Resilient energy infrastructure: A finance view

The energy industry will face many changes over the next few decades, many of them more disruptive than anything the industry has faced over the past few decades.

(For the purposes of this article I will define the energy industry very broadly, covering everything from the exploration and production of fossil fuels to the distribution and transmission of electricity). The reasons for those disruptive changes range from concerns over pollution and potentially catastrophic climate change to the impact of cost reductions in new(er) ‘alternative’ technologies and changing consumer and regulatory preferences.

However, they also include second order effects from the rapid changes in information and communications technology which have the potential to turn passive price-taking consumers into active value-optimising traders, fundamentally changing the nature of energy supply and competition – both in regional energy markets and industries and on a global scale.

These changes have implications for resilience. Resilience, generally seen as the ability of a component or a system to continue functioning after experiencing adverse events, is a complicated topic in itself. My focus is on the interaction between the changes I foresee in the global energy business, their impact on the resilience of that business (or more specifically, on key functions it fulfils like supplying electric power and fuel for heating and transport) and finally on the way a financier might look at the industry and its resilience in the light of current and future challenges.

Financing of resilience

First a comment on ‘financing resilience’. Resilience is an outcome of choices made in building components and constructing systems or networks. Those choices affect the ability of a component to continue to function during and after an adverse event (think of flood defences around a power station or a refinery, or siting such installations away from earthquake zones and building distribution systems to reach consumers), but they can also include redundancy (more and smaller power stations distributed over a greater area and connected by a transmission grid can be more resilient than few larger power stations).

In general, making a system more resilient will tend to increase the upfront investment, and in a world where many of the components are privately financed and operated, that incremental cost will be assessed for the incremental returns it generates. In theory, lowering the risk of a business should lower the cost of financing that business but it in practice the link is not always clear, not least because measuring the risk reduction from increased protection against natural disasters or climate change is not straightforward. Regulators generally demand a level of resilience as part of a license to operate; going above that tends to lead to questions about unnecessary ‘gold plating’ and higher costs to consumers which is never very popular.

The utilities sometimes tell their banks ‘you give us no credit for lowering the risks in our business’. Sadly, that’s probably at least partially true: it is difficult for financiers to identify any incremental cash flow from increased resilience in the base case, or even in a downside case. At best, the financier generally looks to the insurance industry to cover the risks that resilience addresses, and might look to the insurance premium as the mechanism that rewards higher resilience.

Economics of resilience

Back to my main argument about the way changes in the industry will affect resilience and the economics of resilience. The current configuration of the energy industry is that supply is generally the business of a few large companies and a limited number of large components: exploration budgets are measured in billions, and upstream oil and gas platforms tend to cost tens or hundreds of millions of dollars (and more in ultra deep water), putting them out of reach of all but the largest of companies.

In fact, most of the world’s oil and gas reserves and production are under the control of national energy companies, with private companies being relatively small players. There are probably fewer than a thousand companies that make up the majority of the supply side. Consumers, on the other hand, range from very large companies to individuals and number in the billions, globally. Prices do follow demand and supply, but generally in a combination of fast and slow movements: many prices are the result of daily trading of financial contracts linked to physical commodities, but much of the trade in energy is governed by long term supply and off take agreements, often with fixed prices or prices that adjust more gradually, especially for smaller consumers. Improving resilience in such a system means convincing those many consumers to accept a higher average cost in return for a lower likelihood of interruptions, generally with the intermediation of a regulator.

However, the changes that are happening in the energy business change this picture quite fundamentally, both on the supply and the demand side. Supply is changing when individuals and companies can install solar panels on their rooftops and sell the electricity into the grid, when technologies like wind, small scale hydro and more exotic options have very different characteristics like being dependent on wind or sunshine. Demand is changing when smart meters become commonplace, when refrigerators monitor electricity prices to optimise the cost of cooling, when batteries become available either in vehicles hooked up to the grid or installed in homes, factories and shopping malls to perform similar optimisations. Instead of a few large suppliers and many small consumers the world will move to a mixture of a few very large suppliers and many very small suppliers, and not many consumers but many millions of energy traders. The future customer is not a business or a consumer but an algorithm that constantly compares different options for managing the energy demands of a factory, a household or a vehicle.

What does that mean for resilience? Increasing the number of components will tend to increase resilience (the failure of a single component has less impact on the system) unless the complexity of the system introduces unpredictability. Highly dynamic non-linear systems can exhibit behaviour that looks chaotic: certainly financial markets with algorithmic trading show signs of that, and a hyper-connected energy system with millions of algorithms may do the same. Regulators will need to adapt, too, in order to ensure that price signals from a hyper-connected energy system remain relevant both for short term supply decisions and for longer term capacity investments.

The difficult pricing of resilience

Turning to the financing aspects, it remains the case that it is difficult to identify incremental expected cash flows, or reductions in the risks to those cash flows, from improved resilience in a way that can be translated into lower financing costs. In a world of distributed supply and continuous trading by consumers that difficulty only increases. A complex analytical model for a large power station that includes a valuation of the resilience by using state of the art ‘real option’ pricing models is feasible, at least in theory, because the financing is large and the fees to the financier can cover the cost of building such models. Explaining that model to other banks and capital market investors who join in the financing is costly but again, fees might cover those costs. For distributed very small generating systems that cost quickly becomes prohibitive. It is much easier to see how each component is specified at a minimum level of operational resilience, relying on the redundancy of ‘the system’, than to picture a world where each component includes its own contribution to resilience: there is no easy way to provide incremental returns to the more resilient component unless and until an adverse event happens.

So what is the answer? I will not pretend to have an answer but it is possible to identify a few attributes that an answer will have. An approach that collects a small contribution from many components in return for providing the connectivity and backup that generates resilience at the system level looks promising, and looks tailor made for an insurance solution. Financing components like small scale generation, or smart meters, or distributed storage, looks like a classic case for securitisation: take many fundamentally similar loans and bundle them together to benefit from the inherent diversification can lower the financing costs, and requires little more than relatively standardised financing terms. How those two elements interact and cooperate with regulators (and perhaps consumer advocacy organisations or mutually owned trading businesses) is a question that the global finance industry needs to address together with the global energy industry.

Jan-Peter Onstwedder

Managing Director of Global Industries and Regional Markets, Risk Management of Citibank

Jan-Peter Onstwedder has over 20 years experience in risk management across a wide variety of asset classes and global trading markets, both in a banking and a corporate environment. He has held various risk and trading positions for Barclays Bank in London and New York; worked as Head of Market Risk for the Royal Bank of Scotland; Head of Risk for BP’s Integrated Supply and Trading business; and Head of Risk Management for 3i plc.

Currently he works for Citibank in London as Head of Risk Management for the global commodities trading division, and as Head of Credit Risk for four global industries (power, energy, chemicals and mining).

In 2007 he managed the London Accord, the then-largest ever collaborative research project into the financial aspects of climate change. The report, published in December 2007, placed research by leading investment banks, NGOs, law firms and academic institutions in the public domain.