What Really Drives the Performance of Infrastructure Funds?

Written By: William M. Fitzgerald
Researched by Kelsey Smith

Download the pdf of this report
Written: October 19, 2012


Why We Are Curious

Earlier this year, I participated in a conference call with a member of the investment staff of one of the large corporate pension funds in the United States.  We asked him to describe their approach toward investing in infrastructure.  He stated that the investment staff is interested in infrastructure because it expects the performance of the asset class to diversify their equity market risk and to hedge against the risk of inflation.  One of the vehicles that the staff chose to make this investment is a listed infrastructure equity strategy sponsored and managed by a prominent U.S. fund company and he stated that the staff was quite pleased with the results.  As the call continued, I pulled up the returns of the infrastructure fund on my Bloomberg terminal and compared its investment performance to that of the S&P 500 Index.  Curiously, the infrastructure strategy performed nearly identically to the S&P 500 index since the inception of the infrastructure fund in 2007.  At a quick glance, the infrastructure strategy didn’t appear to meet one of the objectives – diversifying equity risk.  We decided to conduct a more thorough investigation.

The Context

We examined the marketing materials for listed infrastructure funds of two prominent investment companies that are leaders in the infrastructure investment marketplace: Fund A is managed by a large U.S. fund company and the largest investor in infrastructure in the United States; Fund B is sub-advised to the largest private equity investor in infrastructure globally.  Each promoted the following characteristics of their investment strategies:

Consistent Demand.  Infrastructure assets and services are necessary and their use remains consistent even when prices change.

High Barriers to Entry. Because they are capital intensive and often encouraged by governments that must approve the projects, infrastructure investments face little threat of competition.

Stable Cash Flows.  Because they enjoy consistent demand and stable or rising pricing, infrastructure investments offers stable cash flows that are less risky that investments in typical companies.

Inflation Hedge.  The contracts that determine the pricing of infrastructure assets commonly tie prices to changes in aggregate pricing indices such as the Consumer Price Index.  Therefore, as prices rise, cash flow rises.

The promoters of these funds distill the above stated characteristics into the following expected investment outcomes:

a)      Low correlation with equities

b)      High correlation with inflation


The Analysis

Using regression analysis, we explored the extent to which the returns of each of these funds reflected these investment outcomes.  As the dependent variable, we collected rolling one-year total returns for each of the funds and modeled the following independent variables to these return series:

  • Rolling one-year returns of the S&P 500 Index
  • Year-over-year Consumer Price Index for All Urban Consumers
  • Rolling one-year returns of the Barclays Capital Investment Grade Corporate Bond Index

From this analysis, we can estimate and model the degree to which variability of these factors explains the variability of the investment returns of each of the funds.  If the marketing messages of these funds is genuine, then we would expect to see a low correlation and poor fit of the of the S&P 500 returns to the fund returns and a high correlation and strong fit of CPI data to the fund returns.

 

The Results

For the returns of each of the funds, the S&P 500 proved to have significant explanatory power, the CPI had weak explanatory power and the Barclays Aggregate Bond Index had insignificant explanatory power.  The specific results are as follows:

 

Fund A

Fund A is an open-ended mutual fund sponsored, marketed and managed by a U.S. fund company whose municipal bond funds as a group hold the greatest dollar value of infrastructure investments in the world.  We tested the total returns of this fund from December 31, 2007 when the track record begins to May 31, 2012 during which the fund produced an annualized return of -0.24% while the S&P 500 Index produced an annualized return of +0.28%.  The regression model with the most explanatory value for this fund is:

 

Fund A total return = [1.05 * S&P 500 total return] - [1.34 * CPI] + 2.82

 

When the S&P 500 returns were tested against Fund A returns, the correlation coefficient was .95, which indicates very high correlation.  When CPI data were tested, the correlation coefficient was .34 which suggests a relatively weak correlation.  The standard error of the test indicates that the S&P 500 accounts for most of the difference between the model estimate of fund returns and the actual fund returns but that including CPI reduces these difference slightly and therefore plays some modest role in explaining returns.

The model indicates that the best estimate of the total return of Fund A is based on a point-by-point tracking with the S&P 500 and that inflation measured by CPI plays a minor role in determining total returns.

 

Fund B

Fund B is a closed-end fund listed on the NYSE and is sponsored and marketed by a U.S. fund company and sub-advised by a firm that specializes in making private equity investments in infrastructure projects and companies.  We tested the total returns of this fund from April 30, 2004 when its track record begins to May 31, 2012 during which the fund produced an annualized total return of +5.89% compared to the annualized total return of the S&P 500 of +3.92%. The regression model with the most explanatory value is:

 

Fund B total return = [1.38 * S&P 500 total return] + [2.84 * CPI] – 1.81

 

The correlation between Fund B returns and the S&P 500 was .86, a bit lower than for Fund A but still suggests a strong correlation with the stock market.  The correlation between Fund B returns and CPI was .18, a very week correlation.  As with Fund A, the standard error indicates that the S&P 500 accounts for nearly all of the difference between the model estimates of returns and the actual returns but the CPI estimator slightly improved these differences.

 

Conclusions

Based on this analysis, we conclude that each of these listed infrastructure funds fails to deliver what their marketing material suggests – diversification of risk of the equity market and a hedge on the risk of inflation.  Instead, the investment strategies that these two funds employ actually provides direct “beta” exposure to the S&P 500 Index – Fund A on a point-by-point basis and Fund B with a multiplier to the S&P 500 returns.  In addition, investors in these strategies should not expect much in terms of a hedge against inflation.

As for why the strategies of these infrastructure funds so badly misses the investment outcomes that their sponsors advertise, this is a question that we will investigate in a future paper.

 

Appendix: Model Results

 

Regression Statistics


Multiple R

0.979

R Square

0.959

Adjusted R Square

0.957

Standard Error

5.137

Observations

43

 

ANOVA

         

 

df

SS

MS

F

Significance F


Regression

2

24562.77542

12281.39

465.4089

1.98814E-28

Residual

40

1055.535352

26.38838

   

Total

42

25618.31078

 

 

 

 

 

Coefficients

Standard Error

t Stat

P-value


Intercept

2.816

1.169

2.409017

0.020692

S&P 500 Index

1.052

0.043

24.51265

1.04E-25

CPI

-1.341

0.634

-2.11402

0.040796

 

FUND B: SUMMARY OUTPUT

Regression Statistics


Multiple R

0.940

R Square

0.883

Adjusted R Square

0.880

Standard Error

10.297

Observations

86

 

ANOVA

 

df

SS

MS

F

Significance F


Regression

2

66226.81181

33113.41

312.3309

2.37E-39

Residual

83

8799.682323

106.0203

   

Total

85

75026.49413

 

 

 

 

 

Coefficients

Standard Error

t Stat

P-value

Lower 95%

Upper 95%

Lower 95.0%

Upper 95.0%


Intercept

-1.811

2.042

-0.88679

0.377753

-5.87175

2.250422

-5.87175

2.250422

S&P 500 Index

1.376

0.062

22.11266

1.53E-36

1.252295

1.499841

1.252295

1.499841

CPI

2.836

0.711

3.986877

0.000143

1.420955

4.250125

1.420955

4.250125