Data on Re-balancing Bands for Taxable vs Non-Taxable Accounts?

[ahhrunforthehills]

Hey everybody,

I know that it is often cited that for a vanilla PP a 20/30 re-balancing bands were ideal for Non-Taxable investment accounts and 15/35 was ideal for taxable accounts... but is that actual data to substantiate that available anywhere? I am sure most people in taxable accounts have a little piece of their soul die when their portfolio crosses the 20/30 threshold then crawls back up.

Anybody ever laid eyes on anything like that? I was unable to find anything via search.

Thanks!

[Smith1776]

I believe that a substantial amount of research was done on this topic in the original epic PP thread over at the Knuckleheads forum.

Do a search there and you'll probably find some useful stuff?

But yeah, your overall assertion that the larger rebalancing bands are more tax efficient is correct.

[ahhrunforthehills]

Thanks for the response. I just headed over there and found a post from "weltschmerz" that I think you were referring to (https://www.bogleheads.org/forum/viewto ... start=3250):

Here's a graph of the total return of the PP, comparing the 15%/35% bands, vs. the 20%/30% bands, vs. no rebalancing:

THIS IMAGE IS NO LONGER AVAILABLE

Go figure!

Anyway, apparently the image was showing that:

There doesn't appear to be huge differences in the total returns, regardless of what rebalancing bands are used (if any). My personal preference would be to use the 15%/35% bands. This would be a simple approach to follow, especially where tax considerations are an issue.

In constructing the graph above, I note that there were:
5 rebalancing events since 1989 for the 15%/35% PP
15 rebalancing events since 1989 for the 20%/30% PP

A few comments on the graph:

Y-Axis represents Total Return since 1989. (1989 starting point was arbitrary; this was the oldest data I could get for the bond portion.)
Data for stocks, bonds, cash from Yahoo! Finance (using the Adj. Close - "Close price adjusted for dividends and splits"). Therefore, all distributions were considered to be immediately reinvested.
Data for gold from usagold.com.
I used a variety of investments to track the asset classes (investment changed as better options for tracking became available):
Stocks = VFINX, then VTSMX, then VTI (at inception)
Bonds = VUSTX, then TLT (at inception)
Gold = daily gold price
Cash = no price tracking until 1996, then VFISX, then SHY (at inception), then SHV (at inception)

Couple other quotes:

Re: Permanent Portfolio Rebalancing Bands
Post by craigr » Fri Jun 11, 2010 12:23 pm

Thanks for the great chart. As Tex stated, Browne said that 15/35 is fine or 20/30 is fine if you are able to handle the additional transactions and are aware of the higher costs. I'm personally a taxable investor and will rebalance in the 30-35% range depending on what I want tax-wise to happen that particular year. I don't allow any asset to ever go above 35% of the allocation or ever under 15% though.

I'd just add that while rebalancing can add returns in some cases, its primary purpose is to control risk which provides a smoother ride for the investor.

by Lbill » Sat Jun 12, 2010 9:47 am

Whether rebalancing adds any value is highly time-period dependent. There doesn't seem to be any consensus among various backtesting studies that rebalancing consistently does anything - but it doesn't seem to hurt either. If it makes you feel better, do it. If you like to set it and forget it, that's probably OK too. Worst thing you can probably do is split hairs trying to figure out if this-or-that rebalancing method is the best. Can you spell "data-mining?"

If anybody saved a copy of that image or has ran the data, I would love to see it

[Mark Leavy]

When you think of rebalancing bands, here are the thought processes that go into it:

There is some volatility harvesting that comes from rebalancing.
True - and it varies slightly with the band width, but not by much - and it varies with the amount of trending that the market is doing. (sequence of returns) And at best case, the rebalancing bonus isn't much. Almost always less than 1% - and can be negative in a trending market. So forget about it. No way to predict whether more or less frequent is better for grabbing a few extra basis points.

Rebalancing reduces your exposure risk in case the market turns against some asset that you are overweighted in.
Yes, this is the biggest factor in a non-taxable account.

Rebalancing too often can turn your long term capital gains into short term capital gains.
This is a big damn deal in a taxable account.

Summary
Given that, it doesn't matter too much when you rebalance. Try to wait at least a year in a taxable account.
Rebalance as much as you want in a non-taxable account. Too frequent won't make much difference and it keeps your risk exposure down.

Given that, you don't want to go rebalancing on a whim. You should have some set rules in place. If I were to pick some good rules, I would probably come up with something like 20/30 in non taxable and 15/35 in taxable. Brilliant!

[ahhrunforthehills]

When you think of rebalancing bands, here are the thought processes that go into it:

There is some volatility harvesting that comes from rebalancing.
True - and it varies slightly with the band width, but not by much - and it varies with the amount of trending that the market is doing. (sequence of returns) And at best case, the rebalancing bonus isn't much. Almost always less than 1% - and can be negative in a trending market. So forget about it. No way to predict whether more or less frequent is better for grabbing a few extra basis points.

Rebalancing reduces your exposure risk in case the market turns against some asset that you are overweighted in.
Yes, this is the biggest factor in a non-taxable account.

Rebalancing too often can turn your long term capital gains into short term capital gains.
This is a big damn deal in a taxable account.

