Why 5% Rebalancing Bands Are Wrong

[Double Dog]

A common rule of thumb is to rebalance your portfolio whenever it gets more than 5% out of whack. If your goal is 80/20 but you're now at 75/25, it's time to rebalance. But after crunching some numbers, I believe 5% rebalancing bands are far too lax. I calculate that 2-3% bands would be more appropriate for most investors. The size of the bands should also vary depending on your specific asset allocation. Please take a look at the numbers below and let me know what you think.

Take this example: Your AA is 80/20. You have $80000 in stocks and $20000 in bonds. Then there's a bull market and stock prices increase by 25%. Would you rebalance? I certainly would - 25% is a big gain! I'd want to take some of that gain and plow it back into bonds. But if you check the math, you would have $100000 and $20000 in bonds, for a new AA of 83/17. A 25% run-up in the price of stocks only results in a 3% change in asset allocation. What?! If you're using 5% rebalancing bands, then you won't rebalance.

Here's a table I made that shows the resulting change to your asset allocation for a change in your stock or bond value, while the value of the other asset type remains unchanged. Changes greater than 5% are shaded red.



The top table is for changes in stock value while bonds are unchanged, and the bottom table is for changes in bond value while stocks are unchanged. Say your AA is 90/10, and your stocks decline 30% while your bonds remain unchanged. You would look at the "stock price change" table for 90% stock allocation and -30% stock price change, and you'll see that would lower the stock portion of your AA by only -3.7%.

Observations: At a 90/10 AA, virtually nothing would ever cause you to rebalance if you're using 5% bands. I'm guessing many investors here have an AA somewhere between 60/40 and 80/20. For them, it takes price movements of about 20-25% to trigger a rebalancing if you're using 5% bands. If your AA is in this range, and you want to rebalance whenever an asset price moves more than about 10%, then your rebalancing bands should be roughly 2.1%.

[siamond]

A common rule of thumb is to rebalance your portfolio whenever it gets more than 5% out of whack. If your goal is 80/20 but you're now at 75/25, it's time to rebalance. But after crunching some numbers, I believe 5% rebalancing bands are far too lax. I calculate that 2-3% bands would be more appropriate for most investors. The size of the bands should also vary depending on your specific asset allocation. [...]

You're on something... Please check this old thread with a similar discussion and a proposal to make things more square:
viewtopic.php?t=186203

[LeslieSmiley]

i don't think that is wrong because people have different risk tolerance. 25% to you is big, but to some is meaningful but not big. so to those people, 2-3% band would be too small.

[corey407woc]

ive wondered what the correct percentage of bands for rebalancing is and apparently its 20%.

https://www.kitces.com/blog/best-opport ... hresholds/

ive also played around with portfoliovisualizer.com with 20 % and seems to be the best

[UpsetRaptor]

The wider your rebalancing bands, the higher probability of a rebalancing bonus. The trade-off is your portfolio can drift further off your desired AA. There's really no right or wrong.

[InvestingGeek]

Great analysis, thanks!

[keystone]

Given all of the recent threads on rebalancing and all of the underlying uncertainty and anxiety, I wonder if most of us (myself included) would simply be better served by investing in Lifestrategy or target date funds to the extent that it is feasible (e.g. in tax advantaged accounts).

What always held be back was the extra fee, but I'm starting to believe that the extra fee will be mitigated by the ease of managing the portfolio and not having to rebalance or second guess when it is a good time to rebalance.

[DB2]

Given all of the recent threads on rebalancing and all of the underlying uncertainty and anxiety, I wonder if most of us (myself included) would simply be better served by investing in Lifestrategy or target date funds to the extent that it is feasible (e.g. in tax advantaged accounts).

What always held be back was the extra fee, but I'm starting to believe that the extra fee will be mitigated by the ease of managing the portfolio and not having to rebalance or second guess when it is a good time to rebalance.

I've come around to believing this more and more as well (for myself included).

