Long horizon investing in infrastructure

Transcription

1 Long horizon investing in infrastructure 1/29 Long horizon investing in infrastructure The journey from investment beliefs to investment delegation and benchmarking Frédéric Blanc-Brude, PhD Research Director EDHEC-Risk A presentation prepared for the June 2015 ICPM Discussion Forum 10 June 2015

2 Long horizon investing in infrastructure 2/29 Seminar structure 1 Investment beliefs 2 Agency 3 Benchmarking 4 Conclusions 5 Table assignment

3 Long horizon investing in infrastructure 3/29 Section outline 1 Investment beliefs 11 2 Questions 12 The infrastructure narrative 13 Known evidence 14 Nature of infrastructure assets 2 Agency 3 Benchmarking 4 Conclusions 5 Table assignment

4 Long horizon investing in infrastructure 4/29 Investment beliefs Two questions: Q1: Are the investment characteristics of infrastructure assets (debt or equity) driven mainly by the industrial function (water supply, transport, etc) performed by the relevant tangible infrastructure? Q2: Should infrastructure assets be expensive (ie low yielding)?

5 Long horizon investing in infrastructure 5/29 Investment beliefs: the infrastructure investment narrative The infrastructure investment narrative (Blanc-Brude, 2013): the set of beliefs commonly held by investors about the investment characteristics of infrastructure assets The role of infrastructure in modern societies and the nature and scale of the required capital investments inform the following intuition: investors in infrastructure are less exposed to business cycle because of the low price-elasticity of demand of the services they provide and because the value of their investment is mostly determined by income prospects that extend far into the future and is little impacted by current events Sunk infrastructure investments thus create long-term value: they are mean-reverting assets par excellence In other words, this ideal-type implies Improved diversification Better liability-hedging Less volatility than capital market valuations

6 Long horizon investing in infrastructure 6/29 Investment beliefs: evidence in the listed space In a new paper (Blanc-Brude et al, 2015), we test the mean variance spanning properties of different stock selection schemes supposed to represent infrastructure We use rule-based schemes (eg stocks in the utilities sector that derive 90% of the income from that sector) and infrastructure indices created by index providers We find zero evidence of improvement of the mean-variance universe of a well-diversified investor: listed infrastructure is already spanned by existing asset classes or factor exposures When it is not, the effect is not persistent after the GFC (so it was just leverage) Other papers support this conclusion and also find that listed infrastructure tends to be more volatile (Newell and Peng, 2007, 2008; Rothballer and Kaserer, 2012), not to have better inflation hedging properties than the stock market (Rödel and Rothballer, 2012) and even to underperform the market after controlling for Fama-French factors (Bird et al, 2014)

7 Long horizon investing in infrastructure 7/29 Standard Deviation Standard Deviation Standard Deviation Standard Deviation Are infrastructure stocks already spanned? see Blanc-Brude et al (2015) 020 Efficient Frontier January 2000 to December Efficient Frontier January 2000 to December Returns 005 Hedge Funds Fixed Interest 90% Utilities Developed Equities Real Estate Emerging Equities Returns 005 Size Term Value 90% Utilities Commodities US Stocks Europe Stocks Default Reference Assets Reference and Infrastructure Assets Reference Assets Reference and Infrastructure Assets 020 Efficient Frontier January 2000 to December Efficient Frontier February 2007 to December Returns 010 High Yield US Equities FTSE Macquarie US Infra Real Estate Returns 010 Term FTSE Macquarie US Infra Market 005 Corporate Bonds Hedge Funds Government Bonds World Ex US Equities Size Default Value 005 Commodities Reference Assets Reference and Infrastructure Assets Reference Assets Reference and Infrastructure Assets

