Just Rebalance: Simple Approaches that Build Long Term Value
With markets currently under pressure from shocks in the energy sector and the COVID-19 outbreak, the Investment Research team reviewed what actions an investor might take to help ensure the short-term market gyrations don’t derail our investors’ long term objectives. What we have found is that that during times of uncertainty the simplest answer is one of the best—just rebalance.
Rebalancing is the process of restoring a portfolio’s current weightings to strategic targets according to a schedule or set of rules. Rebalancing is very important in maintaining target risk exposures, preventing unexpected events beyond the investor’s outlined risk tolerance, and maximizing the probability of achieving financial goals. Rebalancing may also help investors buy at lower prices and sell at higher prices as underperforming assets are adjusted upward and outperforming assets are adjusted downward, thereby locking in gains that can compound the portfolio over the long term.
However, there is a tradeoff, as rebalancing comes with increased transaction costs, monitoring costs, and taxable realizations. Rebalancing as a process can also be plagued by heuristics and suboptimal rules of thumb. In this analysis we hope to shed light on different strategies and how investors can optimize their plan to get the most out of their investment portfolios.
In our analysis, we reviewed a variety of rebalancing strategies and sought to understand the implications of each in terms of varying exposures, costs, and taxes.
To help illustrate our points, we used a simple 60/40 portfolio comprised of 60% US equities and 40% US government bonds, represented by the S&P 500 Index and the Bloomberg Barclays US Aggregate Treasury Index, respectively.
The Optimal Rebalancing Strategy
From our analysis it became clear that any rebalancing is more optimal than simply doing nothing from both a cost-benefit optimization and a risk mitigation perspective. We balance moderate transaction and tax costs with the benefit of a portfolio that has a higher likelihood of behaving according to our expectations while having increased opportunity to capitalize on short-term value opportunities. Of course, any of these strategies will involve some degree of investor cognitive bias, where the investor will be tempted to override the suggestion of the rebalancing signal in favor of their personal views. We hope our insights help alleviate the impact of these biases through a close look at the empirical data.
Overall, the optimal strategy will depend on the investor’s tolerance for risk factor drifts (less tolerant investors should select a strategy that more closely follows policy targets), the investor’s tax status (taxable investors should rebalance less frequently to reduce transaction costs), and the investor’s willingness to invest time in monitoring the portfolio (ambivalent investors should select a lower monitoring cost strategy). Ultimately there is no single approach that is appropriate for every situation, so Athena clients should discuss their needs with their portfolio manager. After a thorough discussion, the manager will develop and implement a plan that considers all relevant factors and outlines a detailed process to monitor, identify, and act upon potential opportunities.
The overarching lesson in this exercise is that by exercising vigilance, remaining aware of behavioral and cognitive biases, and closely following an established plan, investors can maximize their probability of achieving their financial goals.
Why Everyone Should Rebalance
Investing in less-than-perfectly-correlated asset classes over a long horizon will almost certainly lead to a wide variation in portfolio weightings, often to the detriment of the investor’s realized risk and return profile. Take our basic 60/40 portfolio for example—if you had invested $1 in this portfolio in 1973, you would have over $25 at the end of 2019. At first glance, this return is positive, given the variety in economic regimes over that horizon.
However, the issue with this requires a deeper dive. Over this period, the allocation between equities and fixed income varied significantly:
This scenario also has repercussions from a risk factor exposure perspective. By the end of 1999 (26 years after our inception of 1973), our equity exposure fell from its starting position of 60% to 45% then back to 65%. As a result, our portfolio was effectively overweight equities at the market peak, which caused the portfolio to underperform our 60/40 policy benchmark over the following period simply because the portfolio held more equity exposure at a time when equities performed poorly.
Similarly, at the trough of 2009 during the Great Financial Crisis, equities fell to just 35% of the portfolio, which ended up being far from ideal and at the bottom of the trough. As the portfolio rebounded, it lagged the 60/40 policy benchmark simply because it was significantly underweight equities during a time when equities performed quite well.
The lesson here is that by maintaining allocations to target risk factors throughout the business cycle, we are more likely to capture value-additive opportunities, and a portfolio is more likely to behave according to our expectations.
The next step is to identify the optimal rebalancing strategy.
Three Common Rebalancing Strategies
1) Do Nothing
This was the approach in our initial visualization—it is equivalent to setting up a portfolio in 1973 and forgetting about it until 2019. While this strategy has the benefit of zero transaction costs and zero monitoring costs, we have demonstrated the numerous implicit costs in the form of investment goal shortfall, missed value opportunities, and overexposure to certain risk factors at inopportune moments.
