emerge. Expect, in other words, contracts to become less stable as costs of new alternatives get lower and lower.
Will the same happen in other subsectors as technology begins to affect them, and offer cheaper and/or substitute ways of delivering infrastructure services? Too early to tell, but the dynamics are likely to be the same.
What to do? In the short-run, take advantage of the new asset opportunities. But with eyes wide open. Three elements of risk management may be useful:
1. Risk analysis. Most analysis of infrastructure risk concentrates on contracts. Paying attention to underlying fundamentals, especially market dynamics and the direction of technology, will become more important.
2. Diversification. In a world where there are more infrastructure assets in which to invest, on average smaller-sized individual assets, and assets carrying more risk, diversification becomes more important. An asset affected by a decision of a government to terminate a PPA early, for example has a much lower impact on a diversified portfolio. And diversification is easier through scale – so holding portfolios of few infrastructure assets, especially in sectors affected by technology, will not be the best strategy.
3. Exit options. Today’s infrastructure contracts are designed to last a long time, no matter what. When technology moves fast enough that value becomes obsolete, keeping contracts unchanged over a long period may not make sense. Orderly exits are always better than disorderly ones. So developing mechanisms to mutually reflect market realities may make sense. We’ll explore this idea further in future posts.