Rebalancing is considered a key element of buy-and-hold, to the point that buy-and-hold has become a shorthand for buy, hold, and rebalance. Investment advisor Rick Ferri made this clear in his interview with Morningstar: Buy, Hold, and Rebalance Works.
Rebalancing is also said to lead to “buy low, sell high.” The Bogleheads Wiki says:
[Rebalancing] is accomplished by transferring funds from higher-performing classes to lower-performing classes. While potentially counterintuitive, rebalancing ensures that investors “Buy Low” and “Sell High”.
Journalist Linda Stern wrote for Reuters How to force yourself to buy low and sell high:
Investors who wish they were better at buying low and selling high actually have a tool that can force them to do just that. It’s called “rebalancing” … …
Investment advisor Carl Richards wrote in New York Times How Rebalancing Takes Emotion Out of Investing:
Rebalancing is the only way I know of to give yourself the highest likelihood of buying low and selling high in a disciplined, unemotional way.
Just searching for “rebalance buy low sell high” returns many headlines like the ones I cited above.
Buy low sell high sounds good but it’s not true that rebalancing makes you do that. Not if we are talking about rebalancing at the highest level, between stocks and fixed income. Most of the time rebalancing between stocks and fixed income is more like sell high, buy higher.
Suppose you start with a portfolio well balanced between stocks and fixed income. When do you rebalance? When one does better than the other. That will put your portfolio off balance. So you rebalance by selling one and buying the other. When the tide turns, you do the opposite.
Directing new cash to the laggard can be thought of as putting new cash into the preset mix and then immediately selling assets that performed better. Other than trading cost and tax considerations, it’s the same as standard rebalancing.
Consider this hypothetical example for someone who rebalances annually at the end of the year:
| Year | Stocks | Fixed Income | Rebalance |
|---|---|---|---|
| 1 | +15% | +3% | sell stocks |
| 2 | +6% | +2% | sell stocks |
| 3 | +8% | -1% | sell stocks |
| 4 | -1% | -3% | sell stocks |
| 5 | +10% | +15% | buy stocks |
| 6 | +15% | +1% | sell stocks |
| 7 | +10% | +2% | sell stocks |
| 8 | +6% | +3% | sell stocks |
| 9 | -10% | +6% | buy stocks |
You sold stocks most of the time. By year 5, you bought stocks, but only at higher prices, because although stocks did worse than fixed income that year (+10% vs +15%), they were still up. You sold at lower prices in previous years, only to buy back at higher prices when you rebalanced at the end of year 5. That’s sell high, buy higher.
Stocks were down 10% in year 9. When you rebalanced you bought at lower prices than in year 8 but the prices were still much higher than the prices you sold at in years 1 through 7. Again sell high, buy higher.
If you are still saving for retirement by making new contributions every year, most of the time you will sell stocks when you rebalance, only to buy back later at higher prices. Sell high, buy higher.
Rebalancing is still a good way to control risk, but don’t think you are buying low and selling high when you rebalance between stocks and fixed income. When the expected return on stocks is much higher than the expected return on fixed income, more likely you will sell high and buy higher.
[Photo credit: Flickr user Kyle Steed]
Say No To Management Fees
If you are paying an advisor a percentage of your assets, you are paying 5-10x too much. Learn how to find an independent advisor, pay for advice, and only the advice.
Rodger says
Interesting idea but due to compounding, at year five although you are buying stocks at a higher price, proportionally its less of a percentage than what you disposed during years 1-4.
Net net you still sold at the high and this won’t catch up until you have many years of buying stocks through the cyclical low.
KD says
I have truly rebalanced only once and did not like the feeling at all. Since then I just adjust my new contributions to get back in balance. I can stick to this and it does not leave me with an icky feeling. I wish I could verbalize why I got the icky feeling first time.
Harry Sit says
Adjusting new contributions doesn’t get you out of this phenomenon. This post is inspired by this thread in the Bogleheads forum:
http://www.bogleheads.org/forum/viewtopic.php?f=10&t=147704
As I mentioned in the article, directing new cash to fixed income is equivalent to putting new cash into the preset mix followed by selling stocks.
