Hello WSO, I have been a long time lurker, and this is my first post (or maybe I am one of the top rated authors who decided to make a post anonymously). I have noticed a certain lack of information about infrastructure investing on WSO, and those threads that are created from time to time, are long on questions and short on answers, especially on answers from actual industry insiders.
With this in mind, I have decided to make this post and cover the following topics:
- What infrastructure PE is and how it is different from traditional PE
- General info about infra PE landscape
- Key players in infra PE space
- Details about day-to-day job in infra PE
If any of the areas after my post remain unclear, I am more than happy to answer questions in comments. As for my background, I am based in London, and I am senior associate at one of the best infrastructure PE firms.
Here we go.
1. Introduction to infrastructure
When I say investing in infrastructure, I mean investing in such assets as
- Utilities: gas / electricity / water distribution, communications infrastructure
- Transportation: airports, seaports, roads, bridges, rail
- Social infrastructure: hospitals, education facilities, etc.
- Energy related: power plants, oil and gas pipelines, oil terminals, renewable energy assets and such (read on for the discussion about where energy ends and infrastructure begins)
You might want to know why investment community treats infra PE as a distinct bucket and not just as part of "traditional" PE. It is a fair question and there are many reasons for that. Specifically, from investors' (i.e. limited partners') perspective, infrastructure more often than not possesses the following distinct characteristics:
- Low volatility, protected downside, stable cash flow profile - all this stuff basically meaning "low risk" (comparing to traditional PE). We will talk a bit about returns later but I have seen plenty of deals where even in most disastrous downside case the investment would return some 3-5% . How cool is that? Investors value protected downside a lot.
- Strong cash yield. Although this is not an absolute necessity, vast majority of infrastructure assets being purchased by financial investors have some current dividend yield and investors love that. It is quite different from PE (where all your return usually comes on the last day from exit, maybe boosted by one-off dividend recap in the middle) and in this respect it resembles REPE where current yield is more often a requirement than not.
- Infrastructure assets performance is often implicitly or explicitly linked to macro indicators such as inflation, GDP, population growth etc. Therefore many investors naturally see infrastructure as a hedge. For example, if inflation increases or population in a country grows, public pension plan would see the pension payments to people increasing. At the same time, if this pension plan is invested in -linked infrastructure (e.g. regulated gas distribution utility) or population-linked infrastructure (e.g. toll road), you can see how it helps to offset the increasing pension liabilities.
- Defensive resilient performance profile and low correlation with other asset classes (such as PE, equities, fixed income, RE, etc.). No need to explain - private infrastructure provides good diversification benefits to investors' portfolios.
2. What on earth is core-plus? A bit about returns
Infrastructure always consists of physical assets so, unsurprisingly, in many instances it resembles real estate and hence there are many similarities between infra PE and REPE (and both can be brownfield i.e. buying operating business, or greenfield aka development - constructing stuff from scratch). Since I am mentioning real estate here, it is worth noting that infrastructure uses very similar classification of infrastructure assets by risk buckets. Some would hate this classification but it is widely used in the industry and is actually very helpful.
- Core infra: boring operating assets with most/all of the return coming from cash dividends; most likely state-regulated revenue with limited risk on revenue side. All you need to do is not to mess up operating costs. No or very limited growth. Example would be some regulated electricity distribution in the US or Western Europe. As it stands today, you would expect equity IRRs of below 10% for such investments but on the other hand it is still several times above fixed income instruments and risk is much lower than traditional PE. Downside scenario when everything goes perfectly wrong can still yield 3-5% equity IRR. Also, public private partnerships (PPP) are often considered "core" even when they assume greenfield construction - because under PPP model the risks on revenue side are usually very low, theoretically absent, so you just need to manage your costs to get the desired return.
- Core-plus infra: somewhat more risky, will include operating asset with some growth story or growth / expansion capex, or vanilla operating infra asset in questionable jurisdiction (for example, regulated water utility in Poland). In developed market (North America, Australia, Western Europe, Japan) typical equity IRRs would be in low teens, while in emerging markets such as Asia there can easily be premium of 5% or sometimes even more.
- Value add infra: investment in infrastructure requiring some serious operational involvement, potentially substantial business re-profiling. For example, buying regional airport with aim to turn it into super-regional hub, or buying bulk liquid seaport and building container-handling facility (never seen these exact situations in practice but I am trying to come up with something clear for non-infra professional).
