Monitoring Dividend Growth Portfolios
This article describes a bare minimum set of 'things to do' to insure that a dividend growth portfolio does just that, grow its dividends. It does not describe a methodology for picking stocks; although what is contained herein could form an essential part of such a task. It assumes a portfolio of stalwart dividend growers having both exponential as well as linear growth rates. It is geared toward a portfolio manager that needs to insure the portfolio is performing properly over time without a spending a lot of time and effort doing so. It does call for some records to be maintained on spreadsheets. However, spreadsheets with formulas can be constructed in advance (by a more skillful accomplice), so updates are merely inputting data in appropriate cells, not requiring detailed technical knowledge. We have all heard the real estate mantra Location, Location, Location. With dividend growth investing, it is Growth, Growth, Growth. While the thoughts presented here may not be intuitively obvious to the uninitiated, they should not require uncommon talent if you keep an open mind.
This paragraph summarizes the process, which should not be time consuming or distasteful. Details and definitions are presented later. There are two things to do monthly; 1) compare estimated income from the current brokerage statement to the previous month and 2) record dividends/interest received for the month. Annually, there are 7 parameters to record for each stock. Annual dividends/share and EOY stock price go into calculation of Yield and FOM (Figure of Merit). For all stocks (excluding ETFs), Net Income plus Depreciation/Amortization plus Non-Cash Items are used in calculation of ACFO (Available Cash From Operations), which with Dividends Paid determines $Payout. ACFO and Long-Term Debt go into calculation of Years to Payoff Debt. Watching annual trends of these metrics (Dividend Growth Rate, Dividend Increase, Yield, FOM, Payout$, Years to Payoff Debt) will indicate dividend growth performance. If lacking, a short but intense study of company news and Financial Statements, particularly the Cash Flow Sheet, will help determine where the stock is weak, perhaps sell and replace in worse cases. Action to sell a stock should not necessarily be based on a single years' performance. It may take several years of sub-par activity to arrive at a decision. Remember, if the metrics defined above maintain acceptable values over the years, no more need be done.
Perhaps the most useful procedure is easiest; that is simply looking at 'Estimated Income' value on the monthly Brokerage Account Statement (top page). This should be monitored monthly and recorded in a ledger or on a spreadsheet to ascertain that its value is constant or increasing. A value smaller than the previous entry means dividends have been reduced; an event that warrants investigation. To find the affected stock, compare 'Annual Income', current month to last month, for each stock in the monthly statement. Once found, go online to a financial web site for news. Two such sites are: money.msn and finance.yahoo. A new baseline for Estimated Income is set whenever stocks are bought/sold.
If the reduction is small and/or due to a poor economy and the stock is otherwise "OK", one could ignore it. If it is deemed wise, the stock should be sold, either as part of scheduled sales during retirement - with proceeds treated as income or the monies reinvested in another stock, or ETF, held in the portfolio. An effort should be made to retain the original income stream by choosing a replacement having the same or better Figure of Merit, a combination of yield and dividend growth. FOM = Y * DG for stocks/ETFs with exponential growth and FOM = Y * (DG + 10) for stocks with linear growth, where DG is a 1-year calculation. For clarification, exponential growth stocks are regular C Corp equities and linear growth stocks are Utilities, MLPs, REITs and Telecoms. This distinction is made because this is what the better stocks in each category actually exhibit. Part of the exercise here is to insure that they do. Maintaining a portfolio with proper, expected dividend growth of both types is the name of the game.
A second monthly task should be to compare individual dividends paid during the month with corresponding values from the previous year. This can be accomplished either in a ledger or spreadsheet; with months in columns and stocks in rows. The intersecting cell holds the dividend paid in dollars for the number of shares held. At each update, the old value is replaced by the new. Dividend increases should appear 'on schedule' and a quick calculation, (new/old) - 1, provides a quick check on the new dividend growth rate. Thus each stock will be visited just after a dividend payout, in most cases quarterly, insuring tight monitoring. Totaling monthly incomes provides a cash flow indication for the year. Here I also balance the account monthly, tracking money input/outputs. This provides a convenient annual summary of income, expenses, net stock buy/sell totals.