8 Long horizon investing in infrastructure 8/29 Investment beliefs: evidence in the unlisted space The only performance data available is through PE databases, which track PE-like funds including those investing in infrastructure So existing research on private infrastructure funds is biased towards PE-like structures Same issues then other PE funds documented in the literature (Jenkinson et al, 2013; Phalippou and Gottschalg, 2009) Most infrastructure funds today really are large buy-out PE funds and unsurprisingly they are found to perform like PE funds see Bitsch et al (2010), Newell and Peng (2007), Bird et al (2014): Deals are large by PE standards but cash flows are not less volatile: infrastructure PE funds are not less risky than they PE control group Portfolios are limited in size and performance is very heterogenous from one manager to another: asset selection matters Market correlations are found to be low (but not lower than PE) but are shown to increase in time because of the correlation with the credit cycle driven by the use of fund leverage No evidence of inflation protection at best this is a diversifier of other PE investments of course there are other ways to invest in infrastructure equity than PE funds! (but today they are not documented)

9 Long horizon investing in infrastructure 9/29 Investment beliefs: nature of infrastructure assets The performance drivers of infrastructure investment is not about tangible assets, in fact infrastructure investment has nothing to do with real assets! Return to question 1: Are the investment characteristics of infrastructure assets (debt or equity) driven mainly by the industrial function (water supply, transport, etc) performed by the relevant tangible infrastructure? Contracts (and regulation) are everything because they determine the expected value and the volatility of cash flows The different types of risk transfer found in the contractual set up of infrastructure projects are found to explain the cost of capital (cost of debt) These different factors of risk are priced and they are the main source of potential diversification between infrastructure projects see Blanc-Brude and Strange (2007); Blanc-Brude and Ismail (2013); Blanc-Brude (2013)

10 Long horizon investing in infrastructure 10/29 Investment beliefs: the PFI portfolio 5 listed firms mostly occupied with buying and holding the equity and subordinated debt of PFI (private finance initiative) project companies in the UK Table: Total returns of the PFI portfolio, an infra index and the market, pfi portoflio ftse all shares macquarie infra europe ann return ann risk ann Sharpe ratio all price data in GBP, Datastream Table: Maximum drawdown* of annualised returns, pfi portoflio ftse all shares macquarie infra europe Worst Drawdown *Measured as a percentage of maximum cumulative return ie from peak equity

11 Long horizon investing in infrastructure 11/29 Investment beliefs: the PFI portfolio total returns: growth of $1 020 Efficient Frontier January 2000 to December Value Returns UK Gilts PFI Portfolio Hedge Funds World Ex UK Equities UK Equities Real Estate pfi portoflio ftse all shares macquarie infra europe Mar 06 Mar 07 Mar 08 Mar 09 Mar 10 Mar 11 Mar 12 Mar 13 Mar 14 Feb Commodities Standard Deviation Reference Assets Reference and Infrastructure Assets Blanc-Brude et al (2015)

12 Long horizon investing in infrastructure 12/29 What investment beliefs for long-term investors in infrastructure? Asset class or factors? Infrastructure as a heuristic Question 2: Should infrastructure assets be expensive (ie low yielding)? Contracts not concrete Patience is not enough: long-term investment requires a (machine) learning process

13 Long horizon investing in infrastructure 13/29 Section outline 1 Investment beliefs 2 Agency 21 2 Questions 22 The infrastructure PE fund model 23 Market failure 24 Inter-temporal monitoring demand 3 Benchmarking 4 Conclusions 5 Table assignment

14 Long horizon investing in infrastructure 14/29 Agency Two questions: Q3: Amongst external infrastructure managers, have you come across attractive alternatives to the 2:20 PE infrastructure fund model? Q4: Is the information reported today by infrastructure managers (internal or external) sufficient for asset owners purposes?

