Yield Rank's  Instablog

Series Part 3: Risk & Reward - Keep 'em separated

This is the third in my series (on Dividend Growth Stock pricing. In my second Instablog I introduced the concepts of valuing DG (Dividend Growth) stocks compared to bonds and compared to each other.

The focus of my work has been on the pure cash flow analysis side of the question of valuation. I know that so many readers will quickly fill the comment space with alarmist comments of "how can you ignore risk, it is a key part of valuation". Please can you kindly settle down - I am not ignoring risk.

I have intentionally chosen to separate risk and credit analysis from the financial cash-flow analysis, particularly for DG (Dividend Growth) stocks. My reasoning is that millions of words are written on this board and elsewhere discussing the risk factors and prospects of the universe of DG stocks. I am not able to add to that more than a drop in the ocean on that front.

Instead I have chosen to address the following: Once you have done your analysis and concluded an estimate for the forward dividend growth rate of a stock, how can you financially value that compared to other similar or dissimilar assets? A key thing I have learned in investing over many years is to make sure you get enough reward. All rosy stock projections have risk, all bonds have risk, and life is risk. If you buy a handful of securities, especially if they are all stocks, some of the projections will not materialize. The ones that do materialize need to make up for the ones that do not.

In all discussions going forward I will take the DGR (Dividend Growth Rate) as a given. Look at the example used in Part 2 (link):

A) 2% yield, 15% dividend growth rate

B) 3% yield, 10% dividend growth rate

C) 4% yield bond, 4% coupon, 30 year maturity

Assume 15% DGR for A comes true, and it is exactly 15% for your lifetime. Was it better to buy the bond, or the stock? This I will address later. You may ask: How do we know the DGR is 15%? My answer: It does not matter. You can use any method you wish, whether forecast from an analyst, historical rates, projections, newsletters. I often use 10 yr. historical growth rates when looking at 50 year data, since that is all I have available as a forward projection for a point in time. Sometimes I blend it with 5 year and 2 year historical DGR rates. I might include a 2 or 5 year analyst forecast for a present day stock. In the end, you come up with a DGR number and I will calculate for you from a cash-flow perspective which is the best of the three based on current prices (yields).

An interesting note is that the 4% bond appears the easiest of the three to rate. After all, 4% is written into the contract. You know it will not be more than 4%. Of course, if the bond defaults it could be less, and then it depends on the recovery rate. If the bond defaults sooner it matters more than later since you have time to collect a lot of interest in the latter case to offset any principal loss. There are many factors to consider.

My separation of predictions and risk from valuation is analogous to bonds. One important reason for performing credit analysis on bonds is to be able to group them together to compare prices. You are essentially saying "this group of bonds has similar cash-flows, equity and business conditions" and then you can more easily compare their prices based on interest, calls and maturity. I pursue a similar strategy with bond-stock comparisons. If you do not like the number I used for a given DGR, give me another and I will re-run it.

The next question is what should we use as DGR for a stock? Should it be a conservative de-rated growth value that you believe the stock has a high likelihood to exceed? Should it be a mid-point consensus that has about 50% chance of being correct? Or should it be the best-case scenario? There is no right answer, and I can calculate for any value given. So far my thinking is you should tend towards something in the middle. Ask yourself this: If you take ten stocks with similar confidence level, as a group in twenty years they should achieve about 90% of the projected DGR. Those that overachieve can compensate for the laggards. Maybe you take all similar DGR projections and de-rate them by 10% so that you have a high likelihood, in the same region as the bond you are comparing to.

My work so far seems to indicate that historic rates should carry some weight, and that is for good reason. It appears that if a stock has a DGR of 10% over 10 years, it has a high likelihood of growing 10% in year 11. Also, companies that have a long history of dividend growth often recover even if they hit a tough patch. Even though I have used historical rates, I am not a proponent of any method. Pick the projection you like and run with it. But, project you must. My work says that over-paying can severely reduce returns, and you cannot calculate anything without a projection. At least use an historical rate if you do not have a better projection, it will tell you the relative valuation as the stock price moves day-to-day within a quarter.

In summary going forward in the next sections I will always separate risk and reward. Risk calculation will be left to others. I will focus on the relative rewards of various asset classes.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.