[HappyJack]

You might consider using a 20% rebalancing band with a 10% tolerance band. When you re-balance, only rebalance to get inside the tolerance band. This allows an asset class to run a little bit more.

[randomguy]

Take this example: Your AA is 80/20. You have $80000 in stocks and $20000 in bonds. Then there's a bull market and stock prices increase by 25%. Would you rebalance? I certainly would - 25% is a big gain! I'd want to take some of that gain and plow it back into bonds. But if you check the math, you would have $100000 and $20000 in bonds, for a new AA of 83/17. A 25% run-up in the price of stocks only results in a 3% change in asset allocation. What?! If you're using 5% rebalancing bands, then you won't rebalance.

And what difference does it make? Lets see
1987-2020 (i.e. portfolio visualizer dataset), 100k invested 80/20

5% band: 2.053 million, max drawdown of 41.53%
2% band: 1.99 million, max drawdown of 42.11%

Looks to me to be about zero difference. You are looking for too much precision in a problem. The differences between 80/20, 85/15, and 75/25 just aren't big enough to worry about.

[123]

And a contrary view might say that being over regimental about what you do may not necessarily lead to improved results. If someone wants to be regimental and "perfect" so be it. Even the act of rebalancing is somewhat random depending on the calendar day you do it (unless daily). One could even argue that it is better to rebalance using moving averages of some constant duration to remove some of the daily random error of the process.

[Thesaints]

If one rebalances for 5% deviations, the recent market gyrations would have not yet triggered a rebalancing.
This could be good, or bad, but people should be aware of it.

[jjustice]

"I believe 5% rebalancing bands are far too lax. I calculate that 2-3% bands would be more appropriate for most investors." Double Dog

I'm pleased to see your argument, Double Dog. I agree with you. In my own case, for a 47.5% stock target, I sell .5% of my total at 50% stock and buy .5% at 45.125%. (47.5/50 = 45.125/47.5) This policy led me to trade some Traditional for CREF Stock yesterday in my 403(b). The last time I rebalanced to stock was 12/24/2018. I've sold stock a few times in the interval.

John

[jrbdmb]

If one rebalances for 5% deviations, the recent market gyrations would have not yet triggered a rebalancing.
This could be good, or bad, but people should be aware of it.

This is the reason I felt that for my comfort level 5% bands were a bit too wide. Currently at 2.5%, rebalanced once already and not too far off rebalancing again, which seems OK since S&P is about 18% off it's high.

[Pikel]

This equation changes a lot if your money is in taxable accounts. It also ignore that people are generally either accumulating or withdrawing. I would just contribute more to bonds as an accumulator, as a retiree I would make withdrawls from stocks.

[vineviz]

Take this example: Your AA is 80/20. You have $80000 in stocks and $20000 in bonds. Then there's a bull market and stock prices increase by 25%. Would you rebalance? I certainly would - 25% is a big gain! I'd want to take some of that gain and plow it back into bonds. But if you check the math, you would have $100000 and $20000 in bonds, for a new AA of 83/17. A 25% run-up in the price of stocks only results in a 3% change in asset allocation. What?! If you're using 5% rebalancing bands, then you won't rebalance.

And what difference does it make? Lets see
1987-2020 (i.e. portfolio visualizer dataset), 100k invested 80/20

5% band: 2.053 million, max drawdown of 41.53%
2% band: 1.99 million, max drawdown of 42.11%

Looks to me to be about zero difference. You are looking for too much precision in a problem. The differences between 80/20, 85/15, and 75/25 just aren't big enough to worry about.

Absolutely. The the primary function of rebalancing is risk control, and this is equally well achieved with absolute rebalancing bands anywhere in the range of 0% to 12% for most common asset allocations. And in the presence of either transaction costs or taxes, bands in the 8% to 12% range are clearly preferable because the exponentially reduced number of transactions.

[Double Dog]

If one rebalances for 5% deviations, the recent market gyrations would have not yet triggered a rebalancing.
This could be good, or bad, but people should be aware of it.

