Seeking Alpha
Portfolio strategy, ETF investing, dividend growth investing
Profile| Send Message| ()  

This article is a follow-on for one written recently on the same subject except covering stocks with exponential dividend growth. Here we will cover stocks with linear dividend growth. Generally speaking, utilities, REITs, MLPs and telecom companies exhibit linear dividend growth. What this means is that dividends are increased yearly by the same amount in dollars (give or take), rather than by the same percentage (give or take) for exponential growth stocks. Linear growth stocks do not follow the Rule of 72, thus do not compound. Misleading results will follow if they are treated the same.

In this article, I will analyze linear growth stocks in the High Income, Mélange and Bond segments of my portfolio. This will serve to illustrate points being discussed (in a real-life situation), while at the same time, let me update how my stuff is doing.

Stocks in the High Income segment are: CenterPoint Energy (CNP), PVR Partners (PVR), TransMontaigne Partners (TLP), Enterprise Products Partners (EPD), National Health Investors (NHI), ONEOK Partners (OKS), Plains All American Pipeline (PAA), TELUS (TU), Omega Healthcare Investors (OHI), Wisconsin Energy (WEC), Vodafone (VOD), BCE (BCE), Textainer (TGH), Western Gas Partners (WES), Rogers Communications (RCI), Rayonier (RYN), Calumet Specialty Products Partners (CLMT), SeaDrill (SDRL), Shaw Communications (SJR), Sunoco Logistics Partners (SXL), CenturyLink (CTL), Avista (AVA), NextERA Energy (NEE), CMS Energy (CMS), National Grid (NGG).

Stocks in the Mélange and Bond segments are: Verizon Communications (VZ), EV Energy Partners (EVEP), Navios Maritime Partners (NMM), Realty Income (O), Health Care REIT (HCN), Pioneer Southwest Energy Partners (PSE), Energy Transfer Partners (ETP), Martin Midstream Partners (MMLP).

The reason for the separation is that I have a portfolio distribution (withdrawal) plan that, in part, is a takeoff from the Bucket approach. Mélange and Bond segments are in the first bucket. Stocks in these segments, while they have a high yield, also have low (or no) dividend growth. Stocks in the High Income segment are intended to last the retirement time period, with moderate annual sales throughout the period.

In determining a set of potential candidates to buy, it is (obviously) essential that the stocks have a satisfactory history of paying an adequate yield and dividend growth. Thus, all that is needed for due diligence is to determine, if possible, some assurance that this trend will continue. The best way to achieve this goal is to track how profits are spent including insuring some provision to grow the company, to increase profits for dividend growth. While pursuing this task, it is helpful to (try to) read the mindset of the Board of Directors and management regarding their policy for dividends.

In analyzing exponential growth stocks, one general observation was that for the stocks in the sample, nearly all companies were buying back shares. Thus funds were available to pay an increasing dividend, if the company so desired. The situation is different here. Linear growth stocks generally pay a higher yield which absorbs more (or most) of profit. Funds for company growth are raised by issuing (selling additional) shares and/or increasing debt. A challenge here is to determine the ability to repay debt. In my mind, I don't have an issue with issuing more shares. True, there is a temporary dilution of shares, but as the money raised goes into generating more profit, these added funds go to pay the new shareholders, resulting no effect on original shareholders. In fact, with these extra funds, the company grows, which provide some economies of scale that do benefit the original group. It would be another matter if one could make the case that stock dilutions reduce dividend growth. Debt is another issue. I warrant that ever increasing debt levels are not sustainable, even with higher and higher income. What happens when the music stops?

Due diligence is used both in determining if a stock is suitable to buy and in a periodic check to insure all is well. In the following, a series of tests are performed having increasing complexity. The key here is that if the stock turns out to be unsuitable, best to find out as soon as possible. The general format follows the previous work outlined in the due diligence article on exponential growth.

Test #1: Determine current yield and current (one year) dividend growth. I use the dividendinvestor website. Yield, 3 & 5 Year dividend growths are shown in close proximity. This website has up to 5 years of dividend data (click on dividends). The website uses dividend payment date for an annual summation of dividends, which in a few cases slops over into the next year; one year gets 3 payments and the next 5. These anomalies I correct, since it is readily apparent. Here I also eyeball how dividends grow over time to see if dividend growth is reasonable. If you mix sources for dividend data make sure they are compatible. Keep in mind that a big reason for volatility in annual dividend variations is the state of the economy where your prospective company does business.

