Rebalance how often?

[slk23]

What is recommended, rebalancing the portfolio as soon as one of the four categories is >35% or <15%, or rebalancing only annually?  Has anyone done any back testing to compare the two approaches?

TIA.

[Roy]

What is recommended, rebalancing the portfolio as soon as one of the four categories is >35% or <15%, or rebalancing only annually?  Has anyone done any back testing to compare the two approaches?

TIA.

I'm curious also but would be surprised if there was a great difference as the rebalance bands are pretty fat. Someone had a chart showing all the rebalances from 1972 (not many), but I forget where that is.

Roy

[concerned752]

What is recommended, rebalancing the portfolio as soon as one of the four categories is >35% or <15%, or rebalancing only annually?  Has anyone done any back testing to compare the two approaches?

TIA.

I think someone did a backtest of the two approaches in the epic Bogleheads thread. Conclusion: there's a slight advantage to using the rebalancing bands, 50bp or something like that.

Also, frequency of rebalancing with bands was not very different than doing it annually (was in the 11 to 13 month range, I believe).

[Roy]

I did a backtest using the 1972 to 2008 historic data and found something like a 0.5% higher average benefit for rebalancing at the 15/35 weighting bands rather than yearly and fewer rebalances being involved.

With 15/35 rebalancing however only 30% of years require a rebalance on average, so with two buys and two sells to rebalance you're saving on trading costs.  Instead of making two sells and two buy trades each year to rebalance all four back to equal weighting in each of 3 years (12 trade costs), you might make only 2 buys/2 sells once in every three years (4 trade costs).  At perhaps $12 per trade that's a $96 saving across three years or $32 per year average saving.


I've opted to rebalance yearly at a time point near financial year end, as that way I can opt to trade either in the old year or new year according to whichever might be the more tax efficient. 

Clive,

Just so I'm clear, when you rebalanced in these versions, including the once yearly, I assume you restored all 4 asset classes to 25% any time a single asset class reached 15/35? 

Roy

[stylo]

One thing I don't quite understand: I've seen it said that rebalancing is to mitigate risk, but also to make profits (buy low, sell high). Which is it?

If the latter, then wouldn't more frequent rebalancing (only via hitting the bands) make more (less trading costs, of course)?

Relatedly, if you are using the band method, don't you have to recalculate everything daily (or at least weekly) in order to catch any big changes in the market and profit from them through a rebalance? For example, when bonds or whatever it was skyrocketed last year, would it have been wise to rebalance right that week or just ride it out? If only checking monthly maybe you'd miss it.

I do see you've said yearly fixed rebalancing is almost as effective, which seems to show that it has nothing to do with profit making.

So, a little confused.

[slk23]

Thanks for the replies.  I had a few incorrect assumptions such as the idea that the portfolio would be rebalanced annually even if none of the allocations are hitting the 15/35 bands.  I also assumed that the 15/35 bands would be hit several times a year.

In Browne's Fail-Safe Investing book he states:

When you make your once-a-year check of the portfolio’s value, if all four
investments are within the 15–35% range, no rebalancing is necessary.

During the year, if you happen to notice that there’s been a big change in
investment prices, you may want to check the values of the investments. Again,
if any investment has strayed outside the 15–35% range, go ahead and rebalance
the entire portfolio.

With the recent increased volatility in the markets I don't think I'd feel comfortable waiting for a once-a-year check.  I'll be checking my allocation percentages weekly.

.

[craigr]

With the recent increased volatility in the markets I don't think I'd feel comfortable waiting for a once-a-year check.  I'll be checking my allocation percentages weekly.

This is OK as long as you can keep your emotions in check. If you are looking at your portfolio value too often it can cause problems. The human mind is better at taking larger pains all at once than many small bouts of pain very frequently.

But Browne also added a caveat to this check once a year advice that I agree with. That is you can check the portfolio rebalancing bands if you hear about something big in the news that could cause concern. For instance the recent volatility would probably qualify especially with the crazy 30 minute 1000 point drop that happened a couple weeks ago. Or if you hear about the Euro collapsing, etc.

For me the news really has to be on the front page of the paper for it to warrant my attention personally. Even then, I would never trade on reported news because the information is just about worthless. I'd only trade if my rebalancing bands were at, or very near maybe, the trading point. You should definitely rebalance however if any asset is over the rebalancing threshold. It's usually OK to trade a little early if it makes you feel more comfortable (say rebalancing at 20%/30% instead of 15%/35%). However trading too late can be a very costly mistake if the overweight asset crashes.

[slk23]

It's usually OK to trade a little early if it makes you feel more comfortable (say rebalancing at 20%/30% instead of 15%/35%). However trading too late can be a very costly mistake if the overweight asset crashes.

Forgive me if this has been previously discussed, but has anyone back tested 20%/30% vs. 15%/35% rebalancing with the IAU/SHY/TLT/VTI portfolio?

.

[6 Iron]

I like the idea of absolute standards taking the guesswork out of a rebalance trigger. That said, I am more comfortable with a 20-30% trigger band than a 15-35% band. It would push me out of my comfort zone to be sitting on 34.6% exposure to long bonds, or 15.3% exposure to stocks, and say, "let it ride a little longer".

I realize that there is always the potential for rebalancing out of a bubble early, or catching the falling knife, but feel that there is enough upside in terms of avoiding the angst of the above lop-sided weightings, that it more than makes up for it.

Thoughts?

[6 Iron]

.....rebalancing yearly to rebalancing at 15/35 bands using MIDWAY rebalancing.

Clive, could you clarify what you mean by midway rebalancing? Thanks.

[Pkg Man]

I like the idea of absolute standards taking the guesswork out of a rebalance trigger. That said, I am more comfortable with a 20-30% trigger band than a 15-35% band. It would push me out of my comfort zone to be sitting on 34.6% exposure to long bonds, or 15.3% exposure to stocks, and say, "let it ride a little longer".

I share your concern, particularly on the LT bonds, but haven't run across that scenario yet so I don't know how I might feel at the time.  But keep in mind that HB said there was nothing really scientific about his choice of 15-35 bands, it just seemed to be a reasonable point at which to rebalance.  I recall a question on his show about this and he said 20-30 would be fine as well. 

I have toyed with the idea of different thresholds for different investments, say 15-30% for LT bonds, 20-35% for stocks and gold, but I wonder if that might somehow mess with the inner workings of the PP too much.  Still, if it helps one sleep at night, it might be worth considering.