Seeking Alpha
Yield + CAGR at a low RRR
Profile| Send Message|
( followers)

Why does the average Joe believe that equity investing is risky? Investing involves dealing with uncertainty. You do not control the outcome. Even the perception of risk is subjective. We can see risk when it is not there. We can fail to see risk when it is there. The fact that we have experienced two huge "once in a lifetime" corrections in the last 13 years has scared many retail investors away from equities. Like most retail equity investors, I lost over 40% in my retirement savings in the correction of 2008. I never cut my allocation to equities. I have now more than recovered those prior losses. But even with a full recovery, thoughts of those past losses occasionally haunt me. I do not want to be haunted - I want to be educated.

This perma-bull thinks the current equity market is doing the best it can to forget that prior loses ever happened. I am not. While my average loss was 42%, the losses did not fall evenly across each investment in my portfolio. I want to learn and employ the lessons from those years. I want to pump up my risk awareness. I want to be heavily in investments that will lose less come the next correction. I'm not going to pretend that I can time the market and change my allocation come the next recession. I want to be perma-prepped for that event without sacrificing much growth in income that a heavy "equity income" allocation can provide.

In this article, I will throw more data at you than most can handle in one reading. This is done to pump up your risk awareness. I strongly believe there are lessons in the historical data. I will share with you the secrets that the data has whispered to me. My interpretation of those whispers may differ from yours. Be forewarned that some of my interpretations are outside the consensus of opinion. I also believe that there is a strong emotional or psychological component in investment decisions. Unlike most financial writers, I awkwardly attempt to address some of those issues.

I have found an outside consensus definition of risk that works for me. I have found it in the data. That definition: "Risk is primarily found in the uncertainty of earnings." When it comes to MLPs {Master Limited Partnerships}, there is higher risk in owning stocks with the more commodity sensitive contracts. Lower risk is having more fixed fee contracts. Higher risk is having a geographical footprint in a few producing basins. Lower risk is footprint diversity. Lower risk usually involves larger companies, wider and longer pipelines, lower leverage, and longer track records. Those attributes are identifiable and quantifiable. What is the point of using a risk detection system that is nebulous?

I am given uncomfortable pause when I see a riskier than average credit rating. It concerns me. But a credit rating is a debt rating. On a gut level, the risk significance fails to register. Something less nebulous - like a high yield or high CAGR projection - can easily offset that pause.

The metric that is the focus of this article is DCF "projection accuracy." The numbers are calculated by taking the projection at the beginning of the year minus the actual yearly DCF number, divided by the actual number. I also want to briefly mention the "earning projection spread" metric - one that I use in other sectors. That metric is calculated by taking the high projection minus the low projection, divided by the consensus projection. In other sectors, there is a strong correlation between projection accuracy and projection spreads. The inconsistent size of analyst coverage may tend to make projection spreads a less meaningful metric when it comes to comparing MLPs.

The novice might think that all stocks are created equal when it comes to historical projection accuracy and current year projection spreads. This is not the case at all. You may have never heard of these metrics. I know I have never read about these specific terms in any analyst report. I view the market's lack of knowledge of these metrics as a good thing for me. It potentially provides me with an advantage. These numbers provide an objective sizing of risk. This may uniquely help male investors who have difficulty sizing anything (grin). It provides me with a psychological advantage. It makes risk assessment less subjective. It also makes risk assessment an ongoing activity and not an activity that is only done when the market is falling.

As stated before, below average credit ratings give me pause. When I see a DCF projection that I cannot trust going by the historical record, the pause turns in to a halt. This is a number that hits me in the gut. If a MLP lacks a history of strongly predictive DCF projections, I am strongly concerned on both a mental and emotional level. My measure of distribution safety is based on DCF coverage. When I lack certainty of the DCF projection, I lack certainty of the distribution coverage. Take away the comfort that coverage provides to me, and that investment option loses a lot of attractiveness. My key predictor for distribution growth is once again the distribution to DCF ratio. Take some of the confidence in the growth away and the attractiveness of growth recedes. DCF growth is a factor in my CAGR projection. A MLP with poor projection accuracy results in an uncertain picture of distribution growth due to DCF growth. I really hate uncertainty when it comes to DCF projections because it messes up too many ratios that I lean upon in making my valuation assignments.

