This article is about whether and how to rebalance your investments, to keep them aligned with your targets for stocks and bonds. Although many advisers recommend routinely rebalancing your portfolio every year or quarter, our research at able2pay.com finds that other methods may be better. They are explained here, along with step-by-step instructions.
Rebalancing Your Portfolio
When financial advisers and publications recommend routine rebalancing of your portfolio, they typically make two points:
- Annual or quarterly rebalancing may improve your returns modestly, though it's not guaranteed to do so.
- It will keep your actual allocations to stocks and bonds on track with your intentions.
- Do income placement. When you get interest or dividends, don't automatically reinvest them into the funds or assets that generated them. Instead, put the proceeds briefly in a money market fund. Then invest them into holdings that are below their target percentages.
- Check your portfolio for drift, measured by a simple formula (explained below) that summarizes how far your funds have departed from their intended levels.
- If the drift exceeds a trigger, typically 3% or 4%, then rebalance. This tends to happen rarely, about once every year or two. Checking quarterly for excess drift has historically improved a portfolio's returns. Checking more frequently has been even better. However, waiting a year or more may eliminate any benefit. Notably, the potential benefit is likely to increase as your stock-target rises. It's biggest at 80% in stocks, virtually nil at 20%.
- If your portfolio has cash flows, either from new money being added or from withdrawals being subtracted, use them to steer your assets toward their targets. Put deposits into funds that are below-target, and take withdrawals from those above-target. Doing so will reduce drift and limit the need for rebalancing. If you are saving up for retirement or spending down as a retiree, read the Cash Flows section below for more information.
- Finally, if the foregoing steps seem too much, consider investing with a firm that does automated rebalancing. Firms that do so by the best methods, at the lowest fees, are discussed in the section on Best Options.
Practical Advice
Reality Check
Sometimes, rebalancing your portfolio may be harmful to your returns. As Andrew Ang explains in his comprehensive textbook, Asset Management (section 3.2), mathematical analysis implies that rebalancing "profits from reversals" while "increasing the losses during extreme low markets, and underperforming the buy-and-hold strategy during extreme high markets" (emphasis added). The problem, of course, is that you can't tell in advance whether your investments are in a market that's trending persistently up or down (when rebalancing is harmful), or one that's about to change direction (when rebalancing is helpful).
Above all else, then, remember this. Rebalancing can be helpful to your returns only if ...
To appreciate the last point, about broad assets, consider a trenchant example, adapted from Ang's book. Imagine yourself a wealthy Russian in the early 20th century. You systematically and over-frequently rebalance 50% of your money in Russian companies and 50% in U.S. bonds, right through the revolution of 1917 and beyond. As the value of your Russian equities plummets, you rebalance more and more wealth out of your U.S. bonds and into Russian stocks. In the end, your Russian assets, all bankrupt or appropriated by the state, consume 50% of a portfolio whose net worth is mere pennies. To put the point bluntly: Rebalance on boutique assets at your peril.
Best Options
As explained in the rest of this article, your realistic options for keeping a balanced portfolio are as follows, with the best listed first:
Sometimes, rebalancing your portfolio may be harmful to your returns. As Andrew Ang explains in his comprehensive textbook, Asset Management (section 3.2), mathematical analysis implies that rebalancing "profits from reversals" while "increasing the losses during extreme low markets, and underperforming the buy-and-hold strategy during extreme high markets" (emphasis added). The problem, of course, is that you can't tell in advance whether your investments are in a market that's trending persistently up or down (when rebalancing is harmful), or one that's about to change direction (when rebalancing is helpful).
Above all else, then, remember this. Rebalancing can be helpful to your returns only if ...
- Your method waits long enough to catch potential changes in the market's direction, and ...
- You implement the method consistently over a very long time, and ...
- You rebalance across broad assets, such as index funds of all domestic or international stocks or bonds.
To appreciate the last point, about broad assets, consider a trenchant example, adapted from Ang's book. Imagine yourself a wealthy Russian in the early 20th century. You systematically and over-frequently rebalance 50% of your money in Russian companies and 50% in U.S. bonds, right through the revolution of 1917 and beyond. As the value of your Russian equities plummets, you rebalance more and more wealth out of your U.S. bonds and into Russian stocks. In the end, your Russian assets, all bankrupt or appropriated by the state, consume 50% of a portfolio whose net worth is mere pennies. To put the point bluntly: Rebalance on boutique assets at your peril.
