Real Gains to be Made With the Alchemy of Rebalancing
Alchemy, the medieval forerunner to chemistry, was concerned with the transformation of matter, particularly with attempts to convert lead into gold. To some, investment management may appear to be a type of alchemy – a mysterious process that turns money into larger sums of money.
Most investor-alchemists attempt to create wealth by investing in winning securities.
In a recent paper, “Volatility Harvesting in Theory and Practice,” I argue that another way to create wealth is through rebalancing.
After the securities for a portfolio are selected and the weights determined, a rebalancing process will periodically buy or sell securities to re-establish the original weights after price changes have caused them to drift. Securities that have outperformed recently will be trimmed and securities that have underperformed will be added to.
The idea that this simple procedure could produce a higher portfolio growth rate may seem like alchemy but there is really no mystery. Portfolio rebalancing uses market timing instead of security selection as its primary method for adding value. The extra money created by rebalancing comes from the other side of the trade. Namely, traders that are following the current market trend by buying securities that have done well and selling those that have done poorly.
A debate has arisen over this idea. Critics claim this extra growth is illusory or that it requires prices to be mean reverting. Both sides of the debate have used mathematics, thought experiments and empirical examples to try to bolster their positions.
Proponents and critics generally agree rebalancing can create extra portfolio return. The differences of opinion lie in what assumptions are required and in estimating the potential value added and potential risks that such a strategy faces.
Interestingly, recent articles on rebalancing relate to a much older academic debate from the 1950s – detailed in William Poundstone’s book, Fortune’s Formula – about whether there is a scientific formula for making as much money as possible, as fast as possible.
This theory of optimal growth came from an unlikely source. Not from economists, but rather from scientists at Bell Labs. While building the infrastructure of the digital age, they became interested in questions of information transfer, gambling and investments. Later, in the 1980s and 1990s, these ideas were expanded and formalized by Robert Fernholz in his work on stochastic portfolio theory.
Unfortunately, because of its roots in science and engineering, this literature tends to be highly mathematical and is not accessible to many investors.
There are, however, three key findings we can draw from this body of work that relate to a rebalancing portfolio.
First, volatility, specifically the cross-sectional volatility of assets, creates the potential for a rebalancing portfolio to add value.
Second, even though this extra return does not rely on forecasting skill or an informational advantage, the portfolio manager still has a role.
Allowing weights to drift gives away the volatility return, while rebalancing harvests the return. Since this also introduces a friction to the portfolio in the form of transaction costs, careful attention must be paid to implementation details like liquidity, taxes and costs.
Third, for a rebalancing portfolio to outperform a drifting market-weighted portfolio over time, the capitalization distribution of the market must remain somewhat diverse.
For example, if the market eventually becomes concentrated into a single asset, rebalancing will be a losing strategy.
This last point is worth exploring more deeply. The risk and return of a rebalancing strategy are critically dependent on it.
Imagine if Cisco Systems had lived up to the growth projections that the market assumed for it in the year 2000, when it priced the firm at 120 times annual earnings.
From 1990 to 2000 the value of Cisco stock increased by 1,000 times. In March of 2000 analysts were projecting it would double its market cap to $1 trillion within a few years (the market cap was $134 billion as of Feb. 17.)
Set aside history for a moment; if the tech bubble had not burst perhaps Cisco could have acquired Apple and Google in their early years and grown those lines of business, eventually branching out into the energy industry, inventing a new battery and/or solar cell technology.
Maybe it could have continued to grow unabated for 50 years and eventually justified its year 2000 price to become a multi-trillion dollar company representing more than half of the market capitalization of the global stock market.
In this scenario a portfolio that rebalances would have underperformed a portfolio that let the concentration in Cisco build up. We don’t see these types of concentrations in the market. Social forces – such as antitrust laws, market competition and bureaucratic friction – tend to slow down the largest companies.
The requirement for the market capitalization distribution to be diverse is different from assuming that if prices go up they must come down (mean reverting or negatively autocorrelated).
It is also subtly different from assuming stocks have a “small company” risk factor which is the source of the outperformance.
In practice, these distinctions may not be all that important. Portfolios which rebalance have a strong historical track record versus those that let internal concentrations build up.
Of course, rebalancing strategies face a headwind when the market is concentrating, as in the technology and telecommunications bubble of the 1990s; holding a portfolio which is more diverse than the market can be painful at times.
Rebalancing away from outperforming assets and into underperforming assets is psychologically difficult for most investors. However, those who systematically avoid concentration risks can expect rebalancing to produce long-term outperformance.
Rebalancing is not a mysterious alchemical process. It is a simple buy-low, sell-high strategy that trades against the current market trend.
At the end of the story, rebalancing is about strategy, not alchemy. But to take advantage of its simple buy-low, sell-high message, requires manager discipline and experience.