Seeking Alpha
Growth at reasonable price, dividend investing
Profile| Send Message|
( followers)  


  • If you want a portfolio that produces an income flow that keeps up with inflation, then you need some high distribution growth components in that portfolio.
  • All income stocks tend to sell at a reasonable yield plus distribution (or dividend) CAGR (Compound Annual Growth Rate) with significant adjustments for risk.
  • MLP (Master Limited Partnership) GPs (general partners) do not sell at reasonable "yield + CAGR" numbers. Why?
  • MLP GPs are undervalued because they have growth that appears to be too good to be true - and that lack of confidence results in them being undervalued.
  • MLP GPs have growth one can believe in because it is growth that one can forecast. Growth is good. Growth at a discount is great. Predictable growth is fantastic.

Before I start, you should be aware of the following two points: (1) Energy Transfer Equity LP (NYSE:ETE) has had a stock split in 2014 and that split is not accounted for in the data in this article. (2) The Cross-Tex general partner (XTXI) has changed its name and ticker to EnLink Midstream, LLC (NYSE:ENLC) - and the Cross-Tex limited partnership XTEX has changed to EnLink Midstream Partners, LP (NYSE:ENLK).

Some companies like Enterprise Products Partners (NYSE:EPD) and Magellan Midstream Partners (NYSE:MMP) come in one flavor - where retail investors have the choice of only buying one entity. Companies like Kinder Morgan Energy Partners (NYSE:KMP) and Plains All American Pipeline (NYSE:PAA) come in two flavors - where retail investors have the choice of buying the already mentioned MLP, or the incentive distribution rights owning general partners of Kinder Morgan, Inc. (NYSE:KMI) and Plains GP Holdings (NYSE:PAGP). This article focuses on those general partner options. It is the thesis of this article that MLP GPs are undervalued because they have consensus analyst distribution growth projections that appear to be too good to be true - and that lack of confidence in those projections results in them being undervalued based on their growth projection. In this article I will argue that MLP GPs have growth one can believe in because it is growth that one can forecast.

PAGP notes on its web site that incentive distribution rights (IDRs) entitle PAGP to receive increasing percentages of cash distributed by PAA. A different GP says that IDRs entitle the GP to receive an increasing portion of any cash distributed by the MLP. To put this into words that have more meaning to me, when the distribution grows at the MLPs, the cash flows going to the GP increase at an even faster pace, and that most often results in distribution growth at the GP being significantly faster than at the MLP. But that is only a general rule of thumb. There are GPs that own the IDRs to multiple MLPs. For example, Energy Transfer Equity owns the IDRs to Energy Transfer Partners , Regency Energy Partners and Sunoco Logistics Partners. (Technically - ETP owns the IDR's for SXL. But ETE owns a special class of ETP units that gives it the economic interest in 50% of the IDR's of SXL.) SXL has a higher five year CAGR projection than its GP ETE - and twice the last twelve month distribution growth. Williams Companies (NYSE:WMB) owns the IDRs to Access Midstream Partners and Williams Partners. Last twelve month distribution growth has been 23% for ACMP and 18% at WMB.

Those two factoids - about ETE and WMB - should serve as a warning to you. Financial writers provide a service to their readers by providing to them the general rules of thumb to the investment game. And every rule of thumb needs to come with a six-pack of asterisks. In order to provide articles of manageable length, not all of the asterisks get included in our writings. Not all writers know all the asterisks. Everything I am going to ever write is one asterisk away from being wrong.

What have been the distribution growth rates for the GPs?
The spreadsheet below shows the data on GP dividend and distribution growth. Growth between 2010 and 2011 was below trend because we were still crawling out of the credit crisis. The sector average growth between 2011 and 2012 was 48.15%. The high growth from ATLS had the effect on the sector average of (243/11) 22 percentage points. Let's subtract that out from the sector average, to get a run rate of 26%. The growth rate between 2012 and 2013 was 18% - but that was a year when ETE had a growth rate well below trend. The growth rate between 2013 and 2014 was 19.36% - but that stat includes a zero growth rate for PAGP. So casting out those few exceptions, we have a fairly consistent growth rate of around 20% for the last three years.

