I Figured Out My Dividend Risk Using This One Easy Trick (And It Wasn't The Payout Ratio)

|
Includes: PG
by: Investing Doc
Summary

To determine financial metrics that might serve to predict the chances of a dividend cut, I gathered data from about 1000 companies with 10-year operating histories.

Looking at these data, it's obvious that companies that cut their dividends underperform those that preserve them. Is there any way to sort them out quickly and easily?

Looking at core fundamentals, rather than focusing on the dividend itself, appears to be the key.

Like many on Seeking Alpha, I own several dividend-paying issues. Many of these are Dividend Aristocrats, like Procter & Gamble (NYSE:PG) and Johnson & Johnson (NYSE:JNJ). I also own a smattering of MLPs (having now cut my energy exposure to mainly midstream), including Spectra Energy (NYSE:SEP), Magellan (NYSE:MMP), and Energy Transfer (ETP), and REITs, like HCP (NYSE:HCP) and Ventas (NYSE:VTR). My long-term goal -- also like many on this site -- is to assemble, over time, a portfolio capable not only of delivering attractive total returns but also of delivering a growing cash flow that my family will be able to use for income later on. Obviously, for these companies, dividend growth and stability are of primary importance. It would be helpful, therefore, to be determine some metrics that might be able to signal in advance the growing risk of a dividend cut.

While some metrics looking at dividend safety, such as dividend payout ratio, are well known to most investors, I wanted to take a look and see whether other measures of financial health might be helpful in predicting a pending cut to a company's dividend, and then see if this information might be applied to my own holdings.

Experimental Method:

I collated price, income, balance sheet, and cash flow data on each of the components of the S&P 1500. Of these, I screened for companies with 10 or more years' worth of operating history (the better to have additional data points for evaluating company qualities) and those with a history of paying a dividend at the start of the 10-year interval, excluding special dividends. These criteria alone winnowed the list down to 832 total companies. I then computed a 10-year history of the following metrics:

  1. Price,
  2. Dividends,
  3. Current Ratios,
  4. Returns on Invested Capital,
  5. Total Return (defined as 10-year capital gains return plus median yield),
  6. And EPS growth.

I then determined medians of the above metrics based upon observed distributions, and calculated odds ratios for each. Here's what I found.

Results:

Without a doubt, dividend cuts portend poor performance. Across the spectrum of dividend paying stocks, companies that cut their dividends (hereafter, "Dividend Cutters") almost uniformly displayed price declines relative to their counterparts ("Dividend Preservers"). Despite having a generally higher price in 2005, the average Dividend Cutter would have only seen about a 37% total gain in price over the past 10 years, as compared to the 150% gain for Dividend Preservers over the same time period.

Graph produced by author. Price data from Yahoo Finance. Fundamental Data from Morningstar.

And as might be expected, Total Returns from Dividend Cutters were significantly lower over the 10-year period as well, to the tune of 8.4% per annum versus 12.6% per annum over the same time period.

Graph produced by author. Price data from Yahoo Finance. Fundamental Data from Morningstar.

Dividend cuts, then, augur poorly for a company's prospects and prospects for total return down the road. While this seems intuitive, it certainly is nice to see it play out in the data.

When considering the financial health of companies prior to initiating a cut in the dividend, we can look at a variety of metrics. Among these, the payout ratio is one of the most commonly cited. This is appealing intuitively, as it might reasonably be expected that a company cannot pay out more than its net income in dividends to its shareholders for long. This common-sensical finding, though, is not borne out by recent history. Dividend preservers tended to have a higher payout ratio, on the average, with a median ratio of about 28%, versus 16% for Dividend Cutters.

Graph produced by author. Price data from Yahoo Finance. Fundamental Data from Morningstar.

What the data do suggest, though is that negative payout ratios -- that is to say, dividends paid despite negative profitability -- are associated with dividend cuts. This, too, is intuitive, but nevertheless the tendency of the payout ratio to be a poor predictor of future dividend cuts remains. This finding does suggest, though, that other fundamental data related to a company's health -- say, its earnings history and the trajectory and volatility of that history -- have more to do with it.

This is borne out in the data. Dividend Preservers tended to show relative consistent EPS growth with less volatility in year-to-year EPS changes, and the opposite behavior was demonstrated by Dividend Cutters. In fact, a year-over-year decline in EPS of 8% or more was associated with an OR of 2.32 for a dividend cut the following year.

Graph produced by author. Price data from Yahoo Finance. Fundamental Data from Morningstar.

Moreover, not only did negative EPS growth in the previous year suggest an impending dividend cut, but EPS volatility in the years preceding the EPS decline was also associated with a dividend cut as well, suggesting that perhaps companies in cyclical industries might be inclined to proactively manage their dividend policies to adjust to changing environments.

Graph produced by author. Price data from Yahoo Finance. Fundamental Data from Morningstar.

Other measures of financial health seem to predict dividend stability. Dividend Cutters were, on average, more likely to have lower Current Ratios just before their dividend was cut as compared to Dividend Preservers, by a factor of over 400 basis points (1.2 versus 1.6).

Graph produced by author. Price data from Yahoo Finance. Fundamental Data from Morningstar.

In fact, a current ratio of less than 1.2 was associated strongly with an impending dividend cut, with an odds ratio of 1.7.

Finally, looking at Return on Invested Capital:

Lower ROICs were also associated with dividend cuts, with ROICs of less than 0.01 being tagged with an OR of 2.371 for a dividend cut the following year.

Again, while all these factors seem intuitive, the values themselves yield some interesting guidepoints for determining if a company's dividend is at risk. To wit, Procter & Gamble's dividend yield has been climbing of late as the company's EPS growth has slowed and its payout ratio has climbed to a concerning 103% as of last fiscal year. Its y/y decline in EPS from 2014 to 2015 from $4.01 to $2.44 is naturally concerning. But with returns on invested capital still cresting 9%, a current ratio of 1.00, and relatively low volatility in EPS in prior years, the odds are that P&G will find a way to muddle through, and a dividend cut seems unlikely for this Dividend Aristocrat despite the recent high payout ratio.

Compare P&G's situation to that of investment publication firm ValueLine (NVALU), which pays an attractive 4.74% ratio. Most recently, management increased the dividend from $0.15 quarterly to $0.16, which seems reasonable given TTM EPS of $0.75. However, its current ratio of < 1.2 (0.7 currently) and ROIC < 1% (currently -2.3%) would seem to place its dividend at risk, regardless of the sustainable payout ratio, and caution regarding the shares would seem to be warranted.

Other Observations:

Looking at the data by sector, some trends do stand out.

Companies that tend to sustain their dividends over time are unsurprisingly in relatively defensive industries, for the most part: Healthcare and Utilities tended to be stronger maintainers of their dividend policies than, say, companies in Materials, Industrials, or Energy. Perhaps surprisingly, Information Technology companies scored highly with regard to dividend preservation. (To my mind, this speaks to shifts in our economic makeup, away from a manufacturing-based economy to a service and information technology-based one instead). Also perhaps surprisingly, Consumer Products (both Staples and Discretionary) were not as stalwart in maintaining their dividends as one might have thought, though the reasons for this are bit unclear to me currently.

Conclusions:

By no means is this study exhaustive or conclusive; certainly, other metrics exist that may do an even better job at predicting future dividend cuts. However, I am still a bit surprised that payout ratios are perhaps less effective at predicting dividend cuts than might reasonably be assumed. Focusing on other, more core aspects of a business appears to yield better actionable data when it comes to determining dividend risk.

Disclosure: I am/we are long PG,JNJ,SEP, MMP,HCP,VTR, ETP. 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.