Portfolio Boot Camp

by: Roger F. Goodrich

'Oberportfoliomeister' Boot Camp

How would that look on your resume? It takes a little discipline, thus the boot camp, but this discipline is self-imposed - which is better than the other kind. The main item needed is a plan, French for map - as in getting from point A to point B. The plan should have 3 main sections: 1) intent, purpose, goals, major decisions; 2) implementation of #1; 3) verification of #2.

1) A clear, concise statement of intent and purpose is vital. Specify key goals that you can track. Retirement portfolios have 2 phases, accumulation and distribution; they are not independent. The latter phase is income oriented. Calculate the portfolio portion of total retirement income. Determine when to a) retire, b) take Social Security, 3) take withdrawals from IRAs (initial and RMDs). Decide who will manage the portfolio throughout its life cycle.

2) Implementation considerations should include: portfolio segments, portfolio allocation, #stocks/bonds, when/what to buy and sell, dividend re-investing (or not), re-balancing (if & when), multiple accounts (taxable, tax deferred).

3) Verification and portfolio maintenance items are: visual aids, on-line portfolios, web-sites for info, archival data, performance tracking, goal achievements/renewal.

I would urge expansion of this outline until it covers every part of portfolio endeavors. Every time something crops up that is not covered by the plan/rules, add it. What specifically to have in the plan and rules will come from deep thought on your part with help from a world of on-line help. Some of the latter may apply to your situation and some will not. It is important to recognize the two, hence the deep thought. Don't sign up for something that doesn't apply in your situation.

Perhaps the most important advantage to a rule-based plan is that it takes emotions out of the picture. If you have a coherent set of rules, then appropriate action is obvious. That is what the big-boys do, only they program the rules into a computer, then go play golf. If your rules don't work, change them until they do.

At the start, your plan will be skimpy and probably inaccurate. No mind, change it as you go along. My first plan was in 1970 (give or take). I was about 37 and starting to think about retiring, which I wanted to do at age 55. I laid down a plan to accumulate $100K, calling for savings every year assuming a 10% return along with a guess of my future salary, all this calculated on quadrille paper. Did it work out? The assumed return was too high, the dollar goal too low, but my pay raise estimates were also too low. I didn't meet the retirement age goal. My company 'didn't write checks on Monday', so I was a day over 55 at retirement. It happened because I made it happen.

The normalized graph below depicts my net worth after retirement. The S&P500 went from roughly 300 to 1400, a 4 2/3s increase, in the same time period. So I haven't kept up with the Joneses with a times 3 gain, although I have recently reached new highs and the S&P has a ways to go. Some principal has been cashed in, but I have tried to live within the income. Most of the heavy lifting was done by the Market, but at the lows in 2003 and 2009 I made a concerted effort to sell stuff that I felt would not recover properly and buy things that would, even taking heavy losses.

In the graph, the linear regression line was calculated for the first part, then extended. This could be a goal achievement chart where the goal is at some point in the future, the guide line (either linear or exponential) setting the pace, and actual performance plotted each year. This would be a big help in showing how you are doing. Even though I didn't start out with a goal, per se, for net worth, I now strive to keep up with (and surpass) the linear regression line. The last data point is 2012 year-to-date (August).

The general system described here is not one that subscribes to the thought "wouldn't it be nice to…". It is based on logical progressive steps. It is My Plan, not that you have to implement it into yours, but should give you something to think about. My portfolio is divided into segments. Segment yield at retirement and dividend growth are developed from retirement income needs as derived in my previous articles entitled "Life Cycle of Retirement Portfolios", Part 1,2,3. These references show relationships between these parameters and initial portfolio values and payout percentages. FOM formulas, (see here) were derived based on analyses of the loci of yield and dividend combinations to achieve a given return at a future time (ten years). Measurement of dividend growth, (see previous articles "Measuring Dividend Growth, Part 1, 2, & 3), stemmed from determining best fit to theoretical exponential and linear curves. Determining portfolio price performance (Properly Benchmarking your Portfolio Performance) benchmarks against the S&P500. Dividend affordability and sustainability, (Dividend Affordability and Sustainability), were calculated from pertinent company data, redefining metrics where necessary.