Yes, this was my motivation for looking into the question. Like everyone, my stocks are waaaay down these past few weeks. I thought for sure my AA would be far out of whack, and I'd need to sell some bonds and buy some stocks. I was very surprised to find that my 80/20 has only shifted to 77/23.

[Thesaints]

And what difference does it make? Lets see
1987-2020 (i.e. portfolio visualizer dataset), 100k invested 80/20

5% band: 2.053 million, max drawdown of 41.53%
2% band: 1.99 million, max drawdown of 42.11%

Looks to me to be about zero difference. You are looking for too much precision in a problem. The differences between 80/20, 85/15, and 75/25 just aren't big enough to worry about.

Doesn't sound right. Do the number include transaction costs ?
Otherwise, it could be simply the effect of having a higher allocation to stocks in a period of stocks outperformance.
Have you tried running "no rebalance at all" ?

Absolutely. The the primary function of rebalancing is risk control, and this is equally well achieved with absolute rebalancing bands anywhere in the range of 0% to 12% for most common asset allocations. And in the presence of either transaction costs or taxes, bands in the 8% to 12% range are clearly preferable because the exponentially reduced number of transactions.

I'm not sure. Where do those numbers came from ?

[rich126]

Maybe the numbers say different things but it seems like people are over complicating the supposed simplicity of a 2/3/4 fund portfolio. Doing it once a year probably doesn't change things much (returns or max dd), at least based on a few scenarios I glanced at on the portfolio visualizer site.

[vineviz]

Absolutely. The the primary function of rebalancing is risk control, and this is equally well achieved with absolute rebalancing bands anywhere in the range of 0% to 12% for most common asset allocations. And in the presence of either transaction costs or taxes, bands in the 8% to 12% range are clearly preferable because the exponentially reduced number of transactions.

I'm not sure. Where do those numbers came from ?

It's easy to backtest various bands in PortfolioVisualizer to see for yourself.

Here is what two 80/20 portfolios look like with various rebalancing bands (x-axis) and the resulting portfolio standard deviation (y-axis).



If you looked at other measure of volatility (max drawdown, minimum 5-year rolling return, etc.) you'd see virtually identical pictures: across a wide range of rebalancing bands, there's no statistically significant difference in average volatility.

[Thesaints]

It's easy to backtest various bands in PortfolioVisualizer to see for yourself.

It is a back test on a single period ? How do we know that is not a special period ?

resulting portfolio standard deviation (y-axis).

Which standard deviation ?

As rebalancing becomes less and less frequent, how do we say that a given portfolio is, let's say, 60/40 and not 65/35 ? Maybe if my rebalancing threshold is 70/30, a portfolio which was 60/40 at start spends a lot more time being near 65/35.

It is not a straightforward problem; there are many subtleties in play.

[rbaldini]

If one has decided that having bonds anywhere from 15-25% is just fine, then who cares how often one rebalances? You are taking for granted that a big market movement must of course lead one to rebalance. One can disagree with this premise. What if it's about being "close enough" to the target allocation, not how often one reacts to market movements? (Hint: it's the former.)

[dcabler]

A common rule of thumb is to rebalance your portfolio whenever it gets more than 5% out of whack. If your goal is 80/20 but you're now at 75/25, it's time to rebalance. But after crunching some numbers, I believe 5% rebalancing bands are far too lax. I calculate that 2-3% bands would be more appropriate for most investors. The size of the bands should also vary depending on your specific asset allocation. Please take a look at the numbers below and let me know what you think.

Take this example: Your AA is 80/20. You have $80000 in stocks and $20000 in bonds. Then there's a bull market and stock prices increase by 25%. Would you rebalance? I certainly would - 25% is a big gain! I'd want to take some of that gain and plow it back into bonds. But if you check the math, you would have $100000 and $20000 in bonds, for a new AA of 83/17. A 25% run-up in the price of stocks only results in a 3% change in asset allocation. What?! If you're using 5% rebalancing bands, then you won't rebalance.

