The Market

Our market

Demand for Aggreko's services are created by events - our customers generally turn to us when something unusual happens which means they need power or temperature quickly, or there is a requirement which is temporary. Events that stimulate demand range from the very large and infrequent to the small and recurrent.

Examples of high-value, infrequent events or situations we have worked on include:

• Large-scale power shortage - Kenya, Sri Lanka and Venezuela.
• Major sporting events - Olympic Games, Ryder Cup, Super Bowl.
• Natural disasters - Hurricanes Gustav and Ike in 2008.
• Post-conflict re-construction - Middle East, Africa and the Balkans.

Examples of lower-value, more frequent, events on which we might work are:

• An oil refinery needs additional cooling during the summer to maintain production throughput.
• A glass manufacturer suffers a breakdown in its plant and needs power while its own equipment is being repaired.
• A city centre needs chillers to create an ice-rink for the Christmas period.

How big is the market, and what is our market share?

Because we operate in very specific niches of the rental market - power, temperature control and, in North America only, oil-free compressed air - and across a very broad geography, it is very difficult to determine with any accuracy the size of our market. A complicating fact is that our own activities serve to create market demand - Rwanda and the Yemen did not figure highly in our market forecasts in 2003, but are now important customers as a result of our sales efforts. Furthermore, our market is event driven - and major events such as hurricanes in North America, the Olympic Games, or major droughts in Africa can influence market size in the short-term.

As there is no third-party research that exactly matches our business, we have to use a number of different approaches to estimate the size of the global market. All of our measurements of market size relate to rental revenue, as services revenues such as fuel and freight are highly volatile and do not have any reflection on underlying market size.

For most OECD countries in which we operate, we use three techniques:

• Supply-side estimation. We use market intelligence to estimate the supply-side - i.e. how large our competitors are. This is notoriously inaccurate, as competitors often have much broader product ranges. It is extremely difficult to work out how much of their revenue comes specifically from generators and chillers, and how much from the many other lines of equipment they may offer.
• Demand-side estimation. In our Local business, the implementation of our new IT system, and a much sharper emphasis on sector-based marketing, is helping us to develop an improved understanding of our revenue by sector and customer. For our International Power Projects business, we have invested considerable effort in proprietary research with professional economists to develop models which forecast the supply of, and demand for, power.
• Third-party data, where it is available.

By triangulating these techniques, we develop an estimate of market size but the truth is that it is a guess, and probably not a very accurate one. In 2003, we did a great deal of work on market sizing, and came to the conclusion that the market was worth about £1.3bn and was growing at about 5%. Since then, our own rental revenues have grown at a compound annual rate (CAGR) of 22%, which would imply either that our market share has grown improbably fast, or that the original market size was wrong, or that we under-estimated the growth-rate. In all probability, the truth is a mixture of all three factors. Our best guess (and it is just that because we do not feel inclined to spend hundreds of thousands of pounds of shareholders' funds paying someone else to guess) - is that the market in which we operate is worth somewhere between £2bn and £2.5bn per year.

Given our rental revenues of £658 million in 2008, this would imply an Aggreko world-wide share of sales of around 30%. Behind this lies enormous variation. In many developing countries, where the rental market is barely developed, and where we are called in to provide temporary utility power, we may represent 100% of the power rental market for the period of the project but none when it ends. In OECD countries, where the rental markets are better developed, our share of sales will be lower than the 30% we estimate for our global share of sales. However, in nearly all the major markets in which we operate, Aggreko is the largest or second-largest player.

What drives market growth?

Growth in Aggreko's Local business is driven by three main factors:

• GDP - as an economy grows, so does demand for energy.
• Propensity to rent - how inclined people are to rent rather than buy. This is driven by issues such as the tax treatment of capital assets and the growing awareness and acceptance of outsourcing. And in times of recession, people can be more inclined to rent rather than commit capital to purchase.
• Events - high-value/low-frequency events change the size of a market, although only temporarily. For example, the scale of Hurricanes Gustav and Ike in 2008 led to a short-term surge in temporary power demand in the areas affected by the hurricanes.

In seeking to understand the drivers of growth better, we have devised the concept of 'Aggreko GDP'; this is the GDP of a country weighted to account for Aggreko's sectoral mix of revenues. Typically, this means that we are weighted more towards manufacturing than, say, financial services. Over the past few years, we have observed that in countries where the growth rate of Aggreko GDP is below 5%, our revenues tend to grow a few percentage points faster than Aggreko GDP. In economies where Aggreko GDP growth is above 5%, we get an increasingly leveraged effect, with Aggreko sales growth far outpacing GDP growth. This is for a number of reasons, but most notably, simply that when economies are growing fast, customers want equipment quickly, they want high levels of service, and they want to focus on doing what they are good at, rather than owning large amounts of equipment.

The graph below plots this relationship between growth in Aggreko's revenues by country and growth in Aggreko nominal GDP. We would caution that these figures include the impact of the GE Energy Rentals acquisition in December 2006 which will exaggerate the underlying sales growth in some countries, but we feel that the trend they show is directionally correct.

Overall, in times of positive GDP growth, we estimate that the market addressed by our Local business for the short-term rental of power and temperature control is growing at some 2-3% above GDP in developed markets. So, if GDP grows at 3% on average over the cycle, our market should grow at 5%. In countries with rates of nominal GDP growth that are above 5%, the market can grow much faster.

