Wednesday, June 10, 2015

How to value dividend stocks

In my investing, look for businesses I can understand that have some sort of a competitive advantage that translates into consistent earnings power. I try to determine if I believe this business will be around in 20 years, and still have a consistent earnings power, despite obsolescence, competition and regulation. If I believe that to be true, I can then look at trends in earnings, dividends, revenues, returns on equity, payout ratios and make an assumption of what future growth might look like.

This is the part where I also look at valuations. My primary valuation method is often the P/E ratio, relative to 20 times past and/or forward earnings. Of course, regular readers know that this benchmark is not viewed in a vacuum, but in conjunction with past trends in earnings, dividends revenues etc and prospective growth in those earnings and dividends. It is also used as one of the ways I screen for companies for further analysis. P/E ratio is also one of the criteria I use to compare between several companies, before I make my choice.

Over the past few months, I have received a lot of mixed reviews on the type of analysis I perform. I believe the main reasons behind those reviews is the fact that I do a lot of things manually, and have described my outcomes, but never really put everything together in one piece. Actually, a lot of the reasons why I invest the way I invest are also scattered around in a few articles.

Unfortunately, I have found that a majority of readers are not willing to go through the archives or even dig around links provided in an article, in order to find answers to their questions. I know this, because quite frequently I receive questions that could have easily been solved had the reader clicked at a link in the article the hopefully read. In a few scenarios, I have had readers ask me questions that have already been addressed in the article. On the other hand, it is also unreasonable to expect that someone would have the time to sift through 1000 articles I have written on dividend investing.

I think it would be easiest for everyone, if I refer to a few posts I have written on using P/E ratios to value companies.

There are several constraints I face when evaluating a company. I try to discuss them in the following articles:

Not all P/E ratios are created equal

Why do I use a P/E below 20

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

How to read my stock analysis reports

As you can see, I use P/E in conjunction with past and projected growth, when screening for and comparing between dividend stocks. This is further complicated by other constraints, such as portfolio weight. For example, I might like Johnson & Johnson (JNJ) a lot, but I would not buy more, since it is one of my largest positions. In addition, I am intentionally limiting myself to only companies with P/E ratios below 20 for a reason. The most dangerous thing new investors do is see a company that has done well recently, and project recent successes to the sky. Most often, investors overpay dearly for hot growth concepts and expect trees to reach to the sky. Unfortunately, when you purchase a hot growth stock at 30 times earnings, you are essentially paying for the growth in the next 5 or even 10 years and assuming that things will go smoothly in the future. If there is a slight derailment of plans, you will not earn good returns for as long as the first decade. For example, Starbucks (SBUX) is selling for 38 times 2014 earnings and 33 times forward earnings today. In 2014, the company earned $1.36/share, and for 2015, Starbucks is expected to earn $1.57/share. The projected earnings per share for 2018 are approximately $2.50/share. This means the stock is selling today at 20.90 times earnings in 2018. Even if Starbucks manages to earn $2.50/share in 2018, it could still deliver unstatisfactory returns if the P/E contracts to 20. I would be taking a lot of risk when valuation is overstretched. The risk is that paying a high multiple leaves no margin of safety in case the future doesn't turn out as expected. I also ask myself why should I pay for a company at 30 times earnings that doubles earnings every 7 years, when I can find another company that sells for 20 times earnings. A few potential candidates could include Ross Stores (ROST), TJ Companies (TJX) or Ameriprise Financial (AMP).  As you can see from above, I compare different companies that have a P/E below 20, and pick the ones where the P/E and growth combination is best.

Another constraint I have not really discussed in full is expected investment return. I expect equities to deliver an annual return of 9% - 10%/year for the next 30 years. Therefore, when I select companies, I try to determine whether an investment at today’s prices could deliver a 9% – 10% expected returns. This is something I have not discussed exclusively, because I always assumed that reasonable people would understand that a company that yields 3%, grows earnings and dividends by 6%-7%, could deliver expected annual returns of 9%-10%/year. I also sometimes try to stress test results, in order to see how different outcomes can make or break the returns. Check this article on Hershey from a few weeks ago. Let's take a situation where I pay 30 times earnings today, and the earnings double in a decade. If the earnings stream is worth only 15 times to others, then I have not really earned much in terms of a price return. If the dividend yield was approximately 2%/year, my only return would have come from that small initial dividend yield. Identifying the best quality company is not enough – you also have to buy it at a cheap enough price.

