Mutual Fund Research Newsletter
Copyright 2011 Tom Madell, PhD, Publisher
Apr 2011

Don't Allocate Too Much (or Too Little) to Stocks

by Tom Madell

As long-time readers of this Newsletter are aware, at the beginning of each quarter we present our updated Model Portfolios featuring not only which specific fund categories we recommend, but also our recommended overall percentage allocations to stocks and bonds. These percentages normally don't change drastically, but rather in small increments to reflect quarterly changes in how attractive we judge stock and bond (and cash) investments appear to be based mainly on our reading of the economy (ie. prospects for interest rates, growth, corporate profits, etc.).

The assumption is that such perceived changes in outlook should, sooner or later, be reflected in how well stocks do, as a whole; ditto for bonds. Given the apparent high degree of uncertainty of estimating the economy's "state of health," as well as attempting to ascertain if and when such measures will actually effect stock and bond market performance, it would not be surprising if even many of our most dedicated readers tended not to pay a great deal of heed to our modestly changing quarterly overall stock and bond allocations.

But, as this article will show, ignoring the level of and the degree of allocation changes made can turn out to be a huge mistake.

Here's the evidence. We decided to examine whether these recommended percentages are indeed good predictors of how well the stock and bond markets do over subsequent periods. While we were confident that our recommended allocations would turn out to be reflected in the kinds of returns investors could expect in the future, even we were surprised how important these changes proved to be for readers who followed the general thrust of our recommendations to make these periodic changes.

Our findings, below, are highly supportive of using this Newsletter to help make occasional changes to one's weightings of stocks vs. bonds in a mutual fund portfolio. Whether those changes are made quarterly, every 6 mos., or only once a year, if the future is anything like the last 5+ years featuring a high degree of volatility in investment performance from year to year, investors who use this type of modified rebalancing will most likely come out considerably better off than those who merely hold on to their stock and bond funds regardless of the kinds of changing conditions that can drastically effect their funds' potential future performance.


One way to examine if our allocation changes were followed by commensurate changes in stock performance was to see if each quarter's allocation to stocks (and bonds, covered below) show a positive relationship to an appropriate index over the next 6 to 12 mos. If our recommended changes turned out to be helpful, you would expect that the lower our percent allocation to stocks, the lower an index of their subsequent level of performance should tend to turn out to be. Likewise, the higher our relative allocation to stocks, the higher the index should subsequently tend to turn out to be. (Note that we examined our allocations for "moderate risk" investors, not our aggressive or conservative risk allocations.)

That turned out to be exactly what the results of our allocations as compared to the index of subsequent stock performance showed. After 6 mos., the correlation between our percent allocated to stocks and the actual level of the S&P 500 index was +.55. (Correlation is a statistic showing the degree of relationship between sets of two numeric measurements.)

This tells you that whenever an investor mirrored our allocation by raising their own allocation, they tended to find that over the next 6 mos., stocks tended toward relatively higher levels. On the other hand, if they lowered their allocations exactly as we recommended, stocks tended to go to lower levels over 6 mos. time. The degree of relationship was moderately high considering that our allocation recommendations were merely judgments as to potential stock strength, and not some firm quantitative data, although most of these judgements were anchored by objective economic and fund performance data. The relationship between our allocations and actual S&P 500 price levels a full year later was also positive, although somewhat less strong at +.44.

Let's look at this data comparing our allocations vs. subsequent stock market levels in a slightly different and even more compelling way:

Many "moderate risk" investors and advisors prefer having about 60% of their portfolio in stocks at all times. Whenever the figure drops below or goes above 60%, they rebalance back to 60%. However, during the period beginning July 2005 thru Sept. 2010, this Newsletter believed that stocks, on average, were more prone to underperformance than might ordinarily be expected. Therefore, our average allocation to stocks was only about 52%, as shown in Table 1, to reflect that bonds, and perhaps even cash, might temporarily at least, be a better place to be invested. Note that we were not solely considering return potential, but also to mindful of reducing the risk of unacceptable losses during a possible market crash, which indeed did happen.

Table 1. Our Allocation to Stocks Since July 1, 2005

Average Over Entire Period = 52%
Number of Quarters Above 52% = 14 ("above average")
Number of Quarters Below 52% =   8 ("below average")

As we now know, stocks have not performed well on average over the above period, with the S&P 500 Index only averaging about 2.5 percent annualized on a buy and hold basis. However, due to the volatility, for given stretches during this period the stock market did quite well in spite of the overall poor performance.

Normally speaking, we would agree that a 60% overall allocation to stocks for moderate risk investors is an appropriate figure. But because we did not view the excesses of the last five to six years as conducive to "normal" stock performance, our less than 60% average allocation turned out to be moderately helpful in itself, given the shortfall in subsequent performance. And typically, our reduced allocation to stocks meant a higher allocation to bonds, which turned out to have performed better than stocks over the period.

But more crucially, how successful were our allocations that moved either above or below our 52% average for the period? For the 14 quarterly dates that we recommended above that 52% allocation, the S&P averaged 1316 6 mos. later. For the 8 dates that we recommended below that figure, the average level 6 mos. later was 1068. Thus, our own "above average" allocation during the period proved to be relatively bullish; a "below average" allocation proved very bearish.