As a Figure of Merit [FOM], I use a formula developed and reported on in one of my previous SA articles. It is Y*(DG + 10), where Y is yield and DG is dividend growth. For starters, I use current one-year data as in 'what have you done for me lately'. For two companies with different yields and dividend growth; if they have the same FOM, they have equal dividends at the ten-year point. For equal dividends at twenty years, the formula becomes Y*(DG + 5). Again, this formula is for linear dividend growth only. There is a similar formula for exponential growth. For new stocks, I use the first (10 year) formula using a lower limit of 65, which relates to the lowest bond (no DG) yield I would consider owning (6.5%). It is surprising how many hot stock tips fail this test. After purchase, if the stock's FOM drifts lower, I consider other factors before deciding to sell. I don't like strict limits.

Test #2: Read the Profile about the company's business. Here I use money.msn and finance.yahoo websites. It is vital that you understand the company's business model and its product(s). If these products may not be around at the end of your time period (in your judgment), are they adaptable to new ones? Read the news and SA articles discussing the company. This should be a key part of your selection process.

Test #3: Determine an 8-year composite dividend growth. I use 8 years to cover an average business cycle. Shorter periods may miss market downturns and it is harder to find dividend data for longer periods. I use a linear regression analysis using 8 years of dividend data. The technique is described in my previous articles. What this does is calculate a "best fit" straight-line curve to the presented data. I have a spreadsheet dedicated to this task which facilitates the effort. Heavy-duty calculations are performed with the help of a website. If you don't want to go through the regression bit, just enter dividend data in a spreadsheet and graph it. Or calculate the 7 annual dividend growth rates and see how they vary over the time period. Be aware that the economy may affect the dividends a company is willing to pay at any given point. Don't fault the company (necessarily) if the economy tanks. Some of the stocks listed here do not have full 8 years of dividend history (2005-2012). To generate the composite curve, I have feathered-in estimates for missing data such that the outcome is not biased. These stocks are (# years missing): PSE (4), NMM (3), NGG (1), TU (2), TGH (2), WES (4), RCI (3), CLMT (2), SDRL (3), SJR (2), CTL (3), CMS (2).

The graph below depicts dividend history for PVR. The red curve is actual dividend/share data while the green curve shows the best-fit linear calculation with a dividend growth rate of 5.1%. It is easy to see effects of the financial crisis. Dividend growth rate [dgr] is determined from a,b parameters calculated in the regression analysis for the equation y = ax +b, where a is the slope. The value of dgr is 'a' (the slope) divided by the last dividend data point.

I get a much better feel for a company's dividend growth history (and potential dg future) from looking at an 8-year graph than from screening on 1,3,5 year dividend growth data. If an accurate assessment of dividend growth isn't important to dividend growth investors, what is?


(Click to enlarge)

In the spreadsheet dedicated to determining this 8-year Composite Dividend Growth Rate, I color-code the cell containing the stock symbol, green-yellow-red, to reflect level of concern. This provides a quick ID for stocks to monitor closely during the year. While I am interested in growth rates for each stock, at that level, I am as interested in the shape of the curve. Downward trends trigger an investigation. I (try to) manage the High Income segment to be over 7% growth rate (weighted average). Mélange and Bond segments have no dividend growth requirements. Color-coded stocks are: Yellow - SJR, TLP, NMM, NGG, NMM. All have yet to recover to past dividend growth values. Red - EVEP, CTL, NRP, O, MMLP, PSE, ETP. All have little to no dividend growth, but have a high enough yield to score adequately in FOM calculations.

Test #4: Use metrics to determine the worthiness of a company as a dividend growth candidate. I tend to reject traditional fundamental metrics because of inherent flaws in their universal application and, in part, because they are not (may not be) applicable for dividend growth stocks. Besides, many (if not most) traditional metrics are industry sensitive and since we are looking at companies across many industries, their usage is not appropriate. Let me be clear, I use fundamental parameters such as Net Income, Long-Term Debt. They are the data. I choose carefully in using these parameters in formulas (metrics), preferring to find a metric more suitable for dividend growth stocks.

As a replacement for Net Income, I use a simple formula using data that is readily available. I call this new metric OMBA$ (Operations Monies Before Allocation Dollars). The formula is OMBA$ = Net Income + Depreciation + Amortization + Non-Cash Items. I use money.msn website for these data; taken from the Cash Flow sheet, Operations section. It represents the bulk of "profit" in cash money that the company Board of Directors (with Management) has each year to allocate. The units for this metric are $/year. It can be used to calculate a $Payout by dividing "Total Cash Dividends Paid" (Financial Section of Cash Flow sheet) by OMBA$.

