Letters to the Editor

Comments Inspired by Reading "The Story Behind the Numbers"

I received this Letter to the Editor from Rob Bennett. He makes an excellent point.

In your article "The Story Behind the Numbers," this sentence in particular struck me: "The intermediate-term of ten to twenty years is the most important for most investors, especially for retirees."

I agree that retirees need to be particularly concerned about what happens in the first 10 years following a portfolio allocation decision. I have come to believe, however, that what happens in the first 10 years is of particular importance not only to retirees and near-retirees, but to ALL investors (in different ways, of course).

The idea that investors make allocation decisions with the thought of what is going to happen in 30 years uppermost in their minds is a myth. It is a myth without foundation in facts or experience or reason. It is a nonsense assumption. It is as much nonsense as the idea that starting-point valuations have no effect on the long-term safety of a retirement portfolio.

I don't believe that the safe withdrawal rate for an 80 percent S&P500/20 percent short-term Treasuries portfolio in January 2000 was really 1.6 percent. I understand that that is what the numbers say. But the numbers are calculated pursuant to an assumption that the investor will continue to hold as much of that stock allocation as possible while covering his living expenses. This assumption is absurd. I doubt that there has ever been an investor in the history of the U.S. stock market who has done such a thing in worst-case-returns-sequence circumstances. I reject the idea that SWRs calculated pursuant to such an assumption provide fully informed portfolio-allocation guidance.

I believe that the typical investor can reasonably pursue a 30-year buy-and-hold strategy for a PORTION of his portfolio. I am not able at this time to give numbers to indicate the size of that percentage. But I can put forward a few rough thoughts.

I believe that a typical investor needs to have the return from at least 25 percent of his portfolio to be predictable. Thus, as a general rule, the typical investor should never have more than 75 percent of his portfolio in stocks.

I believe that that 75 percent number is the number that applies in low-risk (low valuation) circumstances. In high-risk circumstances, the number drops to 25 percent. In moderate-risk circumstances, it might be 50 percent.

The concept of using a stock allocation that moves from 25 percent to 75 percent, depending on valuation levels, is old news. That's all based on what the historical stock-return data says about 30-year returns.

The item that I would like to add to the mix at some point is the significance of reasonable expectations re 10-year returns to the initial stock-allocation decision. It's not just the 30-year number that affects investing results. The 10-year number affects results too. Why? Because allocation decisions that cause poor results at the end of 10 years cause investors to lower their stock allocations at bad times for doing so.

I believe that to make our analyses truly real-world analyses, we need to incorporate consideration of the range of possible 10-year results into the analytical framework. You have gathered a good bit of data re 10-year results. The numbers are there. But I don't feel that they have been INTEGRATED into the 30-year numbers in an entirely satisfactory way.

Say that a 30-year old investor today wants to know how to invest $100,000. He wants to compare options. One thing he wants to look at is what sort of income stream he could get from the $100,000 if he invested 80 percent of it in an S&P index. He can't use an SWR analysis to tell him this because he is in the accumulation stage of the investing life cycle rather than the distribution stage. But he can't really use numbers indicating the return he will get at the end of 30 years of buy-and-hold investing because it is not reasonable to presume that he will follow a buy-and-hold strategy if he suffers big losses in portfolio value within the first 10 years.

I think it would be helpful for this investor to have a sense of what sort of income stream he could expect from an 80 percent stock allocation (and from all of the other stock allocation percentage options). I don't believe that we can give him a good number today. I don't believe that our analyses on this particular question are real-world analyses.

I think that the answer in conceptual terms is that the 30-year annualized return number MIGHT APPLY. It probably would apply if he had good enough results in the early years to get to a point where he was willing to leave at least $80,000 in stocks even if prices went down at a future date. For example, if his $100,000 grows to $200,000 in the early years, he is probably going to be able to leave the initial $80,000 in stocks regardless of whether there are big prices drops in future days or not. After experiencing that sort of growth in portfolio value, he becomes "immune" from feelings of emotional panic over losses of $40,000 or so ($40,000 represents a 50 percent loss from the original $80,000 stock investment).

But what if he experiences big losses BEFORE experiencing big gains? In those circumstances, he will likely not be able to hold onto the $80,000 invested in stocks long enough for the long-term buy-and-hold magic to happen. This is one of the big downsides of stock investing, and it needs to be taken into account in any truly informed analysis, in my view.

My sense is that the path to answering this question lies in first forming an assessment of what the "immune" amount (the amount for which the investor can reasonably be presumed to stick with a buy-and-hold strategy no matter what) is in various investing circumstances. My guess is that, just as the typical investor always needs 25 percent of his portfolio to be stable (out of high-volatile asset classes), he also feels comfortable having 25 percent invested in volatile asset classes even at times of high risk (high valuation). My further guess is that the typical investor can afford to take on higher stock allocations as the both the long-term AND short-term risk profile of stocks diminishes in the mind of that particular investor.

The point that I am making here is that it is not just long-term risk that matters. Short-term risk matters because it is short-term risk (the possibility of big price drops within 10 years) that causes investors to abandon long-term buy-and-hold strategies.

I don't get the sense that we are in disagreement on the significance of these factors. There is discussion of them running through a good number of your articles. I don't feel that we have incorporated the short-term concerns into the long-term numbers, however. It seems to me that we have not integrated the two different types of concerns into a single number.

I'm putting this forward as a way of possibly generating thoughts for research projects for the future. I obviously don't have the details worked out as to how this research should be done. I continue to think it is important that we venture down this path at some point, however.

MY RESPONSE

Thank you for a fascinating topic.

The Human Element is critically important.

There are some relevant older posts. For example, I was surprised to discover that there have been times when dollar cost averaging into stocks was a losing strategy (when compared to putting money into a savings account).

Here is a new article (and its edited version). This is my initial response.

Recently, we have developed several intermediate-term timing strategies with retirees investing in TIPS up to a decade while waiting for valuations to improve. This addresses alternatives: holding stocks and 2% TIPS during the first decade.

Risky Alternatives
Edited: Risky Alternatives
More About Risky Alternatives

Not many subjects are more relevant to investors than dollar cost averaging. It is a time-tested, successful strategy.

But what about today? Will today's investor stick with it? Maybe not.

Rob Bennett is right. You have to make it through years 10 and 15 to reach 30. There is reason for concern.

Dollar Cost Averaging Today

Relative to deposits, deficits can be HUGE. The point of frustration with dollar cost averaging (DCA) today is 10 or 15 years.

DCA Today: The Point of Frustration

Earlier Letters to the Editor (2006)

Unclemick about the Dow Jones Utilities. BillW (with Thanks!) about getting started. Rob Bennett about Dividend Theory versus Dividend Reality. This led to a breakthrough.

January 11, 2006 Letters to the Editor