I was wondering myself if there was any "magic" to this rebalancing thing. So I setup a quick and dirty spreadsheet to test different rebalance bands on return and volatility. Used CAGR for returns and Std Err of the regression on log price for variability. I just looked at returns with zero withdrawals and a 70/30 stock/bond portfolio.
Unsurprisingly, I guess, over long periods (monthly data from 1900) the less rebalancing the greater the returns. Of course because the stock percentage kept growing. The more rebalancing the lower the returns, there was no magic middle.
Looking at the monthly data from 1900 on, the STE plot was pretty flat with 0% (rebalancing every other month) up to 30% rebalancing band then increased volatility with larger rebalance bands.
Over 15 or 20 year periods there was a bit of bouncing around for both and except for a very few periods such as the 20 year period from 1970, the less rebalancing you did the better in terms of CAGR.
From what I saw you gain a little less volatility from rebalancing but not more often than using a 10% to 20% off band.
No magic from what I can tell. Just seemed to me rebalancing for the most part doesn't do much and rebalancing more often than using a 10% to 20% off band or more than once every year or two would not be worth the trouble.
I was hoping I would see some predictable or repeatable gain to rebalancing but unfortunately I didn't. I guess there is still no free lunch