GP Distribution History
In a programming shortcut - I have encoded dividends for PAGP for 2013 when there were none. This error in history will continue until PAGP has a year over year record

 Distribution / Unit / Quarter% Distrib Growth / Year


I will take a shortcut and use a display of the distribution growth rates for the large cap midstream MLPs as a substitute for the typical MLP in this coverage universe. The sector average growth between 2010 and 2011 was 6.01%. The sector average growth between 2011 and 2012 was 6.47%. The growth rate between 2012 and 2013 was 7.26%. The growth rate between 2013 and 2014 was negative 1.5% but this was a year BWP's fall had a (81/12) -6.75 percentage point influence on that stat. So casting out that one exception, we have a fairly consistent growth rate of around 6% for the last four years.

Large Capped MLPs [ based on Q1 Distributions]

 Distribution/Unit/QuarterPercentage Distribution GrowthAv Growth


Average       8.812.876.016.477.26-1.53

What are the distribution growth projections for the GPs?
In prior articles, I have preached that one should invest with distribution CAGR awareness. Also, one should gather CAGR projections from numerous sources. (The sources are redacted with the numbers 1-4 replacing those names.) One should justify those projections with supporting metrics like the trend in earnings growth; the trend in distribution growth; the distribution to earnings ratio; and the price implied CAGR projection as calculated by using the dividend discount model formula. That information is in the spreadsheet that follows:

  20142015LTM price 
Co.Broker CAGR EstDCF/yrDCF/yrDist/yrDCF/DistimpliedMLPGPExplanation

ATLS34.433. growth spurt of uncertain length
AHGP9. GP - but coal is risky
ETE16.618. is growing again
KMI6. problems with KMP hurt KMI
NSH0. growth MLP results in bad GP
OKE22. div growth history
PAGP21.016.324.70.0na23.9%0.0%1.427.708.018.0too new to get much from the numbers
TRGP19.019.024.427.914.1%18.4%32.8%1.3010.676.918.5good G&P MLP may result in good GP
WGP27. high growth limited partner
WMB15.921.919.020.613.6%12.0%18.8%1.406.995.416.5provided 20% growth guidance
XTXI28. merger spurs growth

I set conservative CAGR projections. And in this case, I probably erred on the side of being overly conservative. The sector average CAGR projection is 15.5% compared to a current growth trend of around 20%. I want you to remember that I low-balled the CAGR projections when you compared the MLP vs. GP "yield + CAGR" numbers.

What are the underlying MLP growth projections?
After showing the GP CAGR projections "that I can believe in" are strongly based on the CAGR projections of the underlying MLP, it is now required that I also display my justification for my CAGR projections for those MLPs. This is shown in the following spreadsheet (with a very long header):

Growth Metrics for CAGR Assignment

The first seven columns for CAGR projections from seven different sources - with the source names redacted. The next four columns show the DCF actual and projected growth for last year and each of the next three years. The next two columns show projected EPS growth over the next two years. The next two columns show the distribution growth over the last two twelve month periods. Those are straight forward stats that require little explanation. This is where the straight forwardness ends.

The projected DCF to distribution ratio for 2014 uses the current distribution - while for 2015, I use the current distribution times its CAGR to get a projected 2015 distribution. And the ratio for 2016 uses the projected 2016 distribution. If the ratio is falling, then it indicates that my CAGR estimate is too high - and that it fails to match the distribution growth expectation that one should have given the DCF growth. And if the ratio is rising, it indicates that my CAGR could be too conservative.

The P-I CAGRs are the 'price implied' projections. I use a version of the dividend discount model formula to calculate the first number. The formula is -100 times [[4 times the Q1-14 distribution divided by the current price] minus the individually assessed required rate of return].
It is logical that CAGRs also influence the price to DCF ratio. The next formula attempts to capture that influence. The formula is 1.1 times [[the current price divided by the 2015 DCF projection] minus an adjustment that captures the influence of varying require rates of return]. The dividend discount model calculation may fail to capture the full effect of varying DCF payout ratios. Having a second price implied CAGR number helps offset that potential shortcoming.
The last column is my CAGR assessment. This is the third of three different spreadsheets I use to arrive at a CAGR assessment. I try to arrive at a CAGR that is aligned with the overwhelming majority of these CAGR indicators. But with that many ducks - or indicators - you are never going to get them all in a row.

Co.Analyst CAGR EstDCF/yr Growtheps/yr growthdist/yr growthProj DCF/DistPI-CAGRMy


There is data in the above spreadsheet that falls into the "missing asterisk" category. Given that this is a very important spreadsheet, let me provide some of those asterisks.