KD says
So what should one do? Let the portfolio risk increase? Keep buying at set percentage of stocks and bonds no matter what allocation you end up each year? I agree that rebalancing is not buy low and sell high as it is sold to be. After you get this point what is actionable here? I am sorry I do not follow.
Harry Sit says
Understanding it correctly is the important first step. If you know you don’t buy low sell high when you rebalance, you won’t be upset or surprised when you sell high buy higher. Accepting it as the cost of controlling risk on the invested sum may very well be the end of the story.
The next step is questioning whether sell high buy higher is the right cost to pay. If you consider the money you will invest in the future, letting the risk increase or keeping it at a high level to begin with isn’t necessarily as bad as it sounds. See my reply to Josh below.
KD says
OK. Got it. I was at 90/10 allocation. I asked for a Bogleheads review. I was convinced a 80/20 would better as far as risk and volatility goes after it. I got a bunch of CDs instead of bond funds as the yield was guaranteed & better. This put me at 77/23 even after the stock market run up of last two years. I guess I am doing opposite of Vanguard glide path and your suggestion of looking at lifetime stock fund purchases and kind of front loading them in initial years. Hmmmm….
Josh says
I don’t really know if there is anything rational you can do about this. I would file this mentally under ‘stuff I have no control over’. No one invests their entire retirement savings all at once at the beginning of their career. Everyone makes ongoing contributions and should try to maintain an asset allocation that suits their needs and risk tolerance – not allocating appropriately or not saving enough makes a much bigger impact than what you describe here. The timing of when somebody makes the bulk of their contributions and when someone retires makes a huge difference in what their retirement balance winds up being, but what can any of us do about that?
Harry Sit says
You can do two things to mitigate this effect.
(1) Don’t rebalance every day (as in all-in-one funds), every time you contribute new cash, every quarter (many advisors do this to show value-add), or every year. Don’t sell so soon when you are still buying.
(2) Hold higher allocation to stocks for longer time in early years. Vanguard’s Target Retirement Fund does it by pegging to 90% stocks for the first 15 years. Focusing on the $50k already invested isn’t very meaningful when you are going to invest another $500k in the coming years.
Jeff says
I think the argument might hinge on the numbers you picked as your example. There are no 40% drops in a year, no multi-year runs of negative results. If we could count on reality being that benign over decades, I think it would make sense to be 100% stocks all the time and so never re-balance.
JohnInIowa says
any thoughts, Harry, re whether rebalancing is defeated by the increasing correlations between asset classes?
It used to be that rebalancing would mean buy low and sell high, but that works best if asset classes sometimes move in different directions. If they all move up together, or down together, then it seems to me that rebalancing won’t accomplish much.
Here’s a JP Morgan report on the increasing correlations of asset classes:
http://bit.ly/j0zais
It seems that everything is more correlated nowadays.
Harry Sit says
I believe rebalancing between equity asset classes such as US and international still works. I’m looking at 1-year return as of 9/30/2014: Vanguard Total Stock Market Index Fund +18%, Vanguard Total International Stock Index Fund +5%. There’s still enough difference. I don’t see any reason why the expected return on international equity must be a lot less than that on US equity.
KD says
I think we had a 30 yr period recently where bonds beat stocks. Wouldn’t such a scenario warrant that one does not follow what you say in your article about front loading stocks in the initial period of accumulation stage?
Harry Sit says
We did. It was long term Treasuries versus S&P 500 from October 1981 to 2011. The realized equity risk premium was slightly negative for one specific 30-year period which started when long term Treasury yield was 15%. If one knew Paul Volker would soon rein in inflation from 13.5% in 1981 to 3.2% in 1983, one would load up long term Treasuries yielding 15%. If someone did, they were very lucky.
If you haven’t read William Bernstein’s book The Ages of the Investor, I highly recommend it.