And 1 year after we recommended an "above average" allocation, the S&P Index ended at an average of 1308; when we recommended below our average, the Index later dropped down to 1092. Table 2 summarizes these results.

Table 2. S&P 500 Index After "Above Average" vs. ""Below Average" Allocations

Above Average 1316 (6 mos) 1308 (12 mos.)
Below Average 1068  (6 mos) 1092 (12 mos.)

If our suggested allocations to stocks was unrelated to subsequent levels achieved by the S&P 500, then there would be no advantage in using them. However, since the results have been shown to be predictive of subsequent improving or worsing stock performance, just as we would expect, an investor who models and modifies their portfolio in according to our stock allocation percentage, much more likely than not will achieve better returns or avoid worse ones over the following 6 to 12 mos.

Now let's look at our stock allocations and see how they compare to actual total returns as opposed to just subsequent levels of the S&P Index. For each 6 mo. period starting July 2005, we took the average allocation to stocks that we operated under during the subsequent two quarters. This 6 mo. average was compared to the total return on the S&P 500 Index at the end of the 6 mos. Specifically, we compared our "above average" stock allocations to our "below average" ones. Result: Over the 11 six mo. periods, the "above average" allocations returned +5.1%, the "below average" -0.4%. Since these results are only for 6 mo. periods, when annualized, they are the equivalent of a +10.2% return vs a -0.8% return. Table 3 summarizes these results.

Table 3. Returns for Our "Above Average" and "Below Average" Allocations

Above Average +5.1% (6 mo.) +10.2% (annualized)
Below Average  -0.4%  (6 mo.)   -0.8%  (annualized)

Now let's look at the "above average" and "below average" allocations for each of the 22 total number of quarters since July 2005 thru Sept. 2010 and see what the mean returns 6 mos. later were for Vanguard 500 Index Fund investors. On a 6 mo. basis, the returns for the "above average" allocations were 3.4% vs. 1.8% for the "below average" ones. Since these returns were for a half a year only, they are the equivalent of 6.8% vs 3.6% for a full year.

But even with these "subpar" returns, one should make no mistake about how important seemingly "small" differences can be to a long-term investor. An investor earning 6.8% annually will double his investment in about 10.5 years (exclusive of taxes). At 3.6%, it will take about 20 years.

Finally, suppose one looks at our average allocation for each calendar year between 2005 and 2010 and compares the "above" and "below average" figures to the total return on the S&P 500 at the end of that year. Once again, the "above average" allocation years beat the "below average" years 8.5% to 1.8%.

Bottom Line: Much Better Returns

It can safely be concluded that for the last 5 and a half years, using our overall stock allocation recommendations proved to lead to better returns over the following 6 and 12 mos. What this means, then, is that investors who wish to either rebalance their stock allocations or to apply a strategy of tactical asset allocation (which regularly takes into account changing market conditions when deciding on how much to currently invest in stocks) can have a reasonable degree of confidence that our recommendations will lead to significantly better returns than choosing to merely hold one's position regardless of what is going on in the stock market.


Similar to the data presented for stocks, the relationship between our level of bond allocation and a subsequent index of bond market level was positive and even stronger than for our comparison for stocks with +.50 relationship (correlation) exhibited after 6 mos. and a +.72 relationship after 12 mos.

Table 4. Our Allocation to Bonds Since July 1, 2005

Average Over Entire Period = 34%
Number of Quarters Above 34% = 11 ("above average")
Number of Quarters Below 34% = 11 ("below average")

Our index of bond fund prices was the iShares Barclays Aggregate Bond Fund (Symbol: AGG). Over the last 5 years or so, this index has ranged between 87 and 107; it is currently around 105.

Six mos. later, our "above average" allocations showed a mean of 104.2; the "below average" ones were at 100. Going out 12 mos., our "above" and "below average" allocations produced mean index levels of 104.9 and 99.9, respectively. However, this data does not take into account dividends which are a large component of returns to bond fund investors.

Using data that does takes into account dividends, after 12 mos, our "below average" bond allocations returned 5.0% if invested in the Vanguard Total Bond Market Fund (VBMFX). Our "above average" allocations returned 6.9%.

To summarize, the above data for both stocks and bonds show that our quarterly Model Portfolio allocations to each overall asset class have been quite successful over the last 5 and a half years in accomplishing what they set out to do. That is, they would have helped investors decide how much to invest in each broad category with an eye toward getting more favorable returns as compared to choosing not making periodic adjustments. To see our latest Model Portfolio allocations, see below.

Important: Mutual Fund Research Newsletter's goal is to provide helpful information and advice to readers. Please take exactly one second (we promise it takes no longer than a single mouse click) to anonymously vote your choice on whether you found this article useful. Just click on one of the following to rate this article:

1. Highly useful

2. Moderately useful

3. Slightly useful

4. Not useful at all

Or, to provide written feedback, email


click here to continue on to our April 2011 Model Portfolios; also for Steve Shefler's latest article.