On an annual basis, in January, the following should be done: 1) record EOY price for each stock/ETF. This can be accomplished either from the December brokerage statement or from a financial internet site. 2) record the year's dividends/share. This is best done from dividendinvestor web site. Be attentive to special dividends (not normally used in calculations) as well as dividends that are not paid in a normal sequence, such as paid in January rather than a normal December. A stock that pays quarterly should have 4 entries/year, not 3 in one and 5 the next. Keep track of these on a spreadsheet (Div Gr History). This, with the formulas copied, will calculate total dividends for each equity (the top rows), Yield and FOM. Insure that the number of shares is maintained up-to-date. Total annual dividends are also calculated for each segment and the portfolio as a whole. I keep an 8 year dividend history.
It is very enlightening to graph annual dividend values by highlighting the dividend range, then click on the chart function in the tool box at the top of the screen. It comes up with a bar chart, but line charts can be selected which are more useful. Looking at the shape of the dividend chart provides a quick way to determine if annual increases are what you expect. Regular stocks should have an exponential increase. Utilities, REITs, MLPs and Telecoms have a linear increase. Also, look at annual values of dividend growth rates, calculated by (current dividend/previous dividend) - 1. Multiply by 100 to get a percentage. For exponential growth (compounding), annual dividend growth rates should be constant, the same or nearly so each year. To look at linear growth stocks, subtract the annual dividends. These differences (dividend increase), in dollars/cents, should be constant. In both cases, yearly values (of dividend growth rate and dividend increase) that are decreasing mean trouble. Reading across a line with annual dividend growth rates and/or dividend increases is an excellent, easy way to determine if dividend increases are acceptable. This is perhaps the most important check on dividend growth. Study the dividend graph, dividend growth rate and dividend increases until you understand. Dividends are one parameter that is pure, not fudged by over-enthused management. They fit in your wallet (purse) nicely.
In all spreadsheets used to facilitate this process of keeping tag on stocks owned, it helps to color-code cells where parameters or metrics indicate cause for concern. I use green, yellow, and red where applicable. When you are looking at the spreadsheet later, it high-lights goodness or not so that you can hone in on those stocks that are marginal.
The question here is to determine if an asset should be kept or replaced. I use a lower limit FOM of 16 for stocks in exponential dividend growth segments and a minimum 55 for stocks in linear dividend growth segments. Dividends paid will vary with the economics governing the stock in question. This is unavoidable and should be taken into account. If, however, a stock accumulates a history of declining dividend growth, or worse, dividend cuts or suspensions, then sale of the stock should be considered. If stock sales are programmed for that segment during retirement, then the obvious is to sell that wayward stock. Otherwise, it should be replaced by buying shares of a company/ETF already held. This will eliminate need for researching new buys.
It is always a good idea to read news about the company to help determine keep or sell decisions. Also, looking at Cash Flow statements helps. These are found on money.msn web-site, last item in the left hand column on the stock main page. Things to look for are: Net Income increasing, Depreciation increasing. Determine in general how the money is spent. Cash Flow Activities balance: start cash; +/- cash flow from operations, investing, financial activities; +/- exchange rate changes = end cash. Note that black numbers mean monies coming in, while red means monies going out. Each company is different and it takes a little practice to ferret out what is going on. A key metric is cash available to spend. This is approximated by adding Net Income plus Depreciation/Amortization/Depletion plus Other Non-Cash Items from the Operations section. I call this ACFO (Available Cash From Operations), previously called OMBA$ in my earlier articles [Note 1]. You can then look to see where this goes; dividends, paying down debt and/or buying back stock, investing as growth-oriented Cap Ex. Generally speaking, if funds are not invested to grow the company either generically or by acquisition, Net Income will lag and funds for increasing dividends will not be available.
