Commonwealth of Massachusetts | Public Employee Retirement Administration Commission
Five Middlesex Avenue, Suite 304, Somerville, MA 02145
Ph 617 666 4446 | Fax 617 628 4002 | TTY 617 591 8917 | www.mass.gov/perac
Domenic J. F. Russo, Chairman | A. Joseph DeNucci, Vice Chairman
Mary Ann Bradley | Paul V. Doane | Kenneth J. Donnelly | James M. Machado | Donald R. Marquis
Joseph E. Connarton, Executive Director
M E M O R A N D U M
TO: All Retirement Boards
FROM: Joseph E. Connarton Executive Director
RE: Portfolio Rebalancing
DATE: December 2, 2008
One of the more frustrating aspects of this year’s epic decline in the stock market is that there’s been virtually no place to hide. Every major asset class, both domestic and foreign, has registered significant declines this year, including bonds, where US Treasury securities rose in value as a result of their perceived safe haven status but all other sectors (corporate bonds, mortgage bonds, etc.) suffered sharp losses. Neither has there been any strategy available that might have minimized losses, except for market timing, a usually futile practice not generally attempted by pension funds. If there is any solace at all to be taken from this year’s market, it is that it has provided a dramatic opportunity for retirement boards to engage in portfolio rebalancing, the practice of realigning investment portfolios back to their target asset allocation range.
During and after major market moves, portfolios deviate from their original asset allocation targets by becoming more concentrated in the best performing asset classes and less concentrated in the worst performing asset classes. A disciplined rebalancing strategy can control risk and possibly enhance performance by taking advantage of the cyclical behavior of capital markets and by eliminating any market timing calls. With rebalancing, investors would have limited the volatility of their returns over the past year by reining in their holdings of equities after the gains of the 2003-2007 bull market. At this time, rebalancing gives trustees an unemotional way to put some money back into stocks at current levels, the lowest in at least five to six years, despite the fact that we are confronting a devastating recession and an equity market that is treacherously volatile and seemingly bottomless.
Transaction costs, including the market impact that may accrue from large funds realigning their portfolios, may offset some of the benefits from rebalancing, but most studies have shown that a disciplined rebalancing program does reduce risk and may improve performance over time. Investors who may have taken money off the table during the early stages of bull markets or who begin adding to their equity holdings well before the bottom is reached during bear markets may have some understandable misgivings, but its never easy to pick market tops and bottoms and most research shows that, over a full market cycle, rebalancing typically does add value.
While cash and US Treasury securities will likely prove to be the only asset classes providing positive returns for 2008, the extent of losses among the other asset classes has varied widely, with domestic equities (-32%), international equities (-48%), and emerging markets (-60%) showing the worst losses through November 25, 2008. Demonstrating the rationale for rebalancing, let’s consider the composite asset allocation of the local Massachusetts public retirement systems that invested predominantly on their own as of December 31, 2007: 38% Domestic Equity, 16% International Equity, 2% Emerging Markets, 22% Fixed Income, 2% High Yield, 8% Real Estate, 3% Alternative Investments, 5% Hedge Funds, 2.5% Other, 1.5% Cash. By late November, the total of 56% allocated to equities would have fallen to 43% (31% Domestic, 11% International, 1% Emerging) while Fixed Income (where returns have been basically flat for the year) rose 9% to 31%. Equity losses have been so bad that asset classes like Real Estate and Hedge Funds actually rose slightly as a percentage of the hypothetical portfolio even though year-to-date losses were estimated at 25%. For those systems where the lowered equity allocation may now be below their target range and the higher allocation to bonds may be above their long-term target, rebalancing may be in order. It would be accomplished by taking money away from one or more fixed income managers and re-allocating it to one or more equity managers.
The concept of rebalancing--- periodically adjusting a portfolio back to within the original asset allocation targets and thereby managing risk by maintaining effective diversification---- is basic yet deceptively simple since developing and implementing a sound rebalancing program can be difficult. First of all, buying low and selling high is an admirable objective but selling “winners” in favor of “losers”, with the fear of perhaps leaving additional future profits on the table or putting more money into a market that seems to have no bottom, can be emotionally very difficult. Most importantly, however, rebalancing is not easy since there are many different ways to accomplish it.
First of all, rebalancing can be periodic, where the portfolio is rebalanced at the end of a particular period (e.g., quarterly). There is also threshold rebalancing, where the portfolio is rebalanced back to its target mix whenever certain asset classes reach the outer boundaries of their rebalancing range. For example, a portfolio with a 60%-40% equity-bond mix and a 3% rebalancing band would move back to 60% equities once that sector moved above 63% or below 57%. An alternative method, sometimes called rebalancing to range, would only adjust the portfolio back to the 63%-57% band. There are also ways to combine periodic and target range rebalancing.
Since different asset classes have different expected volatilities, a typical rebalancing plan might have wider bands for the more volatile asset classes (i.e., equities) than for the less volatile asset classes (i.e., bonds.) Some institutions also extend their rebalancing bands from asset classes down to individual managers within those asset classes.
