Each summer, my wife's family spends a week at the beach, and since taking the horde out to dinner each night isn't exactly economical, we stock up on food and drinks before we leave.
Over the years, the stocking-up process has taken on a life of its own -- competition is second nature in a big family -- and we'll drive to five stores just to see who has the lowest price on things like macaroni and cheese and paper towels. Never mind the gas costs -- we want to save a dime on Sprite!
I'm sure you have similar stories of frugality in your family. I do find it interesting, though, that as consumers, we'll go to great lengths to find the best value, but as investors, we tend to do far less research before making a purchase. As Buffett reminds us, "Price is what you pay, value is what you get," so before buying any stock it's critical to understand just what value we're getting for our money.
For the weekend investor who is juggling life's myriad responsibilities, finding the time to read about or run a valuation on a company may be next to impossible. Fortunately, there's a valuation model available that is both simple to use and is rooted in good financial sense.
More simplicity, less complexity
The "H-Model" was put forth in 1984 by Russell Fuller and Chi-Cheng Hsia in the Financial Analysts Journal, and it's based on the principles of the dividend discount model (DDM), which states that the value of a stock is the present value of its future dividend payments, discounted at the company's cost of equity.
What I like about the H-Model is that it assumes a smooth linear decline in the dividend growth rate over a set number of years rather than the sharp declines assumed in some other models. The H-Model's assumptions also seem to better reflect the natural life-cycle of a business -- as competition sets in and growth harder to come by, all companies will inevitably slow down, but in most cases the decline is gradual and not abrupt.
The H-Model equation can be written as:
Stock value = (DPS * (1+gL)) + (DPS * H * (gS-gL)) divided by (r-gL)
DPS = current dividends per share
H = half-life of extraordinary growth (i.e. if you expect 10 years of extraordinary growth, H = 5)
r = Cost of Equity
gL = Stable growth rate
gS = Initial high growth rate
The dividends per share figure is the only input that doesn't need to be supposed. Starting with the stable growth rate, this is the rate that the company is expected to grow in perpetuity, so it can't grow at a rate above the overall economy (2%-3%). Otherwise, mathematically we'd be assuming the company would eventually become the economy.
For the initial high growth rate, a good starting point is the annualized dividend growth rate over the past three or five years, unless it was a period of exceptional growth.
The next step is to figure out how long it could take for the company's high growth rate to decline to the economy's rate. Surely some companies with strong competitive advantages could grow at above-average rates for the next 20 years (an H-value of 10), but not every company has a wide economic moat. For the latter, a good starting point is 10 years (an H-value of five).
Finally, the cost of equity can be understood a number of ways, but it's essentially the return the company must offer to shareholders to compensate them for taking the risk of owning its stock. For most blue chips today, a good initial estimate is 8%-10%.
The good and bad
Dividend-based valuation models make intuitive sense. After all, as equity owners, the only real cash flows we receive are dividends. Yes, a company may have extra cash flows that it isn't paying out as dividends, but we can't exactly walk into headquarters and demand legal tender unless we care to meet a nice security guard.
On the other hand, dividend-based valuations do have their limitations. First, they work best with, well, companies that pay dividends and ideally those with stated dividend policies or grow their dividends in some relation to earnings growth. Second, they don't give a company credit for the free cash flow it has retained for other purposes, so a dividend-based model may actually undervalue companies that don't pay out most of their free cash as dividends.
Finally, the H-Model assumes a steady dividend payout ratio over time, even though companies tend to pay out a larger percentage of earnings as they mature. It works best, then, with larger companies that already pay out at least 40% of earnings as dividends.
Now that we've kicked the tires, let's take the H-Model for a spin and try it on some real stocks.
For each company in the following table, I've made the following assumptions:
- The initial growth rate (gS) is based on Bloomberg's three-year dividend growth estimate.
- The long-term growth rate (gL) is equal to 3%.
- The period of extraordinary growth will be 20 years (H = 10).
- The cost of equity is 9%.
Dividend Per Share
3-Year Projected Dividend Growth Rate
Philip Morris International
Source: Capital IQ, a division of Standard & Poor's, and Bloomberg as of Jan. 13.
It's important to note that these results aren't meant to be the final word on these stocks' valuations. Some of these companies may indeed have higher or lower cost of equity, for instance. Also, remember that dividend-based models don't give companies credit for retained free cash, so one way to think of the results is as a "floor" on the stock's value.
Valuing a stock may not be as easy (or as fun) as finding value in the supermarket, but it's still a critical step in the investment process. The H-Model has its drawbacks and limitations, but it is a relatively simple and intuitive model that may help you avoid overpaying for your stocks.
Valuation is a key element of Motley Fool Pro's research process. The team has set a lofty goal of closing at least 75% of its positions for a profit (it's well ahead of that at this point), and that means making sure they don't consistently overpay for their investments.
On Tuesday, Jan. 18, Pro will open for just a few days for the first time since June to a small number of new members. If you'd like to learn more, simply enter your email address into the box below.