Using the perspective of "risk is found in the uncertainty of earnings" does cause one to miss the mark in some aspects of risk. Income generating stocks do face a different kind of risk due to changing interest rate environments, a risk that is specific to MLPs in the concern of a change in tax treatment. Those two important issues are not addressed in this article. My focus is on generating a simple risk assessment system that assists in making valuation comparisons between different MLPs.

What are the MLPs that we, the retail investors, believe are the least risky MLPs? Going by both the message board chatter and the valuations, it strongly appears they are large cap midstream MLPs like Enterprise Products Partners (NYSE:EPD), Kinder Morgan Energy Partners (NYSE:KMP), Magellan Midstream Partners (NYSE:MMP) and Plains All American Pipeline (NYSE:PAA). Add Sunco (NYSE:SXL) and TC Pipelines (NYSE:TCP) to the list, and you have the list of the best MLP accuracy ratings. Is that a coincidence? I want to start where I left off in my last article with the "Yield + CAGR Total Return Expectations" spreadsheet that now contain the accuracy ratings - and then show the historical data on which the accuracy ratings are assessed. The data:

A 1 (or best) to 5 (worst) rating is given to each MLP based on the accuracy - or lack of unpleasant surprises - of the DCF estimates. The scoring system is informal. To be a 1 - the MLP will have estimate changes less than 10% over several years - and if there are large changes, then they are upgrades. I never count a large upgrade against an MLP in setting the accuracy rating. A 5 will have four or more negative DCF changes over 10%.

Yield + CAGR Total Return Expectations

CompanyQ3-13ConsensusTotalBondsDCFMyTotal RtnConsensusPr ImplDistrib
YieldCAGRReturnRatingsAccrRRRs- RRRRatingsCAGR/ DCF
Large Cap Midstream
El PasoEPB6.00%5.50%11.50%BBB-1.2010.401.102.94.4092.99
Energy TransETP7.02%3.50%10.52%BBB-3.0011.00-0.482.53.9886.14
Kinder MorganKMP6.43%5.50%11.93%BBB1.0010.001.932.73.5797.78
Plains All-AmPAA4.49%8.50%12.99%BBB1.0010.002.991.95.5172.09

Average 6.40%5.06%11.46% 2.56

CompanyQ3-13ConsensusTotalBondsDCFMyTotal RtnConsensusPr ImplDistrib
YieldCAGRReturnRatingsAccrRRRs- RRRRatingsCAGR/ DCF
Small Cap Midstream

Average 5.16%6.18%11.35% 2.60

CompanyQ3-13ConsensusTotalBondsDCFMyTotal RtnConsensusPr ImplDistrib
YieldCAGRReturnRatingsAccrRRRs- RRRRatingsCAGR/ DCF
Gathering & Processing
Atlas PipelinesAPL6.84%9.10%15.94%B+5.0012.503.441.45.6684.64
DCP PartnersDPM5.83%6.50%12.33%BBB-2.7011.600.732.15.7797.59
Eagle RockEROC13.52%2.50%16.02%B5.0013.003.022.6-0.52111.39
Mark WestMWE4.92%8.60%13.52%BB3.2012.001.521.77.0896.55

Average 6.74%6.51%13.25% 2.07

Changes in DCF estimates by Year: Some MLPs Have Assets That Produce More Predictable DCFs


ACMP 1.651.809%1.972.2012%2.382.7516%1.5
CMLP 0%1.951.66-15%2.252.07-8%2.181.97-10%1.752.0014%2.402.07-14%2.201.94-12%3.7
EPB 0%1.311.418%1.401.6921%1.782.2728%2.532.45-3%2.702.60-4%2.762.71-2%1.2
EXLP 13%2.073.1452%2.902.19-24%2.462.26-8%2.462.502%2.642.46-7%2.582.8510%2.2
NGLS 8%2.303.1437%2.802.55-9%2.572.653%2.893.3014%3.153.171%3.422.88-16%1.7
SEP 0%1.461.6211%1.832.0914%2.212.00-10%2.162.05-5%2.232.08-7%2.132.10-1%2.2
WES 1.401.400%1.651.58-4%1.752.2126%2.152.4012%2.512.676%2.902.52-13%1.7
NMM 1.601.600%1.701.700%1.901.900%3.312.10-37%1.701.8911%1.651.756%2.2

The track record on intra-year accuracy for EPS estimates is tracked in the data below. Given that one is used to seeing data like this in a year over year format, let me remind you again that this is not year over year changes in EPS. This test is done to see if the patterns in DCF data are echoed in the EPS estimates. I do not see a trend of over or under promising in this data. I only see evidence for varying levels of RRRs.