Here's a table I made that shows the resulting change to your asset allocation for a change in your stock or bond value, while the value of the other asset type remains unchanged. Changes greater than 5% are shaded red.



The top table is for changes in stock value while bonds are unchanged, and the bottom table is for changes in bond value while stocks are unchanged. Say your AA is 90/10, and your stocks decline 30% while your bonds remain unchanged. You would look at the "stock price change" table for 90% stock allocation and -30% stock price change, and you'll see that would lower the stock portion of your AA by only -3.7%.

Observations: At a 90/10 AA, virtually nothing would ever cause you to rebalance if you're using 5% bands. I'm guessing many investors here have an AA somewhere between 60/40 and 80/20. For them, it takes price movements of about 20-25% to trigger a rebalancing if you're using 5% bands. If your AA is in this range, and you want to rebalance whenever an asset price moves more than about 10%, then your rebalancing bands should be roughly 2.1%.

There is also a path dependency when you do a backtest with bands. Depending on the exact starting date of the backtest and the exact starting AA, you can get slightly different results. There may be a longer, or shorter, time between the start of the backtest and the first rebalance trigger. You'd probably have to do this manually, but might make sense to do 3 backtests. For example, AA (eg 60/40), max AA(65/35), min AA(55/45) and get something of an envelope of results if using 5% bands...

[dcabler]

If one rebalances for 5% deviations, the recent market gyrations would have not yet triggered a rebalancing.
This could be good, or bad, but people should be aware of it.

That depends on where your AA had drifted to going into the season of market gyrations. Two of my smaller accounts triggered last week, but my larger account hasn't, but it's close...

[tibbitts]

It's interesting that all the Bogleheads who cite Bogle's stance on International not being necessary as justification for domestic-only allocations, don't seem to similary support his position that reblancing is never necessary (and equally likely to be beneficial or harmful if you do it.)

[dknightd]

You can pick any rebalancing bands you want. To consider changing them during market fluctuations is a form of market timing.

[jebmke]

This equation changes a lot if your money is in taxable accounts. It also ignore that people are generally either accumulating or withdrawing. I would just contribute more to bonds as an accumulator, as a retiree I would make withdrawls from stocks.

This is a good point.

We have been in withdrawal mode so sweeping dividends to cash and spending. So every year we are liquidating ~2% of our equity through spending. In the first couple years, I was selling bonds, buying stock (and also harvesting losses) since I had retired in December, 2007 with no pension or SS for at least 10 years.

I'm comfortable with wide bands. My target equity is 40% with -5%/+60% as the range.

[vitaflo]

Many moons ago I read an article that backtested rebalancing bands. The TLDR was that the band that resulted in the biggest benefit was 3.5%. However, this assumed new investment was always put towards whatever assets drifted under your AA and that dividends were also places there as well (not reinvested directly). It also assumed you rabalanced as soon as you hit the rebalancing band (IE, you don't wait until the end of month, etc).

I found the work compelling and have used it since.

[randomguy]

And what difference does it make? Lets see
1987-2020 (i.e. portfolio visualizer dataset), 100k invested 80/20

5% band: 2.053 million, max drawdown of 41.53%
2% band: 1.99 million, max drawdown of 42.11%

Looks to me to be about zero difference. You are looking for too much precision in a problem. The differences between 80/20, 85/15, and 75/25 just aren't big enough to worry about.

Doesn't sound right. Do the number include transaction costs ?
Otherwise, it could be simply the effect of having a higher allocation to stocks in a period of stocks outperformance.
Have you tried running "no rebalance at all" ?

But it is right. Or I haven't found the bug in PV yet:). There are no transaction (selling or taxes) costs. Adding them in would make the 2% case worse. And yes the slight outperformance in this case is by holding a slightly higher average stock allocation (i.e. call it 82.5 versus 81%). If you looked at a dozen 30 year periods, there might be one where 2% gives slightly better numbers. But that isn't the point. The point is that this is a choice that just doesn't matter. You can rebalance at 2%,5%,10%, every 6 months, every year, and you aren't going to end up with a meaningful difference. You are deluding yourself if you think that holding 80/20 versus 85/15 or 75/25 is the absolute right AA for you. It is a level of precision that isn't possible to have. If it satisfies your OCD to rebalance at 2% or every month. Go for it. It is one of those choices that just doesn't matter.

