Spending Principal In Retirement Portfolios

by: Roger F. Goodrich


To sell (or not to sell) for spending purposes, is a private personal issue.

Precluding sale of retirement portfolio shares (live on income alone) works, if you have enough money to start with.

Limited sales in specific instances are appropriate (within strict guidelines) without the risk of dying poor.

Motivation for this article comes from the following:

A) SA articles: Why Selling a Few Shares is Not the Same as Getting a Dividend by David Van Knapp (the Article), found here and Does Higher Dividend Growth Lead to Higher Income by Robert Allen Schwartz (the Comments), found here.

B) Realization that with the Work Participation Rate going from 63% to 58%, millions are not only out of work, but are no longer looking. Many of these are being forced into retirement, some with limited means. What little I can do to offer guidance (that you do not need a large portfolio to retire with dignity) is a step in the right direction.

C) Putting pencil to paper helps me sort out issues in my own retirement world.

D) Trying to provide portfolio management guidance to my daughters so that they can "do" when it is their turn.

These motivators are listed in ascending order of importance (to me).

How can this (spending principal) be done? In a word (or 2); very carefully. Unrestrained spending increases the risk of running short of funds and dying poor. Since risk levels are an individual matter, to spend or not to spend is a question for each of us alone and our judgment is (or should be) of no concern to others. For the retirement community in general any (and all) discussions should be related strictly to assessment of rewards (up or down) versus level of risk. Or, to put another way: if you do this, that will (might) happen. People losing their shirts is only an issue to society if the latter has to support the former. No one has been designated "Trail Boss" and "giving" financial advice without permission is not appropriate. Exception to last sentence: Motivation D.

Those of us who have technical backgrounds like to deal with formulas. The salient one here is:

Equation 1: Total Return = Return from Principal + Return as Dividends/Interests

The adjectives "realized" and "non-realized" can be applied to this formula depending on where you are in the accumulation and distribution phases of retirement portfolios. In the accumulation phase where Total Returns are non-realized, generally dividends/interests are reinvested and "magically" become principal. Stocks/bonds can be sold and replaced. There does not appear to be any disagreement here. In the distribution phase, generally some (or all) dividends/interests are consumed as income; and this is generally accepted. The main bone of contention is where someone desires to sell shares of the portfolio, all designated as principal, using the proceeds as income. The first caution here is to realize that shares sold are gone forever and all future dividends/interests are lost. However, there may be more than enough principal to "go around," or conditions may have deteriorated to the point where action is required.

It is interesting to note that some retirees believe that the equation above has only the last term. This can work, if the are enough funds, but it is not an efficient utilization of resources. In academia, where they invent portfolio withdrawal schemes assuming dividends/interests are reinvested, the equation has only the first term. Are we surprised at the lack of communication? Then there are those of us that think the equation is just fine as it stands. There is a secret to this equation; the more terms there are, the more money you get. Seriously, this equation is vital in understanding portfolio withdrawals. If you try to pull out more Total Return than YOUR portfolio can supply, you lose. That is why the "4% + inflation" scheme does not work. Inflation is poorly linked with market prices. In times of high inflation (more demand) and lower prices (less supply), failure is inevitable. If you exercise the first term, make sure the portfolio can supply the goods. Demand =< Supply.

The whole issue of what constitutes principal is murky at best. Is it original capital invested, and/or the sum of that plus subsequent capital gains? Or that plus reinvested dividends/interests? Capital gains are what I call OPM, the narcotic of investing (Other People's Money). For long-term investors, a good part of the portfolio is OPM. Where I worked, company monies and government contract monies were not of the same mintage, or commingled.

Note that IF reinvested funds were held separately and designated as Un-realized Income, that portion of the portfolio COULD possibly be "plundered" in the future to augment income if the need arose. Not a suggestion, just a thought! Speaking of thoughts; the instant dividends are born, "principal" goes ex-dividend -- meaning dividends are a return of capital. Is there not a theory that says stock prices are the sum the present value of all future dividends? Murky, indeed!

