Not exactly "reversing" the rule - what wasn't discussed much in the article is the *pre* retirement asset allocation and glide path.
Once *in* retirement, it makes a huge difference, especially if there's a down market right after retiring (and in particular, if one doesn't cut back on spending at that time).
The idea is that if you actually have a terminal asset allocation (i.e., long-run target of, say, 50% stocks and 50% bonds), when you start retirement, start *lower* than that allocation to stocks and build up towards it over time.
This is as opposed to either the oldest traditional rule of thumb (bonds by age) wherein one may start retirement with a slightly higher stock allocation and work down towards a lower allocation over time -- or one where one reaches that terminal allocation *at* retirement and just leaves it alone.
I'll repeat the biggest takeaway here: he's saying to start your retirement with *less* stocks, potentially a lot less and increase towards your target rather than starting your retirement with more stocks and decreasing towards that *same* target.
The benefit of this is that (a) you still have a stock allocation, so you're not walking away from all the long-term potential returns; and (b) you are much better insulated against the main cause of failures - stock pullbacks early in the retirement; Moreover, since you have all that insulation, in the event of the pullback, you can actually buy more stocks cheaper - thus helping you recover from that early pullback better than someone who's selling at the same time.
If the market doesn't have ups and downs - if it just goes steadily up - this method does worse than the older ideas. If the market has great years early on but crappy years later on, this method *also* may do worse. But if the market does badly early and well later -- this method does great.
Think of four possible sequences of returns:
good->good
good->bad
bad->good
bad->bad
Nothing saves you in bad->bad (except, maybe a pension with inflation adjustments - like Social Security!)
Traditional glide paths do fine in good->good and good->bad, but may fail in bad->good
This method does *ok* in good->good and in good->bad. But provides much better protection in bad->good.
As I said, though, it doesn't much discuss the *pre* retirement glide path.
If your "terminal" allocation is going to be, say, 50% stock and 50% bonds and you are going to use an upward sloping glide path starting at retirement at, say, 65, then at 65 you might want to have only 25% in stocks and 75% in bonds, increasing your stock allocation by 1%/yr during the course of the retirement.
That would mean, however, *decreasing* your stock allocation from, say, the time you're 30 - where you may have 90% down to that 25% when you're 65 by decreasing your stock allocation by about 2%/year during your working years.
So the long-term glide path is down down down down down down up up up.
The article doesn't address the costs of that downward sloping period - if you downward slope to half the allocation you'd have otherwise gone down to, you may lose out on a lot of potential returns during your working years.
It's fascinating research and Kitces and Pfau are really doing some great work from which we can all benefit - but it's not clear that it's a complete solution, at least not just yet.
[reminder - I'm just making up numbers above for illustration purposes!]
--David