The volatility in this is data scares me. All the brokerages provide P/E stats as one tool in valuations. One side effect of being exposed to this intra-year estimate volatility data - you will never place much value in a G&P EPS estimate (or P/E ratio) again. I skim this data set to see if the accuracy assigned in the DCF estimates is echoed in the accuracy of the EPS estimates. And I believe that it does.

Changes in EPS estimates by Year: Some MLPs Have Assets That Produce More Predictable Earnings
Accuracy Rating is for DCF - this spreadsheet only tests to see if EPS gives the same picture of estimate accuracy.
Negative EPS projections crash this program. All projections have a minimum value of $0.05/share.


ACMP 1.411.35-4%1.521.11-27%1.5
CMLP 0.290.3210%1.280.96-25%1.281.27-1%1.271.23-3%1.451.22-16%1.470.77-48%3.7
EPB 1.090.13-88%1.071.2214%1.381.487%1.681.9516%2.162.12-2%2.272.15-5%1.2
EXLP 1.351.29-4%1.611.610%1.610.77-52%1.040.07-93%0.540.10-81%0.670.8425%2.2
NGLS 0.340.5150%1.071.8371%1.460.66-55%1.540.87-44%1.461.7520%1.921.34-30%1.7
SEP 1.281.19-7%1.291.409%1.461.7117%1.881.68-11%1.841.66-10%1.781.69-5%2.2
WES 1.240.77-38%1.411.21-14%1.341.5818%1.601.674%2.010.84-58%1.7
NMM 1.481.480%1.761.760%1.511.510%1.371.402%1.351.32-2%2.2

Do accuracy ratings and the S&P risk assessments of risk align?
The following midstream companies had their corporate credit ratings equal to BBB: BWP EEP KMP MMP OKS PAA SEP TCP WPZ. Their average accuracy rating is 1.79.
The following companies had corporate credit ratings of BBB-: BPL DPM EPB ETP SXL WES. Their current average accuracy rating is 1.98.
The following companies had corporate credit ratings of BB+, BB, or BB-: ACMP GEL HEP MWE NGLS NS RGP . Their current average accuracy rating is 2.51.
The following companies had corporate credit ratings of B, B- or were note rated: CMLP EROC EXLP TLP NMM TGP. Their current average accuracy rating is 2.72.

Here is one type of question that always bothered me. If MarkWest Energy Partners (NYSE:MWE) has a better CAGR projection than Enterprise Products Partners, then why does MWE sell at a large yield? I was never satisfied with the "because EPD is investment grade" answer. But I can totally buy this slightly different explanation: EPD has better DCF projection accuracy - and that makes the distribution more secure, and any growth projection more accurate. When there comes a time that MWE and EPD have close to similar yields, that is the time that I lighten my heavily over weighted allocation to MWE. I want my portfolio to grow more conservative as I age. I think MWE is a great investment - but it is not a conservative investment.

I do not want to overweight this article with too much data. I want to keep it as simple as possible. So let me save some data for future articles by summarizing some of my findings:

(1) Earnings accuracy explains RRR variances within a sector.
(2) Earnings accuracy explains RRR variances between sectors. When I get around to sharing REIT (Real Estate Investment Trust) or consumer staples data, you will be provided evidence of this.
(3) For those sectors where Yahoo Finance generates meaningful earning projections, the earnings accuracy is reflected in the spreads - between the high and low numbers - of current year earning projections.
(4) Earnings accuracy is reflected in dividend and distribution histories. The three MLPs in my current midstream coverage universe that had prior distribution cuts have poor accuracy records. Those MLPs that had pauses in distribution growth had less than average accuracy records. This is not consistently true in other sectors, but there is still some correlation.
(5) If part of the market is secretly giving value to this accuracy attribute, then the stocks with poor earning accuracy projections should logically have higher RRRs.
(6) The current MLP data confirms the perception that poor earnings accuracy is partially reflected in current valuations. This finding is echoed in the data from other sectors.

