by Charles Rotblut, CFA

Diversification and regular portfolio rebalancing provide measurable benefits to investors. This statement holds true even though the last decade bought two bear markets for stocks and the outlook for the future is currently anything but certain.

I realize that saying diversification and rebalancing help may surprise some of you. The roller-coaster ride stocks have been on and the portfolio devastation experienced by many investors has created a great deal of skepticism about whether the cornerstones of portfolio management still work. Yet the data shows that they do in fact work.

To prove this point, last December I created a simple portfolio consisting of $50,000 in the Vanguard 500 Index (VFINX) and $50,000 in the Vanguard Total Bond Market Index (VBMFX) mutual funds. I ran the numbers as if the portfolios had been held during the lost decade of 2000 2009. As some of you may remember from my commentary in our weekly AAII Investor Update e-newsletter on December 16, 2010, there was an advantage to annual rebalancing.

Though the simplicity worked to prove a point, advocates of diversification suggest that investors hold more than just large-cap stocks and bonds. Plus, I received e-mails from members asking what would happen if other assets were included. So, I decided to expand the number of funds included in the portfolio and rerun the numbers. As you will soon see, a diversified portfolio would have given better returns. Rebalancing on a regular basis (e.g. annually) would have lowered the volatility of the fully diversified portfolio, making for a winning combination.

The Portfolios

The portfolios were designed based on one of the AAII Asset Allocation Models, which can be viewed at www.aaii.com/asset-allocation. I chose the Moderate Investor profile, which assumes an investment horizon of 20+ years. Figure 1 shows the composition of this model. (Allocation ultimately depends on age, health and wealth. The longer the time period before you need to withdraw cash and the less cash you need relative to the size of your portfolio, the higher your exposure can be to risky assets such as stocks and stock funds.)

To conduct the test, I wanted to duplicate what an actual investor would have been able to do. I chose Vanguard index mutual funds for the portfolios to limit the influence that an active manager would have had on the returns. Additionally, using a single fund family limited transaction costs.

An investor might have been able to achieve higher returns by making astute choices about actively managed funds and not sticking to a single fund family. The trade-off, however, would have been higher management fees, transaction costs and the timing issues of switching from one fund family to another. By choosing Vanguard, I avoided the issues of expenses and switching fund families, at the cost of not achieving the most optimal return.

Because I was solely using Vanguard, I had to start the study at the beginning of 1988. This was the first full calendar year that an investor would have been able to own the Vanguards Extended Market Index fund (VEXMX) .

I also was not fully able to exclude active management, even by sticking with Vanguard. In 1988, the only international funds that the company offered were actively managed. Therefore, I used the Vanguard International Value fund (VTRIX). an actively managed fund, until the end of 1996. At the start of 1997, I was able to switch to a passively managed fund, the Vanguard Total International Stock Index fund (VGTSX). This fund was incepted in April 1996.

Finally, the decision to stick with Vanguard meant that funds for all the desired asset classes were not available for the entire period. For example, Vanguard Emerging Markets Stock Index fund (VEIEX) was not incepted until 1994. To get around this, I used the broad-based Vanguard International Value fund (VTRIX) until the end of 1994 and then split some of the portfolio dollars into the emerging market fund (more on this momentarily).

The Mutual Funds

The mutual funds used for the portfolios are listed below. Funds were included at the start of the 1988 or at the start of the first full calendar year after their inception.

• Vanguard 500 Index (VFINX ),
• Vanguard Extended Market Index (VEXMX) ,
• Vanguard Mid-Cap Index (VIMSX ),
• Vanguard Small-Cap Index (NAESX ),
• Vanguard International Value (VTRIX) ,
• Vanguard Emerging Markets Stock Index VEIEX ,
• Vanguard Total International Stock Index (VGTSX). and
• Vanguard Total Bond Market Index (VBMFX) .

The starting portfolios held Vanguard 500 Index, Vanguard Extended Market Index, Vanguard International Value, and Vanguard Total Bond Market Index funds.

The Vanguard Extended Market Index tracks the performance of the Standard & Poors Completion Index. This index contains all regularly traded U.S. exchange-listed common stocks, except those stocks included in the S&P 500 index. The fund was held until the start of 1999, when the holding was switched to two other funds: 67% of the funds value was shifted to the Vanguard Mid-Cap Index and 33% was moved to the Vanguard Small-Cap Index. In order to avoid overweighting small-cap stocks, I kept the small-cap fund out of the portfolio until the mid-cap fund became available.

