Figure 1 shows the historical performance of the considered asset classes for the whole time frame. It becomes evident that apart from emerging markets, the wind farm has outperformed all other asset classes in the analysed time frame. This is mainly due to the sharp decline of all asset classes except for bonds during the sub-prime crisis starting in September 2007.
Using this graph as a first impression, the next step was to have a closer look on the return distributions. The asset classes therefore underwent a further statistical analysis, which revealed that the wind farm undoubtedly offered the best risk-return profile. This was not only due to its stable development as it is depicted in Figure 1. The wind farm also exhibited the highest Sharpe Ratio and relatively low downside potential compared to the other asset classes. The only asset class with lower value at risk and conditional value at risk figures during this time frame were bonds.
We have learned from Markowitz that an investor can mitigate an individual asset’s risk by building a portfolio with low (or in the best case negatively) correlated assets. Therefore, the next step in the analysis was to have a closer look at the correlation structure of the asset classes. It turned out that only bonds offered the desired negative correlation with other asset classes. As expected, correlations between the wind farm and the other assets were close to zero. Apart from commodities with values between 0.3-0.6, all other asset classes had fairly high correlations in the range of 0.7-0.9.
Having examined the statistical properties and correlations of the return series, the asset classes were then undertaken a portfolio optimisation in the style of Markowitz. The results from this optimisation showed that both the wind farm’s weights and its contribution to the optimal portfolios were remarkable. In case of the portfolio with the lowest risk (“minimum variance portfolio”), the wind farm had a weight of 21%. For the portfolio offering the best risk-return profile (“tangency portfolio”), the weight for the wind farm even amounted to an astonishing 40%. Although these optimal portfolios clearly stood against the asset allocation of many institutional investors, the results appeared to be plausible for the chosen downturn period and set of assumptions.
Diversification possibilities within a renewables portfolio
After the diversification potential of wind investments in a multi-asset portfolio has been explored, the next step in the study is to discuss diversification possibilities within a renewables portfolio. This could have been achieved by analysing a portfolio of nine German wind farms, one Italian farm and one from France for a time frame of three years.
Following the Markowitz framework again, the first step was to find out how the returns of the different wind farms in the data set move together. Not surprisingly, the correlations between the German wind farms turned out to be all very high, with values between 0.7-0.9. Only one German wind farm which was located at the edge of the German North Sea had lower correlations of around 0.4 with the others.
The results for the other two countries, however, looked promising. Both the Italian and the French wind farm had extremely low (sometimes even slightly negative) correlations with the German ones and therefore offered a fairly high diversification potential. Having assessed the general diversification potential of the wind farms, the next question was how much risk can be diversified away by holding a small portfolio of wind farms instead of a single one. This question was approached by applying a “CAPM-style” model to the data set and using the resulting betas as a best guess measure for risk. The main challenge in applying this model, however, was to find an appropriate benchmark for the usual market portfolio.
Since there does not exist a total return index for wind energy, something similar had to be created. The analysis followed the idea of Dunlop (2004) and developed a market portfolio for the wind farm market that consisted of all the 11 wind farms in the data set, weighted by their total installed capacity in the respective country and year.
Following the capital asset pricing model (CAPM) approach, a portfolio of five wind farms was then built up step-by-step by adding wind farms with low betas to the riskiest one – the wind farm with variable market prices in Italy. The impact of diversification is illustrated in Figure. Initially, the Italian wind farm had an annual standard deviation of the SAARs of 5.74%. By adding the French wind farm and another three German ones to the small portfolio, the actual risk could have been reduced to 2% – a risk reduction of 65% in relative terms.
The final area of diversification within a renewables portfolio concerned the hedging benefits of wind and solar. Since the wind blows stronger in winter and the sun shines stronger in summer, the returns of these two investments should move in opposite directions, or “counter-seasonal”. Thus, the resulting portfolio when investing into comparable solar and wind parks at once should be naturally hedged, or to put it differently, any seasonal fluctuations within the year can (theoretically) be diversified away.
The relatively small and new asset class of renewables seems to off er attractive diversification possibilities, not only within the asset class itself, but also when including them into a traditional portfolio. It remains to be seen whether these findings will lead to winds of change in asset allocation.
Fredrik Bruns is an analyst at Allianz Capital Partners
1) Bloomberg New Energy Finance (2011): Global Trends in Renewable Energy Investment 2011, www.bnef.com.
2) European Wind Energy Association (2009): Wind Energy – The Facts, Earthscan, London.
3) Markowitz, H. (1952): Portfolio Selection, Journal of Finance 7 (1), 77-91.
4) For an overview, see Inderst, G. (2010): Infrastructure as an asset class, EIB Paper Series 15
Comments