The spreadsheet provides data on CAGR projections from seven different sources. I use five year CAGRs when they are provided. But some of the sources only offer "three year" forward CAGR projections. I know from brokerages that offer both five and three year projections that the mixing of those two numbers is like mixing apples and oranges. I err on the side of favoring exposure to multiple opinions. I lack 2014 sources for some of those projections. That is also a bad thing.

Let's zero in on KMP for a minute. The projections vary from 2.8% to 5.8%. What if you only had access to one or two projections? You who might be buying could have a very different impression than you who might be selling due to having access to very different projections. And that is what makes an unjustified CAGR projection such a bad thing.

I know how this advice is sounding to some of you. You are thinking that I am borrowing investment instruction from the lyrics of a Billy Joel song:

First I'll tell you - you have to find several projections
Then I'll tell you - you have to ignore them away

I want to have CAGR projections in which I have confidence. The most reasonable projection usually lies within the range of projections that already exists. You need access to multiple projections in order to discover that range.

Getting back on topic: There are examples of wide spreads in projections for a lot of other MLPs. What's an investor to do? A good CAGR projection is your best tool in your data arsenal. At the same time, using CAGRs can also result in you using a very bad number to justify what turns into a very bad decision.

Let's continue with KMP and skip to the end. My CAGR projection for KMP is 3.2%. I suspect that some of you are concluding that this was a compromise projection - and others thinking that I chose to side with the low ball projections. In this one case - both of those suspicions would be incorrect. In prior articles, I have talked about "historical DCF projection accuracy" as a metric that I use in assessing my required rate of return. I know from those stats that my consensus DCF projections for KMP have been very accurate. KMP provides very good earnings guidance and strongly tends to deliver the performance to which they have guided. They have assets and a business model that leads to accurate projections. And that results in the "projected DCF/distribution ratios" being a very good tool for confirming a CAGR projection. Using 3.2% for the distribution CAGR leads to an expectation of 106% coverage of the distribution in 2014, 104% coverage in 2015, and 104% coverage in 2016. I would be more comfortable with the 3.2% projection if the coverage ratio was not falling - but it is not falling by much. And it is in line with prior performance. With DCF (Distributable Cash Flow) growth averaging (2 + 8 + 1 + 3) 3.5% over the 2013 to 2016 time period -- my 3.2% CAGR projection is slightly conservative. And I want lightly conservative projections.

In this and prior articles, I have told investors to know their portfolio CAGRs by using forward CAGR projections from multiple sources. Despite that warning, many use distribution inertia as there only guide. KMP had 11.2% last twelve month distribution growth in Q1-2013 and the current stat for 2014's last twelve month period is 5.4%. My CAGR projection is very different from any number that one could derive using inertia.

There was a fair amount of added detail in those last two paragraphs on KMP. That kind of detail could be done for each and every stock in the spreadsheet. But we don't have time for that today. Let's now move on to the other columns.

While I provide a brief description of the calculations for two different "price implied" CAGRs - a full explanation is probably needed. That would require another long article - an article I am sure that next to no one would read. The first calculation used a form of the Dividend Discount Model. But I know that the dividend discount model produces realistic valuations about two thirds of the time. For example, the payout ratio for BWP is so low that the current model says it is currently 63% over valued. Any calculation for BWP based on the DDM would provide a "garbage in - garbage out" price implied CAGR. Due to variances in the payout ratios, there are small problems in leaning heavily on the output of this formula.

There are also problems with a price implied CAGR based on the price/DCF ratio. This is a very experimental metric. The formula is a combination of logic and the result of playing around with the numbers till I liked the results. I was unable to find a second source of someone else doing this. So I lack verification that the formula is anywhere close to right.

What are the growth multiples for the GPs compared to the MLPs?

One of the tools I used to understand and value MLP GPs is the calculation of distribution growth multiples. For example, if the MLP is growing distributions at 4% and the GP at 12%, the growth multiple for the GP is three times (which is express as "3x") the growth rate of the MLP. The spreadsheet below provides details on those stats for the last twelve months.