[Note 1]. Traditionally, Net Income has been (is being) used in many metrics. This is not a good proxy for the amount of cash available to support normal company activities. Available Cash has been expanded for REITs in a FFO (Funds From Operations) metric as Net Income plus Depreciation/Amortization, but even this falls short in many cases. International Accounting Standard [IAS-7] defines Free Cash Flow [FCF] as Cash From Operations minus Maintenance Capital Expenses. This latter parameter is not defined separately so it has become common practice to use all Capital Expenses in Free Cash Flow calculations. Unfortunately, this results in an unwarranted negative value for instances of high discretionary 'Cap Ex' used to expand a company's business. I have chosen to use ACFO (as defined above) as the closest, easily obtainable value for 'profit cash money' with the understanding that Maintenance Cap Ex funds have priority. In other words, all ACFO is not available to support discretionary activities such as paying dividends, expansion Cap Ex, paying down debt, buying back shares, and the like. The other line items in Cash From Operations tend to be volatile and do not represent yearly income amounts. The advantage of this expanded value of cash available is apparent because metrics defined and used herein work for regular C Corp stocks, REITs, MLPs, Utilities, Telecoms and other stocks routinely owned by dividend growth investors. It does not work for BDCs and mREITS because their Cash Flow Sheet is different. One metric used by MLPs is Distributable Cash Flow. Unfortunately, there does not appear to be a standard definition for this; sometime it includes payment to the General Partner and sometimes not. It is left to each company to define this metric, often using parameters not available to the public. Neat!
For all stocks, two additional calculations should be made annually. A) The first is Long Term (LT) Debt divided by ACFO. LT Debt is found on the Balance Sheet; make sure you are on the annual page. This calculation gives the number of years to pay off LT Debt if all available income cash is used. Generally, values for regular stocks and Telecoms is about 2 years and 5 years for others (Utilities, REITs and MLPs). If the number is greater and increasing year by year, that is a bad sign and warrants further investigation. REITs and MLPs payout most of operating income as dividends and use new stock issues and debt (found under Financial Activities on the Cash Flow Statement) to finance growth. LT Debt increases should over time result in new income (increased ACFO) to bring the ratio back in line. Upper limits are 3 (Yellow) and 4 (Red) for regular companies and Telecoms; with 8 (Yellow) and10 (Red) for Utilities, REITs and MLPs. B) The second annual calculation is $Payout Ratio, Dividends Paid / ACFO. Dividends Paid is found in the Financial Activities section of the Cash/Flow Sheet. Upper limits here are 45% (Yellow) and 55% (Red) for regular companies, Utilities, Telecoms; with 95% (Yellow) and 100% (Red) for REITs and MLPs. Both ACFO and Dividends Paid include payments to the MLP General Partner. More on this in my article titled 'What is Due Diligence for Linear Dividend Growth Stocks', where several stocks in each industry are analyzed. I maintain these calculations on a Portfolio Maintenance Sheet (spreadsheet). A good place to see companies activities is on their own web site, found in Profiles (money.msn) and (dividendinvestor) web sites.
There are a limited number of ways a stock can be a cause for corrective action. The table below lists the events, places to discover same, and corrective action (changes to spreadsheets):
Stock Shares Split
A C E H
A B F
A C E H
A B D F G
C E H
A B D E F G
Reduces (cuts) Dividend
B C E H
Reduces Dividend Growth
A B C E H
A) On-line Portfolio (if implemented)
B) Div Gr History (Spreadsheet)
C) Monthly Statement
D) Portfolio (Spreadsheet)
E) Monthly Dividends Paid (Spreadsheet)
F) Div Gr Calc l.r. (Spreadsheet)
G) Portfolio Maintenance Sheet (Spreadsheet)
H) Company News
It helps to have a list of documentation plus other items on a check list so that whenever an event takes place, you run through the list to determine what you need to do for updating. Other items include tax considerations.
To track ETFs annually, one need only to look at 1) the dividend graph, dividend growth rate and 2) Yield and FOM to check to see if all is normal. The economy plays a large role here. Don't fault the ETF is the economy is down. In dire times ETFs may cut dividends, this is normal and something you live through provided the multi-year dividend graph is acceptable.
