In Part 1, we saw how Mr. & Mrs. Income decided to part ways with conventional advice and target their retirement income needs years before they retired. They saved diligently, placing their money into dividend-growth stocks, a general bond fund, and later into a TIPS fund.
They never made it to $1,000,000. But they did not do badly. At the end of 2010, when their retirement started, they had $950,000 saved up. Here’s a snapshot of how each element is expected to perform. The stock numbers are blended, meaning that they indicate total performance across the stock portfolio.
Investment | Principal Amount | 2011 Projected Yield % and $ | 2011 Expected Return % and $ | Expected Annual Income or Return Growth |
Stocks (70%) | $665,000 | 4.5% $29,925 | 6% | |
Bond mutual fund (5%) | $47,500 | 3.6% $1,710 | 1% | |
TIPS fund (25%) | $237,500 | 3.4% $8,075 | 2% | |
Totals | $950,000 | Total Income—see next column | 4.2% $39,710 | 5.5% (weighted) |
They did it! They would expect their first year’s income to be about $39,710. Let’s explain how they calculated these expectations:
- The income from the dividend-growth stocks is obvious: It is the projected yield x the principal amount. When they retire, the Incomes will stop reinvesting the dividends and use them as spending money.
- They will treat the bond funds as sources of growth, almost as if they were stocks. If they held bonds directly, they would just siphon off the income. But since they hold funds, that means that the annual return is comprised of income from the funds’ holdings as well as gains and losses produced by the funds’ managers’ buys and sells during the year. The Incomes plan to take the annual growth in each fund’s value as income. They will do this via redemptions. That means that over time, the base amount in each fund will stay steady (or decline if the fund loses value), so the returns over time will probably lose value to inflation, and their relative weight in the whole scheme will gradually decline.
The Incomes plan to make up for the slight deficit in Year 1 with a little belt-tightening. Actually, with dividend increases during Year 1 not yet reflected in the projected yield, they will probably pull in a little more than their target. As to Target B, they simply want their annual income to grow as fast as inflation. It appears they will do better than that, as the blended anticipated annual increase in their income from the portfolio is 5.5% versus anticipated annual inflation of 3%.
The weighted 4.5% annual yield for the dividend-growth stocks is higher than the simple average yield of 4.0% for this reason: The stocks are not equally weighted. As they kept their eye on the $40,000 target, they were influenced to step up their purchases of higher-yielding stocks. So the 4.5% is a blended yield that reflects their actual stock holdings, which are tilted or over-weighted toward stocks with higher yields.
The flip side of that over-weighting in higher yielding stocks is that they anticipate annual dividend growth will be 6% rather than the 8% simple average. Higher-yielding stocks tend to have lower dividend growth rates. Note that the 8% simple average is itself conservatively estimated as the lowest of the 1, 3, 5, and 10-year dividend growth rates for each stock. (See the stock list in Part 1.)
The Stress Test
For the stress test on the Incomes’ retirement portfolio, we’ll use the hardest stress test from the previous articles about the Growths’ portfolio and even make it a little harder. The return series is based on actual market returns from 2001-2010. In this case, we have not only stocks but also the bond and TIPS funds to consider.
For dividend returns, I have used the total dividend growth rates of the S&P 500 for the same years, which includes the only year ever (according to S&P records since 1988) in which total dividends declined (2009). As has been discussed extensively elsewhere, most serious dividend investors have no trouble beating the S&P’s dividend performance in most years, as they have hand-selected their dividend stocks and watch them like hawks. The S&P 500 is not a good proxy for a dividend-growth strategy. But for that reason—its weakness in dividends—it does make the stress test even more challenging for the Incomes’ retirement years.
With one exception, other assumptions will remain the same here as they were in the scenarios on the Growths’ retirement (refer to the earlier articles on the Growths’ retirement). The exception is that the full volatility of the market is applied. That will make the test more stressful than that applied to the Growths’ portfolio, where volatility was dampened to 2/3 of actual.
Here then is the template or return series for the Stress Test. The series will be repeated three times to total 30 years of retirement.