MLP to GP Distribution History Comparisons

 MLP DistributionGP DistributionGP/MLP Dist
Co.Q1-13Q1-14GrowthQ1-13Q1-14GrowthGrowth Ratio

APL/ATLS58.062.06.9030.046.053.337.73 x
ARLP/AHGP108.5119.810.3774.082.811.821.14 x
ETP/ETE89.492.02.9463.569.39.063.08 x
KMP/KMI129.0136.05.4337.041.010.811.99 x
OKS/OKE71.073.02.8236. x
PAA/PAGP56.361.59.3312.512.50.000 x
NGLS/TRGP68.074.89.9345.860.832.793.3 x
WES/WGP52.060.015.3816.523.140.212.61 x
WPZ/WMB82.889.37.8533.940.318.802.39 x
XTEX/XTXI33.036.09.0912. x

Average 7.28  21.293.22 x

The "last twelve month" picture is a snapshot of the data at one point in time. And the picture you receive from one point in time can be deceiving. In an effort to derive a better picture of the growth multiple, I like knowing the median multiple over one year periods for the past three years. The spreadsheet below does that:

MLP to GP Growth Multiples

For the GPs that paid a distribution in Q1-2011 and had MLPs that had distribution growth in 2013, this spreadsheets is used to determine the median growth multiple - which is GP distribution growth divided by MLP distribution growth. The distribution growth multiple is then multiplied by the MLP CAGR. That result can then be compared to my GP CAGR - and those numbers should be in close alignment.

 2013-20142012-20132011-2012MedianMLPGM BasedActual

APL/ATLS6.9053.337.73 x5.4525.004.58 x48.65242.864.99 x4.
ARLP/AHGP10.3711.821.14 x9.6016.081.68 x15.1220.741.37 x1.
KMP/KMI5.4310.811.99 x11.2119.351.73 x4.503.330.74 x1.
OKS/OKE2.8211.113.94 x16.3918.031.1 x7.0217.312.47 x2.56.716.816.0
NGLS/TRGP9.9332.793.3 x12.8636.062.8 x10.0530.583.04 x3.06.920.718.5
WPZ/WMB7.8518.802.39 x8.5230.813.61 x8.5429.503.45 x3.45.418.416.5
XTEX/XTXI9.0925.002.75 x3.139.092.91 x23.0822.220.96 x2.78.523.021.0

This spreadsheet also takes the median multiple and multiples that times the current CAGR projection for the MLP. There can be no multiple if there has not been distribution growth. And that eliminates ETE and NSH from this data. And there can be no three year history if the GP was not in existence for more than three years. And that eliminates PAGP and WSP from this data.

I want to know if the GP CAGR is in alignment with the MLP CAGR. I see strong alignment in the current data. More to the point, I see strong alignment when using the three year median multiple to determine that alignment. And the fact that there is alignment in nearly all the GPs for which I have data provides some confidence in the CAGR projections for which a lack three year data.

This may be a personal psychological quirk that I do not share with this audience. But I have more confidence in the high GP CAGR when it is justified by a multiple of the lower growth MLP CAGR. My gut reaction to twenty percent plus CAGR projections is that "I can't trust that. It is too dang high to believe". My gut reaction to a twenty percent plus trend in distribution growth is that "It is too good to continue." I need the psychological crutch of using multiples to lean upon before I can step forward to believe in the idea that distribution growth can continue to be this high. Judging by the current valuations - and I am about to show those spreadsheets - the "market" is not that dissimilar to my gut reaction. It strongly appears that they do not trust these high CAGR projections. Lucky for me, I am - sometimes - smarter than my gut.

What are the relative valuations?
There is one data presentation in which I have some lasting pride - and this is it. I give it an A for its unique combination of completeness and brevity. It shows the "yield + CAGR" and two justifying components (bond ratings and historical DCF projection accuracy - with lower accuracy ratings being the better ratings) along with the Required Rate of Return (or RRR) assessment. It offers a valuation judgment based on yield + CAGR (or Total Return expectation) minus the RRR. It sums up everything that one needs to know. In my full data presentation, the CAGR, DCF Accuracy and RRR assessment have their own supporting spreadsheets. And there are components in the CAGR assessment that have their own supporting spreadsheets. But once those metric assignments have been justified, these metrics need to be utilized in combination to arrive at a valuation judgment.