At the start of 1995, 25% of the dollars held in Vanguard International Value were moved to Vanguard Emerging Markets Stock Index to match the models allocation weightings. At the start of 1997, the remaining dollars held in Vanguard International Value were moved to Vanguard Total International Stock Index, a broad-based, passively managed international fund.

Portfolio Construction & Management

Three portfolios were created: static, moderate and rebalanced. All three had a starting value of $100,000. Allocations for the moderate and rebalanced portfolios matched the Moderate Investor profile as close as would have been possible at the start of 1988.

The Static Portfolio assumes that an investor looked at the asset allocation model, developed a basic understanding of the need for diversification and split $100,000 evenly among Vanguard 500 Index, Vanguard Extended Market Index, Vanguard International Value, and Vanguard Total Bond Market Index funds at the start of 1988 ($25,000 in each fund). No changes were made after the funds were originally bought. This strategy may most closely resemble the manner in which many investors have managed their IRA accounts. The Static Portfolio is presented in Table 1 .

The Moderate Portfolio was originally designed to mimic the Moderate Investor profile. Portfolio dollars were allocated to Vanguard 500 Index, Vanguard Extended Market Index, Vanguard International Value, and Vanguard Total Bond Market Index funds in proportion to the asset allocation model at the start of 1988. For example, $20,000 was invested in the large-cap Vanguard 500 Index fund and $30,000 was invested in the fixed-income Vanguard Total Bond Market Index.

After the portfolio was originated, however, it was not rebalanced. The only changes occurred when new funds became available. On those occasions, portfolio dollars were simply shifted from the older fund to the newer fund. As a result, changes in allocations reflected fluctuations in each funds comparative performance, but not any direct involvement on the part of the investor. The Moderate Portfolio is presented in Table 2 .

The Rebalanced Portfolio was also designed to mimic the Moderate Investor profile, with funds proportionately allocated to Vanguard 500 Index, Vanguard Extended Market Index, Vanguard International Value, and Vanguard Total Bond Market Index funds at the start of 1988. This portfolio, however, was rebalanced annually to ensure that it continued to mimic the models allocation percentages as closely as possible. (For example, the large-cap portion of the portfolio was adjusted so that it comprised just 20% of the portfolios total value at the start of every calendar year.)

This meant shifting money out of the best-performing funds and into the worst-performing funds. When a new fund was added, dollars were first shifted from the older fund into the newer fund and then entire portfolio was reallocated. The Rebalanced Portfolio is presented in Table 3 .

Performance Results

Diversification and rebalancing provided clear benefits. The Rebalanced Portfolio, which was rebalanced annually to match the allocation percentages of the Moderate Investor profile, provided the highest annualized return, 9.7%, and the lowest level of volatility, a standard deviation of 12.7%. The Moderate Portfolio, which was originally set up to track the Moderate Investor profile and then mostly left alone, produced a lower annualized return of 9.4% and higher level of volatility (standard deviation of 14.6%). The Static Portfolio fared the worst with an annualized return of just 8.9%. It also experienced the highest level of volatility with a standard deviation of 15.4%.

The Moderate Portfolio and The Rebalanced Portfolio were adjusted to provide increased exposure to Vanguard Emerging Markets Stock Index fund (VEIEX). and this boosted their performance. Emerging markets was one of the best-performing mutual fund categories over the past five years, as was shown in the March 2011 AAII Journal. Furthermore, Vanguard Emerging Markets Stock Index fund had the highest annualized return since 1999 of all funds listed in this article. The trade-off, however, was that the fund also experienced the highest level of volatility. High returns never come without high risk, and Vanguard Emerging Markets Stock Index fund is evidence of this.

It is important to realize that an investor would not have needed foresight about the performance of emerging markets to invest in Vanguard Emerging Markets Stock Index fund when it became available. Rather, an investor would have just needed an understanding of the role emerging markets can play an overall diversification strategy. A big upside of diversification is that it increases the odds of being in the right asset class at the right time.