15 Long horizon investing in infrastructure 15/29 Agency: delegation and the infrastructure fund model Infrastructure fund have proliferated since the 1990s They tend to offer one possible way to gain exposure to infrastructure projects: relatively short-term and exit focused, possibly quite aggressive (extra leverage) and expensive (2:20) This type of product has become notoriously unfashionable with large long-term investors Still, in a world with heterogenous skill endowments, there is value is specialisation, hence, there should be value in delegation as well

16 Long horizon investing in infrastructure 16/29 Agency: a pooling equilibrium (and how to get out of it) Despite the decision to internalise infrastructure investment by very large accounts, the market for delegated service is not changing much Question 3: Amongst external infrastructure managers, have you come across attractive alternatives to the 2:20 PE infrastructure fund model? There are never been so many infrastructure PE funds, with so much dry powder to invest than today With so much competition between infrastructure managers, why aren t new models emerging that offer what long-term investors actually want? this situation can be understood as a form of market failure (a pooling equilibrium) in which all service providers offer the same, low quality, high cost products, because of information asymmetries between clients and managers are large and it is difficult (and costly) for those managers that could offer better products to differentiate themselves from the crowd

17 Long horizon investing in infrastructure 17/29 Agency: inventing new ways to delegate Abandoning delegation is one way to respond to market failure the other is to create revelation mechanism (auctions) by which only the managers that can actually deliver what clients want can be selected (self-select) First investors need to define what they are after: from narrative to objectives, including the recognition that infrastructure is just a heuristic Then mandates can be auctioned so that they can reveal the service providers that can deliver a service at the best cost However, agency theory dictates that there is a cost to make the right agents reveal their type - it should be compared with the full cost of not delegating (including the cost of less diversification) Some managers may also try to signal their difference (eg recent formation of industry associations)

18 Long horizon investing in infrastructure 18/29 Agency: the inter-temporal monitoring demand Long-term investing (buy and hold) increases the demand for monitoring of asset owners Question 4: Is the information reported today by infrastructure managers (internal or external) sufficient for asset owners purposes? Standardised data and performance reporting frameworks are needed Monitoring can be improved through information revelation mechanisms: well-designed delegation or cooperation (data pooling) Monitoring is required whether there is delegation to an external manager or not, including for performance attribution purposes This is especially true if infrastructure investment is lumpy and takes time to build up: asset selection matters but claims of outperformance require a point of reference: a benchmark!

19 Long horizon investing in infrastructure 19/29 Section outline 1 Investment beliefs 2 Agency 3 Benchmarking 31 4 Questions 32 A roadmap 4 Conclusions 5 Table assignment

20 Long horizon investing in infrastructure 20/29 Benchmarking: 4 Questions 1 Asset allocation: risk adjusted performance measures 2 Prudential regulation: extreme risk measures 3 ALM/LDI: duration and inflation hedging 4 ESG: investment impact measures

21 Long horizon investing in infrastructure 21/29 Benchmarking: which we cannot answer today Market proxies are ineffective Existing research using private data is too limited Reported metrics are inadequate Impact metrics are not standardised and aggregated

22 Long horizon investing in infrastructure 22/29 Benchmarking: A roadmap (Blanc-Brude, 2014) 1 Define your terms 2 Design adequate pricing and risk models 3 Determine what data needs to be collected 4 Standardise data collection and populate a global database 5 Design useful reference portfolios 6 Track the performance of these portfolios in time

23 Long horizon investing in infrastructure 23/29 Benchmarking: Recent progress 1 Definitions: discussions with EIOPA 2 Asset pricing and cash flow models (Blanc-Brude et al, 2014; Blanc-Brude and Hasan, 2015) 3 Data collection template (Blanc-Brude et al, 2015) 4 Data collection by EDHEC: currently 300 deals going back 20 years, 1,700+ by Design useful reference portfolios (table assignment)

24 Long horizon investing in infrastructure 24/29 Benchmarking: Recent empirical results Construction risk is idiosyncratic in project finance 70 Cash flow are lognormal in PPPs DSCR lognormal density, contracted infrastructure Density predicted actual Public construction risk - decision to build (Flyvbjerg dataset, n=110, ) Project finance construction risk - financial close (NATIXIS dataset, n=75, ) Greenfields diversify Brownfields Lower credit risk comes at the cost of duration risk PortfolioReturn (percentage) Green elddebt-onlyportfolio Wholelifecyledebtportfolio Brown elddebt-onlyportfolio PortfolioRisk(percentage) Loss(%) Risk Trade off (Rising DSCR family) Duration(%) Loss(%) Risk Trade off (Flat DSCR Duration(%) see Blanc-Brude and Makovsek (2014), Blanc-Brude (2014), Blanc-Brude and Ismail (2013), Blanc-Brude et al (2014)

