Executive Summary
For many planners, passive and strategic investment management is the way to go. As such planners often point out, the evidence is mixed at best that any money manager can ever consistently generate alpha by outperforming their appropriate benchmark. Accordingly, as those planners advocate, the best path is to minimize investment costs as much as possible (since we know expenses we don't pay is more money we keep in our pockets), and investment allocation changes should only occur via a regular rebalancing process. Yet rebalancing does not always improve your returns; sometimes, it actually reduces future wealth. So if you try to come up with a "passive" rebalancing strategy that only enhances returns and doesn't ever reduce them... does that mean you're actually being active after all?
The inspiration for today's blog post comes from an email correspondence I had with a financial planner following the recent New York Times article about my analysis of how critical the final accumulation years are for a retirement goal, and my follow-up blog piece about how portfolio volatility really impact financial planning goals. During our correspondence, the planner suggested that in fact, volatility can be a good thing, as even his completely passive strategic investment approach can be benefited through systematic dollar-cost-averaging of new savings and regular rebalancing when there is volatility. To which I replied: "If you're taking steps to create value by making investment changes in the midst of volatility, isn't that active management?" Yet the planner insisted that his process was purely passive, and not active. I respectfully disagree.
To understand why, it may be helpful to look at an example. The table below shows two investments, A and B. Investment A is a volatile stock fund that generates an average annual growth rate of 10% over the 5-year period; Investment B is a relatively steady bond fund that generates 5% returns. If I invest $100,000 in each and just let the money ride, the stock portion grows to $161,115, and the bond portion rises to $127,570; my total account is up to $288,685.
Scenario 1.
| Inv. A | 22% | -15% | 19% | 2% | 28% |
| Growth of $100k | $122,000 | $103,700 | $123,403 | $125,871 | $161,115 |
| Inv. B | 3% | 7% | 4% | 5% | 6% |
| Growth of $100k | $103,000 | $110,210 | $114,618 | $120,349 | $127,570 |
Of course, this includes no rebalancing! If we're going to include rebalancing at the end of each year - to "take advantage" of the volatility - we end out with very different results, as shown below.
Scenario 2.
| Inv. A | 22% | -15% | 19% | 2% | 28% |
| Growth of $100k | $122,000 | $95,625 | $128,520 | $122,828 | $159,532 |
| Inv. B | 3% | 7% | 4% | 5% | 6% |
| Growth of $100k | $103,000 | $120,375 | $112,320 | $126,441 | $132,113 |
Now, we end out with a total value of $159,532 + $132,113 = $291,645! With annual rebalancing, we created almost $3,000 of additional wealth, with a purely "passive" strategy. Isn't rebalancing great when there's volatility (as we see with the big stock decline in year 2)?
To be fair, it is true that the results here were improved by the volatility. For instance, if the stock and bond allocations generated the same 10% return in stocks and 5% return in bonds and we rebalanced every year, we'd end out with some very different results:
Scenario 3.
| Inv. A | 10% | 10% | 10% | 10% | 10% |
| Growth of $100k | $110,000 | $118,250 | $127,119 | $136,653 | $146,902 |
| Inv. B | 5% | 5% | 5% | 5% | 5% |
| Growth of $100k | $105,000 | $112,875 | $121,341 | $130,441 | $140,224 |
Without the volatility, our results actually got worse! Now our final combined wealth is only $287,126; rebalancing produced results that were below the rebalancing-with-volatility scenario. In fact, we actually finished with almost $1,500 less wealth than the first scenario, where we did no rebalancing at all! So not only is volatility good for rebalancing, but apparently rebalancing is actually "bad" without volatility?
Well, yes and no. It is true that at its core, rebalancing when investments only grow in a straight line actually does produce less wealth. For a very simple reason - if the investments both rise as a steady pace, simply with different returns, then all rebalancing does is systematically sell the higher return asset to buy more of the lower return asset; not surprisingly, then, if you consistently sell the investment that generates better returns for the one that generates inferior returns, you'll end out with less wealth, all else being equal.
But does that mean volatility is "good"? Well, not necessarily. At all, if we look back at the second example - rebalancing with volatility - where did the "value" actually come from? Why is it that the rebalancing with volatility resulted in higher returns, while the rebalancing with no volatility resulted in lower re
turns, compared to no rebalancing at all.
