INEFFICIENT MARKETS - Shaky Foundations

Pension funds are looKing for investments that will perform even when the stock market is doing badly. They also want long-lived assets to match the payments they must make in years to come. Inflation-protected government bonds, especially those with long durations, fit the bill. The trouble is that yields on such bonds have declined dramatically in recent years.

Infrastructure provides the perfect alternative, say those who manage specialized funds investing in toll roads, airports, utilities and other vital services. Not only is infrastructure an inherently safer investment than equities, the returns are significantly higher than those available on government bonds.

This message has not fallen on deaf ears. It has spurred tremendous investment flows into this area in recent years. However, the valuations for infrastructure assets have become so stretched, and the amount of debt employed so large, that investors are now exposed to a wide variety of risks.

Infrastructure constitutes a separate asset class, according to a report issued last summer by David Rickards, head of equities research at Macquarie Securities, the Australian bank that is the world’s leading manager of infrastructure funds. The definition of what constitutes infrastructure, though somewhat hazy, generally comprises businesses and services with monopoly-type competitive positions. Their stable revenues are not supposed to vary much with the business cycle and usually adjust upward in line with inflation.

What’s more, airports, roads and ports don’t disappear overnight. Their secure cash flows can be reliably projected years into the future, making infrastructure attractive to institutional investors, such as pension funds. Rickards suggests that infrastructure constitutes a hybrid asset — somewhere between an inflation-protected government bond and conventional equity — that provides the benefit of moderate to high returns with low to moderate volatility.”

A few years ago, Sydney-based Macquarie launched its own benchmark, the Macquarie global infrastructure index, which originally comprised some 250 public companies from a variety of countries. Analyzing the performance of the index, Rickards observes that the historical correlation between infrastructure businesses and the stock market has declined. In recent years the infrastructure index has also produced better risk-adjusted returns than have global equities or bonds.

Many investors have heeded the message that infrastructure provides, in the words of Macquarie’s Rickards, a “shock absorber to a portfolio during a downturn.” But is this really the case? There are reasons to expect that the future isn’t likely to resemble the past.

There has been a flood of money into this field. Infrastructure pioneer Macquarie faces increasing competition from Wall Street’s finest. Goldman, Sachs & Co. recently closed a $6.5 billion fund. General Electric; private equity firms like Carlyle Group and Kohlberg Kravis Roberts & Co.; and deep-pocketed Middle Eastern investors are also getting in on the game. A Standard & Poor’s report in November by Michael Wilkins, London-based head of the ratings agency’s European infrastructure finance group, estimated that between $100 billion and $150 billion had been raised by infrastructure funds and was waiting to be invested. Add in debt, and this money could be used to finance deals worth $500 billion to $1 trillion.

Yet there is a limited supply of assets that this money can acquire. As a result, the price for infrastructure businesses has been rocketing. To acquire Associated British Ports, Goldman Sachs last summer paid 17 times earnings before interest, taxes, depreciation and amortization. A few months earlier, Spanish construction company Grupo Ferrovial bought U.K. airport business BAA at a similar multiple, and Macquarie and its Spanish partner, Cintra, signed a 75-year lease on the Indiana Toll Road for a whopping $3.8 billion. The price for this 157-mile road checked in at 65 times the prior year’s ebitda.

When investors acquire assets at inflated prices, it’s only to be expected that returns will be lower in the future. So how did Macquarie come to forecast a return of about 12.5 percent for its investment in the Indiana road when the cash flow yield was a mere 1.5 percent? The answer is that this infrastructure deal, like most, was largely financed with debt. Macquarie put up only $385 million of equity for its $1.9 billion stake. Factor in traffic growth and likely increases in tolls, and Macquarie could achieve its projected return.

Borrowing huge sums may offset the consequences of paying inflated amounts for infrastructure investments in the first place. But it also increases the risks. It’s true that the credit markets have been extraordinarily accommodative of late. Infrastructure deals are generally financed with nonrecourse debt. These loans are not only cheap by historical standards but are often hedged against rising interest rates.

Nevertheless, S&P’s Wilkins says the amount of debt placed on infrastructure deals has reached as high as 30 times ebitda. He warns that the “infrastructure sector is suffering from the dual curse of overvaluation and excessive leverage — the classic symptoms of an asset bubble.”

As the infrastructure sector gets more crowded, its definition has been stretched to include telephone directories, airlines, airport trolleys, radio stations, telecommunications networks and car parks. Macquarie even considered a tilt at the London Stock Exchange, whose activity was deemed to be “financial infrastructure.” Cash flows from such businesses are more vulnerable to technological change and the vagaries of the business cycle than those from traditional infrastructure.

Anyone considering an investment in infrastructure might spend a minute perusing the “risk factors” in the prospectus of the Macquarie Infrastructure Group (MIG), a toll road business that is the world’s largest listed infrastructure fund. These include construction risk, operation risk, interest rate risk, inflation risk, business risks, feasibility risks, regulatory risk, government action risk and finally, acts of God (fires, floods, earthquakes, war and so forth).

These are not just template disclaimers. Infrastructure managers derive their valuations by forecasting revenues and expenses far into the future and discounting back to the present. Discounted cash flow models that cover very long periods are highly sensitive to changes in assumptions. A mere 0.5 percent increase in the risk premium used in the MIG models would wipe A$1 billion ($800 million) off the valuation, according to the latest annual report — a sum roughly equivalent to shareholders’ retained earnings.

On top of that revenues may not meet expectations. Infrastructure assets are often held in special-purpose vehicles with several partners who hold minority stakes. The debt in these vehicles doesn’t have to be consolidated on the balance sheet of the fund. Furthermore, the interest rate hedges and refinance risks associated with these positions are not always clear, so investors may have only a vague idea of how vulnerable their equity stakes are.

As leverage rises in the infrastructure field, investors’ exposure to the business cycle and to conditions in the credit markets also climbs. Pension fund trustees looking for secure long-term investments would do well to avoid this field, at least until the current infrastructure frenzy has died down.

Edward Chancellor, an editor at Breakingviews.com, is the author of Devil Take the Hindmost, a history of financial speculation.