- Opportunistic infra: riskiest of all, with limited or no dividend yield, with bigger downside risks. Here we are talking about some 15% equity IRR and higher, depending on jurisdiction. This category is quite small and there is a reason for that - here it becomes too close to the border of traditional PE and limited partners would be inclined to invest into PE at these returns because PE firms are bigger specialists in this high-risk-high-return territory. No firm can invest in infrastructure and consistently earn some 20% IRR for years. If someone tells you that, most likely they are either lying, or they are investing in something that is not infrastructure.
Those of you who work in REPE will definitely recognise this terminology of core, core plus etc.
One infra PE firm can have several active funds in various categories, e.g. one core/PPP fund and another broader core-plus/value-add fund. But there are plenty of fund managers that focus almost exclusively on core infrastructure. I am talking about such names as Amber Infrastructure (with their INPP franchise), John Laing, DIF, InfraRed Capital Partners (and their listed HICL fund), Innisfree and few others, sorry for forgetting.
You normally wouldn't have fund that invests both in core and value add / opportunistic, simply because returns are very different (IRR can be 2x different between core and opportunistic investments).
3. From the beginning of it to today's landscape
Unlike traditional PE or REPE, infra PE is a relatively young game. Some 20 years ago it was almost unheard of, globally, to attract private financial capital for large infrastructure assets. Most infra assets belonged to governments or to corporates who once built them. Infrastructure private equity didn't exist. However, some time ago which I feel was around early 2000s, few countries such as Australia and Canada pioneered attraction of private (financial) capital for both construction of new infrastructure and monetisation of existing assets.is widely credited as the first-mover who actually started raising PE funds with the mandate to invest exclusively in infrastructure.
Currently infrastructure is a crowded space with a lot of fund managers competing for deals in all regions globally. And infrastructure isn't created out of thin air (unlike IT start-ups, for example) hence supply of good deals is growing much slower than demand. And believe me, demand is there, because limited partners throughout the world finally appreciate benefits of investing in infrastructure (meaningful return with protected downside and low correlation to other assets - very good combination).
Speaking about the US market, historically US institutions (such as public pension plans) were comfortable with traditional PE and energy-focused PE. Infrastructure asset class was somewhat overlooked in the United States. While Western Europe, Canada and Australia loved infra for its risk-return profile, the US apparently was finding infra to be a bit too boring. In recent years US has been catching up actively but still it is clear that perception or risk in the US is a bit different. What would be considered value add infra in Europe, would be routinely treated as "core"/"core-plus" in the US. Also US pension plans are in love with energy-related infrastructure so in recent years some asset types that were historically considered "traditional PE" for years (merchant power plants, merchant oil and gas pipelines, even refineries) are now being pushed into infrastructure territory (with IRRs falling correspondingly). To give you an idea, infra PE fund in the US can buy into a construction of a merchant (i.e. no government subsidies or guarantees) gas power plant or even wind park. In Europe, infra funds would only consider it if it is already operating and has some sort of subsidy (capacity payments or else). There are of course exceptions here and there but this should give you a general idea of Europe vs US dynamics. Generally, in the US the borderline between energy-focused PE and infra PE is sometimes blurred. Watch out for such names as ArcLight Capital Partners, Energy Capital Partners - these are on the border between infra PE and energy PE.
Australia is a bit like Europe (or rather Europe is a bit like Australia) - conservative view on risk with sometime unbelievably high valuations (multiple for a seaport or an airport) and fairly low IRRs.
Asia (ex Japan) is still an emerging market and infrastructure IRRs are not very far from traditional PE.
South America is a frontier market with limited number of deals happening (although this seems to be changing recently as competition for assets in developed world has driven valuations to insanely high levels).
Africa is a no at the moment.
Within infrastructure, financial investors can be broadly divided into two groups: fund managers and institutional investors (latter being pension plans, insurance companies and sovereign wealth funds). Until recently the infra PE landscape was dominated by fund managers, with institutional investors supplying passive money to them. However, the trend in recent years is big institutional investors going direct - that is, building their own direct investment teams that would compete for deals with fund managers. Some (especially Canadian pension plans) are ahead of the curve and strong and experienced enough to do deals on their own. Given high quality of people and cost of capital being lower than traditional infra PE fund, they are seen as a very serious contender.
4. Infrastructure PE: industry participants
As I mentioned just couple paragraphs above, the universe of financial players that invest in infrastructure, can be broadly divided into fund management firms (either dedicated asset managers or fund management arms of big diversified banking groups) and institutional investors (pension plans, sovereign wealth funds and insurance companies). Below I provide overview of each of those categories.