What this means is that dividend growth becomes the most important metric in selecting and holding stocks. The five sub-sets of this are: 1) yield (to buy the dividend at a low price), 2) dividend growth robustness, 3) dividend growth affordability, 4) dividend growth sustainability, 5) company products and business model. These are not in any order, all are important and any stock that passes is eligible for purchase. There could be other fundamental metrics that may assist in this process, but their magnitudes are a function of the industry they are in and relationships to dividend growth unknown.

Those of us interested in DGI may have to develop our own DG metrics, the Establishment is not interested. They use a concept known as Total Return. What this means in the real world is that they consider all dividends and bond interest to be re-invested. Stocks and stock segments are characterized for analyses by a total return mean with a deviation assessment. Dividends/interest are out of sight. There may be a time when they come to their senses, but that time is not now. We are on our own - this may be a good thing!

When all is boiled down to the essence, there are only two metrics that count in the distribution phase of a retirement portfolio; capital gains (or sold principal) and dividends (with commensurate growth). While equities can be selected that potentially result in capital gains, that outcome is pretty much left up to the market. We can, however, have more influence picking stocks based on yield and dividend growth. This can be accomplished three ways: 1) do the best you can in selecting stocks that demonstrate these characteristics, 2) replace stocks that under-achieve with better ones (portfolio turnover); 3) sell the laggards (according to the plan) earlier in the distribution program.

The last choice is the easiest. The first selection is vital and much effort should be expended to achieve this goal. The second item may be harder to accomplish for three reasons: a) you may have to buy a higher yielding stock to maintain dividends because the price of the lagging stock has dropped, b) the portfoliomeister is too old to be effective, c) heirs to the portfolio may not have the management expertise or desire to learn.

The following comments expand on the outline above:

1) A retirement portfolio has 2 phases, an accumulation phase and a distribution phase. At the boundary, the time of retirement, there should ideally be a seamless transition. It doesn't make sense to accumulate a bunch of stocks without purpose, then (at retirement) suddenly realize they are not appropriate to get through the retirement years. Basically, in retirement most (or all) income should be derived from stock dividends/bond interest. The traditional 60/40 split was there for a reason. Also, principal should be sold lightly; equity sold cannot pay dividends/interest. There are some that advocate not selling any principal, living solely on dividends/interest. With a portfolio with robust dividend growth (to keep up with inflation), this may make sense. The downside is that the size of the portfolio with a given yield may not provide enough income. In my case it is, but ultimately my portfolio will be split in two and it will not be enough. For this reason and the fact that my two daughters do not have heirs and could (safely) deplete the portfolio through their lifetime, I developed a retirement portfolio distribution scheme since there was nothing suitable out there. This scheme allows selection of principal sales, from none to full depletion of the portfolio within a specified time period. It can be tailored for dividend growth and permits different distribution rates between portfolio segments. There is no risk of running out of money. I have reported on it in previously written articles.

What this portfolio distribution scheme shows is that if segment dividend growth is sufficiently high, little to no principal selling is required. If dividends are allowed to increase without restraint, total income is almost as much as it would be with moderate taking of capital gains. This (not selling) in turn keeps the portfolio safe from severe market declines which, if prolonged, can severely reduce portfolio net value. These interactions are discussed in "Life Cycle of Retirement Portfolios", Part 3.

I have split my portfolio into 5 segments. Two of those segments divide dividend growth stocks separating low yield, high dividend exponential growth (30%) with higher yield, moderate linear dividend growth stocks (30%). This makes sense since linear dividend growth stocks do not compound as inflation does and therefore cannot keep ahead. For this reason, they are 'sold off' at a faster pace. Exponential dividend growth stocks (non ETF) are not sold during the first half of the time period, allowing dividends to grow without limitation. The other 3 segments are: ETFs (15%), Bonds (10%), Melange, a mix (15%). More on my portfolio can be found at "Life Cycle of Retirement Portfolios", Part 2.

By analyzing retirement needs as summarized in the above references, salient portfolio parameters (initial portfolio amounts, yield at retirement, dividend growth for each segment) can be determined, which then become goals to be achieved during the accumulation phase. While individual stock dividend growth is important in the choosing and retaining process, it is segment dividend growth that becomes the metric that is monitored. This is accomplished by a spreadsheet which multiplies yearly values of dividends/share by the number of shares held. Seven years of data are used where available with a minimum of five required. Performance is gauged vis-à-vis goals established by the retirement portfolio distribution analyses.