It should also be noted that the Static Portfolio did not hold Vanguard Emerging Markets Stock Index fund and still had the highest level of overall volatility. An often overlooked benefit of diversification is that adding a volatile asset can lower a portfolios overall risk level if the asset is not highly correlated with the other investments. Emerging market stocks have historically experienced different long-term return characteristics than U.S. and European stocks.

Perhaps the biggest lesson is the importance of a being a proactive investor. Merely rebalancing the portfolio annually increased returns and lowered volatility, as the return data shows. For example, the Rebalanced Portfolio lost 26.4% in 2008 versus 33.3% for the Moderate Portfolio and 32.6% for the Static Portfolio. The Rebalanced Portfolio also experienced smaller losses than the other two portfolios in 1990 and 2002. In addition, it posted a fractional gain in 2001, whereas the other two portfolios lost money.

In exchange for cushioning the blow of bear markets, rebalancing annually will cause returns to lag during a strong bull market for the largest asset classes held in a portfolio. Such was the case in the late 1990s when the Rebalanced Portfolio posted smaller gains than the other two portfolios. The reason is that rebalancing forces an investor to take profits from the best-performing asset class and put the money into the worst-performing asset class. This is the classic buy low and sell high strategy.

Since the money is moved out of the winners and into the losers, annual rebalancing can give the impression that it is not working over short periods of time. This is a misperception because rebalancing is designed to force an investor to sell into momentum. In other words, it reduces downside volatility by lowering the impact that a bubble in a particular asset class would have on the portfolio.

Are There Alternatives?

As the annual results show, neither diversification nor rebalancing will prevent a portfolio from fluctuating in value. They certainly will not prevent a portfolio from losing money in a severe bear market such as 2008.

The alternatives, however, produce even worse performance and more volatility. Failing to rebalance hurt the returns, as shown here. Market timing is intended to signal when to buy and sell a particular asset, but it fails to make correct calls on a consistent basis. Worse yet, if a market timing system causes you to be late with your buy and sell decisions, you could lock in big losses and miss out on big rebounds.

Options could provide a hedge, but they also hurt returns due to commissions and the probability that they will expire worthless. Plus, exercising a put option (a contract to sell a security at a set price within a given time period) subjects you to the risks of market timing.

Conclusion

Both diversification and rebalancing are long-term strategies, and they should be treated as such. A common mistake that investors make is abandoning the strategies rather than sticking with them. Failing to stick with these strategies increases your portfolios exposure to the whims of the market, as the data in this article has shown.

There is no magical formula to building and preserving wealth. Rather, it requires patience, discipline and a sound portfolio strategy.

How Your Age Affects Diversification and Rebalancing

How you allocate your portfolio depends on age, health and wealth. The longer the period of time before you need to withdraw cash and the smaller the proportion of total savings that will be withdrawn, the greater your tolerance for risk.

Investors in their 20s will not need their retirement savings for several decades. Therefore, they can withstand a lot of volatility within their portfolios and should have a very high allocation to stocks.

Retirees, conversely, usually need to withdraw from their portfolios. The amount of the withdrawal will depend on lifestyle, health and other sources of income (e.g. a pension). Such investors may need a higher allocation to bonds then is suggested in the Moderate Investor profile.

Both kinds of investor will need to factor in their age and changes in their lives as they rebalance. A rule of thumb is to increase the allocation to bonds by 1% each year. This measure must be adjusted, however, for any changes in financial needs.

An investor needing to withdraw money should use annual rebalancing to free up cash. Proceeds from the best-performing asset classes should first be used to fund the withdrawal. The portfolio is then rebalanced.

Finally, be sure to consider tax issues when diversifying and rebalancing. The most tax-efficient assetsincluding municipal bonds and stock index fundsshould be held in taxable accounts. Investments that are more likely to incur taxes should be held in tax-deferred accounts. Similarly, income-producing investmentsincluding dividend-paying stocksmay be more suitable for a traditional IRA than a Roth IRA because they produce cash for required minimum distributions (RMDs) .

Charles Rotblut, CFA is a vice president at AAII and editor of the AAII Journal. Follow him on Twitter at twitter.com/CharlesRAAII .