Best Options
As explained in the rest of this article, your realistic options for keeping a balanced portfolio are as follows, with the best listed first:
- Invest at a firm that automatically does income placement plus a percent trigger. Implicitly or explicitly, the firm will charge a modest fee for this service. Only if your stock allocation is 50% or greater is the incremental benefit likely to exceed the added cost. At this writing (December 2015), Betterment appears to be the only retail investment firm that has fully implemented this method, right down to the last dollar in your account, at the lowest overall fees and in a manner that is tax-aware. Wealthfront and Schwab Intelligent Portfolios are possible alternatives, roughly in that order. Their portfolios tend to be weaker or their implementations less complete, compared to Betterment. (For a detailed review of these and other firms, see Best Firms 2016.)
- Do your own rebalancing, using income placement plus a compound trigger of 3% quarterly. You may be able to do this in your 401(k), 403(b), IRA or taxable account. You will have to be disciplined, and check your account every quarter. If you have an account at Vanguard, you may be able to pay one of their financial planners to do this type of rebalancing for you; there would be an incremental fee of 0.3%, however. Provided that your investments are limited to Vanguard's lowest-cost index funds and your stock-allocation is at least 50%, the cost-benefit ratio with the additional advisory fees might be competitive with the other firms mentioned above.
- Don't rebalance at all. Instead, use income placement alone. To implement it, you should not reinvest interest and dividends automatically into the fund that generated them. Instead, have them invested in a money market or short-term bond fund, and periodically transfer the accumulated amount into the funds or ETFs in your portfolio that are below-target. Every month or two may work well.
- Use an all-in-one or target-date fund. Pick one that has low fees and matches your preferred allocations. Allow the fund's managers to do continual rebalancing. This is not a bad option. It's fourth on the list mainly because the financial benefits are limited. But it's better, for sure, than buying a bunch of funds and then ignoring them for years.
- Your target allocation is less than 40% to stocks. A percent trigger or a compound trigger may be extra work with little or no benefit. In this case, option 4 (an all-in-one fund with low fees) would rise to the top as your best option.
- You are taking periodic withdrawals, either as a retiree or for other reasons. Then you should be able to apportion your withdrawals in a manner that keeps your portfolio close to your targets, close enough, in fact, that option 2 (quarterly 3% trigger) will add little value. Options 1 (firms with automated rebalancing) and 4 (all-in-one fund) may be your best choices.
Income Placement
Imagine it's January 1871. You have $100 to invest. You put $50 in a fund that tracks the stocks of large U.S. companies. Over the years, the fund evolves to follow the S&P 500 index (which doesn't exist yet). You invest the remaining $50 in 10-year U.S. treasury bonds. You find rebalancing too complicated, so you never do it. Ever.
Instead, four times a year when your stock fund pays you a dividend, you make a choice. If your porfolio is overweighted in stocks (more than 50%), you invest the stock dividends in bonds. Conversely, if you are overweighted in bonds (more than 50%), you invest the stock dividends back into your stock fund.
Twice a year, the U.S. Treasury Department pays interest on your 10-year bonds. You reinvest the interest in the same way as stock dividends, either in stocks or bonds, depending on what's below your target allocation of 50%. On some occasions, if your current allocation is close to 50-50, you may reinvest your dividends and interest in both stocks and bonds in a manner that sets your portfolio to exactly 50-50. One more detail: You regularly sell your 10-year bonds and buy new ones, to keep your bond-portfolio at a constant maturity of 10 years. In today's world, you would do that by holding a mutual fund or ETF for intermediate-term treasuries.
The essence of your strategy is to place your income, whether it's dividends, interest, or new money, into the investment that's below its target. Look at the chart below, which shows the percentage of your hypothetical portfolio that would have been invested in stocks from your initial investment in 1871 until mid-2014.
Instead, four times a year when your stock fund pays you a dividend, you make a choice. If your porfolio is overweighted in stocks (more than 50%), you invest the stock dividends in bonds. Conversely, if you are overweighted in bonds (more than 50%), you invest the stock dividends back into your stock fund.