[asset_chaos]

Rebalancing bands depend on the cost of rebalancing.

[Thesaints]

I'm not saying there is a bug in PV. I'm saying people misinterpret results.
Aren't you surprised that rebalancing less often gave better results ?

Personally, I'm not, because the portfolio rebalanced less often was heavier on stocks than the other in a period when stocks outperformed bonds.
But that has nothing to do with rebalancing per se. It is simply a consequence of different asset allocations.
I can call a portfolio "50/50", but if I never rebalance over 30 years, on average that will be a 60/40 portfolio, even if it did start as a 50/50.

[randomguy]

I'm not saying there is a bug in PV. I'm saying people misinterpret results.
Aren't you surprised that rebalancing less often gave better results ?

Personally, I'm not, because the portfolio rebalanced less often was heavier on stocks than the other in a period when stocks outperformed bonds.
But that has nothing to do with rebalancing per se. It is simply a consequence of different asset allocations.
I can call a portfolio "50/50", but if I never rebalance over 30 years, on average that will be a 60/40 portfolio, even if it did start as a 50/50.

Nobody is surprised by these results. They are what ever rebalancing study for the last decade has shown. Rebalancing just doesn't matter that matter that much with any reasonable (i.e. no going 20 years without rebalancing) scheme.

[MnD]

Take this example: Your AA is 80/20. You have $80000 in stocks and $20000 in bonds. Then there's a bull market and stock prices increase by 25%. Would you rebalance? I certainly would - 25% is a big gain! I'd want to take some of that gain and plow it back into bonds. But if you check the math, you would have $100000 and $20000 in bonds, for a new AA of 83/17. A 25% run-up in the price of stocks only results in a 3% change in asset allocation. What?! If you're using 5% rebalancing bands, then you won't rebalance.

And what difference does it make? Lets see
1987-2020 (i.e. portfolio visualizer dataset), 100k invested 80/20

5% band: 2.053 million, max drawdown of 41.53%
2% band: 1.99 million, max drawdown of 42.11%

So the thread title should be "Why my hunch about 2% rebalancing bands was wrong".
$63,000 is a lot of money even for those of us with $X million dollar portfolios.
With a 5% rebalancing threshold, I rebalanced twice in 08/09, the 2nd times within a few days of the 2019 bottom. In 2018 I rebalanced on December 24 - the bottom. Lucky in that I could have just missed the 2nd rebalance in 09 or the entire one in 2018 but a 2% rebalancing band IMO will results in a very large number of transactions with most being way too soon and a lot of whip-sawing.

And folks that are just now formulating or reformulating their rebalancing frequency rules on the fly in the midst of a global pandemic are really off the rails. Good prospects for Vanguard Personal Advisory Services.

[Portfolio7]

I'm not saying there is a bug in PV. I'm saying people misinterpret results.
Aren't you surprised that rebalancing less often gave better results ?

Personally, I'm not, because the portfolio rebalanced less often was heavier on stocks than the other in a period when stocks outperformed bonds.
But that has nothing to do with rebalancing per se. It is simply a consequence of different asset allocations.
I can call a portfolio "50/50", but if I never rebalance over 30 years, on average that will be a 60/40 portfolio, even if it did start as a 50/50.

Nobody is surprised by these results. They are what ever rebalancing study for the last decade has shown. Rebalancing just doesn't matter that matter that much with any reasonable (i.e. no going 20 years without rebalancing) scheme.

+1. Of all the things one can be doing to improve their results, or limit their volatility/risk, playing with rebalancing bands and frequency have to be at the very bottom of the list. Make sure your equity/FI AA reflects your risk preferences. That's all that matters.

[BillyK]

I think if you tighten your bands too much for rebalancing, you may potentially lose some momentum in equities.