If I need to provide the data that supports those findings, I am sure the readers will let me know in the comment section at the end of the article.

Valuation lessons I am deriving from the data:

The difference between birds and turtles
In prior articles, I have described yield as one bird in the hand, while growth is the two birds in a bush. I now want to morph that analogy. When there is a higher level of certainty in DCF projections, yield is like one turtle in the hand, while growth is the two turtles that are a few yards away. When there is a higher level of earnings uncertain, yield is like one bird in the hand struggling to be free, while growth is like two quick and nervous birds in a bush. I'm willing to pay a reasonable premium for turtles over birds.

There are wise and wealthy folks who only see one kind of risk. They would tell you that there is increased risk in betting that a high CAGR stock will stay a high CAGR stock. They would tell you that the risk level is the same as betting on a higher yielding stock with increased earnings uncertainty.

I do not see it exactly that way. I see the risk as being different. I still want to invest "in baskets" when I buy growth stocks. One needs to cast a wide net to capture the projected superior performance because there will be some failures in the accuracy of the projections. But the superior performance is out there - waiting to be captured. On the other hand, earnings uncertainty is out there - waiting to bite you. Come the next recession, it will be out there biting somebody. I can handle a future where some of my holdings will once again have dividend and distribution growth pauses with the next crisis. I want to be making the decisions today that create a future where my income flow will be less in danger. I also want to be making the decision today where I'm capturing some of the huge gains that are yet to come in many growth MLPs. The data have pointed me toward high CAGR stocks to capture those potential gains. The data have pointed me toward higher earning projection certainty stocks - and a high weighting in those stocks - to minimize the pain that will find us all come the next season of unpleasantness.

There are many other findings that come with a perception of risk being defined as earnings uncertainty. To keep this article at a moderate length, I will only address four.

The three degrees of falling projections
Earning projections will be volatile - it is their nature. It is my perception that there are three different degrees of falling projections. First - there are small changes that do not influence the long term CAGR. Second - there are larger changes that influence or downgrade the CAGR projections. Third - there is a size of fall that can influence your perception of the accuracy of all future earnings projections for the stock - and cause the RRR to rise for that stock.

The Down-Side of Good Attributes
Earning projection volatility varies by stock. But what happens when historically non-volatile stocks become volatile? There is a triple-whammy to their price. When earning projections fall, prices fall. When CAGR projections fall, prices fall. And when the prior non-volatile stocks become volatile, their required rates of return also rise. This also causes the prices in the stocks to fall. Stocks with higher RRRs are already priced for higher earning projection volatility. Low RRR stocks are not. When really bad things - or significantly falling earning projections - happen to good stocks, the fall in share prices is steeper. This valuation shift happened with Buckeye Partners (NYSE:BPL) in 2012. This is still happening with NuStar Energy (NYSE:NS).

There are problems with other risk or leverage metrics
(1) All debt to market caps are not created equal - even if there was agreement in the debt level - due to varying interest rates. For example, two companies with 40% debt to market caps have different levels of risk to the equity holders due to differences in interest or coupon rates on their bonds.
(2) There is no agreement among the brokerages providing MLP updates in debt to market cap numbers due to complexities of "what is real total equity." The MLP CFOs have become creative in issuing restricted shares - or units - to raise money to fund their expansion via private equity offerings, or to compensate employees with restricted stock compensation. When you notice in an earnings release a significant difference between EPS and diluted EPS, you have just seen a red flag to warn you that this same problem could be echoed in the debt metrics. This potential problem is also echoed in the distribution coverage stat.
(3) There is no agreement among the brokerages providing MLP updates in the numbers because some brokerages use debt to total market capital, others use debt to book cap, and some use debt to enterprise value. These three different metrics provide different pictures of the degree of leverage. I could show you the numbers - but it would scare you into a state of cynicism concerning all metrics from which you might not recover.
(4) There is no agreement among the brokerages providing MLP updates in the Debt to EBITDA numbers. Brokerages have both different debt and different EBITDA projection numbers. Thus the picture of leverage for the same MLP can and do vary significantly.
(5) Some of the brokerages show RRR numbers - often labeled as "discount rates." For those that do show their numbers, there is often huge disagreement in these numbers. That should come as no surprise due to the lack of agreement on key debt metrics.