In reality, the rebalancing "worked" for one major reason: after year 1, we sold down the stocks that just had a 22% return, right before stocks well 15%, and bought a bunch of bonds that returns 7%. Then after year 2, we sold the bonds that had run up 7%, and bought more stocks right before they recovered by 19%! In other words, rebalancing worked because we... sold high, and bought low, and in the next year, we sold (something else) high, and bought (something else) low. Simply put, rebalancing with volatility works because it is a strategy that tries to buy low and sell high, as with any other active management strategy. Similarly, it doesn't "work" when there's no volatility, because now there are no buy-low-sell-high opportunities!
So does that mean volatility is actually a "good" thing? Well, I guess it is, as long as you're willing to try to actively buy low and sell high? Of course, you do have to get the timing right! Otherwise, you may get a scenario like this one:
Scenario 4.
| Inv. A | 22% | -15% | 19% | 2% | 28% |
| Growth of $100k | $122,000 | $103,700 | $127,276 | $129,822 | $157,835 |
| Inv. B | 3% | 7% | 4% | 5% | 6% |
| Growth of $100k | $103,000 | $110,210 | $111,233 | $116,795 | $130,707 |
In this final scenario, the client finishes with only $288,542, slightly less than the original no-rebalancing scenario. Yet in the chart above, there was still rebalancing! It's just that the rebalancing came every other year (i.e., after years 2 and 4), rather than annually. In this case, the client's rebalancing trades didn't occur in a manner that allowed him to sell the stock peaks to buy bonds, and then buy back stocks after the dip; instead, by buying every other year, the client ended out riding the stock returns up and down, and then rebalancing afterwards. In other words, if the timing of the rebalancing doesn't create actual instances of selling high and buying low, there's no value creation to rebalancing, even if the markets are volatile in the meantime.
The point here is not that therefore, you should rebalance every 1 year and not every 2 years. Their returns are just examples. They could have been annual returns, or quarterly returns, or monthly returns; the volatility could have been larger or smaller. The point is simply that in order for rebalancing to generate value, it has to re-create buy-low-and-sell-high opportunities, in the same manner as any other active management strategy. It is sells stocks to buy more bonds in the middle of a stock rally, it still loses; if it buys stocks in the midst of a multi-year decline, it hurts more than it helps. Only a sequence of rises and declines that create buy-low-and-sell-high opportunities allow rebalancing to succeed with volatility, and even then only when you get the "timing" right.
So what do you think? Is rebalancing a form of active management because it only "works" when you buy low and sell high? Does that mean we should spend more effort focusing on the timing of when to pull the trigger on rebalancing? And if we focus on the timing to rebalance, does that make us market timers? Or do you use a different definition for market timing?
I’ve always taken a methodical approach to rebalancing (usually one a year) but your post makes me question what the best approach is. Since I truly believe that no one, including myself, can time the markets effectively over time, I have to consider what is the next best option.
What about rebalancing in pre-set bandwidths, for example, when the stock allocation in the portfolio gets 5% higher than its target? This ensures that there has been some volatility in the market but can’t ensure you haven’t sold stocks that are still on the rise. In addition, in very volatile markets, you could be rebalancing frequently and driving up transaction and tax costs.
Another consideration is that there is some value in rebalancing in order to maintain target percentages that meet the investor’s risk profile.
I’ll have to think about this some more.
Rob,
There was a good article about rebalancing strategies published by Gobind Daryanani a few years ago in the Journal of Financial Planning. You can see a copy of it at http://www.tdainstitutional.com/pdf/Opportunistic_Rebalancing_JFP2007_Daryanani.pdf
The basic gist of the conclusions was that asset classes only tend to deviate “so much” from each other before market forces tend to push them back together again. So from that research perspective, the optimal strategic was not to rebalance on a purely calendar-based time horizon, but instead to do so “opportunistically” at a point where the asset classes had deviated so far the trend would be likely to reverse sooner rather than later (at about 20% relative deviation from target weighting).
Of course, from the perspective I’m presenting here, I think this “opportunistic” nature of rebalancing just emphasizes that in the end, it creates value when it is conducted as a well-timed active management strategy!
Notably, I don’t think that makes rebalancing “bad” in any way. I just think it raises the question about what it means to be “passive” in today’s world, and how you should define “active” and “market timing”.
Thanks, Michael. I’ll check out the JFP article and will reply if I have anything to add.
Yes, this is market timing.
Yes, we should all spend more effort figuring out when market timing is a good thing and when it is a bad thing.
The taboo on market timing is causing financial misery for millions. There is no magic to going with a stable stock allocation. That’s a choice, and like all the other possible choices, it is in some circumstances a good idea and in other circumstances a bad idea.