There are no widely accepted principles either for setting asset allocation bands or for devising rebalancing rules. Some major pension plans explicitly rebalance “whenever the markets suggest it.” Factors affecting the development of a rebalancing plan would include the risk of individual asset classes, the correlations among them, and transaction costs. In markets where volatility is low, wider bands would have helped a portfolio achieve greater returns without much additional risk. In markets that are very choppy, having narrower bands may be more effective in achieving successful rebalancing.
Target bands for rebalancing are typically symmetrical, but since the stock market does trend higher over time and bull markets usually last longer than bear markets, some plans use asymmetrical rebalancing in order to allow a greater upside drift than downside drift. For example, a portfolio with a target of 35-45% domestic equity and a rebalancing range of 6% up and 3% down would allow stocks to go as high as 51% and as low as 32% before rebalancing is initiated. Such a strategy would allow investors to capture more of a bull market.
Rebalancing is basically a market-driven strategy, but other circumstances could also lead to a rebalancing. For a pension fund, revised actuarial assumptions are one of the most likely triggers for asset allocation rebalancing. Funding levels can also determine changes in asset allocation. Systems with a large unfunded liability and a long funding period might justify a portfolio dominated by equities while those that are at or close to full funding might adopt a more conservative income-oriented strategy.
As noted, transaction costs can be a factor in determining the success of a rebalancing program. However, rather than having to simultaneously execute programs of both buying and selling, retirement systems and similar investors have the option to minimize transaction costs by directing cash flows into the underperforming asset classes and focusing withdrawals out of the outperforming classes.
Rebalancing is a disciplined way to trim back over-valued asset classes while adding to undervalued ones. For plans sponsors, rebalancing takes the emotion out of making portfolio adjustments that may seem difficult at the time. It is a very viable and common sense alternative to market timing; since few if any investors are capable of correctly calling the exact tops and bottoms of markets, rebalancing allows portfolios to take advantage of changing market valuations in an incremental, reasonable manner. However, for those plan sponsors that do have strong conviction on the projected performance of different asset classes, tactical rebalancing can be used to incorporate those views by modifying or overriding the changes called for by the conventional rebalancing rules.
Rebalancing should be seen as a way to maintain portfolio risk within targeted ranges. It is not guaranteed to produce superior absolute returns over a buy-and-hold strategy over every possible time period, but it should provide a portfolio superior risk-adjusted returns. The principal justification for rebalancing is that asset classes do have an unmistakable and inevitable tendency to revert to the mean in terms of performance over an extended period. However, it is important to remember that they do not always do so according to our expected timetables. As seen in the late 1990s for stocks and more recently for real estate, the development of market bubbles can throw all the rules out the window. In such cases, aberrations last so much longer than expected that many investors begin to think that a new paradigm is at hand. But, for truly experienced investors, the question is not if but when such bubbles will burst. As seen beginning in March 2000 (the end of the previous bull market) and in the fall of 2007 (the end of the most recent bull market), when the correction finally happens, those who rebalanced by reducing their equity holdings did find decisive vindication. Barring an economic calamity beyond our current expectations, patient investors putting money back into stocks at today’s levels will find satisfaction in the coming years.
If a portfolio doesn’t initiate its own rebalancing, the market will eventually do so on its own, with implications usually less favorable to an account’s performance. By taking money out of the stock market prior to its top and its ensuing major decline and by adding to out-of-favor asset classes that will eventually do very well, it is intuitive to see how rebalancing can make a difference by smoothing returns and possibly enhancing returns over time.
In our first memo on rebalancing issued in February 2003, we observed: “At this time, stocks remain mired in worries over geopolitical and economic concerns, but after underperforming bonds by more than 70% over the past three years, stocks are as attractively valued relative to bonds as they’ve been in several years. For those systems where equity holdings have fallen through the lower end of their target range, adding to equity holdings may prove beneficial in time.” That proved to be a timely observation and, while the economic and market situation at the end of 2008 is considerably more dangerous and frightening than earlier this decade, it is similarly likely that beginning to take advantage of equities at today’s low valuations will prove beneficial going forward.
To the extent that it involves reallocating money among existing investment managers, PERAC approval is not required for a retirement board to implement a rebalancing strategy. Many Massachusetts public retirement systems have rebalancing plans in place and will be discussing implementation at forthcoming board meetings, if they have not already done so. For other systems, the PERAC Investment Unit would be pleased to discuss rebalancing in a more detailed and system specific manner with any interested retirement board.
As described in our 2007 “Investment Best Practices Manual”, rebalancing is one component of the overall asset allocation process, the most important aspect of any investment program. While retirement boards should annually review whether their existing asset allocations remain on track to achieve the board’s investment goals, boards may want to discuss with their consultants whether this year’s expected major hit to funding ratios should hasten their undertaking the comprehensive review of asset allocation that normally would occur about every three years.