By Tim Whittaker, Associate Research Director, EDHECinfra
A review of the paper “Common risk factors of infrastructure investments”, by Semir Ben Ammar, Martin Eling, in Energy Economics 49 (2015) 257-273
Most academic papers continue to use listed equity data to try and identify an ‘infrastructure asset class’ but this quest remains as elusive as ever.
In this paper, Ammar and Eling set out to understand how the infrastructure investment narrative is reflected in the returns of listed infrastructure. They develop several hypotheses of the nature of infrastructure investments: infrastructure assets are expected to exhibit low correlations with other asset classes, acts as a hedge against inflation risk, exhibits low cash flow volatility and possesses a degree of protection during poor market states.
Using data for U.S. listed firms for the period 1983 to 2011, Ammar and Eling identify 396 infrastructure companies in the industries of Transportation, Telecommunications and Utilities. The identification of infrastructure assets follows the approach of Rothballer and Kaserer (2012), in that they use industrial classification. The authors state that the identified infrastructure assets own physical assets and are not service providers or construction companies.
They create 18 test portfolios of infrastructure assets, sorting the infrastructure stocks into size, book-to-market, beta, momentum and industry specific portfolios.
They then proceed to build a number of ’factors’ based on prior academic research in an at- tempt to explain the returns of the infrastructure firms they have identified. The factors consist of:
- The broad market factor from the Capital Asset Pricing Model (CAPM);
- The Fama and French (1993) Size and Book-to-Market factors;
- The Carhart (1997) Momentum factor;
- The Huang (2009) Cash flow volatility factor;
- A leverage factor, of their own construction;
- The Titman et al. (2004) and Xing (2008) Investment factor;
- The Fama and French (1993) Term and Default bond market factors; and,
- The Pastor and Stambaugh (2003) Liquidity factor
Furthermore, Ammar and Eling state that because infrastructure can be considered an alternative asset, then three alternative asset factors for bonds, commodities and currencies from the work of Sadka (2010) should be included to study infrastructure returns.
They employ six different asset pricing model specifications in their examination of the variation in the returns of the infrastructure stocks. These include classical asset pricing models such at the CAPM and the Fama and French (1993) and Carhart (1997) asset pricing models. But also include more recent models such as Sadka (2010) and their own infrastructure asset pricing model, which is an augmentation of the Carhart (1997) asset pricing model.
They find that all models can explain the variation of the infrastructure portfolio returns to varying degrees. The CAPM explains the lowest proportion of infrastructure returns whilst the other augmented models are able to explain up to 80.5% of the variation in the infrastructure portfolio returns. The authors conclude that the addition of further factors to asset pricing models improve on the explanation of the variation of returns (tested using the Hansen and Jagannathan distance).
Finally, Ammar and Eling test other claims of the infrastructure narrative (Blanc-Brude 2013): that it provides an inflation hedge and the natural monopoly characteristics. They find no strong support for the claim that infrastructure investment can act as a hedge against inflation but do find that firms with natural monopoly characteristics (namely a higher level of industry concentration) do produce better risk-adjusted returns. This is consistent with some of the arguments for infrastructure investment.
This paper demonstrates that listed infrastructure is able to be explained by a combination of factor exposures. These findings are consistent with prior research conducted by Bianchi et al. (2014) and Bird et al. (2014). Furthermore, the lack of strong empirical support to the claim that infrastructure provides an inflation hedge is consistent with both Bianchi et al. (2014) and Bird et al. (2014).
Ammar and Eling provide further support to the arguments that listed infrastructure is not a stand alone asset class, rather it is an industry tilted stock picking strategy.
Despite the significance of the paper, there exists some areas we believe that could be improved
- The paper employs an infrastructure identification strategy based on an industrial screen. However, as argued in Blanc-Brude (2014) and found in the recent EDHECinfra/GIH investor survey (Blanc-Brude et al. 2016), investors in infrastructure do not think this way. Instead, institutional investors think of infrastructure in terms of contracts or the natural monopoly characteristics of the assets. In this survey, the industrial classification of the infrastructure asset is ranked third in importance (out of four) for infrastructure investors for the identification and classification of infrastructure assets;
- Ammar and Eling do not explain whether they control for possible infrastructure firms existing in the sample when they construct the leverage and investment factors;
- In the regression results presented, there is no mention of the Carhart (1997) regression results. Therefore, we are unable to compare these results with the other asset pricing models examined; and,
- Ammar and Eling state that the test Hansen and Jagannathan distance test results show that their infrastructure asset pricing model improves on the more traditional asset pricing models. However, this relies on non-traditional measures of statistical significance (10%). If more traditional measures of statistical significance are used (5% or less), then this conclusion cannot be supported.
Overall, this paper provides an interesting addition to the understanding of listed infrastructure investment. In particular, it is in line with our own conclusion in Blanc-Brude et al 2017 (forthcoming), that listed infrastructure is not a stand-alone asset class but can be replicated using existing and known factor exposures.
However, as a result of the weaknesses, we are unable to fully agree with their conclusion that they have constructed a factor model that explains the returns of infrastructure investments.
- Bianchi, R. J., Bornholt, G., Drew, M. E., and Howard, M. F. (2014). Long-term U.S. infrastructure returns and portfolio selection. Journal of Banking & Finance, 42:314–325.
- Bird, R., Liem, H., and Thorp, S. (2014). Infrastructure: Real Assets and Real Returns. European Financial Management, 20(4):802–824.
- Blanc-Brude, F. (2014). Benchmarking Long-Term Investment in Infrastructure. EDHEC-Risk Institute Position Paper.
- Blanc-Brude, F., Chen, G., and Whittaker, T. (2016). Towards better products for infrastructure investors? a survey of the perceptions and expectations of institutional investors in infrastructure. EDHEC Infrastructure Institute Publications.
- Blanc-Brude, F., Whittaker, T. and Wilde, S. (forthcoming). Searching for a listed infrastructure asset class using mean-variance spanning, Financial Markets & Portfolio Management.
- Carhart, M. M. (1997). On persistence in mutual fund performance. The Journal of finance, 52(1):57–82.
- Fama, E. F. and French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1):3–56.
- Huang, A. G. (2009). The cross section of cashflow volatility and expected stock returns. Journal of Empirical Finance, 16(3):409–429.
- Pastor, L. and Stambaugh, R. F. (2003). Liquidity risk and expected stock returns. Journal of Political Economy, 111(3):642–685.
- Rothballer, C. and Kaserer, C. (2012). The Risk Profile of Infrastructure Investments: Challenging Conventional Wisdom. The Journal of Structured Finance, 18(2):95–109.
- Sadka, R. (2010). Liquidity risk and the cross-section of hedge-fund returns. Journal of Financial Economics, 98(1):54–71.
- Titman, S., Wei, K. J., and Xie, F. (2004). Capital investments and stock returns. Journal of financial and Quantitative Analysis, 39(04):677–700.
- Xing, Y. (2008). Interpreting the value effect through the q-theory: An empirical investigation. Review of Financial Studies, 21(4):1767–1795.