Anatomy of a Cash Cow: An In-Depth Look at the Financial Characteristics of Infrastructure Companies

Published:  September 2020
Download the Paper
We examine how infrastructure companies differ from the rest of the economy and in particular whether or not they tend to pay larger and more frequent dividends. We find that infrastructure companies exhibit key systematic differences with a sample of ‘matched’ firms that are otherwise comparable.


Preqin, a fund database, identifies infrastructure as a growing area of alternative investments. Preqin’s database shows that funds under management for infrastructure firms have increased from USD$1,646 million for year 2000 vintage funds to USD$111,247 million for year 2019 vintage funds.

Increasing investor interest in infrastructure as an asset class is justified by what  calls the ‘infrastructure investment narrative’: infrastructure is expected to possesses special characteristics such that “asset owners are expected to benefit from the low elasticity of demand creating pricing power and an inflation hedge, as well as low return covariance with other investments, allowing attractive risk-adjusted returns.”

As such, infrastructure businesses are often portrayed as ‘cash cows’ i.e. the combination of semi-monopolistic conditions, limited opportunities for growth, high leverage and steady revenues should result in significant dividend payouts.

These dividend payouts then are likely to go a long way in explaining why, on a total return basis, infrastructure companies are somewhat less correlated with other asset classes and also provide attractive return.

In this paper, we explore the claim that infrastructure investments are indeed different from other types of firms by asking two related questions:

  1. First, do infrastructure firms exhibit unique characteristics compared to equivalent firms in other sectors?
  2. Second, do these characteristics correspond to a different dividend payout behaviour?

Characteristics of Infrastructure

Today, the academic literature evaluating the veracity of the ‘infrastructure investment narrative’ is scant, as is any testing as to whether identified infrastructure assets possess the characteristics hypothesised.

Papers such as  assume that infrastructure does indeed possess these special qualities, and then goes on to examine asset pricing models in an attempt to explain infrastructure returns.

However, to the authors’ knowledge, no research exists examining whether the characteristics of infrastructure can be measured and whether these characteristics differ from those of other firms in the economy.

We hypothesise that infrastructure has specific characteristics, for example:

  • There is some form of government regulation or input into the operations of the firm;
  • The firm possesses natural monopoly characteristics, through either increasing returns to scale or traffic network effects; and
  • The firm also enjoys large capital investment that is durable and immobile.

These characteristics can be measured using financial variables. As a result, we can firstly assess whether infrastructure assets exhibit these characteristics.

Secondly, we can then gauge whether these characteristics are different to those of other firms in an economy.

Finally, we can discover whether these characteristics contribute to the observed dividend pay-out behaviour of infrastructure firms.

In this paper, we formulate four hypotheses for characteristics of infrastructure, in comparison with non-infrastructure firms.

  1. Infrastructure exhibits higher asset tangibility given its reliance on large capital investments.
  2. Infrastructure exhibits higher asset illiquidity as infrastructure’s large, capital intensive assets are hard to liquidate in times of firm distress.
  3. Infrastructure exhibits higher asset inflexibility due to the firm’s inability to reallocate their assets to other activities.
  4. Infrastructure exhibits lower operating leverage as infrastructure firms have a significant asset base, and the level of operating costs in comparison to the size of these assets would be smaller than other ‘capital light’ businesses.

A High Quality, Handmade Dataset

For this study, we look at the UK as it has the largest and longest history of infrastructure investment. The Companies Act 2006 requires companies to submit yearly company accounts to Companies House, the UK’s registrar of companies. The data we use comes from the FAME database provided by Bureau van Dijk, which provides 20 years of financial data. We extract accounting items from those companies that are the global ultimate owner, report group financials and are incorporated in the UK.

To ensure that there is no overlap between infrastructure and other firms, we employ the firms identified as infrastructure by EDHECinfra to filter out infrastructure companies that appear in the FAME dataset. This list of firms is identified from government and regulator databases as well as infrastructure news services and is cross checked to ensure the firms are conducting an infrastructure activity as defined by EDHECinfra’s TICCS® classifications (see

Regarding the relationship between infrastructure and dividend payouts, only including dividends as shareholder payout will not capture the total component of shareholder distributions for infrastructure as it has excluded the principal and interest components of shareholder loans (). However, FAME does not consistently provide such details on the breakdown of shareholder loans. Hence, we include an additional unlisted infrastructure sample using EDHECinfra data, which incorporates shareholder loans in computing the dividend related measures.