Discussion

James from VA posted over 3 years ago:

Could you please consider having a link to a printable copy? This would make it more readable once printed, and would also save me paper. -)

Nancy from MD posted over 3 years ago:

Have you done this study with a portfolio diversified in alternative assete classes, such as real estate, commodities, long/short funds, infrastructure, convertible and high yield bonds?

You used a very traditional asset allocation. Would like to see how rebalancing would result in with all these other asset classes.

Robert from OR posted over 3 years ago:

Nice work, Charles. In addition to echoing Nancy from Maryland’s comments, I’m curious about what tool(s) you used to do your analysis.

Did you look up the historical quotes and rebalance yourself or use a commercially available tool?

Do you know of a reasonably priced software package that would allow an individual investor to give it a go on his own?

Charles Rotblut from IL posted over 3 years ago:

Thank you.

Nancy — I did think about including alternative investments, particularly the Permanent Fund, but doing so would have increased transaction costs, which I was purposely trying to limit. Plus, there is the problem of determining a proper allocation versus using an allocation strategy that is widely available to AAII members. Diversification does help a portfolio, and depending on what additional assets are included and how much of a weighting is assigned to them, performance should be improved.

Robert — I used Morningstar data, the same data published in our annual mutual fund guides, with an Excel spreadsheet. The advanced portfolio software seems only to be available at the professional level and, unfortunately, is priced accordingly.

-Charles

Patrick from WI posted over 3 years ago:

perfectly stated in the conclusion which i have cut & pasted

Both diversification and rebalancing are long-term strategies, and they should be treated as such. A common mistake that investors make is abandoning the strategies rather than sticking with them. Failing to stick with these strategies increases your portfolio’s exposure to the whims of the market, as the data in this article has shown.

There is no magical formula to building and preserving wealth. Rather, it requires patience, discipline and a sound portfolio strategy.

this is my 1st week in aaii & i want to get as much info as i can for the weekend to review. i have been a member of the motley fools since 12/28/8-08 so i am familiar with this type of thinking. thanks pat

Philip from IL posted over 3 years ago:

Good article, but I have one question. Vanguard would suggest that your US stocks be 65% large cap, 27% mid cap, and 8% small cap.

You used 40%, 40%, and 20%.Why?

Jim from MI posted over 3 years ago:

The period covered by the study had unusually high bond returns. This is common in such studies of diversification because records are more easily available from more recent periods where there were more international funds were available, for example. I would be interested to see what happened when interest rates were going up instead of falling, and how much of the advantage was due to falling interest rates.

Charles Rotblut from IL posted over 3 years ago:

Philip — Portfolio allocations ultimately need to be customized to one’s situation and tolerance for risk. Over the long-term, small-cap stocks outperform large-cap stocks, which is why we give them a higher weighting than other models.

Charles from IL posted over 3 years ago:

H — Standard deviation matters because it measures how volatile the portfolio was. This is very important because highly volatile returns can cause investors to sell at the worst time (e.g. December 2008). So while a single investment in VEXMX would have earned higher returns, many investors would not have been able to tolerate the roller coaster ride it experienced over the last decade.

Gary from IL posted over 3 years ago:

I think that the standard deviation for the Static Portfolio is considerably overstated at 27.1 I get 15.4 when I run the numbers and that seems more in line with the figures from the other two portfolios.

H from TX posted over 3 years ago:

In a 20+ year horizon, you are suggesting that volatility matters in say, the first 15 years? I do not agree. Volatility matters in the last several years prior to removing assets from the portfolio, in my opinion. Many investors may not be able to tolerate buying more of a fund when it has gone down for a year and selling a winning position in another even though that is what possibly should be done. It is my opinion that a young person with a ROTH IRA and a 20+ year horizon would not be smart to have anywhere near 30% of the IRA in a bond fund.

Charles Rotblut from IL posted over 3 years ago:

Gary,

You are correct. The standard deviation for the Static Portfolio is 15.4%, not 27.1% as originally published. The original calculation picked up a negative -100% return figure from a hidden cell, which resulted in the error. My apologies for the error.

Even after the correction, the static portfolio remains the most volatile of the three portfolios and shows the downside of not rebalancing on a regular basis.

-Charles

Dan from CO posted over 3 years ago:

What formula did you use to get the Ammualized Returns? I see IRR was not used to determine the return on the investment.