One thing to remember: Investing is a competitive sport. While the colony of SA dividend growth investors are usually on the same side in a trade, never forget there are big-money bad guys out there who thrive on adding your chips to their pile. I prefer to look at evaluation parameters my own way, having spent many hours in deep thought about what makes sense logically. Many of the metrics commonly used, at best, only apply to a group of stocks that may or may not include those featuring dividend growth. Certainly, they exclude those I call linear dividend growth, which (in my case) constitute half my portfolio. There, in my opinion, is where the challenges are. OK, the down side of my approach precludes screening from a database (origins and accuracy of which may be largely unknown). GIGO comes to mind. As someone said in commenting on one of my articles, 'Maybe we have to go back to analyzing stocks sitting under a tree'. Right on - just make it bird free.
Let's take an example to see how this works; a company that is having trouble paying an increasing dividend is Proctor & Gamble (PG).
EOY Price $
Div Inc $
Div Pd $M
LT Debt $M
The green 8-year non-l.r. exponential curve represents a 10.1% dividend growth rate. For those who use CAGR calculations, the following are applicable: 10yr - 10.6%; 5yr - 8.8%; 3yr - 7.9%; 1yr - 7.0%. It is apparent that yearly dividends for PG have been deteriorating for the past several years. I am not a fan of CAGR calculations; which assumes exponential (compounding) growth and only uses 2 data points regardless of the number available. If expected proper dividend growth is not there, it is hard to ferret out that truth from looking at a series of CAGR calculations. On top of that, erroneous conclusions can be drawn from stocks that have legitimate linear dg growth rates. Here, it is better to observe year-to-year dividend increases. For stocks with linear dividend growth rates, it is meaningless to look at any multi-year growth rates beyond the current year. Mixing the two types of dividend growth using the same calculations is not valid.
The tables and graph above summarize the annual stock analysis described above. Let's take a look line-by-line to determine how PG is doing. The first line offers a clue; for the last 4 years, dividends/share are paid in 1/100s of a penny. That's cutting it thin for a company with a 84 billion dollar revenue for fiscal year 2013. Not to be facetious, but one needs to look for trends and/or changes in data to ferret out what management is doing. In perusing a financial report, don't look for smoothness, look at the wrinkles. Next, note that DG rate is decreasing over the last 5 years. For exponential growth, DGR should remain the same (or nearly so) year-to-year. The test for linear growth is that dividend increases, in dollars, are the same (more or less) year-to-year. We see on that line the increase is flat to down the last 5 years. Folks, dividends for this stock are no longer compounding, haven't for 5 years. We note that FOM has been decreasing the last 4 years. The calculation here assumes exponential growth where it just exceeds the threshold minimum of 16. If we do the FOM calculation for linear growth where yield=2.3665/80.65=0.0293; dg=0.1555/2.3665=0.0657 and FOM = 2.93*(10+6.57) = 49. This value is below the 55 threshold minimum.
From the next table, ACFO is flat to down over the 5 years. This metric needs to increase to pay dividend increases; that is where the money comes from. DG is not sustainable without ACFO leading the way. Next, LT Debt has decreased the last 2 years, monies that could have been used to increase dividends. $Payout is increasing indicating that DG is exceeding ACFO growth. Last, Debt/ACFO has decreased to last 2 years. For companies like PG, where exponential DG is expected, Debt/ACFO generally runs about 2. What PG is doing is paying down debt, in a time when interest rates are at historic lows, rather than increasing dividends. Bummers!
A further look at financial statements of PG show that for the 2009-11 years they were selling part of their operations. In 2013, they had an acquisition. This indicates that some effort was made to better align their products with the market, but this process has not (yet) affected income in substantial amounts. The Cash Flow sheet also shows that in each of the last 5 years they have re-purchased shares in amounts on a par with dividends paid. This has an effect of boosting dividends/share because there are fewer shares to 'share' ACFO. However, to the long-term dividend growth investor, a little more in the dividend pot is better appreciated.
It is interesting to note that AG Lafley was CEO for most of the 2000 decade, retiring in 2010. RA McDonald was CEO from Jul 2009 to Jun 2013. Lafley is again CEO. Ref.: Google 'CEO of PG'. PG stock was trading between Nov 2009 to mid-Jan 2013 in the $60 - $70 range, then rapidly shot up to the $76 - $82 range. This was apparently a result of a good quarterly report. Good article here. It paints a pretty bleak picture of the company as it now is. It may be a long time before things turn around. Key for dividend grower investors is how Lafley will treat dividends versus (say) stock buybacks and debt reductions in the future. At least, get dividends to the nearest penny.