Twice a year, the U.S. Treasury Department pays interest on your 10-year bonds. You reinvest the interest in the same way as stock dividends, either in stocks or bonds, depending on what's below your target allocation of 50%. On some occasions, if your current allocation is close to 50-50, you may reinvest your dividends and interest in both stocks and bonds in a manner that sets your portfolio to exactly 50-50. One more detail: You regularly sell your 10-year bonds and buy new ones, to keep your bond-portfolio at a constant maturity of 10 years. In today's world, you would do that by holding a mutual fund or ETF for intermediate-term treasuries.
The essence of your strategy is to place your income, whether it's dividends, interest, or new money, into the investment that's below its target. Look at the chart below, which shows the percentage of your hypothetical portfolio that would have been invested in stocks from your initial investment in 1871 until mid-2014.
Except for a few historic events, such as the Great Depression and the dot-com bubble, your portfolio stayed fairly close to your 50-50 target. Mostly, it drifted between 60% and 40% in stocks. Not once in this 143-year period did you rebalance by selling stocks to buy bonds or vice versa. A nice benefit of this strategy is that because no sales occurred, tax consequences would have been minimized.
Remarkably, in this portfolio, income placement would have grown your wealth better than if you had rebalanced every quarter. Over the entire period, quarterly rebalancing returned 6.8% per year; income placement, in contrast, returned 7.2%, a benefit of 0.4% annually. During some interim periods, the opposite happened, but over the full 143 years, income placement was the winner.
One caveat: the benefit came with a bias. From 1871 to 2014 with targets of 50% each for stocks and bonds, income placement had, on average, 52.5% invested in stocks. Thus, while it gave you better returns, the method was 2.5% off-plan, overall. This bias is no fluke. In various other portfolios, described later in this article, a strategy of income placement often tilted the investments toward stocks. The bias tended to be larger when the target allocation to stocks was lower. For example, a portfolio with a target of 40% in stocks might average more than 45% over time. If it mattered, you could easily correct the bias by rebalancing if your stock-allocation became persistently and uncomfortably high.
Remarkably, in this portfolio, income placement would have grown your wealth better than if you had rebalanced every quarter. Over the entire period, quarterly rebalancing returned 6.8% per year; income placement, in contrast, returned 7.2%, a benefit of 0.4% annually. During some interim periods, the opposite happened, but over the full 143 years, income placement was the winner.
One caveat: the benefit came with a bias. From 1871 to 2014 with targets of 50% each for stocks and bonds, income placement had, on average, 52.5% invested in stocks. Thus, while it gave you better returns, the method was 2.5% off-plan, overall. This bias is no fluke. In various other portfolios, described later in this article, a strategy of income placement often tilted the investments toward stocks. The bias tended to be larger when the target allocation to stocks was lower. For example, a portfolio with a target of 40% in stocks might average more than 45% over time. If it mattered, you could easily correct the bias by rebalancing if your stock-allocation became persistently and uncomfortably high.
Drift
In words, drift is simply the cumulative amount by which your investments have departed from your target percentages. Mathematically, drift is easy to compute, although some folks get it wrong or make it more complicated than necessary. Here is a worked example.
Portfolio targets:
There's also a shortcut that works well. If you have several stock-funds and bond-funds, you can use the total target for all stocks, rather than individual targets for each particular fund. In the example above, your total stock target would be 60%. Add up the dollar-value of all your stock funds, and divide it by the total value of your portfolio. That's the actual stock-percentage. Subtract your total target for stocks, and there you have it, your short-cut drift.
Portfolio targets:
- 30% U.S. stocks
- 30% international stocks
- 30% long-term bonds
- 10% short-term bonds
- 33% U.S. stocks
- 31% international stocks
- 27% long-term bonds
- 9% short-term bonds.
- Only consider assets that are above target.
- U.S. stocks = + 3%
- International stocks = +1%
- Drift = +4%
There's also a shortcut that works well. If you have several stock-funds and bond-funds, you can use the total target for all stocks, rather than individual targets for each particular fund. In the example above, your total stock target would be 60%. Add up the dollar-value of all your stock funds, and divide it by the total value of your portfolio. That's the actual stock-percentage. Subtract your total target for stocks, and there you have it, your short-cut drift.