Yield + CAGR Total Return Expectations for MLPs with GPs

CompanyQ1-14ConsensusTotalBondsDCFMyTotal RtnConsensusPr ImplDistribution
  YieldCAGRReturnRatingsAccrRRRs- RRRRatingsCAGR/14 DCF

Access Midstream Partners, L.P.ACMP3.78%9.60%13.38%BB-1.5012.001.381.78.2266.27
Atlas Pipeline Partners, L.P.APL7.89%7.00%14.89%B+5.0013.501.391.85.6184.93
El Paso Pipeline Partners, L.P.EPB7.90%2.50%10.40%BBB-1.2010.300.103.12.40101.56
Energy Transfer Partners, L.P.ETP6.59%4.00%10.59%BBB-3.0010.70-
Kinder Morgan Energy Partners, L.P.KMP7.11%3.20%10.31%BBB1.0010.000.312.52.8994.12
Targa Resources Partners LPNGLS5.13%6.90%12.03%BB1.7012.20-
NuStar Energy L.P.NS7.59%1.50%9.09%BB+4.0011.50-2.413.13.91113.77
ONEOK Partners, L.P.OKS5.21%6.70%11.91%BBB1.5010.501.412.15.2990.68
Plains All American Pipeline, L.P.PAA4.37%8.00%12.37%BBB1.0010.002.371.85.6382.27
Regency Energy Partners LPRGP6.97%5.00%11.97%BB-3.2012.00-
Sunoco Logistics Partners L.P.SXL2.98%9.00%11.98%BBB-1.0010.001.982.67.0258.37
Western Gas Partners LPWES3.55%9.10%12.65%BBB-1.7011.201.451.97.6577.67
Williams Partners L.P.WPZ6.91%5.40%12.31%BBB3.0011.001.312.24.0997.01
EnLink Midstream Partners, LPXTEX4.91%8.50%13.41%B+4.0013.50-

Average 5.78%6.17%11.95%    2.33

GP Yield + CAGR Total Return Expectations

CompanyQ1-14ConsensusTotalBondsDCFMyTotal RtnConsensusPr Impl
  YieldCAGRReturnRatingsAccrRRRs- RRRRatingsCAGR

Atlas Energy, L.PATLS4.50%23.00%27.50%B5.0016.0011.502.311.50
AllianceHoldings GP, L.P.AHGP5.17%9.00%14.17%none3.0013.500.673.08.33
Energy Transfer Equity, L.P.ETE2.96%14.00%16.96%BBB-3.0010.506.461.97.54
Kinder Morgan, Inc.KMI4.97%6.50%11.47%BBB1.0010.500.972.35.53
NuStar GP Holdings, LLCNSH6.16%2.50%8.66%BB+4.0012.00-3.343.05.84
ONEOK Inc.OKE2.55%16.00%18.55%BBB1.5010.008.552.27.45
Plains GP Holdings, L.P.PAGP1.83%18.00%19.83%BBB2.009.5010.332.27.67
Targa Resources Corp.TRGP2.30%18.50%20.80%BB2.0013.007.802.510.70
Western Gas Equity Partners, LPWGP1.91%24.50%26.41%none3.0011.5014.911.69.59
Williams Companies, Inc.WMB3.90%16.50%20.40%BBB3.0011.009.402.27.10
EnLink Midstream, LLCXTXI1.72%21.00%22.72%B+4.0016.006.723.014.28

Average 3.45%15.41%18.86%    2.38

Do I need MLP GPs in my MLP portfolio?

In a prior article - How much dividend income growth do you need -
I discussed one of the four keys to investment success. One does not know the amount of the income growth component you need in your portfolio until you know your portfolio income CAGR. That article was written for those investors who plan to live on the income provided by their portfolio - and never touch capital. And there are lots of investors who have that plan. If my portfolio can generate 4% income and grow that income at a 6% rate, then so can yours. And with the BLS telling us that average U.S. Consumer Spending is growing around 5% per year, we need a portfolio with some strong growth components. I am not telling you that inflation is that high. But to keep up with the Jones, our spending needs to grow that fast. In order to have "visiting" children, you have to have a Jones-competitive spending pattern to stay in stock with cool stuff for the grandkids to do during that visit.

I am not going to rehash that article again. Do your homework and know your portfolio CAGR. Buy and plan for the growth that you need - something like 5% growth. And remember the advice of Smokey the financial planner - "only you can prevent CAGR deficiency".

You would also like a tax efficient exit strategy. I hope to stay invested in MLPs for life. But there is no telling what Washington might do to mess that plan up. With K-1 generating MLPs having significant tax deferral until their sale, exiting MLPs can be expensive. I own two "negative UBTI producing" MLPs in an IRA to lessen the pain of any potential exit. I also own some c-corp non-K-1 generating GPs without tax deferral that will also lessen the tax burden if an exit is called for. Invest like an ex-Cub Scout and "always be prepared".