[Steve Reading]

A common rule of thumb is to rebalance your portfolio whenever it gets more than 5% out of whack. If your goal is 80/20 but you're now at 75/25, it's time to rebalance. But after crunching some numbers, I believe 5% rebalancing bands are far too lax. I calculate that 2-3% bands would be more appropriate for most investors. The size of the bands should also vary depending on your specific asset allocation. [...]

You're on something... Please check this old thread with a similar discussion and a proposal to make things more square:
viewtopic.php?t=186203

Bro I used Adaptive Bands I never thanked you for the post!

[siamond]

Bro I used Adaptive Bands I never thanked you for the post!

That's cool! I was wondering if I was the only lunatic doing so... That makes (at least) two of us!

[Steve Reading]

Bro I used Adaptive Bands I never thanked you for the post!

That's cool! I was wondering if I was the only lunatic doing so... That makes (at least) two of us!

Of course you’re not the only one! It just makes so much sense to do it that way. I can’t do the whole 5/25 just because it’s “good enough”. Why would I do “good enough” if I can just do “exactly perfect” for no effort since this one dude named Siamond already figured out the formula I just put into excel?

[vineviz]

As rebalancing becomes less and less frequent, how do we say that a given portfolio is, let's say, 60/40 and not 65/35 ? Maybe if my rebalancing threshold is 70/30, a portfolio which was 60/40 at start spends a lot more time being near 65/35.

It is not a straightforward problem; there are many subtleties in play.

The target allocation is what it is: if the target is 60/40 then what else would you call it. In effect, the rebalancing only needs to happen with enough frequency to ensure that the average asset allocation over the investing period closely approximates the target allocation, and rebalancing bands of 10% or so are typically more than sufficient to accomplish that.

Over the past 85 years, a 60/40 portfolio with 10% rebalancing band would have had an average allocation of 62% stocks and 38% bonds. That's a trivially low difference in risk exposure. Take a look at the following chart of maximum crisis drawdowns for a 60/40 portfolio over the past 50 years with different bands. If there's a case that one band is systematically and objectively superior to another, I don't see it.

[Thesaints]

Nobody is surprised by these results. They are what ever rebalancing study for the last decade has shown. Rebalancing just doesn't matter that matter that much with any reasonable (i.e. no going 20 years without rebalancing) scheme.

Why do you say "it does not matter" ? Clearly it does and rebalancing less often is better, according to that data. I bet, if one had not rebalanced at all it would be even better in terms of return.

The issue is that the advantage does not come from some rebalancing practice, but simply by the fact that in an up market if you rebalance less often your portolio has an effective higher allocation to stocks.
The proper comparison is to calculate the actual, time weighted, asset allocation and compare that one to the same AA with continuous rebalancing. Than we would see that rebalancing as often as possible is better, were it not for additional expenses.

[Thesaints]

The target allocation is what it is: if the target is 60/40 then what else would you call it.

The initial allocation is what it is. Unless you rebalance continuously, your portfolio will deviate from it and the less often you rebalance, the more it will deviate.

In effect, the rebalancing only needs to happen with enough frequency to ensure that the average asset allocation over the investing period closely approximates the target allocation, and rebalancing bands of 10% or so are typically more than sufficient to accomplish that.

I do't think so. Assuming a linear outperformance of one asset over the other a 10% band (do you mean a 60/40 portfolio is rebalanced when it becomes 70/30, or when it becomes 66/34, btw ?) means that on average you are off by 5%. Your 60/40 portfolio is in fact a 65/35 one.

Over the past 85 years, a 60/40 portfolio with 10% rebalancing band would have had an average allocation of 62% stocks and 38% bonds.

How do you know ?

Take a look at the following chart of maximum crisis drawdowns for a 60/40 portfolio over the past 50 years with different bands. If there's a case that one band is systematically and objectively superior to another, I don't see it.

But I want to see all the 50 years, not just 5 data points. Who knows what the other 12,595 show, right ?