Triggers
A trigger is an event that causes rebalancing to occur. We'll consider three varieties:
Research Study
Five sample portfolios were constructed, using historical data obtained from Robert Shiller's research database and from Yahoo Finance. The portfolios started at the earliest date of available data; they ended between June 2014 and January 2015. Additionally, each portfolio was studied at five target levels for stocks: 20%, 40%, 50%, 60%, and 80%. Thus, with five portfolios and five targets, there were 25 scenarios in all. Below are descriptions of each portfolio in the study.
2asset1871. Shiller's historical data on U.S. stocks and bonds.
Calendar Triggers
To evaluate calendar triggers, the study examined monthly, quarterly, and annual rebalancing across the 25 scenarios (five portfolios at five different targets for stocks). The chart below summarizes the results.
- Calendar triggers occur on a schedule, typically once every month, quarter, or year.
- Percent triggers activate whenever drift exceeds a threshold.
- Compound triggers combine the other two. For example, a quarterly 3% trigger rebalances in March, June, September, and December, but only if drift exceeds 3%.
Research Study
Five sample portfolios were constructed, using historical data obtained from Robert Shiller's research database and from Yahoo Finance. The portfolios started at the earliest date of available data; they ended between June 2014 and January 2015. Additionally, each portfolio was studied at five target levels for stocks: 20%, 40%, 50%, 60%, and 80%. Thus, with five portfolios and five targets, there were 25 scenarios in all. Below are descriptions of each portfolio in the study.
2asset1871. Shiller's historical data on U.S. stocks and bonds.
- Assets: Large-cap U.S. companies and 10-year U.S. treasury bonds.
- Timeframe: 143 years, January 1871 to June 2014, monthly
- Funds: Total Stock Market Index and Total Bond Market Index.
- Timeframe: 22 years, November 1992 to December 2014, weekly
- Stock Funds: Equal allocations to the S&P 500 index and small-cap U.S. companies.
- Bond Funds: Equal allocations to long-term and short-term municipal bonds.
- Timeframe: 35 years, January 1980 to January 2015, weekly
- Stock Funds: Within the stock portion of the portfolio, 50% to the S&P 500 index, 10% indexed to small-cap U.S. stocks, and 40% to international stocks.
- Bond Funds: Within the bond portion of the portfolio, 60% to a Total Bond Market index, and 20% each to long-term U.S. treasury bonds and short-term corporate bonds.
- Timeframe: 28 years, December 1986 to January 2015, weekly
- Stock Sectors: SPDR Exchange Traded Funds (ETFs) for U.S. health, technology, energy, and consumer-staples, plus Vanguard's REIT Index for U.S. real estate.
- Bond Fund: Vanguard GNMA.
- Timeframe: 16 years, January 1999 to January 2015, weekly.
Calendar Triggers
To evaluate calendar triggers, the study examined monthly, quarterly, and annual rebalancing across the 25 scenarios (five portfolios at five different targets for stocks). The chart below summarizes the results.
Note that the chart takes a position on a key question: Rebalancing compared to what? It would be unfair to compare rebalancing to doing nothing because, over time, the higher returns of stocks would increasingly bias the portfolio with uncontrolled drift toward stock investments. Therefore, the chart compares income placement alone vs. income placement plus rebalancing. Note that all the lines in the chart are below 0%. Rebalancing had no incremental value. On the contrary, it penalized returns by a small amount at quarterly and annual rebalancing, and by large amounts at monthly rebalancing. The same pattern held up across all five portfolios in the study.
To be fair, income placement creates a modest bias toward stocks, as remarked earlier, thereby giving a slight boost to long-term returns. Thus, a fair reading of the results in the chart is that calendar rebalancing, when done every quarter or year regardless of portfolio drift, is essentially neutral, not slightly negative. It is most certainly harmful, however, if done monthly.
Percent Triggers
The idea of a percent trigger is simple enough. Set your trigger at a certain percentage. When your portfolio drifts beyond that percentage, rebalance your portfolio. But what's the right percentage? To find out, the study looked at various thresholds, across the five portfolios and five targets for stocks. The results are shown in the next chart.
To be fair, income placement creates a modest bias toward stocks, as remarked earlier, thereby giving a slight boost to long-term returns. Thus, a fair reading of the results in the chart is that calendar rebalancing, when done every quarter or year regardless of portfolio drift, is essentially neutral, not slightly negative. It is most certainly harmful, however, if done monthly.