The risk awareness levels for retail investors is low
My statuses as a data provider has given me a license to poll on my favorite message board and generate sufficient replies to my questions. I have done several polls on risk awareness over the years. Some investors are doing several simplistic things in order to lessen the portfolio risk. For example, many MLPs limit their exposure to coal and E&P (exploration and production) MLPs. Some investors - maybe even most investors - are doing nothing. Even if you are doing something simple like limiting your weighting on non-investment grade investments, you are doing more to limit your level of risk than most. You can benefit by taking baby steps in growing your risk awareness. I would be hesitant to suggest that you improve your risk awareness if the opportunity costs were high. But the data suggests that the cost of risk awareness is very low.

A summary of the lessons in the data:
1 - I would be hesitant to but much faith in the DCF or earnings accuracy stats if they were not confirmed by the risk assessments found in the debt credit ratings. I find it logical that the numbers should matter. I see a logical correlation between the low accuracy stocks and their assets or contracts. The accuracy stats help explain some of the exceptions where market implied RRR valuations vary from credit ratings. The numbers make risk assessment objective and quantifiable. That is a huge plus.
2 - Risk is not a potential problem to be pondered when the market is falling. The RRR assignment is always important. You really need objective and quantifiable risk attributes in setting your valuation judgments. You need them all the time. Risk is a key valuation metric.
3 - Most investors who take the time and effort to deal with the complexities of MLPs end up with a significant allocation to MLPs. My personal allocation is right at 25%. So my skill at MLP selection is the key determinant of whether or not my portfolio is a winner - or a loser. If you are going to "go big," then you need to be sure you are doing it right. My recipe for doing it right requires that I have a higher level of metric awareness than the average investor - and harvest an advantage from that awareness. That is not a high hurdle to jump. There are simple tools one can use to accomplish that mission.
4 - The data indicates that there is a small opportunity cost for investing with risk awareness. At the same time, the benefit can be substantial.
5 - I generate my own data to assess risk. It is likely that there are alternative ways to my method that could be superior. I hope others are out there willing to share their own data on how they are assessing risk. SA provides a comments section after each article so that this kind of sharing can be done.
6 - MLPs are almost always priced at attractive "yield plus CAGR" numbers. Much of that premium is merited due to their higher than average earning projection uncertainty. It is not just the fear of K-1s and fear of tax changes that account for the higher total return projections.
7 - Projection uncertainty is another strong reason to have a diversified MLP portfolio. I want a lower risk portfolio. So I use these stats not only in individual decisions, but I also use them in weighting or allocation decisions. It is always good to have a strong weighting in investment grade MLPs with attractive distribution CAGRs.

Let's end with a review of where we have been in prior articles:

In part one of a series on CAGRs (Compound Annual Growth Rate projections of the distribution) I provided data and text on:
(1) Where can I find good dividend and distribution CAGR projections?
(2) Is there evidence that CAGRs strongly influence yields?
(3) Is there evidence that higher CAGRs sell at a discount?
(4) Is there evidence that CAGR awareness adds to total returns?

Part two provided data and text on:
(5) How volatile are CAGR projections?
(6) How many CAGR projections do I need to gather?
(7) What metrics or attributes should I use to double check a CAGR projection?
(8) Are the CAGR projections accurate enough to use in investment decisions?

Part three provided historical data over the last 24 months to answer
( 9) Did the high CAGR choices work to steer you to the best options?
(10) Did the low CAGR choices succeed at steering you away from the worst options?
(11) Did the average CAGR projections also result in close to average unit price appreciation?
(12) Did the Yahoo Finance EPS CAGRs appear to be much help in setting correct CAGRs or in finding the best investments?
(13) Would a consensus CAGR projection be a better tool in selection than "ratings"?

Part four provided data and text on the following question.
{14} How are price-implied CAGRs calculated?

Disclosure: I am long CMLP, DPM, EPB, EPD, GEL, KMP, MMP, MWE, WES. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Source: Overcoming Obstacles: A Unique Method For Assessing Risk By MLP