[Elysium]

It's interesting that all the Bogleheads who cite Bogle's stance on International not being necessary as justification for domestic-only allocations, don't seem to similary support his position that reblancing is never necessary (and equally likely to be beneficial or harmful if you do it.)

I don't think it is totally necessary, unless the portfolio has moved in the extreme, can only happen if you totally ignored it for many years and never had any new money going in. He does make this point, and in fact favors the market doing the re-balancing for you. In anycase it was never intended as a method for creating extra returns, but as a risk mitigation tool. Even at 5% bands, can't say the risk profile has moved too much.

[longinvest]

Given all of the recent threads on rebalancing and all of the underlying uncertainty and anxiety, I wonder if most of us (myself included) would simply be better served by investing in Lifestrategy or target date funds to the extent that it is feasible (e.g. in tax advantaged accounts).

What always held be back was the extra fee, but I'm starting to believe that the extra fee will be mitigated by the ease of managing the portfolio and not having to rebalance or second guess when it is a good time to rebalance.

I've come around to believing this more and more as well (for myself included).

Just in case you're interested...

Last year I moved my entire portfolio into an all-in-one Vanguard "Asset Allocation ETF" very similar to a LifeStrategy fund (except for a different home bias). I was lucky to do the switch early enough before having accumulated significant unrealized gains in my taxable account. This has made investing much simpler for my wife and, surprisingly, for me (which isn't something I had anticipated). I had not realized how simpler a One-Fund Portfolio actually is. It almost feels like a bank account, except for its fluctuating balance, in that I only have to regularly put money in. I don't have to think, anymore, about planning rebalancing moves carefully to avoid (the equivalent of) wash sales, etc.

I think that there's a lot of unfounded fear about the use of a lifestrategy or a target date fund in a taxable account. I think that too few forum members pay attention to the following statement of our wiki's page on Tax-efficient fund placement: "Tax rates and brackets change frequently. What was a logical tax location one year may turn out to be a poor choice a few years later. Consider if it's worth the effort (added complexity) to take this approach."

I've provided, in the One-Fund Portfolio thread, a mathematical proof that a mirrored asset allocation is more resilient in face of an unknown future than prioritizing the placement of a specific asset (stocks or bonds) into specific tax-advantaged accounts:

Here's the thing. If prioritizing bonds in tax-advantaged accounts turns out to be best over one's specific lifetime, a mirrored allocation will turn out to be superior to having prioritized stocks in tax-advantaged accounts. If prioritizing stocks in tax-advantaged accounts turns out to be best over one's specific lifetime, a mirrored allocation will turn out to be superior to having prioritized bonds in tax-advantaged accounts. A mirrored allocation is thus mathematically guaranteed not to turn out to have been the worst location strategy among these three strategies, even if tax laws change in unexpected ways.

Some people might consider this mathematical guarantee, of not being the worst asset location strategy, "not very attractive", yet I have not seen a mathematical proof of a "more attractive" asset location strategy that is guaranteed to always beat a simple mirrored allocation strategy.

It's quite similar to indexing, when you think about it. William Sharpe's theorem guarantees that a simple total-market cap-weighted index investment strategy is guaranteed to never be worse than average (before fees). Some people might consider this mathematical guarantee "not very attractive", yet I have not seen a mathematical proof of a "more attractive" investment strategy that is guaranteed to always beat it.

Now that I have experienced the simplicity of a One-Fund Portfolio (with, obviously, a mirrored asset allocation) and the peace of mind it brings, I'll never go back to holding separate funds or ETFs.

[michaeljc70]

Maybe the numbers say different things but it seems like people are over complicating the supposed simplicity of a 2/3/4 fund portfolio. Doing it once a year probably doesn't change things much (returns or max dd), at least based on a few scenarios I glanced at on the portfolio visualizer site.

+1

[robphoto]

There's a Vanguard white paper talking about various approaches and frequencies for rebalancing, deals with threshold vs. time, or using both, etc.

https://personal.vanguard.com/pdf/ISGPORE.pdf

[pascalwager]

There probably isn't that much performance difference between an 80/20 portfolio and an 83/17 portfolio, so rebalancing at 85/15 seems reasonable.

