Correct me if I'm wrong but the relevant 409A valuation is the most recent one done when you exercise your options and taxes on them become due. I think that's the situation GP was referring to, that this approach might help fix.
Unless you're saying that when Square's valuation suggested a share price of ~$15, a 409A done at the same time would have returned a value closer to the ~$5 that the paper suggests is more reflective of the average employee's situation.
The relevant 409A is the one that is done prior to options being issued. If the company issues options below FMV (i.e. below the FMV determined by an objective third party, "the 409A valuation"), they are giving you in-the-money options, that is, they are giving you something of value at that moment. That's compensation and needs to be taxed as such.
Later, when you exercise, either it is because the company was just acquired (and the share is immediately sold and you pay taxes on the profit) or the company has gone public and the shares are now tradable, so FMV is set by the market not by a 409A report. (I am ignoring for simplicity the rarer case where an employee exercises an option that is not freely tradable).
> when you exercise, either it is because the company was just acquired (and the share is immediately sold and you pay taxes on the profit) or the company has gone public and the shares are now tradable,
Nope. I'd argue the most common case of exercising stock options is actually when you leave a company before it has had a liquidity event (and you have to exercise the options because of a 90-day exercise limit). It's also when you want the lowest possible 409A/FMV, so that your taxes are as low as possible. Because you can't sell the shares to pay taxes.
There's also the rarer case (as you said) of employees exercising options of a private company while still remaining employed there. This is done in anticipation of an IPO or similar, to get a head start on the long-term capital gains tax clock. It's still not that rare though.
You're not wrong, but this isn't just about taxes. In the case of Square, apparently they got a 409a valuation done that set a FMV for common of $9.11 and then issued options with that strike price as is standard practice.
Anybody who exercised immediately at $9.11 probably wasn't subject to any tax liability (assuming appropriate 83b election). Employees exercising at that price may have fooled themselves into thinking they were already $6/share in profit-land, which would be incorrect and is the point of the article.
But the bigger problem to me is that employees were fooled into paying $9.11 for something that was actually worth far less.
Unless you're saying that when Square's valuation suggested a share price of ~$15, a 409A done at the same time would have returned a value closer to the ~$5 that the paper suggests is more reflective of the average employee's situation.