Building benchmarks: A global index for private infrastructure debt and equity

A standardised database

To create this sampled universe of investable infrastructure companies, data is collected about each individual firm. This requires a standardised framework that captures relevant aspects of infrastructure companies' finances such as their financial structure, cover ratios, free cash flow, etc, so new valuation models can be developed (see A modern approach to asset pricing, below).

The Global Infrastructure Investment Data Standard (GIIDS) is part of the effort to standardise performance reporting for the infrastructure asset class. The standardisation of the data collection allows for more transparent comparisons of the performance of unlisted equity and debt infrastructure investments between asset classes and, as a result, the increased transparency created by this data standard will strengthen the development of the infrastructure asset class.

For this purpose, the GIIDS identifies information that is:

• reasonably available in a standardised format
• a subset of all information available to investors that can be collected most of the time
• focused on the key processes of a private infrastructure company's performance, i.e. it is not overly specific to different types of infrastructure.

This approach has led to the creation of the largest database of infrastructure investment data in the world, containing data taken from hundreds of firms over the last 20 years. The project was undertaken in coordination and collaboration with the infrastructure industry, banks, asset owners and managers – a clear indication that that all stakeholders are receptive to a more transparent benchmarking process.

A modern approach to asset pricing

Measuring the value and risk in unlisted infrastructure investment can raise significant methodological issues. The capital asset pricing model uses stock market indices as listed proxies, but fails to capture the specific characteristics of infrastructure companies (bit.ly/EDHECfakeinfra). Also, it is almost impossible to find direct equivalents (or comparable information) for either unique assets or assets that rarely trade.

The Unlisted Infrastructure Asset-Pricing Methodology has been developed to respond to this challenge by deriving a 'fair value' framework to measure the performance of unlisted infrastructure.

Each quarter, implied excess returns are derived from actual transaction prices and forecasts of dividend or senior debt payments. These expected returns represent the aggregate market price of risk for a cross-section of equity or debt investments at that point in time. They are then decomposed into multiple risk factor premia (e.g. the size or leverage risk premia) using a cross-sectional regression, a type of regression in which the variables are associated with a sequence of points in time.

Each risk premia in a given period can then be applied to all relevant investments at that time, whether they are traded or not, using each company's factor loadings (e.g. its actual size or leverage, the correlation coefficients between the variables and factors) to derive a market-implied discount rate for each investment.

As long as there is enough data, this method of factor pricing addresses any biases in transaction data and the uniqueness of each transaction by deriving average valuations from the risk factor price of actual transactions, in line with IFRS 13 guidelines.

Risk-adjusted indices of performance

Finally, the global index for private infrastructure debt and equity will produce indices by using data from the individual performance of each asset captured in the sampled universe. The data and technology used allow the re-pricing of hundreds of individual assets through time, using actual transaction prices to recalibrate expected returns and discount rates. This approach uses market inputs, thus avoiding the smoothing of returns caused by appraisal valuations and providing a genuine fair value assessment of performance at the index level.

Unlike other private asset indices, which only report an average internal rate of return, these new indices allow for the computing of a return covariance matrix and the effect of diversification and can provide advanced risk metrics such as index volatility, value-at-risk and risk factor prices.

Much work remains to be done, but the strong support of EDHECinfra's partners in more effective documenting of the performance of infrastructure investments means that momentum for this benchmarking project is expected to continue into 2020, and beyond. As more infrastructure investors pool their data and there is clearer understanding of the infrastructure investment sector, benchmarking results will improve and more valuable investment solutions in infrastructure can be designed.