Building a Strong Dividend Growth Portfolio

When it comes to indexing I usually have two views: it's a great tool if one doesn't have the time or inclination to analyze stocks, but it's also a wide catch-all net that brings in the tuna along with the catfish.

So, what should a time-constrained investor do if one doesn't want to invest in the entire market or a sector at once and get the chaff along with the wheat? Answer: Buy the strongest dividend payers from selected industries. And to find out who these dividend payers are one can do their own detailed analysis or read this blog.

Selected pharmaceutical index

The list below is a non-diversified list of strong pharmaceutical companies that have a history of paying dividends. Not all are cheap, but the point of indexing a sector is to get the benefit of the industry as a whole without having to guess or investigate which company has the best pipeline, the best strategy or the best cash position to go after acquisitions.

If one believes, like I do, in the future of the pharmaceutical industry -- which is home to great companies, many international, and robust dividend payers -- then buying the strongest companies is a wise strategy.

Dividend yield-based weighting

Should one buy an equal-weighted index of these companies? Well, first, if by equal-weight one means same number of shares, the answer is no. Share price is too dependent on number of shares, which is arbitrary and as such this index would be arbitrary as well. If one buys an equal number of dollars per company, that's a more suitable strategy but it would still assume that all companies are equivalent, which is not a good assumption.

A better way to index is to use a dividend yield-based indexing: buy more dollars worth of companies with higher dividend yields. This approach has two benefits: 1) one buys more of the higher-dividend companies which in turn boosts the yield of the portfolio and 2) higher yielding companies in the same industry are typically considered cheaper than peers since the yield is higher because the price is lower, so one ends up buying more of the cheaper companies.

However, this pure yield-based approach also has drawbacks: 1) high yielding companies can be cheaper for a reason, and a thorough investigation of these reasons defeats the purpose of this lazy indexing approach; and 2) high current yields could mean the companies have little growth ahead of them and as such the dividend may not keep up with inflation or growth elsewhere.

So, what's the alternative?

Dividend growth-based index

Factoring in current dividend yield as well as dividend growth should provide a much better index because current yield plus dividend growth equals total dividend returns. So, this indexing method is equivalent to saying "buy more of the companies that will return more money to you in form of dividends".

Of course, dividend growth involves making predictions about future dividends. However, armed with fundamental analysis and a long history of dividend growth one can make informed assumptions about future growth and thus minimize the risk of making a bad call.

With this, here's my dividend growth-based index of pharmaceutical companies based on 10-year historical growth rates and current yield and their respective weights.

Ticker	10-yr div. growth	Curr yield	Curr Price	P/E	Weight
NVS	13.74%	3.93%	49.55	12.07	2.36
JNJ	12.84%	3.57%	60.54	12.72	2.19
LLY	7.86%	5.57%	35.17	9.07	1.8
ABT	9.26%	3.66%	48.03	14.08	1.73
PFE	7.87%	4.87%	14.77	11.89	1.7
GSK	5.53%	5.28%	34.84	10.2	1.45
BMY	3.73%	5.00%	25.6	14.71	1.17
MRK	3.29%	4.19%	36.3	7.57	1

Putting it all together

The "weight" column in the table above indicates a dollar multiplier for each ticker. That means one should buy 2.36 times more Novartis (NVS) than Merck (MRK).

If one has a budget of $10,000 to allocate to this index, and rounding the number of shares to their nearest whole number, one should buy $1783.8 (36 shares) worth of NVS and only $762.3 (21 shares) of MRK.

The final portfolio worth approximately $10,000 in today's prices would look like this:

	Shares	Cost $
NVS	36	1783
JNJ	27	1634
LLY	38	1336
ABT	27	1296
PFE	86	1270
GSK	31	1080
BMY	34	870
MRK	21	762
	Total	$10,031

Of course, one should only follow this approach if one doesn't have time to do a thorough due-diligence fundamental analysis and after consulting with their financial advisor.

Disclosures: Long JNJ at the time of writing.