I believe it's still quite vivid in Dr. Bernstein's mind how many of his followers - who'd read all of his books and done their risk tolerance profiles and signed on to stay the course with their asset allocations - threw in the towel during the '08-09 market meltdown and sold significant chunks of equities at or near the bottom.
He's not talking about 20-25 years in cash but rather in short-intermediate bonds (most likely Treasuries and TIPS given his stated preference for them). Here again looking at Tyler's phenomenally-useful charts tells the tale about what it is like to actually live through huge stock market losses that go on for years. Very few can stay the course, and based on my own experience having ER'd in 2002 and lived through the '08 crash with a very sophisticated slice-and-dice DFA portfolio (Bob Clyatt's RIP) that backtesting showed "couldn't" lose more than ~8% but actually lost ~25% (this with only 40% in equities mind you) I have a great deal of appreciation for the PP, Golden Butterfly, Swedroe's most recent Reduced Fat Tails approach, etc.
Just in case it's of interest, John Greaney, pretty much the "old dog" of early retirement finance writers, has an excellent update of his periodic real life retiree returns series here:
http://www.retireearlyhomepage.com/reallife16.htmlThe whole article is worthwhile, but IMHO the most interesting comment is pretty far down where he points out that if you'd retired in 2000 rather than 1994 as Greaney himself did:
"If you happened to retire in January 2000, the last sixteen years haven't been pleasant. Only the Warren Buffett portfolio has a value appreciably exceeding its $100,000 starting balance. The 100% fixed income portfolio is underwater while the MPT portfolio, Larry Swedroe Portfolio, Harry Browne Portfolio, and Harry Dent Portfolio are all 8% to 14% in the black. The other two portfolios both show losses. The worst performer was the 75% S&P500/25% fixed income portfolio which is now less than two-thirds of its starting value."
Getting back to Golden Butterfly, the PP and non-correlated assets, Craig Rowland (co-author of the essential book on the PP and source for a lot of the material in the Deep Risk pamphlet) flatly states that no portfolio composed only of stocks and bonds can be said to contain uncorrelated assets, and the behavior of the PP, GB and other such portfolios during the '08 crisis lends a great deal of credibility to such views (as Treasuries and to a lesser extent gold saved them while complex slice-and-dice MPT porfolios saw all of their noncorrelated assets suddenly correlate and tank in unison).
Re: arebelspy's point about gold as an inflation hedge not mattering, I think it DOES matter that Harry Browne was dead wrong about allocating 25% of his portfolio to something that doesn't work to combat the exact problem it was included to address while failing to offer a sophisticated, internationally diversified, small cap and value tilted equity allocation that is the actual tool for the job in question.
While I followed the PP myself for about 4 years and was lucky to have done so timing-wise I came to see that pretty much all of the components of that allocation deserve to be revisited due to changes in the investing landscape since Browne's time.
Paper gold, much of which isn't backed by bullion, along with behind-the-scenes manipulation of gold prices are a far cry from keeping your gold coins and bullion in your bunker along with your MRE's and firearms, as Browne and many of his followers surely envisioned. It isn't just that gold has no inherent value nor rate of return: it's that paper gold isn't gold.
Half the portfolio is in Treasury Bonds, but I don't think Browne ever envisioned not only 30 year Treasuries offering yields below CPI inflation rates and - more importantly - a Congress composed of sociopaths who would willingly cause "full faith and credit" to be called into question and cause credit downgrades for the sake of political brinksmanship. Treasuries may still be (in the inimitable words of Rowland's co-author J.M. Lawson) "the best horse at the glue factory" when it comes to bonds, but the allocation of half the portfolio to them is so obviously outdated that even PP loyalists at least tweak the cash portion.
Tyler's own tweaking of the PP stock allocation is a welcome breath of MPT fresh air, but when I look at constructing portfolios from truly non-correlated asset classes I'm a lot more intrigued by Larry Swedroe's more recent iterations (in "Reducing Black Swans") of the so-called Larry Portfolio, in which a 30-35% equity allocation is made up entirely of such volatile exotica as International Small Cap Value and Emerging Markets Small.
A last comment since it's getting late is that I find it interesting how all of the "sweet spot" (risk:real return) on Tyler's site are pretty much iterations of the classic retiree 40:60 equity:bond allocation but with gold thrown into the mix. And meanwhile Vanguard's Wellesley Fund - actively managed, 35:65, and owning only a relative handful of mostly corporate bonds and dividend stocks - has outperformed most if not all of these backtested wonders for decades. If nothing else this certainly suggests that if you're going to have a bond-heavy portfolio (and I realize few on this board are likely to do so) bonds are the one are, given the size and relative illiquidity of the market, where professionals can actually deliver alpha. Wellesley's 9.95% CAGR 1970 to the present (and only ~18% drawdown during the '08 crisis) puts it in a league of its own among managed funds.