The Power of Mean Reversion
by: James Picerno
February 10, 2011
The more you look at it, the larger it looms. Yes, rebalancing is an old idea—buy low, sell high. But it's forever new for at least two critical reasons.
One is simply bound up with the recognition that rebalancing works, or at least it has a long history of working by enhancing return while keeping a lid on risk. But it's not always clear what constitutes "optimal" rebalancing, or even if such a state of financial nirvana exists. That keeps analysts turning over rocks in search of progress.
As for real-world results, much depends on the definition of rebalancing. Factors such as timing, the composition of portfolio assets, and the rules that guide rebalancing activity vary quite a bit. Accordingly, so too does the degree of risk from rebalancing strategies. Yet it's also true that rebalancing in one form or another is behind a number of the new-fangled attempts to mint alpha.
The rise of rules-based indexing products in particular rely on rebalancing, even if that's not always clear. Some products emphasize technical signals, for example, that offer insight on when to rebalance.
Other funds focus on changes in accounting-based metrics, such as sales, earnings, etc. But at the heart of many strategies is a common denominator: rebalancing.
Some of this bleeds over into other risk factors, which can blur the lines for deciding where rebalancing's influence ends and something else begins. But no matter how you dress it up, rebalancing is the foundation for most strategies that have a decent chance of delivering alpha over the long haul. The lesson applies to many strategies within a given asset class, and it's true for most multi-asset class strategies.
For the latter, efforts at earning alpha by dynamically managing a set of betas range from the routine to rocket science. But you don't need to a Ph.D. in engineering to earn a modest premium over a passive asset allocation. What you do need is a fair amount of backbone to buy out-of-favor asset classes and sell the winners. That's a dangerous pursuit with individual securities, but history and a small library of empirical research suggests this is a worthy strategy for most investors at the asset class level.
As a review of the possibilities, consider how the major asset classes stack up against one another in recent history and how those returns compare to some simple forecast-free rebalancing strategies (see table below). Note the wide range of return results over the past five years within markets, ranging from a 1% annualized return for a broad mix of commodities up to nearly 10% a year for emerging market stocks. Holding all asset classes in a passive, unmanaged mix in weights that approximate their respective market values earned 4.6% a year (BIR Global Market Index). If you rebalanced this mix every December 31 back to the allocation from the previous year, you earned more: 5.7% .
Equally weighting the same portfolio by rebalancing to equal weights every December 31 did even better, delivering 6.7% a year. By comparison, the U.S. stock market returned an annual 2.5% over the past five years, based on the Russell 3000.
.
.
The lesson is that by diversifying broadly and dynamically managing the mix in a simple but straightforward manner, you can earn competitive results. That's hardly a secret. As I explain in some detail in Dynamic Asset Allocation
, the finance literature has been telling us no less in recent decades.
Results will vary, of course, although a fair amount of the variation is under your control. If you can stomach more than average risk, you might consider an aggressive rebalancing strategy.
Instead of mindlessly rebalancing on pre-set calendar dates, you can be more opportunistic by taking advantage of volatility as it comes. Rebalancing in the fall of 2008, for instance, rather than waiting for the markets to calm down. You might also rebalance in something more than mild form, or more frequently, or allocate among a more granular definition of asset classes.
In fact, the possibilities are endless for boosting the rebalancing bonus, but so too are the risks. Earning higher returns by systematically adjusting asset weights isn't a free lunch. You have to be willing to assume a higher risk than the average investor. The main hazard is linked to mean reversion in prices. History shows that this phenomenon tends to persist over time, particularly with broad asset classes.
But there's always a question of whether it'll prevail the next time, or how soon prices will revert to their mean after a large change in the market.
Such questions help explain why most investors find it hard to rebalance when market volatility is high. But that's why the expected return from rebalancing looks so rewarding.
As for real-world results, much depends on the definition of rebalancing. Factors such as timing, the composition of portfolio assets, and the rules that guide rebalancing activity vary quite a bit. Accordingly, so too does the degree of risk from rebalancing strategies. Yet it's also true that rebalancing in one form or another is behind a number of the new-fangled attempts to mint alpha.
The rise of rules-based indexing products in particular rely on rebalancing, even if that's not always clear. Some products emphasize technical signals, for example, that offer insight on when to rebalance.
Other funds focus on changes in accounting-based metrics, such as sales, earnings, etc. But at the heart of many strategies is a common denominator: rebalancing.
Some of this bleeds over into other risk factors, which can blur the lines for deciding where rebalancing's influence ends and something else begins. But no matter how you dress it up, rebalancing is the foundation for most strategies that have a decent chance of delivering alpha over the long haul. The lesson applies to many strategies within a given asset class, and it's true for most multi-asset class strategies.
For the latter, efforts at earning alpha by dynamically managing a set of betas range from the routine to rocket science. But you don't need to a Ph.D. in engineering to earn a modest premium over a passive asset allocation. What you do need is a fair amount of backbone to buy out-of-favor asset classes and sell the winners. That's a dangerous pursuit with individual securities, but history and a small library of empirical research suggests this is a worthy strategy for most investors at the asset class level.
As a review of the possibilities, consider how the major asset classes stack up against one another in recent history and how those returns compare to some simple forecast-free rebalancing strategies (see table below). Note the wide range of return results over the past five years within markets, ranging from a 1% annualized return for a broad mix of commodities up to nearly 10% a year for emerging market stocks. Holding all asset classes in a passive, unmanaged mix in weights that approximate their respective market values earned 4.6% a year (BIR Global Market Index). If you rebalanced this mix every December 31 back to the allocation from the previous year, you earned more: 5.7% .
Equally weighting the same portfolio by rebalancing to equal weights every December 31 did even better, delivering 6.7% a year. By comparison, the U.S. stock market returned an annual 2.5% over the past five years, based on the Russell 3000.
.
.
The lesson is that by diversifying broadly and dynamically managing the mix in a simple but straightforward manner, you can earn competitive results. That's hardly a secret. As I explain in some detail in Dynamic Asset Allocation
, the finance literature has been telling us no less in recent decades.Results will vary, of course, although a fair amount of the variation is under your control. If you can stomach more than average risk, you might consider an aggressive rebalancing strategy.
Instead of mindlessly rebalancing on pre-set calendar dates, you can be more opportunistic by taking advantage of volatility as it comes. Rebalancing in the fall of 2008, for instance, rather than waiting for the markets to calm down. You might also rebalance in something more than mild form, or more frequently, or allocate among a more granular definition of asset classes.
In fact, the possibilities are endless for boosting the rebalancing bonus, but so too are the risks. Earning higher returns by systematically adjusting asset weights isn't a free lunch. You have to be willing to assume a higher risk than the average investor. The main hazard is linked to mean reversion in prices. History shows that this phenomenon tends to persist over time, particularly with broad asset classes.
But there's always a question of whether it'll prevail the next time, or how soon prices will revert to their mean after a large change in the market.
Such questions help explain why most investors find it hard to rebalance when market volatility is high. But that's why the expected return from rebalancing looks so rewarding.
0 comments:
Publicar un comentario