Implementing Rebalance Band Rules - Does it Really Move the Needle?

[DawgMan]

During the accumulation stage, I practiced once a year (end of year) rebalancing and never applied any rebalance bands. I would just ride out down markets, short or long, "as is" until my rebalance date. As I have transitioned to the withdrawal stage (2022 being the first year), current negative market conditions have me wondering if I should tweak my approach and employ rebalance band rules to better manage my portfolio and improve my long term returns? Obviously, this long bull market has helped all our portfolios, but I am curious of those of you who have practiced rebalancing band rules for many of these cycles... have you found that they really made a measurable difference over time (as opposed to the once a year rebalance)?

Also, for those of you who have been employing the bucket approach, since your AA will fluctuate during different market cycles, what kind of spread have you experienced in your AA after refilling your buckets (i.e. 70/30 in a bull market to 50/50 in a bear)?

 

[easysurfer]

I've done both (bands and yearly) but haven't really compared the difference. I prefer yearly (the lazy in me ) and that's what I do presently.

 

[sengsational]

I've defined my bands as a fraction of each target. So, for instance, if I had 15% target on international equity, then +/-10% error bars are at 16.5% and 13.5%. But I usually rebalance in December and often don't hit the error bars. I wish everything wasn't so correlated, but because they all tend to move in the same direction, the 10% error bars are not often violated.


As to whether it's moving the needle, I've read it doesn't make much difference. There is a slight advantage in simulations, but I do it to "keep busy" and so not to do something rash. When the rule is "don't just do something... sit there", this keeps me out of trouble.

 

[SecondCor521]

I think of rebalance bands as helpful to managing risk, not really improving returns.

Although I have rebalancing band rules, in practice they don't come into play that often because they're fairly wide: My target is essentially 97/3 at the moment and I allow my bond allocation to be too high or too low by 25% (so between 75% of 3% and 125% of 3%) before I require myself to rebalance. I'm allowed to rebalance within the band if I want to, and I do sometimes if I feel like it.

My rebalancing band rules induced me to reallocate from bonds to stocks as the market was falling about a year ago, so that was helpful in theory but in practice the dollar amounts were not really all that much.

I think the studies I've seen summarizing rebalancing is that it's a marginally useful idea if done once every 12 to 18 months. More frequently doesn't help, and less frequently doesn't hurt too much.

I don't use the bucket approach so I can't comment on that part of it.

 

[DawgMan]

My target is essentially 97/3 at the moment...

Are you drawing part/all income from other sources (i.e. pension, RE, working) or 100% from your assets? Just curious as 97/3 feels very aggressive if you were drawing 100% from assets.

 

[SecondCor521]

Are you drawing part/all income from other sources (i.e. pension, RE, working) or 100% from your assets? Just curious as 97/3 feels very aggressive if you were drawing 100% from assets.

I've described before how I get to the 97/3 number. I'll briefly reprise it here and would be happy to go into more detail if you want it.

I start with a FIREcalc run using all my actual numbers, which has given me a 100% success rate for the past 8 years or so.

I then use the investigate tab to figure out what a 95% success spending rate would be in order to "stress test" my situation. Historically this amount is somewhat more than 2x my current spending.

I then update my spending in FIREcalc to whatever that 95% success spending rate is and use the investigate tab again to figure out how varying the AA affects the success rate. With my particular numbers, the maximally successful AA is usually 90/10.

I then apply that AA to the portion of my net worth which represents 25x my annual spending. Any excess amount above that gets allocated 100% to equities because that excess amount is what I expect my kids to inherit 30 years from now. Currently, my 90/10 portion and my kids' 100/0 portion results in an overall 97/3.

The above is with my particular situation, which includes being 53 years old with a gross WR of about 2% and a net WR of about 1%. The 1% difference between those two numbers is a result of a number of small income streams that are not based on my portfolio, so I call them my "non-portfolio income" or NPI. These include things like gifts, tax refunds, income from a small side gig, and a few other miscellaneous things.

I also include my SS income in my FIREcalc runs, which I plan to claim at 70. At this point looks like it will actually more than cover my current spending (so at age 70 I will have a negative spending rate unless I start spending more).