Independent asset managers
These can be:
- Pure play infrastructure fund managers (e.g. Global Infrastructure Partners who are on track to close their monstrous GIPIII fund at well above $11 billion; also keep eye on such names as Alinda Capital, Highstar Capital, AMP Capital, IFM Investors - all managing billions of infrastructure equity)
- Fund managers focused on broader real assets (Brookfield being the most notable firm focused on real estate, infrastructure and timber - their latest dedicated infra fund BIFIII is record-breaking $14 billion - largest infra PE fund in history)
- Diversified asset managers ( Fund II closed at $3.1 billion last year, EQT Infrastructure is targeting $3+ billion for fund III, etc.). used to have infrastructure PE business but then the team spun off to create an independent firm called Stonepeak Infrastructure (Stonepeak closed their second fund at $3.5 billion earlier this year). Carlyle raised $1 billion infrastructure fund in 2007, probably investments didn't go well - they never followed up with fund II. Apollo, TPG, other - haven't heard about dedicated infra platform inside these.
Many of the biggest banks that have merchant banking or private equity business, would have infrastructure PE franchise. For example:
- (named West Street Infrastructure Partners, third vintage at $3 billion). I understand that for regulatory purposes US banks have to come up with different names for their private fund management business hence Goldman becomes West Street etc.
- (named North Haven Infrastructure Partners, second vintage at $3.5 billion)
- - these guys are investing primarily through their open-ended Fund and separately-managed accounts. It is hard to estimate the amount of infra capital they manage.
- Deutsche Asset and Wealth Management (on track to close their second vintage of Pan-European Infrastructure Fund II at $2+ billion)
- Macquarie Infrastructure and Real Assets - infra PE business of Australian Macquarie Bank. With some $100 billion of infra assets under management, they are probably the largest infra PE firm out there. Unlike many other investors mentioned in this topic, these guys prefer to raise regional funds that tend to be smaller in size (e.g. Korea focused fund, Mexico focused fund) but aggregate amount they raise is outstanding. Their flagship funds are quite big though, e.g. Europe-focused MEIF5 closed earlier this year at $4+ billion hard cap.
- have some infra funds but not very big, as far as I know.
- I can't recall doing anything in the space.
Canadian pension funds (CPPIB and OTPP being the strongest, also Borealis and several others) are known for being well mature in a sense that they don't need fund managers anymore - for infrastructure they almost entirely replaced their fund investment program with direct investment program. They have strong teams and fairly low return requirement (to some extent because they invest their "own" money so there is no management fee or performance fee layer) which makes them very strong competitors to fund managers. Many high profile assets were won by the Canadians in last few years. They would normally focus on bigger deals, because they have a lot of money to invest and teams are generally small, so they can't afford chasing some $ deals.
In Europe Dutch pension managers (APG and PGGM) are reasonably mature to do deals on its own, also USS in the UK, also they invest in funds as well.
In Australia many superannuation funds are capable of investing directly on their own in Australian infrastructure, but outside their home turf they are usually quite helpless and would only invest directly alongside a more experienced fund manager.
Sovereign Wealth Funds
Similar story to pension funds - historically SWFs have been limited partners but now are actively building direct investment teams, poaching bankers from BBs and other infra PE firms. However, the industry needs few more years to understand how it plays out.
These days most active are SWFs from Middle East (ADIA, ADIC, Mubadala, QIA, KIA through their infra investment arm called Wren House), China (CIC), Singapore (GIC). They also have fairly low return requirements, same as many pension plans, but they also have more capital, virtually unlimited given that government can print money (simplifying here, but only to a degree).
Investment arms of insurance companies
For relatively low risk, resilient performance and link to macro indicators, insurance companies also have come to like infrastructure quite a lot. Some invest purely money (e.g. Allianz), some set up fully fledged fund management platforms and manage both money of the insurance parent and of third parties (e.g. Swiss Life). Nevertheless it is still quite rare for insurance companies to invest directly and most remain passive limited partners / investors in third party funds.
5. On the job
Day-to-day work and deal making is very similar to traditional PE with majority of deals being fairly classic LBOs.
In terms of deal sourcing, it is a typical combination of proprietary sourcing and competitive auctions (which become increasingly common, at least in developed markets). Privatisations are still common but so are secondary deals (financial investor selling to financial investor) and buying non-core assets from corporates (for example oil giant Total selling pipeline system in North Sea to an infra PE fund).
With that being said, public-private partnerships (PPP) is a world of its own within infrastructure. This is when, for example, private sponsor constructs a toll road with government guaranteeing $X million revenue per annum or, even better, guaranteeing Y% return on construction costs. This is just one example, PPPs can take a lot of various forms, but common theme is private capital + certain guarantees and support from the government. In today's infrastructure PE world PPPs are not very big part - note that PPPs do not always require financial investors, often a construction company would finance PPP with their own balance sheet (e.g. Vinci).