2) An initial look at dividend growth curves generated by the application of linear and non-linear regression techniques described in Measuring Dividend Growth, Part 1, 2, & 3 shows stocks could be separated into two populations, one adhering closely to the 'best-fit' curve and the other, not so well. There are enough of the first category in my portfolio that could warrant their inclusion in a separate segment. Stocks in this new segment would never be sold, the assumption being that the company Boards of Directors appear to be doing all possible to maintain dividend growth at these rates. Stocks in the 'other' category appear to have dividend growth affected by the financial crisis and subsequent slow recovery. The list of stocks in my Dividend Growth segment that demonstrate this 'best-fitness' are: Procter & Gamble (NYSE:PG), Time Warner (NYSE:TWX), Becton, Dickson & Co. (NYSE:BDX), Darden Restaurants (NYSE:DRI), Kimberly-Clark (NYSE:KMB), Owens & Minor (NYSE:OMI), Target (NYSE:TGT), Colgate-Palmolive (NYSE:CL), Abbott Labs (NYSE:ABT), Accenture (NYSE:ACN), General Mills (NYSE:GIS). Corresponding equities in my High Income segment are: Sunoco Logistics Partners (NYSE:SXL), Enterprise Products (NYSE:EPD), Buckeye Partners (NYSE:BPL), NextEra Energy (NYSE:NEE), Omega Healthcare Investors (NYSE:OHI), Rogers Communication (NYSE:RCI), CMS Energy (NYSE:CMS), Western Gas Partners (NYSE:WES). A more scientific approach to separate the stocks would be to use the 'r' (correlation coefficient) and/or 'rss' (residual sum of squares), which is automatically calculated as part of the regression process.

3) Re-balancing portfolios is traditionally meant to be a buy low - sell high scheme. A portfolio of stocks and bonds would become unbalanced at market peaks and bottoms, so re-balancing to a given allocation forces one to take profits. With current levels of bond interest, many of us are reducing bond allocations. A re-balancing concept may still make sense if you consider doing it with low and high beta stocks. It does not apply in my retirement portfolio distribution scheme because the various segments are depleted (or not) at different rates and re-balancing would upset that process. I do re-balance in the accumulation phase to keep segment allocations in line. This is accomplished by buying and selling equities or by transferring them between segments. I use my Melange segment as a repository of equities that do not perform to set standards in their original segments, but are not 'bad' enough to sell.

4) Buying stocks: I get candidates of stocks to buy from articles in SA and other internet sites. Generally, the first thing is to calculate Figure of Merit (FOM) as derived in article entitled "Dividend Growth As A Figure Of Merit". There are 2 FOMs, the product of Yield (NYSE:Y) and Dividend Growth (NYSE:DG) for stocks with exponential growth and (DG + 10) * Y for stocks with linear growth. The latter is useful because it can be used for stocks with no dividend growth, even bonds. Dividend growth is determined by the linear and non-linear regression techniques described in "Measuring Dividend Growth", Part 1, 2 & 3. If I have several candidates, I prioritize them by FOM, and use the Affordability Ratio described in the article entitled "Dividend Affordability and Sustainability". In addition, I examine Financial Statements as outlined in that reference. Included in this analysis is a clear understanding of what the company does and how it does it with attention to future company longevity, no buggy-whip companies here. I do not try to rank order stocks, other than FOM. They either qualify for admission to the portfolio or they do not. Other than a distinction based on their dollar weight in the portfolio, they contribute equally. Specific stock selection, from the list of eligible candidates, is based on portfolio segment balance or other like considerations.

Most of my stock buying is tied with a concurrent sale (to raise the money). This is done to upgrade the portfolio segment, usually to improve dividend growth, but sometimes to increase dividend flow. Since I am buying and selling at the same time, price is relative and not a prime factor since generally 'all boats rise and fall with the tide'. I find it very difficult to sell one stock at the 'right' time, then hold funds to buy another stock when it has reached its proper value. To quote an old saying, I am in it for the 'long run'. Usually, I sell from the Mélange segment, transfer another stock in, then buy into the segment where the transferred stock came from. Quite often, I will initially buy a stock with a half position, purchase the rest later when it is depressed, a double down. If you are in the process of accumulating stocks over a long period, then picking a best-value stock from a list of acceptable candidates makes sense. There are more good stocks out there than we have money to spend.

