Is anyone proposing a trading strategy of selling all of your Berkshire holdings near the close of the 7th trading
day of each month, and buying them back near the close of the 5th-last trading day of that same month?
“Proposing” a strategy…well, that’s a bit strong.
It’s a thought experiment.
Imagine you had the current stats on the strategy, and a time machine, and a broker that charged virtually nothing.
(the latter, at least, is not so hard to achieve)
Since Jan 2000, the buy-and-hold guy would have made 10.05%/year with a risk metric of 9.85%.
(using a risk relatively reasonable metric, not monthly standard deviation of returns.
I use annual downside deviation: basically, the probability of your portfolio not making 10%/year
in a given rolling year, with a squared penalty on the size of the shortfall)
A portfolio using a switch between holding the stock in the lucky days, and cash earning zero, would
have made 14.66%/year (4.62%/year more for 22 years), with a risk metric of 4.43%, about 45% of the risk of buy-and-hold.
So, over 4%/year more returns for under half the market exposure risk.
(half the risk makes some sense, by the way, since you’re in cash almost exactly half the time)
For example, the buy-and-hold fella would have had a worst rolling year of -50.1%.
The crazy monthly timer had a worst rolling year of -12.7%.
The buy-and-hold guy had a probability of a positive rolling year of 75.1%.
The “lucky days” timer had a probability of 90.7%.
Am I seriously proposing that you do this? No.
But hey, if you’re thinking of buying some BRK some time soon, you might as well do it at the start
of the next “lucky” stretch rather than at the start of the next “unlucky” stretch.
Can’t hurt.
The strange thing is, I’ve been looking at and commenting on this effect for Berkshire for a very long time now.
It hasn’t gone away.
Jim