Tuesday, May 27, 2014

Why I don’t do discounted cash flow analysis on dividend stocks

Imagine that you have two companies that are selling for the same valuation, which are also in the same industry and have similar current yields. Both companies would be characterized as “quality dividend paying ones”, as they each have strong branded products that consumers buy, good pricing power, and opportunities for growth. Coca-Cola (KO) and PepsiCo (PEP) fit my idea of what a quality dividend growth company is.

The question to answer is, if you wanted to buy the shares in just one of the companies, which one would you purchase?

For me, the answer is always to look for the one with the higher growth potential. It is a no brainer decision to select the company with the highest expected growth, particularly if the valuations are similar.

In reality however, investor expectations are just that, expectations. They do not become tangible fact, until enough time has passed in order for you to determine if those expectations occurred or not. In other words, if I expect Coca-Cola to double earnings per share every decade for the next 30 years, I would have to wait until 2044 until I can say if my expectation was right. One can expect a lot of things changing from now until 2044, many of which might not be even comprehensible for someone thinking about the future in 2014.

Therefore, in my stock selection process, I focus on the quantitative factors and the qualitative factors, but I always try to leave room for error. This is why I never really put exact fair values on stock prices, based on coming up with a future growth rate, and then discounting it back. At the end of the day this method could provide the investor with a fair value and a premium or discount relative to current prices for companies they are analyzing. However, this method would also provide you with a precise numbers, which in reality are just an illusion that is dependent on investors assumptions. If you change your assumptions, and the “precise numbers” changes as well. This is why I merely bump companies against each other, and select the ones that look best. I described how I compare between dividend companies in a previous article.

The precision behind those fair values is an illusion that is derived from making expectations on future growth rates in earnings or dividends, and discounting them back using an estimated rate. I believe that investors who only look at that are essentially outsourcing their ability to think critically to a formula.

Instead, I believe that the better way to do things is to keep it simple. If you have companies with similar valuations but higher growth rates, you might want to go for the growth one. However, you might also decide to hedge your bet and purchase the lower growth one, just in case. It doesn’t make sense to make a lot of projections, do excess calculations, and discount those amounts to the present, when you can merely compare the expectations at the same levels of valuations and be done with it. However, if you can find a company that sells at a lower valuation, and can still provide consistent growth, I would say you might have found yourself a winner. In the case of Coca-Cola versus PepsiCo, I would say both companies might be good acquisitions today, even if one has different growth expectations behind each.

For example, Coca-Cola (KO) sells for 19.40 times forward earnings. The company yields between 3% and has a five year dividend growth rate of 8.10%/year. The ten year dividend growth rate is 9.80%/year. In comparison, the company has managed to grow earnings per share by 7.90%/year in the past decade. Check my analysis of Coca-Cola.

PepsiCo (PEP) sells for 18.90 times forward earnings, yields 3.10%, and has a five year dividend growth rate of 7%/year. The ten year dividend growth rate is 13.70%/year. In comparison, the company has managed to grow earnings per share by 7.70%/year in the past decade. Check my analysis of PepsiCo.

I also believe one needs to do the actual work in looking at companies, researching past data, and making some assumptions about the future. However, as we know that assumptions are tough to figure out exactly in advance, it is important to try and focus on best valuations available at the present time. This could be evidenced by a low price to earnings ratios for example. After all, sustainable earnings per share, coupled with a moderate but consistent growth in those earnings per share will result in quiet but sure compounding of wealth and dividend income over time. In other words, why swing for the fences, chasing a percent in yield or growth, when you can do reasonably well and protect and grow wealth and income? The risk in chasing growth at any price is that if growth does not materialize, you may be stuck holding on to a company that doesn’t deliver much in dividends and dividend growth, while your capital contracts in value. In contrast, for a company with a dependable earnings and dividends growth, purchased at an attractive valuation, a decline in growth would not affect an investor as badly.

