This is part 2 of the post I wrote last month on popular misconceptions in the MLP space. The idea came from Ken Fisher’s recent book Debunkery where he debunks all of the popular investing maxims. He had 50 in his book, I’m trying for just 10. So here goes round 2…
MLPs Have Limited Correlation with Other Asset Classes
From an RIA that sells his services as an MLP focused asset manager: “MLPs have strong historical returns and a low correlation to other asset classes”.
That’s just not true anymore. The day to day correlation is very high between MLPs and the major stock indexes, particularly in times of stress, as I discussed here. All risk assets are correlated these days in crisis situations. What makes the longer term correlation break down are the distributions and the long term out-performance of the MLP sector. MLPs have outperformed the S&P 500 in each of the last 11 years on a total return basis. From 2001 to 2010, the MLP Index averaged annual total returns of 14.4%, compared with 1.7% for the S&P 500. The chart below is from my post in late September, correlation has dropped somewhat since then, but should still finish the year very high relative to pre-2008 levels.
Correlation will continue to inch higher for MLPs as more and more of the sector is held by investors through easy to exit exchange-traded pooled investment vehicles (ETFs, ETNs, closed-end funds, open-end funds). And with so much of the returns in the MLP index coming from NGL-related midstream services (NGL prices tend to move with crude prices), and with 7 E&P MLPs now in the Alerian MLP Index (50 MLPs total), MLPs are going to trade more inline with oil prices going forward.
Best Way to Value MLPs is Using Relative Yield
Relative yield was a really good way to value MLPs in the mid 1990s, when interest rates generally weren’t manipulated to extreme lows, before MLPs really started to grow distributions aggressively. Back then, they acted more like tax deferred bonds. Relative yield is a leftover metric from that bygone era. Certainly investors still look at and like to quote yield as it relates to the income they receive. But to take the yield of a high growth MLP and compare it to yield on Ferrellgas or some other very low growth MLP, and try to make a judgement on how far apart the yields should be and therefore decide one is cheap compared with the other is arbitrary. It’s just a way for research analysts to subjectively say one MLP should have a higher value than another MLP. There are only a few situations where looking at relative yield makes sense: like when comparing a GP’s yield to the yield of its subsidiary MLP.
MLP Cost of Capital is Equal to its Equity Yield
As mentioned above, when MLPs first came out, most of the total return of MLPs was driven by the distributions holders received. In the mid-2000s, that began to shift and since then an increasing amount of the total return of MLPs is derived from distribution growth and capital appreciation. As a result, analysts and investor have turned to distribution discount models as the primary means of valuing MLPs (with relative yield and EBITDA / DCF multiples as sanity checks).
Distribution discount models are very sensitive to distribution growth estimates, terminal growth rates, and the rate at which you discount the future distributions back to the present, also known as the equity cost of capital. Each research analyst and each MLP manager has a different metric for cost of capital it seems, and some analysts use yield as the starting point for their cost of capital estimates, adjusting the yield to reflect high distribution coverage and high GP takes.
This and similar methodologies tend to understate cost of capital for high growth MLPs that carry extra coverage. My version of cost of capital is more traditional, focused on cash flow risk (both of distribution cuts and cash flow volatility). My version also accounts for GP take and high distribution coverage ratios, but it also makes accretion more challenging / realistic, and doesn’t result in 75% of the sector looking undervalued at any given time.
MLPs With High GP Takes are Inherently Unattractive
There are 19 MLPs current trading that are currently in the 50% IDR tier (meaning 50% of all incremental distributed cash flow goes to GP). Of those 19, 7 are currently paying their general partner more than 25% of all cash flow (PAA, ETP, KMP, SXL, ARLP, WPZ, NRP). MLPs that are receiving more than 25% have certain common characteristics. They are typically mature, with a track record of growth in the past that has allowed them to reach the 50% IDR tier and beyond.
A high GP take makes growth more difficult, because to grow, the MLP must combat the large of large numbers (growing a larger MLP takes more cash flow to grow DCF/unit than does a smaller one), but also a growing cost of capital in the form of more cash to the GP. So, all else being equal, those MLPs that have lower GP takes are more attractive.
All is not equal, however. MLPs that have grown into high GP takes have typically done so by outperforming their peer MLPs, by executing better and growing smarter than their peers. The assets that they own and the management teams that have brought them into the 50% tier can make up for a more expensive cost of capital. Proof is in performance, as shown below.
These 7 MLPs have historically done better than the sector as a whole. There are exceptions (like ETP and NRP so far this year), but overall, the high GP take has not negatively impacted this group.
MLP Can’t be Purchased by Foreign Investors
For most MLPs, there are no restrictions on foreign investors. There are definitely complications with regard to taxation, including potential federal income tax filing requirement and distribution withholding at the highest effective tax rate (read here for more), but foreigners can buy MLPs. I have been corrected by a reader on my earlier example with regards to interstate pipeline assets: there are certain MLPs with interstate pipeline assets that cannot have tax-exempt investors (like BWP), but foreign investors are ok.
Related: Five Prevalent Misconceptions About MLPs
Thanks for reading. Feedback is welcome, let me know if you come across any other prevalent misconceptions with regard to MLPs.
Disclosure: The information in this article is not meant to be financial advice, we are not your financial advisor and I am posting my comments for informational purposes only.