In my column last month, I started a discussion about rebalancing, noting that virtually every study about the sustainability of withdrawals from a portfolio has some rebalancing assumption imbedded within it. I focused on the importance of buying when markets are down and the difficulty practitioners often encounter in getting clients to take such action. Further, I showed fears of making things worse by being "early" when one rebalances may be overblown. Today I hope to show that promises of outsized additional return may be exaggerated as well.

The basic scenario consisted of a portfolio that targets a 50/50 split between A and B with $100,000 in each. A earns 4% while B loses 20%. The total is $184,000 (A=$104,000 and B=$80,000). To rebalance, $12,000 is moved from A to B so each has $92,000. If A again rises 4%, it will be worth $95,680. The portfolio will be back to $200,000 when B rises 13.39% ($104,320-92000/92,000). At that breakeven point, the actual price of B would still be off by 9.29%. If no rebalancing had occurred, A would be worth $108,160 and B would have had to add 14.80% to get the portfolio back to $200,000 (91,840-80,000/80,000).

In the scenario above, there is no indication of the time in between transactions. This has no bearing on the premise that buying when down can get the portfolio closer to whole faster than if no buy occurs even if there is more to the downside.

For simplicity, assume A earns nothing but B ultimately drops 40% from its start point. Rebalancing when we did (after a 20% decline) would result in a lower portfolio value than if no rebalancing occurred, but not much. In this case, just $2,500 ($157,500 vs. $160,000). The difference between rebalancing and not rebalancing is only the $10,000 buy after the 20% drop in B.

If you hadn't rebalanced at -20% but did at -40%, B would need to increase by 50% to get to whole if the only rebalancing happened at -40%. If rebalancing had occurred at the -20% point and again at -40%, B would need to increase by 54%. Either rebalancing choice requires less of a bounce than the 66.67% required with no rebalancing at all.

Clients that want their portfolio managed over the long-term will be faced with a never-ending series of chaotic changes in the value of their holdings. Looking beyond this isolated sequence illustrates other issues.

Bear markets occur on average every five years. When markets go lower, we are more inclined to be buying. When does inclination turn to action?  Does rebalancing really add to portfolio returns?

A number of studies over the years have indicated that rebalancing adds to returns in any periods and that there is no big difference between doing so quarterly, semi-annually or annually, in cost-free environments. In practice, particularly in taxable accounts, it makes little sense to incur costs associated with making trades simply for the sake of trading. Instead many practitioners use their judgment to override the calendar's prompt.

For instance, if we start with $100,000 each in A and B and when we get our prompt from the calendar, A=$100,000 and B=$103,000. Most would agree with the portfolio only up 1.5% and an A to B ratio of 50.74% to 49.26%, the situation does not beg for action.

Gobind Daryanani's 2008 Journal of Financial Planning contribution, "Opportunistic Rebalancing: A New Paradigm for Wealth Managers" provided a view of what varying thresholds yielded in the way of return enhancement. Daryanani established bands around each asset class within which the asset class could fluctuate without prompting a trade. For instance, if a target allocation to an asset class was 10% and a 20% band was set, that asset class could comprise between eight and 12% of the portfolio and no rebalancing would be indicated.

Daryanani, the developer of the highly regarded iRebal rebalancing software, wondered if how often one looked to see if the portfolio was out of balance affected his results. Among his conclusions were that rebalancing enhances returns and it was good practice to look more often but trade less frequently than one would with a traditional quarterly, semiannual, or annual rebalancing process.

Later in 2008, Marlena Lee at Dimensional Funds Advisors wrote "Rebalancing and Returns," in which she found a returns improved from rebalancing in many scenarios but less reliably than Daryanani's paper suggests. She found no optimal rebalancing rules that produced the highest returns on all portfolios in all time frames.

In the example where we start with $100,000 each in A and B, we keep a 50/50 target, and A earns nothing while B sinks 20%, the ratio of A to B is 55.6% to 44.4%. If having 55.6% is the trigger point, we would rebalance when A was 44.4% of the portfolio. Since we sold $10,000 of A to rebalance and A earns nothing, the portfolio would reach a trigger point when the total balance reached $202,702.70 ($90,000/.444).

A couple of paragraphs earlier, there was little reason to trade a $203,000 portfolio yet we're trading at just under that to restore the 50/50 mix and that action seems reasonable based on how far out of balance the portfolio became. That may be curious on its face but we are looking at two different time frames. In the first, B has risen 3% between looks, but we don't have any information about what happened in between the start and end points.

In the second there has been a notable drop in the value of B, but we aren't yet to the point where B has reached a value 3% higher than the start point when we rebalance. In fact, B is only about .18% higher than the start point, making an AB mix that did not rebalance worth roughly $200,184. By the time B is up 3%, the rebalanced portfolio would be worth $205,551 and a 49.31% to 50.69% ratio of A to B.  

Whether rebalancing like this is a plus for a client depends on several factors. It is true that the rebalanced mix produced a profit while both A and B had no change. In a taxable account, the ignored mix has a basis of $200,000, split evenly between A and B. The rebalanced portfolio has the same basis prior to the rebalancing at $202,720 but it is divided $90,000 to A and $110,000 to B. The amount of a gain incurred for tax purposes upon rebalancing depends on the accounting method used at that time, and the tax due is dependent on what else goes on the client's 1040 that year. Of course, if B continues to run, the rebalancing done at $202,702 would hurt returns from that point, as fewer shares would experience the rise.

Clearly, time frames affect one's view of the return benefits of rebalancing. I believe an increase in return is possible but I'm not sure that is the benefit that is the most reliable. I think the risk management elements of rebalancing are more compelling.

Whatever set of rebalancing rules one employs, it can encourage a behavior that makes sense. One largely maintains the exposures needed over the long term while buying a little of things at relatively low prices and selling a little at relatively high prices. The logic in that has great appeal but as important, it is a concept one can actually execute without depending on accurate market forecasts.

Having a rebalancing discipline can help address some of the emotional aspects of managing a portfolio. Properly educated about the matter, clients may be less inclined to get swept up in times of hype or dragged down in times of despair.

When conditions are particularly chaotic, many clients crave control. We cannot control when, or even whether, we are rewarded from risk taking but we can control the types and levels of risks that we bear and what we do when those risks reward or punish us.

Clients expected us to rebalance because we asked them to have that expectation. They knew that like good times, bad times don't last forever. In some cases, by sharing with them the mathematics I outlined, they could see that by buying they would get back to whole faster than sitting tight or worse, selling.

I think Lee summed it up well in her paper: "The primary motivation for rebalancing should not be the pursuit of higher returns, as returns are determined through the asset allocation, not through rebalancing. .... Rebalancing decisions should be driven by the need to maintain an allocation with a risk and return profile appropriate for each investor. The optimal rebalancing strategy will differ for each investor, depending on their unique sensitivities to deviations from the target allocation, transaction frequency, and tax costs."

Next month: Why what A and B actually are can be critical.

Dan Moisand, CFP has been featured as one of the America's top independent financial advisors by most leading financial advisor publications. He has spoken to advisor groups on five continents on topics such as managing investments and navigating tax complexities for retirees, retirement readiness, and most topics relating to the development of the financial planning profession. He practices in Melbourne, Fla. You can reach him at (321) 253-5400 or [email protected]