Another metric, useful in analyzing linear growth stocks, as a replacement for Debt/Equity, is: Long-Term Debt / OMBA$. This gives the number of years it would take to pay off debt using the entire cash flow. Long-Term Debt is found on the Balance Sheet. It does bring up another issue in using metrics of this sort. Suppose a company borrows money to build a new plant. At first, debt goes up, taking this ratio with it. It may take a few years before the new plant is operational and contributes to cash flow. Rather than compare these metrics across different companies, it is more meaningful to track the metric over time in a given company. In this case, for example, if the ratio was trending higher, it might indicate the company was losing its grip in utilizing debt to boost income.

As a test for new potential buys, do the $Payout and Debt/OMBA$ just to get calibrated. Both $Payout and Debt/OMBA$ center around values depending on the industry. If a company fell outside the normal range for that industry it could be used as a possible rejection criteria. The tables below list these, and other, metrics for the stocks analyzed, grouped by industry:

Utilities

******

2012

Yield

%

****

1 Yr

FOM

2012

*****

# Years of 5 Issuing

Stock/Debt

**********

5 yr avg

$Payout

%

******

Debt/OMBA$

2008-2012

Avg.: 6.3

******************

CNP

4.2

53

3/2

23

8.5-8.0-6.8-6.7-5.4

WEC

3.3

83

0/4

36

11.9-9.7-6.0-7.6-5.4

AVA

4.8

75

5/5

24

5.0-5.1-5.7-4.8-5.6

CMS

3.9

96

4/5

19

8.2-8.3-8.7-6.8-5.9

NGG

5.4

89

1/2

21

5.0-6.7-5.7-4.3-4.7

REITs

******

2012

Yield

%

****

1 Yr

FOM

2012

*****

# Years of 5

Issuing

Stock/Debt

**********

5 yr avg

$Payout

%

******

Debt/OMBA$

2008-2012

Avg.: 6.2

******************

NHI

5.0

130

0/3

96

0.1-0.0-0.5-1.2-2.3

OHI

7.1

135

4/4

82

4.6-5.5-7.2-8.2-7.8

O

4.4

53

4/3

87

5.6-6.1-7.1-7.2-9.2

HCN

4.8

70

5/4

92

8.6-6.0-12.6-12.8-10.9

Pipelines

******

2012

Yield

%

****

1 Yr

FOM

2012

*****

# Years of 5

Issuing

Stock/Debt

**********

5 yr avg

$Payout

%

******

Debt/OMBA$

2008-2012

Avg.: 4.2

******************

EPD

5.1

77

3/4

31

5.9-5.8-5.3-4.4-4.0

OKS

4.8

103

4/4

74

3.8-4.9-4.0-3.5-4.5

PAA

4.7

84

5/4

68

4.5-5.0-5.4-3.3-3.7

WES

3.9

110

5/4

47

0.9-0.9-2.3-2.7-4.8

SXL

3.7

89

2/4

51

2.8-2.8-4.2-3.2-2.6

ETP

8.3

83

5/5

88

4.7-5.4-6.0-6.3-9.7

Telecom

******

2012

Yield

%

****

1 Yr

FOM

2012

*****

# Years of 5

Issuing

Stock/Debt

**********

5 yr avg

$Payout

%

******

Debt/OMBA$

2008-2012

Avg.: 2.4

******************

TU

3.7

76

2/2

21

2.4-2.3-1.9-2.1-1.9

VOD

8.3

477

2/2

31

1.7-2.2-1.9-1.6-2.0

BCE

5.1

98

2/2

24

1.8-2.0-1.8-2.2-2.1

RCI

3.4

72

0/5

18

2.9-2.8-1.9-2.2-2.2

SJR

3.8

57

3/3

28

2.3-2.4-3.1-4.5-3.5

CTL

7.4

74

4/2

32

3.7-4.7-3.0-4.6-3.3

VZ

4.7

59

2/2

15

1.3-1.7-1.4-1.4-1.3

Transportation

******

2012

Yield

%

***

1 Yr

FOM

2012

*****

# Years of 5

Issuing

Stock/Debt

**********

5 yr avg

$Payout

%

******

Debt/OMBA$

2008-2012

Avg.: 4.6

******************

TLP

6.7

89

1/2

61

3.3-3.2-2.1-1.7-2.8

TGH

5.1

190

3/4

31

4.6-5.7-5.1-5.8-8.1

MMLP

9.8

101

4/4

81

4.0-5.1-5.9-5.9-5.7

SDRL

11.7

556

3/4

76

11.5-4.4-5.4-4.1-5.9

NMM

14.4

164

5/3

67

4.7-3.4-2.9-2.2-1.6

Natural

Resource

*******

2012

Yield

%

***

1 Yr

FOM

2012

*****

# Years of 5

Issuing

Stock/Debt

**********

5 yr avg

$Payout

%

******

Debt/OMBA$

2008-2012

Avg.: 3.4

******************

PVR

8.1

147

3/5

81

3.9-3.8-4.2-4.4-9.4

NRP

11.9

135

1/4

91

2.0-3.6-3.1-3.0-3.4

RYN

3.2

150

5/3

41

2.2-1.4-1.7-1.9-2.4

Oil/Gas/

Refinery

*******

2012

Yield

%

***

1 Yr

FOM

2012

*****

# Years of 5

Issuing

Stock/Debt

**********

5 yr avg

$Payout

%

******

Debt/OMBA$