Some Specific Examples of Spending Principal

1) I have sold shares to pay off a mortgage and to pay cash for a car. This permitted us to go through both market downturns debt free. I consider this prudent money management. Perhaps my worst sin was to sell shares for a trip to Europe. During the Dot-Com bubble, I paid around $8000 for shares in Sun Microsystems. In August 2000, the value of those shares increased by $10000. I sold enough shares ($10000) to take my wife and I on a trip to Europe. The remaining shares were sold in 2003 for $3000 after a significant drop in value. So I spent $5000 of my money and $5000 OPM. I could have sold all shares in August 2000, kept the cash for re-investment in late 2002, early 2003, but that did not happen. It would not surprise me when my daughters inherit, they will sell shares and go to Europe. OK once, but not every year. I have never sold shares for daily expenses, should I starve to death? Neither have I sold shares at market bottoms that were not immediately re-invested in equity. I am currently reinvesting some dividends in one account and am income level with RMD in the IRA, putting me in both the accumulation and distribution phases after 25+ years of retirement. It is fine to have guidelines in investing, but you have to treat each occasion by itself and make the best decision. Good Heavens (pun intended), even the Commandment "Thou Shall Not Kill" has loopholes.

2) One should always know how each holding is doing with respect to dividend growth. "How to" is beyond the scope of this article. I have written articles suggesting how this might be done. This section addresses situations where one of the holdings is not performing (low/no dividend growth) and action is required. The first most logical action would be to replace it with another in the portfolio, either in the same segment or another with higher yield, even at the expense of dividend growth. Included in this process is the idea of replacing it with an ETF. If you do not already own ETFs, the next paragraph describes a possible starting point.

The article "ETFs for Dividend Growth Investors, Re-visited," found here, features a 6 ETF group that has an 8 year (2006-13) non-l.r. dividend growth of 11.5%, average yield: 2.8%. It does show a fairly significant drop in 2009, but I have an "app" for that. A small portion of dividends are reinvested each up-market year into a commission-free intermediate treasury bond ETF. The proceeds are used to fill dividend voids in major down markets. It is my feeling that an appropriate ETF group will maintain a dividend growth pattern (commensurate with the economy of global segments covered by the group) over multiple business cycles requiring less portfolio maintenance. As individual stocks in the main portfolio deteriorate in dividend growth, those stocks can be replaced with one or more ETFs. Thus, research for new companies to invest in is not required.

If none of the above potential actions is deemed appropriate, there is always Plan C. Plan C is where the holding is sold, proceeds put in a Saving Bank, Credit Union or even buried in the back yard. The idea here is to maintain the same dividend flow using these funds until all sold principal is consumed. The table below depicts the number of years it would take to deplete the principal amount, rounded to the nearest year.

Yield >
































Obviously, caution is advised, but if you were 85 years old and did not expect to live over 100, an initial yield of 3% growing by 10%/year (risk free) for 15 years might not look so bad. Example: $10,000 start principal. First year, withdraw principal times yield ($10,000 X 0.03 = $300). In subsequent years, multiply the previous take by one plus dividend growth ($300 X 1.1 = $330; $330 X 1.1 = $363; etc. for 15 years per the table)

Another option here is to "amortize" the wayward stock. Those over 70½ with IRAs are familiar with Required Minimum Distribution [RMD]. RMD divisor for age 80 is 18.7, for 81 it is 17.9. RMD is found by dividing principal by the yearly divisor, in our example $10,000/18.7 = $535 for the first year. Taking 60% of this amount is $321. Assuming no change in share value, the next year calculation is 0.6 X $9679/17.9 = $324. This calculation is for sold principal alone, any dividends would be added to the pile per Equation 1. Also, the multiplier 0.6 can increase year-to-year up to 1.0. This procedure would provide a stable income while assuring some principal is left, RMD for age 100 is 6.3. In the example given, amounts sold each year are small and, with commissions, may not be cost effective. If this were the case, replacing the stock with a broad-based commission-free ETF might be a better choice.

If you were wondering how I insure, in applying Equation 1, that Demand =< Supply, I use (as applicable) the amortization approach described above. For the multiplier [Z] and RMD divisor [T], the variable Z/T multiplied by principal [P] is the dollar amount of principal sold each year. As long as Z/T is less than 1.0, there are remaining shares. If no principal is sold, Z = 0. If the portfolio is divided into segments, individual Z/T functions can be tailored to each segment. There may be segments where Z = 0, absolutely guaranteeing the portfolio is never totally depleted. Normally Z/T is much less than one, particularly at the beginning of the distribution time period

It would appear that after age 80 (or so), options open up for maintaining adequate income for cases where dividend growth decreases below minimum guidelines. This at the expense of part of the principal, but you cannot take it will you.

