May 2005 Highlights

Here are some of our latest findings about retirement portfolios.

1. We can now identify three good strategies for retirement portfolios. Our baseline consists of nothing more than TIPS or a TIPS ladder. This is an amazingly effective approach that beats almost all alternatives hands down. A good alternative is a dividend-based strategy. You focus on income streams, never intending to sell shares. This protects you by isolating you from the wild price fluctuations of the stock market. The third approach is to vary your stock allocation according to how attractive stocks happen to be. We have shown that Professor Robert Shiller’s P/E10 measure of valuation (which is the price of the S&P500 index divided by the average of the previous ten years of earnings) can help us boost retirement income substantially. At this moment, we would be out of stocks entirely, possibly invested in TIPS. We would be very close to adding a small stock allocation as valuations fall.

2. We can combine these strategies. In fact, they go together naturally. If you do not put your money in the stock market as a whole, where do you put it? Perhaps, TIPS. Perhaps, higher dividend stocks. Perhaps both.

3. TIPS work well in retirement portfolios because you can draw down principal as well as receive interest. You can use the mortgage formula to calculate withdrawals (in real dollars, after adjusting for inflation). The amount that you can withdraw is a little bit more than 1 divided by the number of years plus one-half of the interest rate. If TIPS were yielding 1.6% interest, you could withdraw more than 4.13% (=3.33%+0.8%) for 30 years or 3.3% (=2.5%+0.8%) for 40 years.

4. Dividend-based strategies are especially attractive to those who plan for a long retirement. Most of the time, dividend amounts keep up with inflation, even if only in an erratic manner. You must allow for a small amount of attrition. Otherwise, dividend-based approaches last far into the future.

5. If you are planning a dividends-based strategy, make use of the Dividend Discount Model, an idealized formula. Your overall return equals the initial dividend yield plus the annual growth rate of dividends.

6. Be careful about dividend growth projections. They are not very reliable, especially after four or five years. Start with a high enough dividend yield, just in case dividend growth falls short. There are many quality companies yielding 3.0% to 3.5% and even more.

7. What kills retirement portfolios is NOT SIMPLY selling stocks when prices are down. It is selling stocks while they are down during the short-term (of three or four years). You do much better if you can tap into something to tide you over for a few years.

8. It is a good idea to have a buffer that gets you over the normal rough spots. But if stock prices are high, you are better off taking your lumps and dumping your stocks at bad prices than holding onto them for a decade.

9. [It is best not to act rashly. Use your cash buffer to time your sales to get reasonable prices. Don’t insist on a high price. Just get out of your bad position without too much damage.]

10. We have discovered that rebalancing your portfolio is a BAD idea. Rebalancing offers a small improvement on the downside at a huge penalty on the upside. For rebalancing to work, you must be unable to discern when investments are cheap and when they are expensive. Although very few people can make such distinctions in the short-term, there are many measures of valuations that work well in the intermediate-term. The best that we have found so far is Professor Robert Shiller’s P/E10.

11. In terms of numbers, varying allocations according to P/E10 historically would have allowed us to increase the amount that we could withdraw safely from 4.0% to 5.0%+ (of the portfolio’s initial value plus inflation), when compared to a fixed allocation of stocks and bonds. An alternative approach that rebalances many uncorrelated asset classes, each with returns equivalent to the S&P500, strains to do better than 4.5%.

12. Today’s stock market is still in bubble territory and outside of the historical range. It is reasonable for us to try for a withdrawal rate of 4.0% (plus inflation).

13. We have found that withdrawing a constant percentage of your portfolio’s current balance does poorly. What happens is that the amount that you withdraw can fall much lower than you wish, especially around year 20. The old rule of withdrawing 5% of a portfolio’s current balance may cause your income to dip to 2.8% of the initial balance (plus inflation). A better approach is to withdraw the larger of 4% of your portfolio’s current balance or 3.5% of your portfolio’s initial balance (plus inflation).

14. We have found that P/E10, which is calculated on the S&P500 index as a whole, also works very well for the Large Capitalization Growth, Small Capitalization Growth, Large Capitalization Value and Small Capitalization Value slices as well. You do not need a separate indicator.

15. Regarding P/E10, its inverse 100% / [P/E10] is the percentage earnings yield of the S&P500. Because earnings are averaged over ten years, it turns out to be more reliable than dividend yield for projecting market behavior. This is because dividends come out of earnings and using a decade of earnings protects us from surprise dividend cuts.

16. For those still accumulating assets, your best approach is dollar cost averaging into stocks WITHOUT rebalancing. This has always been best in the past. As you approach retirement and after you are well on your way to funding your nest egg, it is a good idea to take some money off of the table, especially at today’s valuations. Invest what you take off the table as if it were in a retirement account, temporarily with no withdrawals.

Have fun.

John Walter Russell
I wrote this on May 11, 2005.