Robust Controls for Endogeneity

The decision to set up an infrastructure firm is an endogenous decision which can result in firms exhibiting certain ratios and sizes. This endogeneity limits the ability to draw conclusions from the analysis unless it is explicitly controlled for.

As a result, we employ propensity score matching to attempt to control for endogenous differences between infrastructure and non-infrastructure firms and then conduct tests on differences again. By matching firms that are most alike in size, leverage, revenue growth and profitability, we are then able to determine if infrastructure firms are the only firms that possess the unique attributes hypothesised.

Significant and Systematic Differences

We then test our hypotheses and analyse whether infrastructure exhibits different characteristics from non-infrastructure assets. We first assess whether there are differences in the mean and median of the variables of interests.

The table below provides a summary of the findings and implications. We are able to conclude that infrastructure does exhibit different characteristics compared with other firms. For example, using FAME data, we observe that infrastructure indeed exhibits lower operating leverage compared with non-infrastructure firms. For inflexibility, we find that unlisted infrastructure has a statistically significant higher mean than unlisted non-infrastructure firms, which is in line with our hypothesis.

These characteristics are as a result of the nature of infrastructure businesses, specifically the requirement to invest in large, highly specialised assets that cannot be re-purposed easily.

Hypothesis Finding Implications
Infrastructure exhibits higher asset illiquidity Evidence that infrastructure has lower asset liquidity compared to non-infrastructure Infrastructure has large capital investments which are relationship specific and hard to liquidate in times of firm distress
Infrastructure exhibits higher asset inflexibility Strong evidence for unlisted infrastructure having higher inflexibility compared to non-infrastructure Infrastructure has large and durable assets with sizable sunk costs and cannot adjust production to adapt as well as other firms in response to market shocks
Infrastructure exhibits higher asset tangibility Strong evidence for unlisted infrastructure firms having a higher intensity of Property, Plant and Equipment compared to non-infrastructure Infrastructure assets have high physical tangibility though they may not necessarily be mobile
Infrastructure exhibits lower operating leverage Strong evidence for unlisted infrastructure’s operating leverage being lower than that for non-infrastructure Infrastructure is different from other firms by having an asset base so significant such that the level of operating costs in comparison to the asset size is small
Infrastructure has special characteristics which can explain dividend payout behaviour Strong evidence for infrastructure paying out more dividends as a proportion of their revenues compared to non-infrastructure; strong relationships found between payout behaviour and characteristics such as operating leverage and asset illiquidity Infrastructure does pay higher dividends than other firms and these higher payout ratios correlate with the identified characteristics

The Uniqueness of Infrastructure

We then turn to examining whether these characteristics are able to explain a firm’s dividend payout behaviour using various regression models. Three dividend related measures including the dividend payout ratio are used. We also examine whether the unique characteristics of infrastructure firms go some way to explaining the level of dividends for unlisted infrastructure firms. We find that:

  1. Both listed and unlisted infrastructure firms do pay out a larger proportion of their revenues compared with other comparable firms; and unlisted infrastructure firms pay out more dividends relative to asset size, in comparison with unlisted non-infrastructure firms.
  2. The higher the operating leverage of an infrastructure firm, the higher its dividend payout as a proportion of total assets. An explanation for this observation is the documented negative relationship between operating leverage and leverage. With lower leverage, a firm is able to pay out more free cash to shareholders.
  3. Asset illiquidity has a positive relationship with dividend payout for non-infrastructure firms but a negative relationship with dividend payout for infrastructure firms. This means that an infrastructure firm with more illiquid assets pays out higher dividends as a proportion of its revenue and assets. This may be due to brownfield infrastructure companies being in a better position to pay out dividends than greenfield infrastructure companies, which have more liquid assets like cash.

One major issue with infrastructure investment is the lack of a commonly agreed definition. The characteristics identified and examined in this paper can go some way to understanding what makes infrastructure different as an investment.

Furthermore, it is possible to employ these characteristics to provide a check on whether firms classified as infrastructure, actually are infrastructure.