Year | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 |
Year# | 1, 11, 21 | 2, 12, 22 | 3, 13, 23 | 4, 14, 24 | 5, 15, 25 | 6, 16, 26 | 7, 17, 27 | 8, 18, 28 | 9, 19, 29 | 10, 20, 30 |
Actual S&P 500 Price Change | -13% | - 23% | + 26% | + 9% | + 3% | + 14% | + 4% | -38% | + 23% | + 13% |
Actual Div. Change | + 1% | + 4% | + 9% | + 13% | + 12% | + 11% | + 10% | + 0% | - 20% | + 5% |
VBMFX Actual Return | + 8% | + 17% | + 8% | + 8% | + 3% | 0% | + 12% | - 3% | + 11% | + 6% |
VIPSX Actual Return | + 8% | + 8% | + 4% | + 4% | + 2% | + 4% | + 7% | + 5% | + 6% | + 6% |
[Data sources: S&P 500 price changes: S&P 500 data. Price-only changes, not total return changes, are used, because the Incomes will be taking the dividends each year. Dividend changes: Actual S&P 500 total dividend changes each year, as determined from S&P website accessed 1/5/2011; note that this annual data has been removed from the website, but quarterly figures are still available. Bond fund changes: Actual annual returns from Morningstar for Vanguard Total Bond Market Index (VBMFX) and Vanguard Inflation-Protected Securities (VIPSX) rounded to nearest whole percent.]
In the table below, please note:
- There is no “Amount Withdrawn” column, as the Incomes never withdraw (liquidate) any of their stock assets. The heading of that column has been changed to “Amount Needed.” This is based on a $40,000 need in Year 1 incremented by 3% per year for inflation.
- “Total Income” shows the income produced by the portfolio. This is the sum of the dividend income from the stock portfolio plus the redemptions from the two bond funds. As stated above, the redemptions equal the funds’ actual returns for each year, so the base amount stays steady (and loses to inflation as a result). The three individual contributors to the total income are shown in the Appendix.
- “Cumulative Excess/Deficit” shows how much the income exceeded (or fell short of) the required amount for that year. No assumption is made that this is reinvested. The Incomes put it under their mattress as a buffer against poor years.
- The columns have been rearranged to place the income columns first. This underscores the primary importance of retirement income versus withdrawals and balances. In fact, for fun, I have left the two columns for portfolio balance blank. The income does not depend on those balances, which is a dramatically different situation from the “4% rule” withdrawal strategy. The stock market has been eliminated as a factor; we won’t even look at it. (OK, it’s presented in the Appendix.)
Stress Test Results
First Decade:
Year Number | Amount Needed $ | Total Income $ | Excess/Deficit $ | Cumulative Excess/Deficit $ Under the Mattress | Beginning Portfolio Balance $ | End Portfolio Balance $ |
1 | 40,000 | 52,025 | 12,025 | 12,025 | 950,000 | ? |
2 | 41,200 | 75,297 | 34,097 | 46,122 | ||
3 | 42,436 | 54,823 | 12,387 | 58,509 | ||
4 | 43,709 | 59,233 | 15,524 | 74,033 | ||
5 | 45,020 | 51,008 | 5,988 | 80,021 | ||
6 | 46,371 | 49,556 | 3,185 | 83,206 | ||
7 | 47,762 | 84,246 | 36,484 | 119,690 | ? | |
8 | 49,195 | 47,671 | (1,524) | 118,166 | ||
9 | 50,671 | 70,128 | 19,457 | 137,623 | ||
10 | 52,191 | 60,707 | 8,516 | 146,139 | ? |
Second Decade:
11 | 53,757 | 66,704 | 12,947 | 159,086 | ||
12 | 55,370 | 89,217 | 33,847 | 192,933 | ||
13 | 57,031 | 70,746 | 13,715 | 206,648 | ? | |
14 | 58,742 | 77,300 | 18,558 | 225,206 | ||
15 | 60,504 | 71,668 | 11,164 | 236,370 | ||
16 | 62,319 | 72,725 | 10,406 | 246,776 | ||
17 | 64,189 | 108,878 | 44,689 | 291,465 | ? | |
18 | 66,115 | 73,372 | 7,257 | 298,722 | ||
19 | 68,098 | 89,757 | 21,659 | 320,381 | ? | |
20 | 70,141 | 81,701 | 11,560 | 331,941 |
Third Decade:
21 | 72,245 | 87,774 | 15,529 | 347,470 | ||
22 | 74,413 | 110,530 | 36,117 | 383,587 | ||
23 | 76,645 | 94,705 | 18,060 | 401,647 | ? | |
24 | 78,944 | 104,445 | 25,501 | 427,148 | ||
25 | 81,312 | 102,492 | 21,180 | 448,328 | ||
26 | 83,751 | 107,160 | 23,409 | 471,737 | ? | |
27 | 86,264 | 145,926 | 59,662 | 531,399 | ||
28 | 88,852 | 111,456 | 22,604 | 554,003 | ||
29 | 91,517 | 119,322 | 27,805 | 581,808 | ||
30 | 94,264 | 113,116 | 18,852 | 600,660 | ?? |
Totals:
30 years | 1,903,028 | 2,503,688 | 600,660 | ?? |
Discussion
The income covered the Incomes’ needs in every year except Year 8, when the TIPS fund had a negative return. By Year 18, when the TIPS fund had another negative year, the dividend income had grown sufficiently to more than make up the difference. By Year 30, the excess cash from annual income surpluses had grown to more than $600,000. The total income needed by the couple in 30 years of retirement was a little over $1.9 million. Their portfolio actually delivered more than $2.5 million.