Percent Triggers
The idea of a percent trigger is simple enough. Set your trigger at a certain percentage. When your portfolio drifts beyond that percentage, rebalance your portfolio. But what's the right percentage? To find out, the study looked at various thresholds, across the five portfolios and five targets for stocks. The results are shown in the next chart.
Three points stand out. First, there's a peak, positive benefit at thresholds around 3% or 4%. Second, the benefit is greater, the larger the allocation to stocks, with a miniscule benefit at 20% and a substantial one at 80%. Third, a very low threshold of 1% may be harmful.
For portfolios with 50% to 80% invested in stocks, a percent trigger set at 3% or 4% tended to boost returns by 1% to 2% per year, on average, compared to income placement alone. That's a big benefit. The same pattern occurred for all five portfolios in the study.
An important qualifier is that in this chart, drift was checked weekly. If done less often (monthly or quarterly), the benefit tended to diminish, but not to vanish entirely. There's an irony here: percent triggers work best if checked frequently, even if action is taken rarely.**
Compound Triggers
If checking your portfolio weekly seems onerous, perhaps a compound trigger would be more attractive. It would check for drift less frequently while invoking a threshold percentage to decide whether to rebalance. To investigate what might work best, the study examined 450 scenarios. That's five portfolios combined with five target levels for stocks, across triggers of 1%, 2,%, 3%, 4%, 5% and 10%, for monthly, quarterly, and annual rebalancing ( 5 * 5 * 6 * 3 = 450).
As illustrated in the next chart, the best combination was a quarterly 3% trigger, provided that the target for stocks was 40% or more. At targets under 40% in stocks, or at monthly or yearly rebalancing, or at triggers other than 3%, the benefit tended to be smaller, negligible, or inconsistent.*
For portfolios with 50% to 80% invested in stocks, a percent trigger set at 3% or 4% tended to boost returns by 1% to 2% per year, on average, compared to income placement alone. That's a big benefit. The same pattern occurred for all five portfolios in the study.
An important qualifier is that in this chart, drift was checked weekly. If done less often (monthly or quarterly), the benefit tended to diminish, but not to vanish entirely. There's an irony here: percent triggers work best if checked frequently, even if action is taken rarely.**
Compound Triggers
If checking your portfolio weekly seems onerous, perhaps a compound trigger would be more attractive. It would check for drift less frequently while invoking a threshold percentage to decide whether to rebalance. To investigate what might work best, the study examined 450 scenarios. That's five portfolios combined with five target levels for stocks, across triggers of 1%, 2,%, 3%, 4%, 5% and 10%, for monthly, quarterly, and annual rebalancing ( 5 * 5 * 6 * 3 = 450).
As illustrated in the next chart, the best combination was a quarterly 3% trigger, provided that the target for stocks was 40% or more. At targets under 40% in stocks, or at monthly or yearly rebalancing, or at triggers other than 3%, the benefit tended to be smaller, negligible, or inconsistent.*
One final chart gives additional perspective on the quarterly 3% trigger. Look first at the middle panel below, which shows the returns for a compound trigger at 3% quarterly. To the left are results for rebalancing every quarter, unconstrained by a percent threshold. To the right are results for rebalancing any week when a 3% trigger was met, unconstrained by the quarterly calendar. Clearly, the compound method is a compromise. It generated returns more favorable than simple quarterly rebalancing, but less favorable than rebalancing whenever the 3% threshold was crossed.
Cash Flows
You're Saving Up ...
What if you regularly add new money to your account, as you would when making scheduled deposits into a retirement plan? You could (and should!) direct your deposits toward the funds or ETFs that are below-target. Would you still need to rebalance?
To investigate the matter, our study added a stream of monthly deposits to the five portfolios. For portfolios that started in 1980 or later, the deposit was exactly the amount necessary to invest the maximum that the IRS allowed in 401(k) and 403(b) accounts in that year. The IRS has regularly increased the maximum level, to adjust for inflation. For the portfolio that started in 1871, the deposits started at $75 per month and increased with inflation, so that by 2014, they equaled the maximum level then permitted for 401(k) and 403(b) deposits. Finally, all the portfolios started with a zero balance in the very first month.
The chart below shows results for two methods of rebalancing. The value plotted in the chart is the additional annual return from income placement plus rebalancing, as compared to income placement alone.
