In order to figure out what fair value you ought to pay for a stock, you need to know four things: the current profits, the number of shares outstanding, the capitalization rate, and the expected growth rate. You already know two of those things: the current profits and the number of shares outstanding. In other words, the value of any security comes down to figuring out the relationship between the capitalization rate and the expected growth rate.

Let’s imagine then, if you are familiar with Coca-Cola’s history of dividend growth dating back to 1963, know that it sells over 10,000 products per second in over 200 countries, and derive some solace from the fact that the most famous investor in the world has made a lot of money owning Coca-Cola for the long term. But even though overpaying for great assets can eventually work itself out over an extremely long holding period, you still want to follow Benjamin Graham’s advice to go through life without overpaying for an asset.

What do you do? You take the current profits, and then divide that figure by the capitalization rate – growth rate. The difficulty of investing is trying to get the capitalization rate and growth rate right because it is more art than science.

When calculating the capitalization rate, you ask yourself the question: If there were no growth from the business, and the business instead chose to pay out all earnings to me instead of doing anything that would grow the earnings per share, how much income would I demand from the stock on annual basis? Because this is a fiction that involves a no-growth universe, investors tend to look to the historical returns of 12% on cost to arrive at the initial capitalization rate. This assumption assumes that rational investors, in times of ordinary interest rates, insist on $120,000 in cash for every $1,000,000 they deploy into a no-growth business. As a result, 12% gets treated as a rebuttable presumption in capitalization rate analysis.

Then, the next thing you do is figure out the weighted average of expected growth rates. You ask yourself what the best case, moderate case, and worst-case returns for earnings per share growth might be.

You might look at the lowest analyst predictions and figure there is a 25% of 5% annual earnings per share growth happening. You might look at the ten-year track record of 8% earnings growth and conclude there is a 50-50 chance of that happening (50%). And, if you look at the strong dollar and figure that Coca-Cola has a 25% chance of exceeding expectations due to a weakening dollar and strong operational performance, you might figure there is a 25% that the company will grow earnings at 11%. This gives you an expected growth rate of 8%.

Now, you have the numbers you need to figure out the per share price that reflects the value of Coca-Cola. If we had to represent it as an equation, it would be V= P/ (C-G). P refers to current profits, C is the capitalization rate, and G is the growth rate. The P at Coca-Cola is $8.6 billion in profits per year. The C is the 12% capitalization rate we presume. The G is the 8% growth we just calculated. And V is the overall value of Coca-Cola stock.

Plugging in the numbers, we get V= $8.6 billion / (.12-0.08). Or $8.6 billion / 0.04. That gives us a value of $215 billion for Coca-Cola as a whole. But of course, Coca-Cola isn’t a private business owned by one person. It has a corporation that has divided itself into 4,342,000,000 ownership slices that collectively represent the entirety of the common stock ownership. So we take the $215 billion value and divide by 4.342 billion shares, and we get a price of $49.51 per share as the fair value of Coca-Cola.

Now, of course, this is a simplification of the process. There are all sorts of variables that would affect the process. If you believe the U.S. dollar is stronger than where it ordinarily will be, the the current profits of Coca-Cola are understated and the fair value of the stock is going to be even higher than $49.51 per share.

On the other hand, if you believe that interest rates are going to be higher over the measuring period than they are now, then you would have to tinker with the default presumptions of the capitalization rate.

And, if you analyze the spectrum of potential growth options for Coca-Cola across different probabilities than the three I provided, then your returns will be affected depending on whether you are more optimistic or pessimistic than the projection probabilities that I provided. Also, debt on the balance sheet may rise or fall, and this may also affect the capitalization rate as total leverage has a proportionate relationship to bankruptcy risk.

In other words, I’m not writing this article to convince you that the value of Coca-Cola is somewhere between $49 and $50 per share right now. Instead, it is to lay out an approach to calculating fair value that involves the capitalization, growth rate, and profits of the business. It also should explain why non-cyclical industrials make such great investments during the down part of the business cycle.

It’s not just the reinvested dividends of Emerson Electric at $24 per share during the financial crisis that drives returns. It is also the fact that the investor community struggles to predict the earnings of companies that have unpredictable earnings. If we went back to 2010, and had to make five-year projections about Exxon and Coca-Cola, the projection for Coca-Cola is going to be a lot more accurate than the Exxon prediction because foreseeing the oil price decline is difficult.

But you can use this unpredictability to your advantage. When a business cycle starts bottoming out, the growth rates attached to industrial and cyclical stocks tend to become unduly pessimistic. Even if you can’t figure out whether Emerson Electric will grow at 8% or 12% long-term from the lower base, it doesn’t matter if you are trying to pick up undervalued securities. If the market is anticipating 4% long-term growth from that vantage point, your knowledge that the growth is somewhere above that 4% by a decent amount is enough to give you great returns even if you can’t pinpoint what the figure may be.