Investment banks are active advising buyers and sellers in infrastructure. Some banks use energy and natural resources teams to advise on infra deals, some use industrials teams, some are advanced enough to have dedicated infrastructure M&A advisory groups (e.g. JPMorgan and ).
Working in infra PE
As an investment analyst or associate at an infra fund you will do the same stuff as you would do in any PE firm: modelling and deal execution, , market research and reportage to limited partners. Some associates are also involved in launching new funds and fundraising.
Investment process is also typical. You would normally expect seller to release Info Memo and some initial materials and then require first round indicative bid for an asset within 3-5 weeks, and then second round (assuming you made it through the first round) binding bid 4-6 weeks after. A buy side team will hire a small army of advisors, will have several meetings with their Investment Committee including final price-setting meeting couple days before the binding bid. When deal is negotiated privately, more variations are possible but the pattern would often be similar.
If you have read it by here, you should have realised that protecting downside is very important when structuring and executing an investment in infrastructure. I think this is largely a reason why modelling is often insanely detailed here. You would have a model horizon of at least a decade (often longer, for core assets till the end of concession which can be 50+ years). As if it wasn't enough, having quarterly model isn't uncommon. Read it again: models are insanely detailed. Revenue, opex and capex are modelled in great detailed and is broken down into line items, and almost on each item, even the smallest one, you need to take a view backed up by thorough due diligence, research or legislation. Simple "let's grow this line at 5% per annum" is not acceptable. Even if revenue is based on the government-determined tariff (sounds simple, eh?), the official formula for tariff calculation is often ridiculously complicated, depends on opex, capex, macro assumptions and takes hundreds or or thousands of rows in. Sometimes I think that going that complex increases the probability of a mechanical error in Excel and also probability of making a wrong assumption as well. Unfortunately simplistic approach about growing something at random 5% can sometimes be more accurate than super-duper detailed bottom up modelling with wrong assumptions (infamous GIGO effect).
With that being said, the above information about modelling and investment process is only true for fund managers and some of the top pension plans. There are other participants in the market - less sophisticated, sort of, and they will rely on more simplified assumptions and valuation techniques or would just trust their M&A advisor on modelling and valuation completely. Read on for more details.
To give you sense of the scale of infra PE space, see some approximate figures on fundraising:
- By infra PE funds: $50+ billion per annum
- By real estate PE funds: $100+ billion per annum
- By traditional PE funds: $200+ billion per annum
It is somewhat unsurprising that infra is lower as there are much less airports than shopping malls than cloths retailers! $50 billion every year is still quite huge though. Add another $100 billion of debt that can be raised against this equity (infra assets are usually very leverage-friendly thanks to stable cash flow profile) and you have some $150 billion of new financial capital chasing infra deals every year.
Here are few examples of infrastructure deals in last year or two:
- IFM Investors buying Indiana Toll Road (US) for nearly $6 billion
- CPPIB and Hermes GPE buying Associated British Ports (UK) for $2+ billion
- OTPP and Borealis buying London City Airport (UK) for $2+ billion
- Global Infrastructure Partners buying 20% of Gas Natural SDG (Spain) for $4+ billion
- Brookfield, CIC and GIC buying 90% in Nova Transportadora do Sudeste (Brazil) for $5+ billion
- Macquarie MIRA buying Cleco Corp. (US) for $4.5+ billion
Just because of the nature of the asset, small and mid-cap infrastructure doesn't really exist. Although at my firm we have a very flexible mandate, I rarely see opportunities at below $200 million equity (which usually means $400+ million EV). Anything below is often a joke - subpar asset that requires a lot of work that will not adequately reward the fund manager.
Fund manager economics
Comparing to traditional PE, you would expect slightly lower management fee (mostly 1%-1.5%; it is very rare to see 2%), and 20/8 carry structure. However, there are two things that need to be noted. First of all, holding periods for infrastructure assets are much longer than in traditional PE. No one buys a toll road or water utility to flip it on in two years. Typical investment horizon is 5-10 years and becoming longer, with some funds in the market having life of 12 years and more. Therefore one infra fund will generate management fees for a longer period of time than a PE fund of the same size would. Downside of it, of course, is that you will wait longer to get your carry. In absolute size, carry is probably somewhat comparable to a decently performing PE firm (lower IRR + longer holding period = approximately same money multiple and therefore approximately same carry)
Here I seem to be running out of things to tell you about. But it is also possible that I forgot something very important. Please ask in comments.
Mod Note (Andy): Best of 2016, this post ranks #6 for the past year