2008-2012

Avg.: 2.9

******************

EVEP

5.4

57

4/3

61

4.6-2.7-4.9-5.6-4.2

PSE

9.1

109

3/3

57

0.0-0.7-0.8-0.3-1.3

CLMT

7.6

216

4/3

55

4.6-3.6-3.4-3.9-2.6

Some comments on the above tables: Regarding $Payout - The magnitude of OMBA$ which is used in the calculation of $Payout is actually larger than it should be. According to the "rules," the amount of CapEx needed to maintain the company at its current level of activity should be subtracted from OMBA$. Since this parameter is not a line item in the Financial Statements, we cannot account for it. Nevertheless, that fact should be remembered. Thus, when $Payout exceeds 100%, the 'extra' dividends have to come from somewhere. In the current group of companies being analyzed, there were 6 companies that had $Payouts of 100% and above. These are listed below with some explanation.

PVR, 2012, 111%: There was a lot of activity in write-offs and acquisitions, but there were adequate funds raised via issuing stock and debt to cover. Dividend growth for the past (recent) several years has been lagging the 8-year Composite Dividend Growth rate. It looks like the company wanted to show things were getting better.

SDRL, 2008, 109%: This was the first year the company paid a dividend, which was cut the next year. Go figure!

SDRL, 2012, 119%: There were extra funds from the sale of investments and financing cash flow but these normally occur in some years. The strange thing is, the overpay was due to a special dividend. Granted, the dividend would have been cut without it, but the total dividend was larger than necessary. I think JF (mover & shaker) likes dividends.

NHI, 2008, 122%: This made the dividend higher than normal, which was cut the next year. There were positive cash flow items which, apparently, were used for extra dividends, to pay down debt and buy back shares. The cash balance EOY also increased. Party time?

HCN, 2010, 104%: Apparently to maintain the dividend level.

NRP, 2009, 107%: Apparently to raise the dividend to maintain dividend growth of previous years. The next two years were flat.

ETP, 2010, 100%: To keep from cutting the dividend. The next two years were flat, making the dividend flat for 4 years.

It was also interesting to look at a potential metric, Long-Term Debt / Market Cap. Stocks will values over 80% are: ETP - 102%; CMS - 89%; CTL - 82%; TGH - 101%; CNP - 80%. Wouldn't it be fun to buy a company for the sticker price, then find you owed again as much in debt?

Test #5: Visual check of the company's Financial Statements. I use money.msn; they present 5 years of data, which are conveniently laid out. Hint: more data are available under some line items, click on the blue arrow. Also, when going to the Balance Sheet, it comes up with quarterly data instead of annual. Companies are required to submit these statements to SEC on a quarterly and annual basis. This agency does not pass judgment, merely makes them available to the public.

Financial Statements consist of an Income Statement, Balance Sheet and Cash Flow Sheet. Cash Flow is broken into 3 parts: Cash from Operating Activities, Cash from Investing Activities and Cash from Financing Activities. Values for Net Cash Beginning and Ending Balance are also provided. Line-items within these categories that I use are: Operations - Net Income, Depreciation, Amortization, Non-Cash Items; Investing - Capital Expenditures, Other Investing Cash Flow Items; Financing - Total Cash Dividends Paid, Issuance (retirement) of Stock, Issuance (retirement) of Debt. All these data balance. If you start with Beginning Cash, then algebraically add the 3 cash groups, you end up with Ending Cash. In my view, the remaining line-items vary throughout the year and are covered by the cash balance; i.e., not relevant to dividend growth considerations.