What if you regularly add new money to your account, as you would when making scheduled deposits into a retirement plan? You could (and should!) direct your deposits toward the funds or ETFs that are below-target. Would you still need to rebalance?
To investigate the matter, our study added a stream of monthly deposits to the five portfolios. For portfolios that started in 1980 or later, the deposit was exactly the amount necessary to invest the maximum that the IRS allowed in 401(k) and 403(b) accounts in that year. The IRS has regularly increased the maximum level, to adjust for inflation. For the portfolio that started in 1871, the deposits started at $75 per month and increased with inflation, so that by 2014, they equaled the maximum level then permitted for 401(k) and 403(b) deposits. Finally, all the portfolios started with a zero balance in the very first month.
The chart below shows results for two methods of rebalancing. The value plotted in the chart is the additional annual return from income placement plus rebalancing, as compared to income placement alone.
With a steady inflow of new money, the essential findings depicted in the chart were these:
You're Spending Down ...
What about the opposite scenario, when you are taking regular withdrawals from your savings, and not adding any new money? This would happen, for example, if you were retired and living off your savings. To investigate, the study examined the same portfolios, but instead of investing new money every month, 1% of the portfolio value was withdrawn every quarter. That's a total of 4% per year, a rate of spending that might apply to many early retirees.
The dollar amount was not adjusted for inflation; instead, the withdrawal was simply 1% of the then-current value of the portfolio every quarter. In most cases, the principal continued to grow modestly, resulting in a tendency for the dollar-value of the withdrawals to increase over time.
- On average, rebalancing with either a simple 3% trigger or a quarterly 3% trigger improved returns over income placement alone.
- The compound quarterly 3% trigger had modest benefits. Even so, over many years, the cumulative benefits could be substantial.
- The simple trigger, which rebalanced in any week when the drift exceeded 3%, had bigger benefits.
- The benefit was negligible for portfolios invested 20% in stocks, but grew as the allocation to stocks increased.
You're Spending Down ...
What about the opposite scenario, when you are taking regular withdrawals from your savings, and not adding any new money? This would happen, for example, if you were retired and living off your savings. To investigate, the study examined the same portfolios, but instead of investing new money every month, 1% of the portfolio value was withdrawn every quarter. That's a total of 4% per year, a rate of spending that might apply to many early retirees.
The dollar amount was not adjusted for inflation; instead, the withdrawal was simply 1% of the then-current value of the portfolio every quarter. In most cases, the principal continued to grow modestly, resulting in a tendency for the dollar-value of the withdrawals to increase over time.
The chart above shows the results:
- On average, both methods of rebalancing improved returns over income placement alone.
- A compound trigger, which rebalanced only if the drift was 3% or more at the end of a quarter, had limited benefits. In this case, rebalancing was not worth the extra effort.
- A simple trigger, which checked weekly and rebalanced if the drift exceeded 3%, had much bigger benefits.
- The benefit was smallest for portfolios invested 20% in stocks, but grew as the allocation to stocks increased. Note that when taking withdrawals during retirement, stock-allocations under 60% might be the norm.
Disclaimer: Historical data cannot guarantee future results. Although a mixture of bonds, stocks, and other investments may be safer than investing exclusively in one class of assets, diversification cannot guarantee a positive return. Losses are always possible with any investment strategy. Nothing here is intended as an endorsement, offer, or solicitation for any particular investment, security, firm, or type of insurance. You are responsible for your own investment decisions. Please read our full disclosures and Fiduciary Oath.
* The research department at Vanguard also studied compound triggers, looking at monthly, quarterly or annual rebalancing conditioned on hitting a threshold of 1%, 5% or 10% drift from a target allocation. Theyused a single portfolio invested 60% in U.S. stocks and 40% in U.S. bonds. Their analysis led them to recommend quarterly rebalancing with a 5% threshold as a good strategy for most investors.
** Gobind Daryanani reported a systematic study in 2008, where he evaluated daily, weekly, and longer periods of checking a variety of percent triggers. He, too, found that returns improved with more frequent checking.
** Gobind Daryanani reported a systematic study in 2008, where he evaluated daily, weekly, and longer periods of checking a variety of percent triggers. He, too, found that returns improved with more frequent checking.