[Triple digit golfer]

I have an AA set to manage risk. To that end, being 75%-85% equities is fine for my 80/20 desired AA.

Also on a $1mm portfolio at 80/20, a drift that would trigger rebalancing at 5% bands is a 25% drop in equities. Rebalancing would be a $40k sale of bonds to equities.

On a 2.5% band, rebalancing would be done at a approximately a14% drop and would require a roughly $22k exchange.

So the amount left in the "wrong" asset class, for lack of a better term, is $22k, potentially, if you don't rebalance at 2.5% bands and wait for a 5% band that is never hit.

Doesn't seem like a big deal to me to leave 2.2% of the high point in your portfolio alone. For most people, particularly young investors, 2.2% of their portfolio is made up in no time through contributions anyway.

Plus, it can work against you. Maybe you rebalance at 2.5% bands, then the market keeps falling, but never triggers again, but would have triggered once at the 5% band, which would have you buying much lower and for a larger amount.

Small bands may intuitively seem more favorable, but with an unpredictable market, we don't know which will be best. So I chose the 5% band because it just seems the most reasonable to me.

Considering that Vanguard's LifeStrategy Funds are only offered in 20% equity increments, it seems to me that such precision probably really isn't that important.

[randomguy]

Nobody is surprised by these results. They are what ever rebalancing study for the last decade has shown. Rebalancing just doesn't matter that matter that much with any reasonable (i.e. no going 20 years without rebalancing) scheme.

Why do you say "it does not matter" ? Clearly it does and rebalancing less often is better, according to that data. I bet, if one had not rebalanced at all it would be even better in terms of return.

The issue is that the advantage does not come from some rebalancing practice, but simply by the fact that in an up market if you rebalance less often your portolio has an effective higher allocation to stocks.
The proper comparison is to calculate the actual, time weighted, asset allocation and compare that one to the same AA with continuous rebalancing. Than we would see that rebalancing as often as possible is better, were it not for additional expenses.

And in a down market, rebalancing more often results in a higher allocation to stocks. The slight difference in AA result in slightly different numbers. You can't predict which approach will win. That is all really basic rebalancing stuff. At the end of they day it just doesn't matter if you hold 60/40, 61.1/38.9 or even 65/35. Pick one and get on with life. You aren't going to gain or lose enough with any of the choices to matter.

[Tamarind]

If one rebalances for 5% deviations, the recent market gyrations would have not yet triggered a rebalancing.
This could be good, or bad, but people should be aware of it.

Yes, this was my motivation for looking into the question. Like everyone, my stocks are waaaay down these past few weeks. I thought for sure my AA would be far out of whack, and I'd need to sell some bonds and buy some stocks. I was very surprised to find that my 80/20 has only shifted to 77/23.

For me, I find this to be an indication that 5% bands are just right. All I want is a reminder not to give in to fear or greed during a meltup or 2008-style crash. Obviously your mileage may vary.

I'm also accumulating so 5% is even harder to trigger for me than it would be for a retiree with the same AA.

[Triple digit golfer]

If one rebalances for 5% deviations, the recent market gyrations would have not yet triggered a rebalancing.
This could be good, or bad, but people should be aware of it.

Yes, this was my motivation for looking into the question. Like everyone, my stocks are waaaay down these past few weeks. I thought for sure my AA would be far out of whack, and I'd need to sell some bonds and buy some stocks. I was very surprised to find that my 80/20 has only shifted to 77/23.

For me, I find this to be an indication that 5% bands are just right. All I want is a reminder not to give in to fear or greed during a meltup or 2008-style crash. Obviously your mileage may vary.

I'm also accumulating so 5% is even harder to trigger for me than it would be for a retiree with the same AA.

Right, because new contributions go into the lagging asset? I'm in the same boat as you. I don't think I've ever rebalanced in my 13 years investing.