5) This leads the discussion to the question of how many equities should be in a portfolio. IMO, the answer hinges on the amount you are willing to lose if the tide turns against you in one stock, a total loss. I think each of us has a feeling for that number. Mine is roughly my yearly salary as a freshly minted engineer in 1958. With this number in hand, the number of equities in a portfolio will grow with the size of the portfolio. I have >100 stocks in my portfolio, but with 5 segments, the number in each (which is a mini-portfolio) is about right to achieve diversity. Another point, the plan in the distribution phase is to sell a whole position of the poorest stock, thus improving the portfolio. Having each position about 1% is a ballpark amount to do this.

6) A gimmick I sometimes use when I have a high yield equity that needs to be sold is to pair it with another with low/no yield, then buy a stock with intermediate yield. One rule I have is to avoid reducing the portfolio dividend flow when you are buying/selling. Another rule is to improve segment dividend growth while in these maneuvers. As I stated previously, I have yield and dividend growth goals for each segment, so all portfolio transaction activity should morph in that direction.

Speaking of gimmicks, I have been known to pair the selling of an equity that has a capital loss with one that has a gain, thus avoiding the situation where I 'lost money'. I am getting better at being mature and taking losses in stride. My overall batting average is good but you can't 'win them all'. One needs to recognize situations where you have a loser and it may never get back to where you want it. Whining is not permitted at Boot Camp. Better to sell, take the loss and replace it with something that has a better chance. I have done just this at both the 2003 and 2009 market bottoms with excellent results.

7) I do not have a fast and hard rule about selling equities that have cut dividends/interest. Generally, I will transfer them to my Mélange segment. It all depends on the reason for the dividend decrease. If it was due to a general market condition or the economy, and the stock has a potential future, I keep it. If the stock dividend was suspended, I would probably sell, blaming management for not paying attention. I would put the large bank stocks (and General Electric (NYSE:GE), for example) in this category during the recent financial crisis.

8) Portfolio maintenance is an area that needs attention. For portfolios heavily addicted to dividend growth stocks and/or retirement portfolios where income is important, then keeping track of dividends is vital. One web-site with copious dividend data is dividendinvestor.com. Many of its features are free, including a section that not only has 4-5 years of dividend data but declaration, ex-dividend, recording and payment dates as well. A good way to keep up-to-date on dividends goings-on is to make up a calendar with declaration dates for all equities you are interested in tracking. Use past declaration dates as a clue. Looking at the calendar will remind you to check just after the declaration date for the latest info. You will then be current on changes, good and bad, in a timely manner.

A yearly (at least) scrubbing of all equities in the portfolio is vital. Starting in the early fall, I start collecting EOY dividend/share data. By January of the following year, I update these data to recalculate yield/dividend growth metrics for each segment. When end of fiscal year Financial Statements are updated, I collect data for the Affordability Ratio and review these documents for changes in company activity.

9) On-line Portfolio Hosting: Both Morningstar and Finance-Yahoo (among others) have provisions for inputting your portfolio. I use this feature to keep tabs on my portfolio segments, using it to track segment value, performance relative to the S&P500, individual stock performance (including % change from 200 day moving average). This last metric provides a quick determination of whether a stock is over or under valued. If you input cost info, you can monitor gains/losses. A click on the stock symbol gets you to the quote page which has a news section where you can ascertain if anyone at SA is talking about your stuff (among other news). I also use the web-site for watch list portfolios of potential buys.

10) Data Archival: I am a great believer in recording vital data for later analyses and future projection to discover trends. These data include: portfolio value, dividend amounts at both stock and segment levels, company metrics. Monthly account statements have projected dividend amounts, both at the stock and at the portfolio level.

I hereby declare this article as part of My Plan. Now, it's your turn.

Disclosure: I am long PG. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: I am long all stocks mentioned except GE and CMS.