Another thing is that in these FIREcalc runs, I use a spending horizon out to age 90, so a 37 year planning horizon tends to favor equities in the above process.

Finally, because of a variety of factors, I have both high capacity and high tolerance for risk. So 97/3 doesn't bother me a bit even though I can see how people would view it as aggressive.

I don't have a pension, and my only real estate is my paid off home.

I'm heading out to do volunteer tax prep for several hours, but feel free to ask questions about any of the above.

 

[Midpack]

I use the 5/25 rebalancing guidance to alert me, but I may delay rebalancing due to taxes, especially now while I am doing large Roth conversions and I don't want any CG's I can avoid. YMMV

 

[Out-to-Lunch]

In Bogleheads fashion, my IPS says to rebalance yearly and if my AA exceeds bands of 5% absolute, 25% relative.

HOWEVER, if you crunch the numbers, you are not likely to hit the rebalance band within a year. The "worst case" (i.e., most prone to needing to rebalance) is for a 50/50 AA. In that case, the market must be down 19% (and bonds unchanged) to hit your rebalance band.

 

[Chuckanut]

I re-balance at the end of the year. At that time I will often take out next year's withdrawal in total if the current year has been good to me.

There are a lot of theories on re-balancing, including some that indicate re-balancing is not a good idea. Set your AA at day one of retirement (or whenever) and let the index funds do their thing over the next few decades.

 

[audreyh1]

I don’t know if they really move the needle. They can make you catch. Filling knife, ha ha.

I widened my bands after a tough experience in 2008, although I eventually came out the other side with plenty of gains. Not too frequently is the only rule I know.

These days I’m pretty lazy and do it once a year if needed after I withdraw funds. If it close enough in balance I don’t bother. At the end of the year, several of my mutual fund pay out capital gains which I take in cash. That’s often the only rebalancing needed other than topping off fixed income.

 

[DawgMan]

I've described before how I get to the 97/3 number. I'll briefly reprise it here and would be happy to go into more detail if you want it.

I start with a FIREcalc run using all my actual numbers, which has given me a 100% success rate for the past 8 years or so. Do you run a basic all inclusive needs/wants/wishes annual spend, or just a general needs spend? In other words, do you include projected lumpy expenses and gifting or does that come out of the legacy bucket?

I then use the investigate tab to figure out what a 95% success spending rate would be in order to "stress test" my situation. Historically this amount is somewhat more than 2x my current spending.

I then update my spending in FIREcalc to whatever that 95% success spending rate is and use the investigate tab again to figure out how varying the AA affects the success rate. With my particular numbers, the maximally successful AA is usually 90/10. Do you mean you plug in your actual planned spend here which gives you the 90/10 at 95%?

I then apply that AA to the portion of my net worth which represents 25x my annual spending. So I read this as you are underwriting effectively the 4% rule at age 53, which you are comfortable to ride at 90/10?Any excess amount above that gets allocated 100% to equities because that excess amount is what I expect my kids to inherit 30 years from now. Currently, my 90/10 portion and my kids' 100/0 portion results in an overall 97/3.I have thought about something similar here where excess is left to run as probable kid inheritance, but I suppose my logic has been the core portfolio (your 25X) should be a more conservative 60/40-ish AA.

The above is with my particular situation, which includes being 53 years old with a gross WR of about 2% and a net WR of about 1%. The 1% difference between those two numbers is a result of a number of small income streams that are not based on my portfolio, so I call them my "non-portfolio income" or NPI. These include things like gifts, tax refunds, income from a small side gig, and a few other miscellaneous things. I'm assuming your 1% - 2% WR is based on your entire portfolio?

I also include my SS income in my FIREcalc runs, which I plan to claim at 70. At this point looks like it will actually more than cover my current spending (so at age 70 I will have a negative spending rate unless I start spending more).

Another thing is that in these FIREcalc runs, I use a spending horizon out to age 90, so a 37 year planning horizon tends to favor equities in the above process.

Finally, because of a variety of factors, I have both high capacity and high tolerance for risk. So 97/3 doesn't bother me a bit even though I can see how people would view it as aggressive.

I don't have a pension, and my only real estate is my paid off home.

I'm heading out to do volunteer tax prep for several hours, but feel free to ask questions about any of the above.

See above in red or if easier, point me to the thread that dives further into this?