It’s important for investors to recognize that not rebalancing is an active decision in itself—by taking no action we are making a deliberate decision to minimize costs at the expense of allowing the portfolio to drift from its targets without restraint. The risk and return profile of the portfolio will change dramatically and, without exact monitoring, we could end up with unintended outcomes.
This strategy involves resetting our portfolio weightings on a regular schedule, such as quarterly or annually. While benefitting from limited monitoring costs, the strategy may suffer from increased transaction costs, depending on frequency. This strategy is agnostic to market events and helps avoid behavioral biases involved with trying to time the market. It also benefits from dollar cost averaging as valuations tend to mean-revert over a long-term horizon. For those who are indexing their portfolio to a calendar-rebalanced benchmark, this strategy can help mimic the benchmark’s rebalancing schedule.
If we had followed this strategy instead of the do-nothing strategy, we could have realized returns closer to our expectations, and maintained risk factor exposures closer to our targets. As an added bonus, our strategy of maintaining target allocations even outperformed the do-nothing strategy over this specific horizon, which is directly attributable to having maintained our equity exposure over time rather than letting it drift.
We can see that our portfolio weights were maintained throughout the investment horizon with only a handful of intra-quarter slips. This is clearly a more digestible outcome from a risk-factor exposure perspective.
However, following this strategy may cause investors to transact at the wrong time. For example, a quarterly rebalanced strategy would miss an instance where weights shifted dramatically within the quarter but returned to targets on their own by quarter end. For example, in the first quarter of 2009, if we had started with a 60% allocation to equities, equity underperformance would have sent that weight to below 55%, but the ensuing rebound returned equities back to around 58%.
What looked like only a 2% weight drift over the quarter (hardly a significant divergence at first glance) actually ended up being a missed value opportunity. If we had rebalanced at the bottom of that trough, we could have helped maintain our risk exposures while adding excess return. This is the major pitfall of calendar-based rebalancing—while efficient from a monitoring perspective, this strategy may rebalance when it’s not needed and ignores economic events that could lead to value generation in practice.
These strategies involve only changing portfolio weightings when weights change beyond a certain threshold. This strategy benefits from the fact that it is responsive to economic forces, only triggers transactions when necessary, allows some pre-defined target drift, and helps avoid behavioral biases due to its use of objective rules. It has the added benefit of potentially requiring fewer transactions to maintain policy targets than a calendar-based strategy, especially if the calendar strategy is frequent.
This strategy suffers from the need to pick a threshold before a rebalance is triggered—the decision to choose a +/- 2%, 5%, or 10% threshold (or anywhere in between) is largely a function of the investor’s tolerance for policy drifts, the investor’s tax status, the correlation of the portfolio’s investments, and trading costs (both explicit and implicit). Generally speaking, when trading is more expensive and when the investor is more tolerant of weight drifts, the allowable drift policy should be larger than if trading were less costly and the investor were less tolerance of weight drifts. In our example, we selected a threshold of +/- 5% as a reasonable starting point.
This strategy added a slight performance advantage over a simple quarterly rebalancing strategy due to the increased responsiveness to economic forces that caused weights to drift.
When we compare the intra-portfolio weight dynamics with those of the quarterly rebalanced portfolio, we can see that there are indeed fewer transactions—specifically there were 23 rebalances for this strategy vs. 187 for the quarterly strategy over a 47-year horizon. This will help alleviate the effects of costly transactions and realized gains by only rebalancing when it is economically necessary.
It is important to carefully consider the likelihood of the portfolio reaching the pre-set threshold. This is a function of how volatile and correlated the portfolio’s assets are; more highly correlated assets are less likely to result in large portfolio deviations, while highly volatile assets are more likely to meet a higher threshold. In our example portfolio, a 5% threshold resulted in 23 transactions over 47 years—about once every two years—while a 10% threshold resulted in only 6 rebalances over 47 years—once every 7-8 years, on average. This is likely too high a threshold as many opportunities within that horizon would have been missed.
Rebalancing is even more important during periods of extreme volatility as market participants begin to act irrationally. Unprepared investors can easily be caught in the confusion. We believe that a well-designed rebalancing strategy provides a solid framework for responding to market events in the most informed and disciplined manner, alleviating the impulse to overreact while helping the investor maximize the probability of financial success.