Since stocks whose dividends grow consistently over time are generally presumed to be good investments, we consider an annually updated portfolio of dividend growth stocks wherein the criterion for selection is the rate of growth of dividends over the prior five years. Based on the results of back testing, a *permanent portfolio of dividend growth* *stocks* may also be constructed as we describe below.

The basket of stocks from which the selections were made at the end of each year for the following year consists of the 2013 dividend champions. Obviously our calculation suffers from survivor bias, but the results are nevertheless quite interesting. For a given year the portfolio consists of the ten stocks whose rate of growth of dividends during the prior five years was the highest.

The main performance metrics based on the returns for the period 1997-2012 are:

CAGR 14.1%

Maximum Annual Drawdown: 23%

Number of years of Losses: 3

-23%[-2008], -8.5%[2002], -0.3%[2001]

Sharpe Ratio .66

Kelly Fraction .72

2000-2012 CAGR 11.3%

The portfolio growth curve is shown in the following figure:

The following figure shows the allocation diagram for the years 1997-2013.

For the current year 2013, the portfolio consists of NC, LANC, TMP, TGT, WEYS, TR, MCY, LOW, EV and MDT, and has returned 10.6% to date (2/12/2013).

**The Permanent Portfolio of Dividend Growth Stocks**

On the basis of these results, we can choose ten stocks which appeared most frequently in the annual selections, thus forming a *permanent portfolio of dividend growth stocks*. This portfolio consists of the following: TROW, MDT, EV, MCD, AFL, CTAS, NUE, WMT, SYY, MSA.

The performance metrics of this portfolio for the period 1997-2012 are:

CAGR: 14.7%

Maximum Annual Drawdown 25%

Number of years of losses 2

-25% [2008], -10.5% [2002]

Sharpe Ratio .65

Kelly Fraction .78

This portfolio has returned 8.2% YTD.

**Disclosure: **I am long MCD.