It goes back to the realization that no investor can predict the peaks or troughs of the market, so getting out of the market is effectively trying to time it. One may be lucky and get out at the peak and then buy back in at the bottom, but on average you will lose money this way.

What about the last 100 years?

(But since you speak of a diverse basket, you probably know this already.)

His main argument was rebalancing effectively switches high-returning assets for lower-returning ones. If one part of your portfolio did better than another part, why would you get rid of it just to bring the asset allocation back to your target?

Rebalancing will also generally generate capital gains in taxable accounts. If you're aiming for 60/40, but your stocks are now up and it's 70/30, selling those stocks to bring it back to 60/40 again means you will pay capital gains on whatever you sell.

The data shows you can get better returns by rebalancing weekly, but this is really too much for most investors, unless you use something like M1 where rebalancing is a single button click.

Of course, there's more complexity to this question. The above mostly applies to the accumulation phase. To someone in, or close to, retirement, for example, having a portfolio that's drifted too far into stocks can be a risk.