Changes to accounting etc are dwarfed by the simpler fact that stuff is just plain a lot more expensive than it used to be. People are simply willing to pay a whole lot more for a dollar of honest annual owner earnings than they used to.

There have been changes to how earnings have been booked and what the precise current mix of firms is, and changes to tax rates and interest rates, but those are quibbles by comparison. What good reason is there to think that the doubling in the P/E ratio of a boring refrigeration firm is justified by anything other than people with cash in hand being happy to pay more for the same old stream of true owner earnings? Adjusting the E a bit isn't going to change the conclusion.

This does NOT mean that broad US market equity valuation multiples will fall again. Now, soon, or in fact ever. Nobody knows.

It does mean that you don't get many dollars of future owner earnings per dollar invested at today's prices in the typical firm, so broad market returns will have to be low even if valuation levels stay high. For the typical US equity, good returns from here can come only from even higher multiples.

Jim

PS
As for the thesis of the article about R&D expensing, I remember being trapped in that mire in the 1980s when I ran an R&D department--expense or asset, expense or asset? Every year the same debate. My simple observation is that, for most firms, most "R&D" really is a current expense and has no lasting value. Most of it is more like a depreciation cost, or at best running in place to keep the competitive position from eroding. Sure, there's a remainder that has a long life, but I suspect it's material for only a tiny number of firms. And in any case it's dwarfed by a simple change in prices.