On the Balance Sheet, key items under Assets are: Property/Plant/Equipment; Long-Term Investments. Goodwill and Intangibles are not positive items. Under Liabilities: Total Long-Term Debt is of interest.

In the previous article on due diligence for exponential dividend growth, the focus was on how OMBA$ was allocated. This made sense because dividends paid was a smaller portion of that parameter. Here dividends play a much larger role in OMBA$ allocations. Interest then centers on how remaining OMBA$ plus any positive cash flow from issuing stock and/or debt is distributed. These funds will flow through the Investing part of the Cash Flow Sheet onto the Asset portion of the Balance Sheet. You are encouraged to read this article, it provides more detail on how to account for cash flow. Those considerations are valid here, the only thing that (generally) changes is the sign attached to stock/debt issuance (retirement); negative for fund outflow and positive for fund inflow.

A good quick look at these fund flows can provide a feel for allocation of funds that provide new added capability to the company. Sometimes this is done through acquisitions as reported in the Investing section of Cash Flow. In any event, the following parameters should increase over time: Depreciation, OMBA$, Property/Plant/Equipment. Long-term Debt may increase but should at a slower rate because the remaining OMBA$ and issued stock help boost these other parameters. Long-Term Debt may increase independent of Debt Issued because of acquisitions. Goodwill and Intangibles may increase due to acquisitions, but these are non-productive items. If you can ferret out management strategy vis-à-vis dividend policy and company growth tactics, 2 thumbs up.

There is no easy way to score this effort. I record a "yellow tag" to any indication that something is not as it should be. These can be investigated more thoroughly, if possible, or just attached as a negative indicator in the evaluation process. Remember, you know dividend history and here are only looking for positive evidence that the future will be as bright. Don't sell the chickens until you see them in the fox's mouth.

In performing this test on these stocks, the following observations were made: For all utilities, revenues decreased during the 5-year time period, but this was offset by higher reductions in fuel costs, so Net Income was not as effected. All stocks exhibited a feedback of funds to "grow" the company by various combinations of issuing stock/debt and the residual (after dividends) of OMBA$. In some cases, a subsequent increase in Net Income was not (yet) forthcoming. These were reflected in higher values of 2012 Debt/OMBA$. There was no evidence that any company was in trouble, although several need watching more closely. As a group, Telecoms look good in terms of Debt/OMBA$ compared with other linear growth industries. Many companies listed here have raised large amounts of debt in the last 1-2 years. These should be monitored for increased Net Income in the near future.

A word about CTL. In early 2013, dividends were cut from $0.725/Q to $0.54/Q. The stock price tanked but has recovered about half the loss. The company stated they wanted to improve the balance sheet and were planning some stock buyback. In 2011, CTL acquired Qwest, a larger company. They borrowed money in 2011 to close the deal, but paid it back in 2012. They did inherit a lot of debt, but as the table shows, the Debt/OMBA$ ratio didn't change. It was, however, higher than other Telecoms (see table). In my opinion, they would be better off repaying debt vs. share buybacks. I plan to wait and see how it shakes out before considering selling my position. As I indicated in the last article, I may sell some shares to use in keeping the dividend level unchanged for a few years to delay the sell decision.

In terms of new buys, I would rank-order stocks by their 8-year FOM using the 8-year Composite Dividend Growth and 2012 Yield, start at the top and work down, cutting off at FOM 65. This rank-order is: NMM, SDRL, NRP, VOD, MMLP, ETP, PVR, OHI, CTL, TLP, CLMT, TGH, CMS, EVEP, AVA, BCE, NGG, EPD, WES, OKS, PAA, SJR, RCI, CNP, NHI, WEC. Eliminated (in order) are: SXL, VZ, TU, HCN, RYN, O, NEE. I would further cut stocks with low dividend growth and high trending Debt/OMBA$. Lastly, run any remaining test outlined above. Keep in mind, this ranking includes effects of special dividends paid in 2012 by SDRL and VOD. Recognize that NMM and possibly SDRL are cyclical and probably should not stand the whole cycle.

Suppose you don't do as I suggest here but rather come up with a different list using some other approach. There will undoubtedly be some overlap. What we don't know is which portfolio will perform better in the future in terms of portfolio turnover. Since we will, in all likelihood, be using dividend levels as the prime criteria for determining sell decisions, I believe my portfolio will prevail since that parameter is dominant in the process outlined here. Time will tell!

Source: What Is Due Diligence For Linear Dividend Growth Stocks?

Additional disclosure: I am long all stocks listed in the article.