For example, if you purchase a company that yields 2% and grows dividends at 15%/year, you might do well for many years, while getting an increasing yield on cost. However, if growth slows down, you might be left holding a low yielder without the benefit of fast growth. In comparison, a 3% yielder that grows by 7% is not too bad of a choice, even if growth drops to say 3- 4%, because the yield is sufficient to cover a higher portion of expenses in retirement. Of course, those growth rates are not known in advance, as only the yield is locked at time of purchase. However, even assumptions about its sustainability are subject to folly as well.

In reality, each company will have different yield/growth characteristics, which is why you need to assemble a diverse portfolio of quality businesses, purchased at attractive valuations. That safety in numbers can be very beneficial in guarding the investor against folly. I have chased high yielding stocks before, and have gotten burned in the process. I have also chased high dividend growth stocks which stopped raising dividends as fast. But by assembling a mix of the three types of dividend growth stocks I focus on, I ensure that my diversified portfolio can withstand different pressures on its components, and still provide me with the growing income in retirement.

Full Disclosure: Long KO and PEP

Relevant Articles:

The Tradeoff between Dividend Yield and Dividend Growth
How to read my stock analysis reports
My Retirement Strategy for Tax-Free Income
How to be a successful dividend investor
Should I buy more high yielding stocks in order to retire early?


  1. By doing a DCF with solid assupmtions (not optimistic) you can get a good estimate and you can see if you are overpaying, also you can apply a margin of safety to your valuation. There is no need to be too precise. Also with a DCF you divide the sum of FCF and perpetuity FCF with the number of shares of the company, which can make you realize if you are getting a good portion of those future cash flows. This is good to decide in which company you should invest if you are comparing more companies in the same industry. The mathematical part of this game is very important. Regards,

  2. DGI,

    I use a Dividend Discount Model analysis to value stocks, which is a close cousin of the DCF model.

    I use it as a general measure to get a rough idea of what a stock might be worth. Without some idea of what a company is worth, I find it nigh impossible to know if I'm underpaying or overpaying for equity. However, as you mention changing the input even slightly will give you different output. As such, any analysis or model must be used with a grain of salt. You're not looking for an exact number, but rather a rough guesstimate as to what equity might be worth on a per share basis. From there you can determine if there's a large enough margin of safety to make up for incorrect input or growth rates that don't live up to expectations. I generally take my own analysis and blend it with analysts' valuation estimates to get a general idea of what a stock might be worth at any given time.

    Best wishes!

  3. DGI,

    I'm in agreement with you. DCF models contain far too much uncertainty to be reliably credible - while numbers provide a sense of comfort in doing "analysis" really it can mislead as much as it can inform. I prefer analyzing a company's moat to determine how it can avert downside scenarios. Remember, companies are made of people and reducing an organization down to "an average growth rate" doesn't make sense.


  4. great article. Now u have me rethinking everything lol.! like you said a right mix of all kinds of dividend growth stocks is neccasary because every company is so unique.

  5. Hi DGI,

    I need to use the DCF very often in my current job as a 'valuation specialist' - usually valuing equity in businesses for accounting / tax / transaction purposes for corporates, which I've been doing for around 10 years now. The DCF is a huge part of what we rely on (but always with a multiple of earnings cross-check to other market benchmarks!), but despite all the complexity that goes into these valuations, I'm in agreement with you about keeping things simple - particularly for us investors. I think the DCF is better as a useful 'sense-check', as Dividend Mantra mentioned, or to test the impact of possible future scenarios. For example, if your DCF suggests the current share price is assuming negative 3% growth in earnings forever, that might indicate a good margin of safety.

    No one can predict what events will happen, even to the most stable companies, so it's much more important to use these techniques to ensure a reasonable margin of safety and avoid huge losses. Some great food for thought here - thanks DGI!


Questions or comments? You can reach out to me at my website address name at gmail dot com.

Popular Posts