Dave from WA posted over 3 years ago:

Charles,

Where do you find this data in Morningstar going back to 1988 and is this just year end price data with a % return for the year?

Would this data account for re-invested dividends? In general it seems to me that my Morningstar portfolios don’t accurately account for re-invested dividends.

Thanks for a great article.

Charles Rotblut from IL posted over 3 years ago:

Dave — We receive a data feed from Morningstar with the fund information. Total returns are used, which include price appreciation and dividends. No cash was withdrawn from the portfolio.

James from OH posted over 3 years ago:

Charles is to be commended for the effort and analysis he did for this article. However, because of the constraints he was trying to operate within (in particular, his attempt to match the Moderate Investor portfolio), he ended up constructing an “apples to oranges” comparison with the Static, Moderate, and Rebalanced portfolios that, at least in my mind, invalidate the results and the conclusions.

Why is it an “apples to oranges” comparison? The Static portfolio did not include VIMSX, NAESX, VEIEX, or VGSTX which were in the Moderate and Rebalanced portfolios. These funds all materially out-performed the mutual funds (in particular, VFINX and VBMFX) which were in the Static portfolio over the comparable years. You can see that by a casual comparison of the Static and Moderate portfolio spreadsheets in the article.

In simple terms, the Annual Returns of the Moderate and Rebalanced portfolios are primarily higher than the Static portfolio because of the addition of mutual funds which out-performed the funds in the Static portfolio.

Please note. I’m not trying to criticize Charles’ work. No doubt, he bent over backwards in order to present a reasonable comparison. I am merely offering a warning on the degree to which you accept the conclusions. It is difficult to construct an “apples to apples” comparison over the long term using several investment options, given the fact that historical data on most mutual funds (and different markets) is only readily available for the last 20+ years or it is expensive to acquire.

I have done my own back-testing of asset allocation and rebalancing. The common allocation that the big financial institutions describe in their literature and web sites is between the stock market, bond market, and money market. I tested 5 different allocations across those three markets. I tracked the results day-by-day to determine what the largest drawdowns were that an investor would have experienced. I compared the results for No Allocation, Allocation without Rebalancing, and Allocation with Rebalancing. What I found was that the best Annual Returns are achieved with No Allocation and that Allocation with and without Rebalancing produce about the same level of Annual Returns (for each of the 5 allocation methods). No Allocation produces the highest risk. Allocation without Rebalancing produces somewhat lower risk and Allocation with Rebalancing produces somewhat lower risk than that. (The results hold whether you reinvest dividends or not.)

I will acknowledge that over the long run, Rebalancing is somewhat better than not Rebalancing since it produces about the same level of returns with slightly less risk. (However, I don’t mean to imply that rebalancing is the best alternative for an investor.) However, an investor would have endured drawdowns of 22%-39% in his/her portfolio from time to time, depending on the level of allocation. If this is acceptable to you, then your investment expectations are lower than mine. I do recognize that each investor needs to determine an investment strategy that fits his/her “sweet spot” that balances off higher returns with lower risk, which are two conflicting goals.

I offer the following additional thoughts. (1) It is only implied in the article, but Rebalancing presumes you are reinvesting the dividends. If not, you should expect materially lower Annual Returns and seriously consider No Allocation(i.e. No Diversification). (2) In these days of very high correlation across all types of investments, asset allocation is not as effective. (3) As Warren Buffet is credited with saying, “Diversification [like allocation — JG] is for those people who don’t know what they are doing.” I would add, “Or for those people who just don’t want to spend time managing their money more than once a year.” (4) There is a large consensus in the investment community that higher risk produces higher returns. If the stock market didn’t produce higher returns to offset its higher risk, eventually everyone would catch on and stop investing in it. (5) When someone tells you about an investment that will allow you to make more money for less risk, you should inhale and count to 10 (3-4 times).

Jim Grant

Solon, Ohio

JWGrant@AOL.com

Charles from IL posted over 3 years ago:

Thanks James. I’m working on a follow-up article that provides a direct apples-to-apples between rebalancing and not rebalancing. It also factors in retirement withdrawals and long-term data on rebalancing. No set date for publication, but hopefully by the end of the year. I will be talking about it at our national conference in November, however. -Charles Rotblut