# Identifying Undervalued And Overvalued Stocks

|
Includes:
by: Ironman at Political Calculations

How can you tell if a given company's stock is overvalued, undervalued or priced just right?

If you're the kind of trader who likes to try their hand at picking winners and avoiding losers, that's the million-dollar question, isn't it?

But for stocks of firms that pay dividends, there may be a tangible way to determine the answer to that question, provided that you know how those firms will act to change their dividends in the future. And as it happens, we have a speculative forecast for how five S&P 500 firms are expected to do just that in 2017, but who, as yet, have not declared what they will do. The following list summarizes the forecast for dividends for each of the companies:

• Comcast (NASDAQ:CMCSA) - 15% increase
• Home Depot (NYSE:HD) - 14% increase
• Cisco Systems (NASDAQ:CSCO) - 15% increase
• TJX Companies (NYSE:TJX) - 16% increase
• Vulcan Materials (NYSE:VMC) - 25% increase

Being able to do something useful with this knowledge requires some familiarity with the concept of the dividend discount model, where stock prices are recognized as representing the approximate net present value of all dividends that will be paid to shareholders in the future, and also the part of the efficient market hypothesis that says that stock prices will almost instantaneously incorporate information about the future as it becomes known.

[In reality, the relationship between the current prices of stocks and their expected future dividends per share is more complex than these two simple assumptions suggest, but we can still put them to work to do something interesting. (Remember the motto: All models are wrong, but some are useful!)]

The way we'll do that is to chart the relationship between the stock prices at the beginning of the last several years and the actual annual dividends per share that each company would go on to pay out during each of those years.

For the historical data shown in this chart, what we're effectively doing is pretending that investors had perfect knowledge of what dividends each firm would pay out during the year to come, which they incorporated into the closing stock price for each company on the first trading day of each year. We then calculated linear trend lines to approximate the relationship between the first trading day of the year's stock price for each company and the forward annual dividends per share that each would pay, extending the trend lines out past the point of the forecast dividends per share for 2017. We then added one last point, indicating the forecast annual dividends per share for 2017 for each company along with its stock price as of the close of January 6, 2017.

As you can see, we obtained some pretty interesting results. For three of the firms, Comcast, TJX and Cisco Systems, we find that the forecast data points for 2017 are all within a fairly small margin of error of the trendlines projected from the historic data. This result suggests that the stock prices of these firms are priced about right.

But we have some significant deviations when we look at the forecast values for both Home Depot and Vulcan Materials, where Home Depot's stock would appear to be undervalued and Vulcan Materials' would appear to be overvalued.

Are they, though? Could VMC be experiencing something of a speculative bubble that could soon pop? Or could investors be expecting the board of Vulcan Materials to boost the company's dividend by more than what has been forecast for 2017?

Similar questions arise about Home Depot. Is the stock price being unnaturally depressed by investor pessimism about the company's future prospects? Or does HD's board have plans for a much smaller dividend increase than the forecast anticipates?

And how might potential changes in either Home Depot's or VMC's stock buyback programs change the dividend per share math (and stock prices) for the firms?

All you need to do to pick winning trades is to answer questions like these! We should know most of the answers to the questions we've asked in this post by the end of 2017-Q1, if not early in 2017-Q2.

On a final note, the kind of analysis we just showed isn't new by any stretch, but more often than not, when it has been presented, has typically been purely backwards-looking. What we're doing that's new is applying it to look forward in time with information that is still potentially actionable. We're looking forward to finding out how it all plays out.