As the global economy slows, and GDP growth in some countries goes negative, the question arises as to what happens to demand when the markets we operate in are in recession. Since none of the markets we operate in have been in recession since we re-organised our Local business in 2003, we do not have the richness of data to analyse what happens to demand when GDP growth goes negative. Our working assumption is that the relationship between Aggreko GDP and our revenues also works in reverse - in other words, in times of negative economic growth, our revenues will fall faster than the economy as a whole. The impact on profits could be material, as operational gearing and the margin enhancement we see when revenues are rising would likely reverse. In these situations the benefits of geographic diversity are especially important, as we can shift investment and fleet from under-performing to over-performing economies and sectors. In addition, our broad mix of customers spread across multiple industries means that we are not overly reliant on any single sector.

The factors which drive the growth of our International Power Projects business are different. The main trigger of demand is power cuts: when the lights go out in a country, people want power restored as quickly as possible. It is a perverse fact that people value power most when they are without it. We believe that in many parts of the world, and most particularly in many developing countries, there will be increasing numbers of power cuts caused by a combination of burgeoning demand for power and inadequate investment in new capacity. We believe that demand for power is going to grow much faster than is commonly believed: working with a leading group of professional economists at Oxford Economics, we have built a model which takes data on GDP and population growth, power consumption, and power generation capacity for 120 countries over the last 10 years. Using this historical data, it then projects future power demand based on forecasts of population and GDP growth. Our model predicts that world-wide demand for power will grow by around 5% per annum between now and 2015, compared with forecasts by the International Energy Agency (IEA) of around 3.5%. Our model reflects the sharp divergence between the growth in power consumption between OECD and non-OECD countries in recent years, as shown in the graph below.

The rapid growth in power consumption in developing countries is driven by industrialisation and by the growing number of consumers having access to devices which consume electricity, such as fridges, televisions and mobile phones. Between 2000 and 2006, the number of people whose power consumption is growing faster than per-capita GDP increased by 1.3 billion to 4.3 billion, representing nearly 70% of the world's population. The vast majority of these people live in developing countries, where investment in the acquisition of new generating capacity and maintenance of existing capacity has been woefully inadequate.

To make this situation worse, by 2015, 25% of the world's installed power-generating capacity will be over 40 years old, which we believe is a reasonable proxy for the average life of a permanent power plant. So the next ten years will see the beginning of a replacement cycle during which the majority of existing power-plant construction capacity will be dedicated to replacing existing plants in OECD countries, rather than building replacement or additional capacity in developing countries.

Our models predict that the combination of these demand-side and supply-side factors will increase the world-wide shortfall of power generating capacity from about 70 gigawatts in 2007 to between 400 and 1,000 gigawatts by 2015. The ultimate size of the shortfall will depend on both the rate of increase in demand, and the net additional generating and transmission capacity brought into production during the period. However, even if the shortfall is at the bottom end of our forecasts, it will still represent a level of global power shortage many times larger than today's. We are confident that such a level of power shortage will drive powerful growth over the medium and long term in demand for temporary power.

What impact could a recession have on demand for our International Power Projects business? It is certainly likely that lower rates of per-capita GDP growth will lead to slower rates of growth in demand for electricity in developing countries. However, we believe that, unless there is a prolonged economic catastrophe, the market for temporary power in developing countries will continue to grow. Our reasoning is that the current macro-economic environment will create a number of conflicting forces. On the one hand, lower rates of economic growth will reduce demand for power in developing countries, and particularly in countries where mining activity represents a high proportion of grid consumption (a small part of our customer base). Also countries which are highly dependent on oil revenues will find themselves poorer and less able to support high levels of infrastructure investment. On the other hand, our customers, who are the power utilities, will find that the lower cost of fuel makes temporary power generation much more affordable. On the supply side, reduced demand for replacement generation capacity in North America and Europe may shorten lead-times for equipment. However, we believe that this improvement in the supply-side equation will be dwarfed by the very severe impact that the current macro-economic environment has had upon the ability of developing countries to raise finance for major infrastructure projects. At a time when many businesses in OECD countries are finding it hard to raise funding, it has become almost impossible for most power utility businesses in developing countries to contemplate raising the finance for sizeable power generation projects, which require hundreds of millions of pounds of upfront investment and generate returns over 20-30 years.

We therefore believe that, even in the face of challenging macro-economic circumstances, many power utilities in developing economies facing chronic supply shortages will continue to fi nd Aggreko's offering compelling. Their choices are limited: continue to suffer chronic power shortages; wait until financing new permanent power becomes possible - and then a further five to ten years to construct the plant before the first watt of electricity can be generated; alternatively, contract with Aggreko and have the lights back on in a matter of weeks.

Probably a greater risk than lack of demand in our International Power Projects business in times of recession is that some of our customers' behaviour may become more unpredictable, and their payment habits more volatile. Historically, we have made very high returns on capital from our International Power Projects business, partly because we tread where others are fearful to do so. We always advise investors to regard these returns as a reward for the fact that their assets are at greater risk of loss or impairment than they would be if they were sitting in the suburbs of London or New York. At some point, the risks may crystallise into an actual loss of monies and, or, equipment, and that point is more likely to arise at a time of economic stress.