MLPs Extremely Overvalued Based On Historical Relative Yield

|
Includes: CQP, EEP, KMP
by: James A. Kostohryz
Some investors are prone to jump to the conclusion that if the distribution yield of an MLP is higher than the dividend yield of a particular stock that the MLP is a better investment. This is not necessarily the case. A higher distribution yield does not imply superior relative value in isolation. Yield must be combined with other factors to determine relative value.
Investors need to remember that there are reasons why MLPs have high distribution yields. First, as shown here, MLPs such as Enbridge (NYSE:EEP), Cheniere (NYSEMKT:CQP) and Kinder Morgan (NYSE:KMP) pay out more than 80% of available cash flow as distribution yields. By contrast, the average utility or telecom stock, for example, pay out only 24.9% and 36.7% of their cash flows, respectively. Amongst other things, this implies that the latter will be able to sustain cash flow and dividend growth at higher rates on a per share basis than MLPs going forward. Thus, it is total free cash flow from operations and growth of total free cash flow – not just cash distributed to shareholders – that should be the primary consideration in valuation analysis. In the long run, reinvestment drives per share growth of dividends.
Second, even when comparing yields in isolation, it is a mistake to compare them only in a particular moment in time. Relative yields must be compared on a historical basis. For example, all other variables remaining equal, MLP distribution yields are only attractive to the extent that they are higher than they have been historically compared to other sectors. While it is true that MLPs currently offer higher yields than are currently offered by the S&P 500, Utilities and Telecoms, the fact of the matter is that the distribution yield of MLPs are relatively unattractive in historical terms, compared to these other sectors.
Historical Analysis of Alerian Index Yields
I have downloaded the price and yield data for the Alerian Index directly from the Alerian website. The data for the dividend yield of the S&P 500, S&P Utilities Index and S&P Telecoms Index is taken directly from Bloomberg (unlinked). The present analysis covers the entire period since the inception of the Alerian Index in January of 1996 through August 30, 2011.
Comparison to S&P 500
The evolution of the historical relationship between the distribution yield of the Alerian Index (AMZ) and the S&P 500 is provided below.
[Click all to enlarge]
AMZ/SPX Yield Ratio
AMZ/SPX Yield Spread
In historical terms the relative yield of the Alerian Index compared to the S&P 500 is near all-time lows. The current yield ratio is 3.11 versus an average of 4.61 since 1996. This is historically in the 90th percentile of relative overvaluation for the Alerian Index since inception. The current yield spread is 4.49% versus n average of 5.97% since 1996. This is historically in the 82nd percentile of relative overvaluation for the Alerian Index since inception.
Based on this analysis, the MLP sector is substantially overvalued in historical terms compared to the S&P 500 based on relative distribution yields.
Comparison to S&P Utilities
The evolution of the historical relationship between the distribution yield of the AMZ and the S&P Utilities Index is provided below.
AMZ/Utilities Yield Ratio
AMZ/Utilities Yield Spread
In historical terms the relative yield of the Alerian Index compared to the S&P Utilities Index is near all-time lows. The current yield ratio is 1.56 versus an average of 2.17 since 1996. This is historically in the 93rd percentile of relative overvaluation for the Alerian Index since inception. The current yield spread is 2.38% versus n average of 4.10% since 1996. This is historically in the 93rd percentile of relative overvaluation for the Alerian Index since inception. Based on this analysis, the MLP sector is substantially overvalued in historical terms compared to the S&P Utilities Index based on relative distribution yields.
Comparison to S&P Telecoms
The evolution of the historical relationship between the distribution yield of the AMZ and the S&P Telecoms Index is provided below.
AMZ/Telecoms Yield Ratio
AMZ/Telecoms Yield Spread
In historical terms the relative yield of the Alerian Index compared to the S&P Telecoms Index is near all-time lows. The current yield ratio is 1.24 versus an average of 2.81 since 1996. This is historically in the 91st percentile of relative overvaluation for the Alerian Index since inception. The current yield spread is 1.26% versus n average of 4.38% since 1996. This is historically in the 91st percentile of relative overvaluation for the Alerian Index since inception. Based on this analysis, the MLP sector is substantially overvalued in historical terms compared to the S&P Telecoms Index based on relative distribution yields.