25 Long horizon investing in infrastructure 25/29 Section outline 1 Investment beliefs 2 Agency 3 Benchmarking 4 Conclusions 5 Table assignment

26 Long horizon investing in infrastructure 26/29 Conclusions Long-term investors in infrastructure are faced with a large and complex information problem We can create the right tools and framework that can answer the most relevant questions This monitoring and benchmarking framework can be smart and learn sequentially as and when new information becomes available But it requires poling information that is otherwise scattered amongst investors This is what EDHEC is currently implementing, but your investor input is extremely valuable

27 Long horizon investing in infrastructure 27/29 Section outline 1 Investment beliefs 2 Agency 3 Benchmarking 4 Conclusions 5 Table assignment

28 Long horizon investing in infrastructure 28/29 Benchmarking Table assignment: 1 What role do privately-held infrastructure assets play in the portfolio of a large well-diversified, long-term investor? Discuss both equity and debt 2 Given the lack of immediate investability of reference portfolios of private infrastructure equity or debt, what should we try to benchmark? Should we focus on capturing the performance of an infrastructure asset class or try to document something else? (and if yes, what could that be?) 3 Benchmarking privately-held infrastructure investments requires collecting information scattered amongst long-term investors on an ongoing basis What will make it worth investors while to participate to such an exercise?

29 Long horizon investing in infrastructure 29/29 References Bird, R, H Liem, and S Thorp (2014, May) Infrastructure: Real Assets and Real Returns European Financial Management 20(4), Bitsch, F, A Buchner, and C Kaserer (2010) Risk, Return and Cash Flow Characteristics of Infrastructure Fund Investments EIB Papers 15(1), Blanc-Brude, F (2013) Towards efficient benchmarks for infrastructure equity investments EDHEC-Risk Publications, 88 Blanc-Brude, F (2014, June) Benchmarking Long-Term Investment in Infrastructure EDHEC-Risk Position Paper Blanc-Brude, F, R Delacroce, M Hasan, C Mandri-Perrot, J Schwartz, and T Whittaker (2015) Data collection for infrastructure investment benchmarking: objectives, reality check and reporting framework EDHEC-Risk Position Paper Blanc-Brude, F and M Hasan (2015) The valuation of privately-held infrastructure equity investments EDHEC-Risk Publications January Blanc-Brude, F, M Hasan, and O R H Ismail (2014) The valuation and performance of unlisted infrastructure debt EDHEC-Risk Publications July Blanc-Brude, F and O R H Ismail (2013) Who is afraid of Construction Risk? Portfolio Construction with Infrastructure Debt EDHEC-Risk Publications, 128 Blanc-Brude, F and D Makovsek (2014) How much construction risk do sponsors take in project finance? EDHEC-Risk Publications Blanc-Brude, F and R Strange (2007) How Banks Price Loans to Public-Private Partnerships: Evidence from the European Markets Journal of Applied Corporate Finance 19(4), Blanc-Brude, F, S Wilde, and T Witthaker (2015) The performance of listed infrastructure equity: a mean-variance spanning approach EDHEC Business School Working Paper Jenkinson, T, M Sousa, and R Stucke (2013) How Fair are the Valuations of Private Equity Funds? SSRN Electronic Journal Newell, G and H W Peng (2007) The significance and performance of retail property in Australia Journal of Property Investment & Finance 25(2), Newell, G and H W Peng (2008) The role of US infrastructure in investment portfolios Journal of Real Estate Portfolio Management 14(1), Phalippou, L and O Gottschalg (2009) The performance of private equity funds Review of Financial Studies 22(4), Rödel, M and C Rothballer (2012) Infrastructure as Hedge against Inflation Factor Fantasy? The Journal of Alternative Investments 15(1), Rothballer, C and C Kaserer (2012) Is Infrastructure Really Low Risk? An Empirical Analysis of Listed Infrastructure Firms