Michael Kitces says
Harry,
Nice post on this. I wrote something similar a few years back, which stirred up a great deal of Bogleheads discussion as well – see http://www.kitces.com/blog/is-rebalancing-supposed-to-enhance-returns-or-reduce-them/
Given the general positive tilt of equity drift, rebalancing will/”should” regularly REDUCE returns by pulling back on equities. That may be comfortable to fit someone’s risk tolerance, but it’s not necessarily a return enhancer. It only “works” for returns if the rebalancing trades actually go off at a “good” time, which ironically means that return-enhancing rebalancing is basically about getting the “market timing” right! 🙂 http://www.kitces.com/blog/is-passive-rebalancing-a-form-of-active-management/
Keep up the good work!
– Michael
Harry Sit says
Thank you Michael. It should be obvious, but I guess the volatile sideways “lost decade” together with good fixed income returns from declining interest rates created the impression that rebalancing between stocks and fixed income makes you buy low sell high.
Michael Kitces says
Harry,
Indeed. There was some interesting analysis by Michael Edesess (see http://www.advisorperspectives.com/newsletters14/Does_Rebalancing_Really_Pay_Off.php ) that makes the point that rebalancing IS expected to add value WHEN THE ASSET CLASSES HAVE SIMILAR LONG-TERM EXPECTED RETURNS IN THE FIRST PLACE (and thus volatility of non-perfectly-correlated assets creates buy-low-sell-high opportunities without introducing a return drag).
Those ‘comparable’ returns were true for the past decade (given low equity returns in the ‘lost decade’) but are obviously NOT the generalized case!
– Michael
K says
The article ignores the fact that over time, the price of stocks trends upward. For an investor making periodic contributions over a 40 year career, you can say it is “Keep buying higher and higher”. Yet it still works out in the end as long as prices continue to trend upward.
Toward the end of the accumulation phase, you usually don’t get to buy at the same prices as you did in the beginning except during market downturns. During those times, rebalancing does allow you to sell bonds and buy stocks at lower prices.
Harry Sit says
Just the opposite. It reminds people that because over time stock prices trend upward more than fixed income, when you rebalance you are selling stocks at prices you will buy at later. When a downturn finally comes, prices wind back a few years but not lower than the prices you sold at many years before.
John says
Harry,
Wouldn’t re-balancing help you lock in the gains? So during the Japanese run up in stocks or 2000 US dot com bubble one would have come out ahead by selling the stocks and having cash in bonds to buy during the fall.
I do see the challenge is that the crazy bull run lasted for several years so annual re-balancing would not have been a good idea. Not sure if re-balancing needs to also take into account some Additional factors like the the Shiller CAPE?
czeckers says
The “buy higher” phenomenon applies to all investors making periodic contributions over time, and is not just limited to those that rebalance. Assuming that security prices have an overall upward trajectory, an investor who is 10, 20, or 30 years into their investment carreer, will very rarely be able to buy securities at the same price as they did when they first began investing.
The only way to side step this issue would be to front load one’s retirement savings with a lump sum, but very few people have the funds to do that, thus we continue to make our contributions at higher and higher prices. The thing that matters most, is the price of the securities relative to the purchase price when it comes time to draw down the portfolio.
Y.G. says
“Sell high, buy higher”, while true, may seem to imply that it guarantees loss (or reduces gain). But it’s not the case.
Consider two asset classes, A and B, for simplicity held in a 50-50 portfolio. Suppose A gained 20% in year 1 while B stayed flat; A and B both gained 20% in year 2; A stayed flat while B gained 20% in year 3. If the portfolio is rebalanced after each year, A is first sold at 20% gain and then bought at 44% over its original price. This is, of course, “sell high, buy higher”. But meanwhile B is first bought at the original price and then sold at 44% gain. This is “buy low, sell high”, resulting in overall gain of 45.2% vs. 44% without rebalancing.
I can not rigorously prove to myself right now that “sell high, buy higher” is always accompanied by “buy low, sell high” — much less whether the gain from “buy low, sell high” always exceeds the loss from “sell high, buy higher” — but intuitively it rings true.
Did you consider both sides of rebalancing when you wrote that “it’s not true that rebalancing makes you [buy low, sell high]”?
Harry Sit says
The first order of rebalancing is between stocks and fixed income, with different expected returns. As a mental exercise consider rebalancing between stocks and cash with zero returns. There is no other side. Sell high buy higher clearly isn’t good. Now replace cash with fixed income. The other side still isn’t much.