Instead of asking “Is this market timing?” we should be asking “Is the reason for engaging in market timing here a good one or not?”
It’s not possible to hold intelligent discussions on asset allocation until we get over the idea that there is something “bad” about market timing just because it is market timing. We should always aim to be going with the best stock allocation, whether doing that requires market timing or not.
Rob
I think the problem originates with having imperfect/incorrect definitions of both passive and active management. They are generally thought of as polar-opposites, but it might be better to think of them as being on a spectrum with various shades of of each other.
I see pure active management as the belief that one can outperform a particular market index because they believe that the market is inefficient AND they can exploit this inefficiency. The focus is not on the investor’s risk/return needs, it’s really all about beating the given market/index.
Pure passive investing, IMO, is a belief that it is a losing bet to try and pick underpriced securities and to know when to buy and when to sell them. Passive investing is more focused on getting the investor close to market returns by lowering investment costs and accounting for the investor’s risk/return needs.
If stocks perform well and equities make up, say, 80% of a portfolio that has a target of 70%, then it’s not unreasonable to take some risk off of the table with a rebalance. I think this strategy as being client-focused. It is does not necessarily come from a conclusion that equities are overvalued.
Is rebalancing market timing? In a sense, I suppose it has some elements, but I think the key difference is that the decision is being made from the standpoint of the client’s overall risk/return profile and not from an analysis of of the equity markets at that moment.
Michael:
We’ve been thinking about the same topics quite closely, though sometimes you’ve been a couple weeks ahead of me. Though I knew about them before, I didn’t get around to read all your recent blog posts about market timing and active management and tactical asset allocation until last week. I think they are really helpful and I mostly find myself in agreement with your conclusions. I’m writing a research paper now in which I plan to cite several of your blog posts related to this topic, as you help to clearly define a number of important issues.
On this issue of rebalancing becoming an active strategy when it seeks to sell high and buy low, this is a close corollary of a discussion we’ve been having at the Bogleheads Forum about valuation-based tactical asset allocation for the long-term investor. Though the thread now has 159 posts, I would invite you to have a look:
http://www.bogleheads.org/forum/viewtopic.php?t=67093
As well, let me take your analogy one step further and ask: “Is Doing Anything Other Than Buying the Riskless Asset A Form of Active Management?”
If someone buys stocks, they are taking on risk. But they are making some type of active decision that over a sufficiently long period of time, they will be compensated for this risk by earning a higher return. I’m not sure yet, but I think that this idea can be attributed to Edgar Lawrence Smith’s 1924 book, _Common Stocks as Long-Term Investments_.
If one decides in advance, to rebalance at some fixed time interval (or perhaps at a percentage deviation from a target), then IMHO there is simply no way that a reasonable person could call that *timing*, in any sense of the word.
It is in fact, the very opposite of market timing; it is a priori strategic PLANNING, and doing so without a view to trying to figure out what the market will do, or when, but merely RESPONDING post facto to whatever reality already occurred, and bringing one’s allocation back to the point that planned for in advance (or the new point required by life circumstance — but again, with a view towards working a pre-existing plan, and not by trying to predict the market.
Michael, I can easily see the scurrilous and disreputable Robert “Hocus” Bennett making such a ridiculous claim just to make a splash and generate some eyeballs, but I honestly thought you were above such sophistry.
C’mon. You know better.
DripGuy,
I actually had started writing a response to you here, but found it was running so long, I just turned it into another blog post itself, which you can read at:
http://www.kitces.com/blog/index.php?/archives/114-Is-Rebalancing-Supposed-To-Enhance-Returns,-Or-Reduce-Them.html
Thanks for the dialogue!
I can easily see the scurrilous and disreputable Robert “Hocus” Bennett making such a ridiculous claim just to make a splash and generate some eyeballs
Oh, my!
Rob
Given that the title of your new article itself is a non-sequitur (The purpose of re-balancing as practiced by Passive Investors is NEITHER to ‘enhance’ nor certainly is it to ‘lower’ returns, Michael.
It is to return the allocation to the PLAN, such that a relatively consistent, controlled risk profile is maintained as intended, prior to ANY particular moves up or down by the markets.
You are much too intelligent, well-schooled, and experienced to be playing such games as “Are you still beating your wife?”
I am terribly disappointed in you, not that that probably means anything to you, especially compared to generating some page views and perhaps other sorts of notice.
It’s just sad to see, especially in an era of too much bombast and misinformation already.