Overall, interesting approach. I'm not sure I have the risk tolerance on my core portfolio you have, but definitely have been thinking about letting the excess run 100% equities. My WR is running between 2% - 2.5% and is a relatively large planned spend which is extremely discretionary fat-filled and grows annually. Realistically, short of big annual gifts to the kids, I don't see myself spending at this rate until the dirt nap, but planning none the less.

 

[SecondCor521]

See above in red or if easier, point me to the thread that dives further into this?

Overall, interesting approach. I'm not sure I have the risk tolerance on my core portfolio you have, but definitely have been thinking about letting the excess run 100% equities. My WR is running between 2% - 2.5% and is a relatively large planned spend which is extremely discretionary fat-filled and grows annually. Realistically, short of big annual gifts to the kids, I don't see myself spending at this rate until the dirt nap, but planning none the less.

I don't think I have a thread for it anywhere. I just spell it out periodically if it seems relevant.

I clipped out your red stuff out and putting it in quotes so you can see what I'm replying to - hopefully that's OK:

1. "Do you run a basic all inclusive needs/wants/wishes annual spend, or just a general needs spend? In other words, do you include projected lumpy expenses and gifting or does that come out of the legacy bucket?" For my base FIREcalc run, I use my recent actual spending, so it's everything I've spent in Quicken the last six months multiplied by 2 to make it annual. I do subtract out the "College" category because that comes from a separate dedicated bucket of funds.

I lead a fairly boring life so I don't really have large lumpy expenses. I actually and somewhat uncharacteristically handwave at any analysis there and rely on having a 1% WR 99% of the time to cover the potential of those occurring.

2. "Do you mean you plug in your actual planned spend here which gives you the 90/10 at 95%?" No, if I use my actual spend, I get 100% success with any AA. The idea here is if I went hog wild and started spending a lot more, there would be a point where my spending would pass the point of 100% safety at all AAs and enter a region where the safety of my spending would depend on my AA before entering a region where my spending would be unsafe regardless of AA. So I arbitrarily chose a 95% safe spending amount (which would be at the lower end of that middle region). That way I could see how the variation in AA impacted my safety. I then wanted the safest point (which is the "mountain top") on that AA / safety curve. Which, for my particular set of actual inputs but with a single change to my spending to be slightly unsafe, ends up at about a 90/10 AA.

3. "So I read this as you are underwriting effectively the 4% rule at age 53, which you are comfortable to ride at 90/10?" I'm comfortable as long as I'm spending less than 4% of my portfolio at age 53, yes. Since I'm spending approximately 1% of my portfolio, I think my AA really doesn't matter. If I were to spend more to an a marginally unsafe degree, I would want my AA to be the most survivable one at that higher spending level, which turns out to be 90/10 as described above.

4. "I have thought about something similar here where excess is left to run as probable kid inheritance, but I suppose my logic has been the core portfolio (your 25X) should be a more conservative 60/40-ish AA." I agree that the core portfolio should be the most conservative, but I interpret conservative to be the most historically safe WR given the planning horizon.

Somewhat counterintuitively, with a 37 year planning horizon (and my other inputs), historically a 60/40-ish AA is less historically safe than a 90/10 portfolio. At least according to FIREcalc with my inputs. I think it's because 40% bonds is more susceptible to inflation risk than the 60% stocks are to sequence of returns risk, basically.

5. "I'm assuming your 1% - 2% WR is based on your entire portfolio?" Well on my FIRE stash, yes. Which as is typical, excludes my home and possessions, and also excludes my kids' college funds and a potential inheritance. When I calculate that WR number, it does include the NPV of a derated amount of my SS.

I should add that I do my numbers two different ways - one way is via FIREcalc, and the other way is via my homebrew all-purpose Excel spreadsheet. Depending on context this may be a bit confusing. For my WR%, I rely on a calculation in my Excel spreadsheet because it's just a static snapshot based on current balances, current spending (well, last six months in Quicken annualized), and SS estimates and projections.

 

[VanWinkle]

I re-balance on 5% bands and made additional gains in 2020 due to re-balancing. I just re-balanced 4 days ago based on the same 5% band, and it may add to gains also. If not, I'm ok with just reducing risk.

VW