Chasing growth is a dangerous game. What truly matters to the investor is good returns, not overpaying for future growth. Actually, if you pay too much for expected future growth, you might not do too well. For example, between 1957 and 2003, IBM had much better growth prospects than Exxon. IBM had higher revenue, and earnings growth and even dividend growth than Exxon. However, Exxon returned more than IBM, because it always had lower P/E ratios and higher dividend yields. When you have low expectations behind a business, it usually sells at cheaper valuations. This also allows the reinvestment of dividends at those cheaper valuations, which turbocharges returns.

I use earnings per share, but also normalize it for one-time events. For example, in 2012, Coca-Cola appeared cheaper than it was, because of one-time accounting items. In early 2013, Johnson & Johnson appeared more expensive than it really was. For those like me, who like going through annual reports, and press releases, I can identify some of the reasons for annual EPS fluctuations outside the norm. For others, looking at forward earnings and applying a forward P/E ratio could have been a very good approximation of the intrinsic value of the company. In fact, many times I have found that looking at a forward P/E ratio is a good enough shortcut to compare P/E ratios between tens or hundreds of stocks, rather than poring over hundreds of press releases.

For my analysis, it is helpful to look at trends in earnings per share over the preceding decade. However, while I look at the past years earnings, I also look at the prospect for earnings growth in the next two years or so. Purchasing a company at a cheap valuation is helpful, but I also want to see prospects for increase in earnings per share. Otherwise, there would be no future fuel behind future dividend increases. In many cases, looking at the past year earnings and the future year earnings paints a better picture. For example, oil companies like Exxon Mobil (XOM) sell at pretty cheap valuations when you look at past year’s earnings of $7.59/share ( equivalent to a P/E of 11 times earnings). The earnings from 2014 however do not accurately reflect the decline in oil prices. Therefore, one needs to look at estimated 2015 and 2016 earnings per share of $4.26 and $5.34/share ( for an equivalent forward P/E of 20 and 16 times earnings). Those paint a more accurate picture of the near term earnings power for Exxon Mobil. While forecasts are not 100% accurate, I have found them to be close enough for my investing needs. When I buy stocks, I never expect precision. If I demand precision in past and future prospects for earnings and dividends, then I set myself up for disappointment. This is where the margin of safety principle comes in handy.

I apply a set of quantitative criteria on the list of dividend champions. Assuming I have analyzed those companies, I would compare them against each other. I do not assign fair values, like everyone else however. I look at inputs such as P/E ratio, earnings growth, dividend growth, yield, dividend coverage, and my expectations for the future, before choosing an investment from the pack. This is more manual than merely comparing to a calculated fair value, but I think it is easier to think through all the numbers in detail, rather than create a shortcut and miss thinking about an important item.

Overall, screening the list of dividend growth stocks, comparing between dividend stocks, analyzing dividend stocks and hand selecting companies to invest in is a very manual process. Unfortunately, in order to learn how to be a good investor, you need to do the work to have a right to an opinion. In order to succeed in any activity in life, you need to spend thousands of hours and several years learning, perfecting and adapting your knowledge. Continuous learning is important. At the end of the journey however, the reward would be a well maintained dividend machine, which will take care of your needs forever.

Full Disclosure: Long XOM, JNJ, KO, IBM, ROST, TJX, AMP

Relevant Articles:

Stress Testing Your Dividend Portfolio
Buying Quality Companies at a Reasonable Price is Very Important
Dividend Investing Knowledge Accumulates Like Compound Interest
The work required to have an opinion
How to never run out of money in retirement

Popular Posts