In the first two articles in this series about the 4% rule, I had no preconceived notions of how any trial would end up. I simply wanted to illustrate the dangers inherent in a withdrawal strategy for funding retirement, including the fact that even a “successful” retirement can get pretty scary at the end.
But when I began this article, I had a strong suspicion that it would end up OK. I was actually surprised by how well the strategy performed. I was really surprised by how well the bonds funds delivered in the “lost decade” for stocks. I usually pay little attention to bond funds or ETFs, so I had to research those as if exploring a new topic. When I first conceived the article, I was just going to SWAG returns of 3% or 4% per year for the bond funds. But then I decided to use actual results after discovering that the Vanguard total-bond fund was available when Mr. & Mrs. Income began their serious investing (it kicked off in 1986) and that the TIPS fund became available soon enough to be meaningful (2000).
I did not try to cook the outcome by starting with the result and working backwards. I began with what I consider to be sound investing principles, and I tried to use unfavorable assumptions that would lower the likelihood of success. Such assumptions included:
- The failure of the Incomes to save as much money as the Growths did.
- Keeping stock market volatility at full strength rther than reducing it to 2/3 of its actual. (It turns out that this was irrelevant, as the results had no dependence on stock market performance.)
- Not overloading the stock portfolio with super-high yielders. The stocks are representative of common dividend-growth stocks. Many have modest yields of 3.5% or less. I included no REITs or MLPs.
- Using the S&P 500’s rate of dividend increases. Most well-selected dividend-growth portfolios would have done better.
- Having the Incomes do nothing to increase the value of annual excess income. In real life, they would probably do something more intelligent than stuff the money under their mattress.
Here are my main takeaways from working on this article:
- Dividend growth investing is a viable strategy for funding retirement.
- If you manage your dividend-growth portfolio well, the compounding of growing dividends during retirement may simply overwhelm your income needs even if you started a little short. That is in stark contrast to the withdrawal method, where the compounding effects of inflation can overwhelm the growth portfolio’s ability to keep up.
- The reason that #2 is true is that over time, dividend increases outpace inflation. See “Do Dividend Increases Keep Up with Inflation?”
- It was striking how the $90k + withdrawals needed in the final few years seemed so daunting in the withdrawal strategy, but they were not scary with the income strategy.
- The more you can eliminate the effect of market fluctuations from your retirement funding plan, the better off you will be. There is simply less risk in the prospect of good companies continuing to raise their dividends than there is in the prospect of the market continuing to raise stock prices in an orderly fashion.
- Income-based strategies can eliminate or reduce the worry that comes from depending year in and year out on market returns rather than on the relatively more dependable delivery of dividends. The perils of total-return sequencing disappear, along with the steep plummet towards zero that can occur even with a withdrawal strategy that “succeeds.” Since dividend income is not a function of what the market is doing, the luck-based coincidence of your retirement age with a good, flat, or lousy market is close to irrelevant.
- Bonds may have more value than I gave them credit for, although the declining interest rate environment in which bond funds thrive cannot go on forever. 2001-2010 was actually a good period for the total return on bonds, but it is unlikely to be repeated as interest rates rise, which they eventually will.
Appendix
Sharp-eyed readers will note that the decade I used in this Stress Test is shifted by one year from the decade used in the Growth articles. This was not a deliberate attempt to tilt the results. I simply remembered incorrectly which years I had used in the earlier articles. Correcting this would have required a great deal of work, and I don’t think the results would have been significantly changed.
In the earlier articles, I used total returns (including dividends) to give the Growths the maximum opportunity to succeed. In this article, since the Incomes are drawing off the dividends, the correct figures to use for market returns are price-only. Note that price-only returns are always worse than returns that include dividends, since there is no such thing as a negative dividend.
