Would I, could I, predict dividend growth?

I was working on a contribution to another guy's idea; it involves picking stocks likely to clock high dividend growth.

Here's the catch; I'd to choose from stocks that already have at least a 3% dividend yield. Oh, and I'd also have to choose some stocks that will have the highest 5 year average dividend growth in 2020.

So, I started thinking about how to put a fence around that task.

First thing is easy;  find equities that have a 3% dividend yield. I decided to limit it to common stocks and ownership units that pay distributions. No funds, no preferreds, no ETFs, ETNs, etc. I also decided they should trade on US or Canadian exchanges. It turns out there are about 17,000 names to sort to start with, and the criteria above bring it down to 2900 some-odd options.

Predicting dividend growth is about the future, but the only data we have is present and past. Now, analysts spend a lot of time attempting to predict earnings growth. They set expectations and companies either meet them or don't meet them. I don't understand much of what they do and how they do it.  I'm left with my own experience and common sense to puzzle through this assignment.

First; any company that already pays a 3% dividend yield has a significant commitment to dividend policy. Companies in the rapid growth mode usually retain all earnings in order to reinvest in the business. As business value rises, stock price rises and executives get rich. The company can raise money by selling more stock, or by borrowing, but the cleanest way to grow is to retain earnings.
I think a young, growing company will rarely initiate a dividend, much less raise a dividend significantly if it has to borrow to fund growth. Once a company has a history of paying dividends and owners expect to be paid, freezing or cutting a dividend has an immediate and negative effect on stock price, so executives are loathe to provoke the market in such a way. It can hurt the executive directly, if he or she has stock options that vest at a certain price. There's another target for excess earnings; stock repurchases. In fact, the combination of stock repurchases and dividends is often called shareholder yield, as if the stock price puts money in the pocket of the owner. In fact, it does drive up earnings/share, and that may drive stock price, but it deprives the owner of determining how his/her personal stake in the company's capital is allocated, so I don't see stock repurchases as equivalent to dividends. Since I routinely reinvest dividends within a tax-protected environment, they are essentially equivalent, but I still claim the right to make that decision, rather than have the management make it for me.

The 3%+ dividend payer  is either a mature company that takes a more measured approach to growth, doesn't see adequate opportunity to allocate capital to growth, or is simply producing such a prodigious amount of cash that it can afford to pay that dividend and still have lots of cash to invest in growth.  Some companies pay routinely high dividends; they have invested what they need in infrastructure, produce substantial cash flow, but have limits on their ability to grow; They will pay a high dividend but won't raise the dividend significantly.

So, in order to find a 3+% dividend payer that is also growing the dividend significantly, it has to either be in that enviable place where the is a ton of cash flow, it is growing rapidly and can afford to pay a significant dividend that is still only a fraction of the cash it produces each year. If the cash is growing rapidly, so can the dividend.

One can also see significant dividend growth off a 3% base if the company is in transition from rapid growth to slower growth, but produces significant excess cash flow. It has targeted a significantly higher payout ratio than current and is incrementally marching towards that higher payout ratio that will then track in parallel to it's terminal rate of growth.

So, my task is to identify 5 companies that have 3%+ dividends and will raise them the most in 2016.
Also, I am tasked to identify 5 companies that will have the highest 5 year dividend growth rate in 2020 off of that 3% base.

Let's get some things straight from the outset:  Price only matters now. Price matters now because it defines the dividend yield.  There are some companies that I can't consider because their recent price action has driven their dividend yield below 3%. If their dividend is hiked later or their price drops later, too bad; they can only be considered based on their current dividend yield. Second; dividend growth is not the same as a rising yield. A rising yield means that the price is not keeping up with the dividend, or the price is dropping, or the company is hiking the dividend disproportionately to  rising price.  Dividend policy essentially never follows price, if I could venture a guess. It is much more likely to follow earnings, unless the company has come incentive to continue hiking the dividend in a stagnant earnings growth environment. That would be reflected in a rising payout ratio.

The company that is growing earnings rapidly, raising the dividend proportionally and experiencing capital appreciation to match earnings and dividend growth will have a very steady yield.

Where am I going to look to find companies that have significant growth in earnings, significant increase in dividends, starting from a 3% dividend yield?

First: I can see what the analysts are predicting.
Second, I can look to the past, see how these parameters have been trending, make an educated guess about whether business conditions are getting better or worse, and try to predict the dividend behavior based on that analysis. It's also good to remember that we're talking about dividends/share. A company spending a bunch on share repurchases can drive dividends per share by reducing share count.  That could be another clue to a company that can raise dividends per share, even if earnings growth is blunted. 

I took two approaches;
     The first was to examine my own portfolio, which is chock full of companies paying 3% dividends with which I am already familiar.
     The second was to use the screening tool in F.A.S.T. Graphs.  That required that I define the screening parameters I would use to narrow my choices from 17k to 3k, down to about 50 companies.
I'm not likely to be able to discern the dividend growth prospects of a company that is paying a dividend for the first time in history. The history of dividend payments and changes in those payments lends confidence to the observer in predicting how things might look next year and in 5 years. If the company has a long track record of behaving one way or another, that lends a degree of reliability in predicting they will continue to behave that way in the future. One can't know how a company will respond to a recession, unless you look back to how they responded in the last recession. Limiting the search to companies who grew earnings and dividends in the last recession would be a reasonable screening tool.  Since the last recession hit in 2008, and the one before that in 2002, one would need a 10 year history to catch the major recession and a 15 year history to catch two recessions.

Looking forward, one has to ask what industries are likely to grow earnings at an accelerated rate, remain central in consumer demand and have room under the earnings tent to push the dividend?
One sector where my crystal ball is REALLY fuzzy is tech. It's easy to look back and see who has survived the wars in Tech, but far more difficult to look forward and ask who will still be performing in 5 years. I don't invest in those companies that depend on being at the front of the Zeitgeist in personal tech appliances, so I'm not likely to choose one of them for my list of darlings.

What about market cap? In general smaller companies grow faster than large ones. That means growing revenues, hopefully also earnings. One won't pay a dividend without becoming profitable, or having positive earnings. One won't pay a rising dividend without positive and rising earnings. Is there likely to a 3%+ dividend candidate amongst smaller companies? If you include REITS and MLPs and you consider companies less than 5Billion in revenue to be small, absolutely.

My real problem with this assignment is that I'm not motivated to find 5 companies paying more than 3% dividends who will grow dividends the fastest and have the highest 5 year dividend growth rate 5 years from now.  I want companies that are most likely to grow steadily, without volatility, and least likely to have a melt-down. My view of performance is more about predictability, protection of capital, credit quality, conservative and prudent management than it is about rates of growth. My target total return is around 10% per annum.  Sure, I'll take more, but I'm not willing to bet the farm on flyers. I also hold 50+ equities, so I will be bouyed and anchored by averages.

It turns out I understand the tools to do the screening, but have no stomach for guessing which companies that screen well are likely to perform best. Furthermore, I may not be inclined to buy them for reasons stated above, so I'm not highly motivated to rank them at the top. So, as I probe the depths of my soul, I'm not the right guy for the assignment.

The challenge was not entirely lost on me, however. It got me to thinking about looking forward with the stocks I own. Surely, amongst the 50 issues I already own, I can stratify for those most likely to out-perform in the next 3-5 years and then think about why I'm holding the laggards. Maybe I could trim that list to 30, or even 25 companies...hmmm...