At some point, there is a breakeven between the likely amount which could be won by keeping the chips of the roulette table, verses redeploying the funds if the market drops. While predictions are difficult to make, especially when it comes to the future (thank you Yogi), I'm guessing the numbers hit an inflection point.

Yup. Investing is hard. Buy or wait?

But one can try a back of the envelope for two scenarios: buy typical firms now, or wait in cash till a better opportunity comes along. How soon and how good would the opportunity have to be in order for the ten year returns to be better by waiting for now?

Pencil in some assumptions:

Let's assume that in the average year the T-bills make nothing after tax and inflation. (at the moment it's moderately positive, but that's a bit rare historically). So there is no expected return during the waiting time.

Let's assume that brilliant stock picking isn't really going to rescue us here. The returns on offer will be roughly those on typical large US firms at contemporary average valuation levels.

Let's assume that the average US stock in the average year at the average valuation gets you inflation + 6.5%/year, Siegel's constant. So, if you were able to buy a middle-of-the-pack firm today at an "average" valuation level, you'd make inflation + 88% in the next decade in today's money.

...but you can't. The S&P 500 is 50% more expensive right now based on smoothed real earnings than its average in the last 20 years. This is presumably why not a lot of capital is getting deployed into US equities right now. Let's simply assume that the valuation level drops to that "recent typical" some time in the next decade, but value generation continues at the usual inflation+6.5% rate. Thus a holder of the S&P might expect a real total return of around inflation + 2.3%/year over the next decade: the 88% usual total gain minus a 1/3 one time drop, annualized over a decade.

Let's say that Berkshire could find a way to deploy money at a long term rate of return of inflation + 8%/year during a market disruption when better opportunities appear. Not a very aggressive number because the amount of cash to deploy is so large.

With those assumptions (which you can change at will), Berkshire would be better off sitting on zero-real-return cash for up to 7 years before buying stuff offering 8%, than to buy the average large US firm now at today's general valuation levels. If they had to wait for 8 years before getting the buying opportunity, the ten year return would be 0.7%/year worse than buying average stuff at current valuation levels today.

Jim