A recent article on dividend-growth investing myths by David Van Knapp brought up the ever-controversial topic of yield-on-cost (YOC), as evidenced by the comment thread. Though I'm somewhat reluctant to rehash this topic yet again, after some reflection on the metrics I use to evaluate my portfolio performance, I see similar issues for total-return investors, and therefore hope to present some new thoughts into this discussion.
To be transparent, let me say upfront that I have been one of the skeptics of YOC, however, I have also been open to ideas expressed by other SA contributors regarding their uses for the metric. My objections usually occur when I hear someone state something like "why would I sell my stock when it has a YOC of <insert large number>%??" Also, I believe that ANY metric that is cumulative or an average tends to hide information that could be valuable for decision making, therefore as Geordy Wang's recent article stated, YOC should not influence the ultimate buy/sell decision, but I believe it can signal a potential problem. I tend to worry more that these types of metrics will cover up poor performance. As I reflected on growth-oriented metrics, I observed similar issues with metrics such as cumulative total return and compound annual growth rate (CAGR); the same issues apply to both styles of investing.
Appropriate Uses of YOC
Let's start with two ideas presented by fellow SA contributors that I believe are appropriate uses of YOC. In a previous article's comment thread, David Crosetti mentioned using YOC to determine an entry point for a stock. For example, suppose you want to buy Intel (INTC) but you require a 3.5% minimum yield. Therefore, at the current 84 cent dividend, you would need to buy the stock at $24/share, not the current $27.75. This is the equivalent of setting a limit price for growth investors. I wonder if perhaps this should be called "Entry Yield-on-Cost" or just "Entry Yield," to avoid confusion with the more traditional use of YOC, which tracks yield after the purchase has been made. Regardless, using this tool to determine one's entry price is certainly valid and important to dividend investors.
The other use of YOC that I agree with is using it to set a target yield goal. David Van Knapp has written articles discussing his "10x10" strategy, which seeks to identify stocks and dividend growth rates (DGRs) that can yield 10% on the initial cost within 10 years. While this requires estimation of future DGRs, it is really no different than a growth investor setting a future price target using either earning growth projections and P/E ratios, or a discounted cash flow analysis. The investor can then track the stock's performance against this benchmark to see if it is meeting expectations. If it isn't, this could lead to a sell decision, but additional information should still be considered. From a naming standpoint, I think of this more as "Target YOC," with the goal of my actual YOC reaching it in the desired time frame.
Challenges of YOC and Total Return Metrics
Much of the debate over YOC focuses on how useful the metric is, as it is based on the original cost, not the current value of the stock. As the dividend increases, YOC keeps going up, which looks great, but it ignores the current price (and yield) and the time needed to get to that yield on cost. From a decision-making standpoint, this information is very important. A high YOC looks good, but if it took 15 years to get there, the dividend gain may not be as impressive as it appears. If the current yield is low, the investor may earn more income by switching to a higher yielding stock, regardless of YOC. It is important to not misjudge a stock's performance or become "married" to a stock because of this metric or others like it.
- Suppose I purchased McDonald's (MCD) on March 15, 2002, at $27.75 with an initial yield of 3.4% (94 cents/share). At today's $2.80/sh dividend, my YOC is 10.1%, but the current yield is 2.9%, which is lower than the original starting yield. While the high YOC confirms our target YOC and our excellent stock selection, the decision to keep or sell this stock should not be based on the YOC, a backward-looking metric. An income investor might be concerned about the 2.9% current yield and seek a suitable replacement stock with a higher yield. Another investor may consider MCD a core holding and be perfectly happy to hold on to it, provided that other metrics (PE, growth projections, etc) are acceptable. Calculating a new Target YOC might help the decision-making process.
- Suppose I purchased Leggett & Platt (LEG) on March 1, 1996 at $11.435 (split-adjusted) with an initial yield of 1.9% (22 cents/share). Today, at a $1.12/share, my YOC is 9.8%, but it took 16 years to get to this level, whereas MCD reached the same YOC in 10 years. This highlights the lack of a time component in YOC, and a reason to consider other metrics such as the DGR, which would be lower for LEG.
The same issue occurs for total return investors who simply look at their cumulative capital gains or total return on a stock, without consideration of the time variable. For example, I currently have a 195% gain on my Apple (AAPL) shares. Even if the stock's price stayed frozen at $600 for the next three years (or heaven forbid, it declines!), I bet that this cumulative gain would still be the highest in my portfolio. Looking only at this metric could lead me to keep AAPL, as it is my best performer, despite no gain in three years. Without noting the time component, I fall into the same potential trap as with YOC.
Adding in the Time Component
To address the time component issue, I prefer to look at the CAGR for total return and the dividend growth rate (DGR). These numbers provide average growth rates to help investors monitor the growth of their income or total return. However, averages pose problems as well, since they smooth out data that may not be so smooth.
- Suppose I purchased CenturyLink (CTL) in March 2004 at $27.49/share with a 23.2 cent/share dividend. The initial yield was 0.84%. Today my YOC is 10.5% and the current yield is 7.4%. The compound DGR over these 8 years was 37.1%. Overall, this looks pretty impressive. However, what is hidden is that CTL had a significant dividend policy change in November 2008, raising its dividend from 6.8 cents to 70 cents per quarter. Since then, the dividend has barely changed.
Again, the same problem occurs when looking at CAGR, as a stock may have a large gain one year, and then no gain or even declines the following years. If I bought Merck (MRK) way back in March 1992 (adjusted price of $12.45), my cumulative return over 20 years would be 205.7% (current price = $38.06). My CAGR would be 5.75%. What this doesn't show is that the stock went ballistic between 1992 and 2001, and then has generally declined for the last decade. Somewhere around 2004-5 the cumulative CAGR would have dropped below 8%, which would have been a warning though.
Looking at Different Time Periods
These examples illustrate the need to examine both long-term and short-term rates to observe trends, which may support a sell decision. Personally, I find this data more relevant than my cumulative returns. Looking at the one-year or three-year CAGRs for Merck would have signaled potential concerns back in 2001-2002, allowing me to perhaps decide to exit the stock and lock in my gains. Likewise, for dividend growth investors, the one-, three-, and five-year DGR values combined with the current yield, provide more decision-making information than YOC or cumulative DGR. Fortunately, if your stocks are on the CCC list, this information is already computed for you each month (thank you, David Fish!). Using this data, an income investor may decide to sell a stock due to a slowing DGR, low earnings growth projections, a higher current yield opportunity, or because a new Target YOC based on the current data is not as promising as another prospect. Then again, perhaps the investor is comfortable with the current yield and views the stock as a core holding, so s/he keeps it. The point for all investors is to utilize more detailed data instead of the metrics that may conceal valuable information.
I know the YOC topic has been beaten to death, but I hope that this article adds some new perspective to the discussion, particularly by showing the parallel issues that total return investors face when using comparable metrics. Like some other authors, I've reached a point where I don't care as much whether people find YOC useful or not - to each his/her own - however, I do feel the need to say something when it appears someone is not considering the full picture when he does use it or another metric. Once a teacher, always a teacher! But now, I can direct the investor to this article and a few others first! It's not so much whether metrics such as CAGR, DGR, or YOC are useful; it's more how they can be used and recognizing their respective limitations.