I listed the "pros" - what are the "cons"?
Share price appreciation is highly correlated to distribution growth as long as the projection for forward distribution growth is in alignment with historical growth. And that has been great news for the GPs. But high growth projections can dissipate. Your typical GP sells at a 3.5% yield because of the high growth projections. Reduce that growth rate to zero, and the price will fall to result in at least a 6% yield. You need diversification in your high CAGR investments. The cost of owning them is a near constant (or quarterly) CAGR awareness vigilance. You should be doing that already - and for all your investments. But high CAGR investments are like a Christmas gift with an "open me first" sticker.

So which GPs should you be buying?

First off, history has shown the MLPs and GPs have had a "sell in May and go away" pricing pattern over the past several years. I don't argue with history - even when it lacks sense. Second, valuations show the sector to be at least fairly valued at the present time. I see no urgency in buying today. You have time to do an even better job of due diligence. Use that time. Buy in slowly. And be diversified in your purchases.

I first purchased shares in TRGP in August of 2011 at $28.42. I purchased more in January of 2012 just under $40. After such a bullish run in such a short period of time, most investors would have thought that they were too late to the party to be buying more. Those shares closed Friday at $104.51. The talking heads on CNBC would tell me to sell half of my shares because you would then be playing with "house money" going forward. It is my opinion that such advice would be good for those retail investors who fail to grasp the metrics used to determine valuations - meaning something like maybe 80% of all retail investors. But I know the growth multiples for the GPs. I know the projections for the MLPs. I am comfortable with a growth projection of at least 20% per year for TRGP over the next four years. I should come close to doubling my investment over that time. The rest of my portfolio is not going to come close to doing as well. So until 2018 (or if the operating metrics significantly change), the only way I am going to let loose of my shares is for them to be pried from my cold dead fingers. Behavioral finance tells us that retail investors tend to sell their winners too quickly and hold their loser too long. I fully expect to be suboptimal in my timing decision. No one buys at the bottom and sells at the top. (That task will be done by some rapid trading computer.) So when I err, I will err by holding too long.

The two other GPs that I own are companies that have MLPs which have investment grade rated debt - and they are both GPS to midstream MLPs. TRGP is GP to G&P (gathering and processing) NGLS which does not have investment grade debt. I strongly believe that it is important to have a MLP portfolio that most heavily weighted in midstream companies - and companies that have investment grade rated debt. While it is important to invest with CAGR awareness, one can do that while also having RRR awareness.

Having expressed my (somewhat extreme) bullishness on TRGP, I need to inform you of one significant bearish note. The yield on TRGP is only 44.88% of the yield for its MLP. The sector average "percent of yield" is around 63%. This means that the relative valuation of TRGP to NGLS is high. Let me provide that information for my full coverage universe. You should also remember that this and other MLP to GP comparisons only uses data from the primary MLP in those comparisons. In other words, it is a little bit more complicated than this - while at the same time, this content that I am providing to you is already complicated enough for this expected audience.

'What if' Yields

What if the CAGRs are accurate and the distributions increase as projected - then after projecting that distribution growth, what would be the yields of the GPs and MLPs based on today's price? The spreadsheet below does those projections. It takes the current distribution and multiplies that by one plus the CAGR percentage to get a projected 2015 distribution - and then displays the current yield based on that projected distribution - and then produces a ratio between the GP to MLP yield. And the spreadsheet continues that projection into years 2016, 2017 and 2018.

Projected GP to MLP Yield Ratios [Using CAGRs] Based on Current Price

 Current YieldsProj Q1-15 YieldsProj Q1-16 YieldsProj Q1-17 YieldsProj Q1-18 Yields


Average3.806.4862.77  67.64  73.03  79.02  85.66

My suggestions would include any of the companies that have near 20% total return projections in my "Yield + CAGR" spreadsheet. But once again, my full advice is a little bit more complicated than that. If you are pondering a decision between KMP and KMI, you can use the above data sets to justify your selection for KMI. On the other hand, if you are pondering a decision between ETP and ETE, you can use the above data sets to justify your selection for ETP (by using the by using the project Q1-17 yield ratio number). On the third hand (grin), ETE is also GP to SXL and RGP - so that complicates or taints the use of that metric in your decision process. My strongest suggestion is to diversify and own at least three - and have most of your investment dollars in the lower RRR companies.

Source: MLP GP Distribution Growth Tells Me They Are Undervalued