The top graph of PG dividend history shows data from 2006-2013 which is consistent with the procedure outlined in this article. While a close look indicates the problem of reduced dividend growth, it is enlightening to look further back in time.
The graph above shows PG dividend history from 2002-2013 (red curve), green curve is the 8-yr l.r. calculated curve using the 8yr span from 2002-2009. Here the calculated curve fits the actual data almost exactly in that time frame. Beyond 2009 you would expect dividends to grow as indicated by the green line. What happened is that growth followed the red line. As time goes on, the situation gets worse. At the end of 2013 the gap in dividends is over 32 cents/share, 12% below expectations. How long does one wait before taking action? As I write this, PG stock is trading at an all-time high. At the very least, this procedure gives a heads-up look, hopefully before bad news becomes common. TYMAR (Take Your Money And Run)???
An alternative to PG is the Vanguard Consumer Staples ETF (VDC). The table below has dividend information:
The graph above gives dividend/share comps for PG and VDC. Current prices and yields are: PG: $80.61, 2.9%; VDC: $108.88, 2.4%. VDC has 113 holdings with PG the largest at 12.28%; Expense Ratio is 0.14% and the Turnover Ratio is 7%. With a linear dividend growth of 7.0% and yield of 2.9% for PG and an exponential non-l.r. dividend growth of 18.7% and yield of 2.4% for VDC, the crossover point is less than 3 years. The 2013 fiscal year is over for PG (Jul - Jun). My guess is that VDC in December will have a dividend growth rate less than last year but greater than the 7.0% fielded by PG. Many of the companies in VDC do not (yet) pay a dividend. They could be acquired and thus contribute to an existing dividend flow with their new products or grow and initiate a new dividend stream. Key is a low Turnover Ratio. Impossible to predict the future, but seedlings are there to potentially host a long term dividend growth equity, longer than what one might expect from any single company.
In the case of PG, with smooth transitions in dividends, it is more difficult to determine to what is happening from looking at the dividend graph. However, a close look at dividend growth rate and dividend increases does provide that determination. One can easily see (from those metrics in the PG data table) the transition from exponential growth to linear and then lower increases in dividends. For the VDC case, where year to year dividend changes are more erratic, looking at the dividend graph paints a clearer picture. Analyzing both the graph and calculated data should provide an easy, clear understanding of dividend growth in sufficient detail to make proper decisions. Reading the fine print may put you ahead of the pack. I have included the 8yr non-l.r. curves in the graphs above to facilitate this learning process. I am not recommending those calculations in the procedure outlined herein because that complexity should not be necessary; decisions required are hold or sell.
To gauge the affects of a linear DG, we can compare it with an exponential inflation curve.
If the inflation rate is 3.3% or less, then a linear dividend growth rate of 2 times the inflation rate will provide more than adequate protection for inflation over a 40 year time span. Said another way: to achieve a curve pattern as shown above; if the inflation rate is less than 3.3%, the corresponding linear growth rate is less than two times the inflation rate. It takes a linear growth rate more than twice the inflation rate for inflation rates more than 3.3%. Applying this to PG: if the linear DG for PG and current inflation rates remain the same (or move in sync), PG dividends will outdo likely inflation for 40 years, providing you find the 3% yield acceptable.
So what is my take on PG. At best, it is marginal. One could wait until Apr 2014 to see what the next dividend increase is. I would like to see a special dividend using funds now spent on reducing debt. This would indicate to investors that management is positive about the company's future. I don't have a warm feeling they are positioning to grow Net Income, a precursor for dividend growth. It begs the question whether a company this size can maintain sufficient growth to sustain exponential dividend growth. And the yield is too low for acceptable linear growth, less than that borders on the unacceptable. A viable option is to replace it with VDC; the dividend increase there (coming in Dec 2013) may tell. I am putting PG on the chopping block, but have not (yet) raised the axe.
Additional disclosure: I am long PG and VDC.