The Globe & Mail recently ran an article on rebalancing, featuring comments and opinions from some industry giants. The article ends with advice from John Bogle, who suggests that rebalancing is a personal choice. He discourages the practice because of the costs involved (i.e. brokerage fees + taxes). But perhaps it’s worth adding some context to this often-discussed topic.
The rebalancing challenge
The question of portfolio rebalancing presents a significant challenge. When a financial asset rises in price in the short-term, that asset has tended to continue rising for a period of time. (The same effect has persisted on the downside.) This is called the momentum effect and it has persisted for generations. Since stock and bond markets tend to spend more time going up than falling, this momentum effect can be a significant benefit over time.
Rebalance too frequently and you risk cutting off the benefits of this momentum effect. Failing to rebalance enough could result in overexposure to certain asset classes or sub-classes. The challenge, then, when faced with developing a rebalancing method is to capture as much of this momentum effect while keeping risk within a range that is suitable and reasonable.
Developing a rebalancing philosophy
Two years ago, HighView Financial Group tackled the task of formalizing a firm-wide philosophy around rebalancing. Our process began with a discussion of how each of our partners had previously implemented rebalancing for client portfolios. We then debated the pros and cons of various methods. And our goal was to agree on a given method and take it away to back-test the method – along with a couple of other alternate methods that we thought might be equally valid.
Never did we seriously consider rebalancing based on some pre-determined time frame. We quickly concluded that rebalancing should be triggered by deviations away from the target allocations of investments held in client portfolios. But how wide or narrow should we set these tolerance bands? And should they be based on a set number of percentage points away from the target or should ranges be relative to the size of the target allocation?
Finally, we concluded with an agreement on a methodology that we thought made the most sense. We thought that allowing an investment to move a full 20% from its target allocation struck the right balance. In other words, if you invest 70% in A and 30% in B, investment A would be allowed to fluctuate between 56% and 84% of the portfolio ( low: 70% x [100%-20%]; high: 70% x [100%+20%] ) while investment B would be allowed to fluctuate between 24% and 36% of the portfolio. And the ranges could be wider for taxable accounts to reduce the tax cost of rebalancing.
We quickly discovered that rebalancing is not a long-term return enhancer. And this makes sense. If the momentum effect persists over time and markets rise over time, the highest return is achieved by never rebalancing. But in practice, where investors are sensitive to risk this is not a realistic option for the vast majority of investors.
It’s important to understand that what works in the long-term doesn’t work in all periods. For instance, rebalancing every month didn’t fare as well longer-term but did work for the five years through February 2009 (the recent bear market bottom for most markets using monthly data). Using a five-component 60% stocks/40% bonds portfolio, not rebalancing at all resulted in a 5-year return of -1.3% per year through February 2009. Monthly rebalancing produced +4.5% per year. And yet longer term, not rebalancing produced the highest raw return for the thirty years through 2009.
For those who can tolerate higher risk and are comfortable underperforming for years in exchange for the benefit of outperforming over the long-term, perhaps no rebalancing at all is best. But for the rest – which make up the vast majority of investors – prudent rebalancing that strikes the right balance (between risk control and return potential) is key to achieving long-term investment goals. Just understand that it’s a risk-control strategy that may or may not boost returns.