The following table shows how the “Total Income” column in the Stress Test was calculated from the contributions of dividend income and bond-fund returns. It also shows the missing results for the two blank “Balance” columns. After Years 8, 18, and 28, when the TIPS fund had a negative return, the remaining years take into account the reduced size of the TIPS holdings. The balances include the excess cash under the mattress. The balances for the total bond and TIPS funds are held constant (except for the reductions noted for the TIPS fund), because the Incomes drew off the funds’ returns each year, leaving only the original seed money invested. Over long periods of time, that reduces the relative contribution of the bond funds to total income.
Year | Income from Dividends $ | Income from Total Bond Fund $ | Income from TIPS Fund $ | Total Income | Beginning Balance $ | Ending Balance $ |
1 | 29,225 | 3,800 | 19,000 | 52,025 | 950,000 | 875,575 |
2 | 31,122 | 3,800 | 40,375 | 77,297 | 875,575 | 764,581 |
3 | 33,923 | 1,900 | 19,000 | 54,823 | 764,581 | 858,696 |
4 | 38,333 | 1,900 | 19,000 | 59,233 | 858,696 | 912,351 |
5 | 42,933 | 950 | 7,125 | 51,008 | 912,351 | 921,170 |
6 | 47,656 | 1,900 | 0 | 49,556 | 921,170 | 1,006,592 |
7 | 52,421 | 3,325 | 28,500 | 84,246 | 1,006,592 | 1,068,628 |
8 | 52,421 | 2,375 | (7,125) | 47,671 | 1,068,628 | 746,705 |
9 | 41,937 | 2,850 | 25,341 | 70,128 | 746,705 | 867,104 |
10 | 43,034 | 2,850 | 13,823 | 60,707 | 867,104 | 930,233 |
11 | 44,474 | 3,800 | 18,430 | 66,704 | 930,233 | 850,965 |
12 | 46,253 | 3,800 | 39,164 | 88,217 | 850,965 | 743,032 |
13 | 50,416 | 1,900 | 18,430 | 70,746 | 743,032 | 835,040 |
14 | 56,970 | 1,900 | 18,430 | 77,300 | 835,040 | 888,794 |
15 | 63,807 | 950 | 6,911 | 71,668 | 888,794 | 899,171 |
16 | 70,825 | 1,900 | 0 | 72,725 | 899,171 | 983,831 |
17 | 77,908 | 3,325 | 27,645 | 108,878 | 983,831 | 1,045,936 |
18 | 77,908 | 2,375 | (6,911) | 73,372 | 1,045,936 | 726,712 |
19 | 62,326 | 2,850 | 24,581 | 89,757 | 726,712 | 844,177 |
20 | 65,443 | 2,850 | 13,408 | 81,701 | 844,177 | 905,881 |
21 | 66,097 | 3,800 | 17,877 | 87,774 | 905,881 | 828,814 |
22 | 68,741 | 3,800 | 37,989 | 110,530 | 828,814 | 724,668 |
23 | 74,928 | 1,900 | 17,877 | 94,705 | 724,668 | 815,184 |
24 | 84,668 | 1,900 | 17,877 | 104,445 | 815,184 | 869,979 |
25 | 94,838 | 950 | 6,704 | 102,492 | 869,979 | 882,863 |
26 | 105,260 | 1,900 | 0 | 107,160 | 882,863 | 967,793 |
27 | 115,785 | 3,325 | 26,816 | 145,926 | 967,793 | 1,030,983 |
28 | 115,785 | 2,375 | (6,704) | 111,456 | 1,030,983 | 721,085 |
29 | 92,628 | 2,850 | 23,844 | 119,322 | 721,085 | 826,157 |
30 | 97,260 | 2,850 | 13,016 | 113,116 | 826,157 | 886,636 |
TOTAL | 1,946,325 | 76,950 | 480,413 | 2,503,688 |
[Source of data: Income from dividends was calculated using the 4.5% beginning yield, then applying the actual dividend change of the S&P 500 and compounding that (including the 20% drops in Years 8, 18, and 28) for each subsequent year. Bond fund returns are as shown by Morningstar for each year. Total income is the sum of those three sources. The total portfolio balance starts at $950,000, made up of $665,000 in stocks + $47,500 in total bond-market fund + $237,500 in TIPS fund. The balance in stocks is changed each year by the percentage change in the S&P 500. The balance in the total-bond fund stays constant, as the return was redeemed each year. The balance in the TIPS fund stays constant, except in Years 8, 18, and 28, when it had a negative 3% return. That took the amount in the TIPS fund down from $237,500 in Years 1-7, to $230,375 in Years 8-17, to $223,464 in Years 18-27, and to $216,760 in Years 28-30. The total year-end balance is the sum of the stock and fund balances at the end of each year + the cumulative excess under the mattress.]
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