3) I have an article titled "Trade Down On Yield Without Losing Income" found here, wherein it describes a scheme for maintaining dividend flow when you sell a high yield, low dividend growth and buy a lower yield, high dividend growth stock. You take up to about 20% of the original principal to buy a no-commission equity ETF such as the Russell 1000. This ETF is sold off in pieces over a, say, 4-year period to compensate for dividend loss. After the 4-year period, the high dividend growth stock is paying the original dividend level. I started doing this about 2 years ago and sold the first batch in 2 accounts with an extra 25.4% and 27.2% capital gain, not including dividends from about a 2% yield from the ETF. Currently, unrealized capital gain for the remaining shares is 35%.

The reverse of this is also possible, sell a lower yield, spend fewer dollars buying a higher yield for the same dividend flow, pocket the difference.

4) The vogue approach to retirement portfolio withdrawals is a bucket scheme where portfolio assets are distributed into 3-4 "buckets" and consumed serially and/or monies from assets transferred from bucket to bucket. The first bucket is short-term bonds, CDs and the like. These can be consumed entirely over 2 to 3 years until the next bucket "matures." The last bucket could contain low yield-high dividend growth stocks where dividends have been re-invested during the time earlier buckets are in play. There is a large group of people that are not averse to consuming principal. Indeed, this is the mainstream approach in the retirement community. I have seen a reference (2009) where over half of financial advisors are recommending to their clients some variation of a bucket approach.

If it is desired to leave something to heirs, these funds could be separated into another account for a more appropriate investment approach commensurate with the difference in time spans. I have just finished a book, [Rational Expectations: Asset Allocation for Investing Adults by William Bernstein] where the author suggests accumulating assets (high reward-high risk) until you have sufficient funds to purchase a TIPS ladder (low risk) to be applied during the distribution phase. In this approach, again, the portfolio is consumed. Often, it is nice to hear someone else expound on the trials and tribulations of investing. Cleans out the clogged pipes. Note: the book is a good read on retirement investing in general. There is a review (pg. 63) in the 19 Jul 2014 issue of The Economist; and, of course, Amazon.

I have published an article on this subject titled "Dividend Growth Investing, the End Game", found here. It utilizes a parallel bucket approach. Share sales are managed such that remaining shares are predetermined assuring that you do not run out of funds. Shares are sold in 2 of 4 segments, one completely and the other partial. End result, over half of original shares remain at the end of a presumed 40-year period. Twenty percent of the original portfolio was consumed in the first twenty years, mostly at the beginning. This was the result even though I was willing to deplete the entire portfolio to get acceptable results. So the concept of selling shares does not have to mean wall-to-wall party time.

5) To address the case of low dividend growth, this article can provide some insight: "Total Returns from Retirement Dividend Growth Portfolios", found here. Again, shares are sold, but in a metered way. The figures below provide a way to visualize the problem.

In this case, exponential dividend growth was the full whammy, not tapered down over time. Since the portfolio depicted here is assumed to be actively managed, results can be predicted because lagging stocks are sold and replacements bought to keep dividend growth as indicated. The start point on the Total [blue] curve is 16; the initial portfolio value was 400, equally split between the two investments. The time span is 40 years. If we assume an inflation rate of 3.6%, then the Rule of 72 predicts a doubling of dividends every 20 years. Thus, to keep up with inflation, the Total payout would have to exceed 32 at midpoint and 64 at the end, which it barely does. Note that initial payout is 4%.

Dividend growth rates shown in Figure 1 are about the lowest one should consider before looking at alternatives, such as selling shares. The article mentioned above shows that total output increases if the amount of share selling is limited. Look at it this way. With large dividend growth, a dividend only payout might make sense, if you have the down payment to reach the desired income amount. If there were no dividends paid, then obviously shares would have to be sold to generate funds needed. In between (the case taken in the referenced article), some principal can be advantageously used to augment income.

In Figure 2 all that has been changed is dividend grow rates (to more realistic attainable values) plus the exponential dividend growth rate has been tapered (linearly) going from 11% at the start to 5.5% at the end, cut in half. Assuming the portfolio manager can keep this pace, one can note that there is more money available than needed at the end. If we assume the start of this 40-year period is when the retiree is 60 years old, then the quarter of time when she is in her 90s there are surplus funds which may not be needed. We see that there are enough funds (around this time) from the exponential growth segment alone that can supply the needed funds. Thus principal in the linear growth segment could be spent, either to shore up exponential segment payout if dividend growth is lagging, or to provide extra funds in case of a (say) medical emergency. This works if one has not adjusted their lifestyle to the blue curve.