Summary
Given that, it doesn't matter too much when you rebalance. Try to wait at least a year in a taxable account.
Rebalance as much as you want in a non-taxable account. Too frequent won't make much difference and it keeps your risk exposure down.

Given that, you don't want to go rebalancing on a whim. You should have some set rules in place. If I were to pick some good rules, I would probably come up with something like 20/30 in non taxable and 15/35 in taxable. Brilliant!

Great points.

It seems like for taxable accounts that re-balancing is almost a bad thing. The proper asset allocation is the PRIMARY goal... so do everything you can to help keep everything within the bands.

So for taxable, the best rules to play by are pretty much:

1. Re-balance at 15/35
2. ALWAYS being on the lookout for anything that can be tax-loss harvested for a loss. Swap the asset for a similar (but not too similar) asset and pocket the tax deduction.
3. Never auto-reinvest dividends. All interest/dividends should go to lagging assets to lower the chance of a re-balancing event.
4. New money should always go to lagging assets.
5. If you sell anything for a gain, do everything you can to sell a lot that was purchased over a year ago to avoid short-term capital gains.

Sound about right?

[Mark Leavy]

So for taxable, the best rules to play by are pretty much:

1. Re-balance at 15/35
2. ALWAYS being on the lookout for anything that can be tax-loss harvested for a loss. Swap the asset for a similar (but not too similar) asset and pocket the tax deduction.
3. Never auto-reinvest dividends. All interest/dividends should go to lagging assets to lower the chance of a re-balancing event.
4. New money should always go to lagging assets.
5. If you sell anything for a gain, do everything you can to sell a lot that was purchased over a year ago to avoid short-term capital gains.

Sound about right?

Looks solid.

[LittleDinghy]

you don't want to go rebalancing on a whim. You should have some set rules in place. If I were to pick some good rules, I would probably come up with something like 20/30 in non taxable and 15/35 in taxable. Brilliant!

Great points.

It seems like for taxable accounts that re-balancing is almost a bad thing. The proper asset allocation is the PRIMARY goal... so do everything you can to help keep everything within the bands.

So for taxable, the best rules to play by are pretty much:

1. Re-balance at 15/35
2. ALWAYS being on the lookout for anything that can be tax-loss harvested for a loss. Swap the asset for a similar (but not too similar) asset and pocket the tax deduction.
3. Never auto-reinvest dividends. All interest/dividends should go to lagging assets to lower the chance of a re-balancing event.
4. New money should always go to lagging assets.
5. If you sell anything for a gain, do everything you can to sell a lot that was purchased over a year ago to avoid short-term capital gains.

Our portfolio is about 1/4 taxable. And our GB is spread over the taxable, tax-deferred, and tax-free portions. And since creating the GB last June, we've not hit rebalance bands (but are coming close!). So, it seems that it will be pretty easy to just re-balance in non-taxable accounts. My sense is that the taxable rebalancing recommendations don't really apply to us right now. Do you agree?

[ahhrunforthehills]

So for taxable, the best rules to play by are pretty much:

1. Re-balance at 15/35
2. ALWAYS being on the lookout for anything that can be tax-loss harvested for a loss. Swap the asset for a similar (but not too similar) asset and pocket the tax deduction.
3. Never auto-reinvest dividends. All interest/dividends should go to lagging assets to lower the chance of a re-balancing event.
4. New money should always go to lagging assets.
5. If you sell anything for a gain, do everything you can to sell a lot that was purchased over a year ago to avoid short-term capital gains.

Sound about right?

Looks solid.

Okay, so if you are investing in a fully taxable account, we came to the conclusion that you should rebalance everything back to 25/25/25/25 after hitting a 35% band.

But now I am wondering if this is sound logic after watching this video: https://www.youtube.com/watch?v=MhszdAX9Leg

Wouldn't we be much better off just bumping that 35% asset down to 34% or 33%. Afterall, if the name of the game is to "trigger a taxable event as a last resort", going all the way from 35% to 25% is the nuclear option tax-wise.

To take it even further, the guy has me wondering if a 65% stock and 35% Gold allocation is a much better option than a PP for a person in taxable accounts.

He discusses here how bonds being added reduced risk by only a very little bit: https://www.youtube.com/watch?v=LA2Yr6NoZyA (around 5:05).

However, those bonds are taxed as ordinary income, which destroys performance overall.

Meanwhile stocks not only qualify for capital gains, but can also receive a step-up basis if you die holding them for your kids.

Since you are no longer holding all of those fixed-income treasuries, you now have less taxable income. This in turn further reduces your tax rate that you would owe on Gold sales as reflected in this video: https://www.youtube.com/watch?v=_aZqiP4gtZQ

Meanwhile, if compare a 65%/35% Stock/Gold to a Vanilla PP at Portfoliocharts, it appears that the 65%/35% outperforms in terms of Withdrawal Rates in both a US and Japan environment. Everything seemed pretty good in the Simba Spreadsheet as well.

In summary, it seems like if someone is in taxable investment accounts, a 65%/35% stock/gold allocation is superior to a standard PP.

Is there more volatility? Sure. Investments CAN go up. Investments CAN go down. But taxes on those treasuries are basically GUARANTEED to occur.

Has anybody kicked the tires on a strictly stock/gold allocation?