As a long-term fundamental investor, who bothers to value stocks using DCF (with full consideration to how arbitrary this is, but accepting that this is your way of understanding the investment world), why would you consider rebalancing without further consideration of the current valuation of the investment and the current yield vs what you are rebalancing into?
From memory rebalancing has its best effects when there is a massive differential between current yield in current portfolio vs rebalanced portfolio.
Buffett, never did it, and that says a lot. He made plenty of comments on the concept too.
His Ves, nice to hear from you.
This is an absolutely stunningly great question. Thanks for raising it.
The short answer is that, if you have a reliable insight, like Buffett, you should take it into account for rebal. In prior posts, I have mentioned that rebal is cream on cake and can be applied over whatever portfolio of stocks you think is right for you. What matters is that you rebalance against the drift.
Since you do DCF, I'm going to dive in.
The purest theory of rebal goes like this...
For today, find me the best portfolio that
Maximises [Expected return from stock selection over the next period until rebal + Rebal Profit Expected over that same period - LAMBDA x Risk - Transactions Costs of moving from the current portfolio] subject to restrictions
LAMBA is just a parameter to say how much you dislike whatever measure of risk is right for you.
Notice the Expected Return from Stock Selection part. That's the part where you think stocks are cheap or rich.
But, note that there is room for rebal profits, which will tend to want to make the portfolio more diverse, move it to volatile stocks and find ones which are less correlated.
Risk is anything...risk of losing capital, volatility, less than inflation, less than your friend next door, less than market...
Then there are trading frictions which may or may not include tax, brokerage, spread and market impact.
Restrictions may be things like not sort selling, maximum exposure to cash, industry exposure limits.... whatever you can dream up which is important to you.
If you change your view about stocks, it will affect the portfolio in the direction you anticipate. Rebal is not an excuse to turn your brain off.
Buffett has earned the right to place enormous weight on the Expected return from portfolio selection bit so it completely dominates considerations related to rebal. But he's the only Warren Buffett there is.
As we move from Grand Master to mortal, the Expected Return from Stock Selection becomes much more like the expected return from the share market and rebal profit becomes more important no matter what your definition of risk. One way to consider this, if your case is that your portfolio can be considered as the sum of two parts, is as follows:
+ Just being in the market at all; and
+ Your skill in beating the market.
As a buy and hold the above is possibly the most useful way of thinking, although other ways are reasonable too. So, if you could otherwise just buy an ASX 200 ETF super-cheaply and it is the average of all investors in the market, but choose to manage actively, it is reasonable to say that "why bother" risk is performance relative to the ETF. In that case, for reasonable levels of forecasting power, even enough that - if you worked in some major investment shop - would make you a centi-millionaire, you'll find yourself being asked to produce a fairly diverse portfolio where the lion's share of the skill in beating the market is derived from your stock picks, but where maybe 10-20% or so over time is derived from rebal depending on how tolerant you are of underperforming the ETF. The impact of rebal is not overwhelming, but nothing to sneeze at either and probably would be worth $50m to you in career earnings given how pointy pay for performance is. Cream on the cake.
The less forecast skill you have, the greater the relative importance of rebal and the less that changes in your view will affect the portfolio....hope that makes sense. In the absence of any stock selection skill, you just move straight to the pure rebal portfolio which tends to produce around 2% per annum outperformance in general over time and would be responsible for 100% of outperformance. On average, skill is zero. Hence, on average it is better for anyone to just hold a diverse portfolio and rebal and be done with it. But almost everyone considers themselves to be above average and, in aggregate, places too much emphasis on the portfolio expected returns part and this is a probablistically bad thing to do.
To more directly answer your question, you would think about your view on stocks before rebalancing. If a stock rose, and your view became that is still cheap because of some development, the outcome would most likely be that you wouldn't rebalance the stock down or down by not as much. It all depends on the strength of your view and, ideally, some notion of your abilities as a forecaster. You may even increase the weight. In a prior post, I called rebalancing just a point of departure. It isn't necessarily your destination.