Challenge to all income-only investors: Take a careful look at your portfolio. Separate those stocks that are not really pulling their weight and put them in a segment of their own. Calculate dividends coming from this segment compared with the total. This is the amount you would lose if you sold all the stocks in this segment. At 3%, for every $3 of dividends lost in share sales, you have $100 of principal to "play" with. Surely, the portfolio can be improved by some combination of buying better stocks, doubling down on higher yielding stocks that you own, freeing up enough principal to treat the gang at Starbucks and take your family out to dinner. You end up with a better portfolio with money to spare. Welcome to the club!

Looking at just the blue curve, it is enticing to ponder if there were a mechanism to pull some of those late cycle funds for use in an earlier time. This would provide a larger than 4% initial return, using principal that is not needed to provide increasing payout demanded by inflation. If the price/future dividend theory is correct, one should be able to extract from current principal those future dividends that are not needed. This is what I attempted to do in the End Game article listed in 4) above. In the example shown there, the initial withdrawal rate was 6.7%. Thus, for $40,000 desired income, less than $600,000 is needed as an initial portfolio value. One would need $1 million with a withdrawal rate of 4% to provide this amount. The approach I took has significantly less risk than trying to invest in stocks/bonds that yield 6.7% and grow as well. For me, it is an engineering challenge to deal with the variables: smaller initial principal, larger initial payout, payout increasing more than inflation, never run out of funds, minimum active portfolio management (for you and/or your replacement).

There are many options: one could separate available funds, one portion to provide income shown in Figure 1, which would provide the basic payout. This is relatively easy to manage and low risk. Remaining funds could be managed more aggressively, including dipping into principal, to provide a larger range of desirable outcomes.

6) I have not addressed here the issue of when to take capital gains and spend them because, frankly, I do not know. It is everyone for themselves. Obviously, it is tempting to eat the apple (pun intended) when it falls from the tree, or maybe before it falls. On the other hand, maybe it is more prudent to balance windfalls with disaster losses. Right now my portfolio is split, half in low yield and the other high yield. I tend to keep it that way. On the other hand, if a stock is running, I tend to leave it alone. You can not win a horse race by pulling on the bit. If you find a stock that has recently pulled away (up or down) from its 200 day moving average, to around 20%, odds are that it is in for a correction to the mean and you might take advantage of that. Each case is a judgment call based on salient factors at the time. Case in point, Time Warner (NYSE:TWX) recently popped up (31% above 200dma) on a buyout/takeover bid that was rejected. What to do?….

There is another way to look at this issue. We are all familiar with the concept of leverage when, for example, you borrow a large sum (mortgage) to buy a house and live there while you pay a relatively small sum every month to pay off the loan. We think this is a good thing. Investing in dividend growth stocks is (sort of) like the reverse; you put a large sum out there and someone pays you a small sum back periodically. We think this is a good thing. It is, but you are tying up large amounts of assets for long time periods. There is more; there is a chance for capital gains. But if you don't take advantage, it is of little benefit (disregarding passing it on to heirs). There is also a risk of losing principal, or watching it go up, then down again.

Principal out there is like a 800-pound gorilla in the room. It grows (more or less) by the same percentage as the dividends. For a 10% growth rate, dividends will double in about 7 years (exponential growth). At the same time, the principal also doubles, a much larger pot of money. Disrupting this process by selling shares will interrupt the exponential growth, but holdings in a portfolio are not all created equally. Maybe let the good ones grow and if the ones lower down have exceptional capital gains, realize them and consider spending part.

I have noticed on more than one occasion where one of my dividend growth stocks lagged a bit, but on further analysis noticed that the market liked the stock and the price had appreciated quite nicely. These may be the times to take advantage of capital gains.

I invest in dividend growth stocks because it is easy. You cannot fake dividends or their growth, so picking stocks is not hard and you are not directly competing with the big boys in the market. When elephants are mating, stay out of the corral. Dividends tend to be more stable (and predictable) than stock prices. Changes in dividend flow is easy to spot for possible corrective action. Selecting potential large capital gain stocks that do not pay dividends is a very much harder row to hoe. Even if you pick one with good value and potential, the market has to agree with you for it to pay off.

Not withstanding all of the above, I will accept monies of all colors (interest, dividends and capital gains). Put them under my bed and I cannot tell the difference. Rather than declare selling shares for income is never acceptable, the general dividend growth investing/retirement communities are better served with discussions about when and how that might be useful, but (again) that is just a thought.

Disclaimer: I am an engineer, not a financial advisor. Any (and all) discussion presented herein is for your use if it applies to your situation, as determined by you.

Disclosure: The author is long TWX. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.

About this article:

Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500. Become a contributor »
Problem with this article? Please tell us. Disagree with this article? .