Lighting up the dark – new tools to value mature unlisted equity: the particular case of long-term unlisted infrastructure equity

Various assets, from artworks to intangible assets and financial contracts, are hard to value (HTV). Unlisted infrastructure is one such HTV asset. Unlisted infrastructure is hard to value due to a lack of transparency, liquidity, data and information sharing among unlisted infrastructure equity participants. The market has relied on limited tools to make investment decisions without accurate valuation techniques. Simple historical and forward multiples and the Capital Asset Pricing Model to determine the cost of equity combined with discounted cash flow models have dominated valuations. We know these are very blunt techniques, but in the absence of anything better, they continue to be widely used, even if often not widely followed in practice. Challenging these approaches and assumptions could have material impacts on the valuation of unlisted infrastructure investments, measuring risk and return, building investment portfolios and establishing infrastructure as a distinct asset class. This research aims to enable market participants to more accurately and transparently measure and price equity risk of unlisted infrastructure equity using data available to investors. This research attempts to develop a clear and transparent model of ex-ante unlisted infrastructure equity valuation, risk, return and correlation that translates directly with listed equities utilising information available to public markets. The research will utilise information sets and assumptions of greater transparency, liquidity and certainty to estimate markets with lower transparency, liquidity and certainty synthetically. Specifically, the research utilises data on critical attributes of infrastructure assets from listed markets to develop estimates and assumptions about unlisted infrastructure equity. If the model parameters are accurate, the data relevant and available to investors in more liquid and complete markets should enable us to get better pricing for those assets in incomplete markets, especially data on actual and forecast cash flows. The thesis develops a model for valuing unlisted infrastructure that builds on recent innovations in asset valuation, especially the work on the implied equity risk premium as a forward-looking proxy for the expected rate of return. The model achieves various conceptual and empirical goals challenging for researchers and practitioners in the infrastructure space.  One achievement is that the model seeks to “solve” equity risk premiums rather than excess returns. The position of this thesis is that equity risk premiums, proxied by the Implied Equity Risk Premium (IERP), are not bound to the restriction of equivalence to actual returns. Instead, the IERP is viewed as a “charge” or “cost” imposed on expected cash flows for risks related to those cash flows. In the model, all predictors are derived from information sets available to unlisted infrastructure equity investors (i.e. do not require market pricing) and are forward-looking. The model is developed and, therefore, integrated and consistent with valuation models proven to be less reliant on hazardous Terminal Value calculations outside the realistic forecast horizon (building from the AEG model of Claus and Thomas 2001). A vital feature of the model is the use of information about attributes of liquid assets with similar characteristics to unlisted infrastructure assets to derive valuation information. Therefore, any regression approach using attributes of listed market data must be consistent with that goal. The approach adopted here is cognizant of this class of unlisted assets' valuation challenges, and the thesis outlines how these challenges are incorporated into the valuation model. The model is tested on Australian equities using a non-parametric Generalized Additive Modeling with data supplied by Refinitiv Eikon. The testing results suggest that the model significantly reduces absolute value-weighted errors of unlisted assets such as infrastructure. The thesis concludes with some comments about model development and suggestions for further empirical research.