The Difference Between Yield Ratio and Yield Spread
Some analysts prefer to analyze yield ratios and some prefer to analyze yield spreads. There are advantages and disadvantages of each. A yield ratio analysis compares the evolution of the ratio of yields between two stocks or sectors. Yield A is divided by Yield B. In this case, we track the ratio between the distribution yields on MLPs and the dividend yields of three groupings of stocks: S&P 500, Utilities and Telecoms.
The advantage of comparing yields in terms of a ratio is that it expresses a common understanding of relative value. For example, all other factors remaining equal (which as I have pointed out, they never are), an MLP that provides $10 in distributions should be worth twice as much as a stock that provides $5 in dividends. When we say that the MLP is worth twice as much we are expressing the relationship of relative value in terms of a ratio.
Some analysts prefer to analyze the relationship between yields in terms of a spread. A yield spread is calculated by subtracting Yield B from Yield A. A spread provides information regarding the difference between the rates of return on investments. However, spreads do not indicate the proportions of relative value. For example, if stock A yields 10% and Stock B yields 8%, the spread is 2.0%. If stock C yields 4% and stock D yields 2%, the spread is also 2%. But note that, in theory, all things remaining equal, stock C should be two times more valuable than stock D, while stock A should only be 25% more valuable than stock B.. This fact is not reflected in the spread analysis whereas it is reflected in the ratio analysis.
Why not compare the Alerian yield to Treasury bonds? Some investors like to compare MLP distribution yields to the yield on the 10Y Treasury bond. Such an analysis is inferior to the analysis of relative dividend/distribution yields provided above for three reasons. MLPs are equities, not bonds. MLP distribution yields are best compared to the dividend yields of stocks. This more of an apples-to-apples comparison. Relative yields of C-Corps more attractive. While MLP distribution yields are currently attractive relative to 10Y Treasury bonds, the yields of the equities in the sectors analyzed above are even more attractive relative to Treasury bonds in historical terms. Thus, comparing MLPs to Treasuries is a pointless exercise in this context. 10Y Treasury Bond is an overvalued benchmark. Currently, it is a fundamental mistake to utilize the 10Y Treasury Bond as a benchmark for value. The yield on the 10Y Treasury Bond is being intentionally repressed by the US Fed. For this reason, the yield is artificially low. Furthermore, for reasons that I discussed at length in previous articles, yields on 10Y Treasuries are not sustainable in the long term at current levels. For various reasons, Treasury bonds are experiencing a value “bubble” at the present time. Thus, saying that MLPs are cheap relative to US Treasury Bonds is a bit like saying in 1999 that a bricks-and-mortar stock with a PE of 70 was “cheap” because internet stock PEs were over 200.
Conclusion
MLPs have rarely been as expensive as they are now in terms of a comparison of their distribution yields in relation to the dividend yields of stocks in comparable sectors. MLPs are also overvalued in historical terms on a PE, P/CF, EV/EBITDA basis. I will show this in future reports.
One of the advantages of performing the sort of historical analysis demonstrated here is that factors discussed in parts two and three in this series such as steadiness of cash distributions, tax advantages, business risk and beta are rendered mute. For the most part, these factors have not changed significantly for well over a decade. Thus, there is absolutely no fundamental reason to believe that MLPs should command higher premiums versus the S&P 500 Utilities or Telecoms than was the case a decade ago.
Why and when this relative overvaluation of MLPs will correct itself is the subject of future pieces. For now, potential buyers and owners of MLPs simply need to be aware that stocks in this sector are generally not attractive in terms of historical relative valuations. Long-term total returns look more promising in the utilities and telecom sectors, to name just two pertinent examples.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.