William Bernstein provides the first part in The Intelligent Asset Allocator in Table 2.3 (reproduced below), copyright 2001.

OK...and when was the data taken? the book was published in 2001. Can we assume the data is from 1999/2000?

So if we are down from the peak by 40-50%, then we have to assume there is no more down, that 'risk' has been removed? Bear market happened (worst case) and it is over? or another bear market in XXX years, where XXX is to be determined by divining?

And it still seems the best asset classes to be in is stocks and REITS, not CDS like YKW claims.

And the 'worst bear market' loss is insteresting....but then you have to start applying other 'factors'...like the lenght of the bear market, the recovery....most bear markets don't last more than 18 months. So how long will his 'bear market' last (or have we had it?)

Your quoting of a data set from year 2000 leaves a lot to be desired in 2003, especially since the markets of 2003 are at signicant different valuation and P/E ratios.

There are planning tools out there that let you plug your numbers in....expected rate of return...for each/all your assets. 2% growth, 3% growth, whatever you want to use. Simulation tools. Monte Carlo.

And since no one knows the future, or year by year gains/losses in each asset category, all we have to look at is historical data and trends. You can simulate all you want. That gets you lots of answers, none of which might be right.

FWIW, here's my actual FIRE port's allocation:

08% MMF (expected real return 1%)
17% 2y CDs (expected real return 2%)
50% Lg cap stocks (expected real return 4%)
25% Sm cap stocks (expected real return 6%)

Expected real return of blended portfolio: 3.94%

OK...and when was the data taken? the book was published in 2001. Can we assume the data is from 1999/2000?

The data he cites runs through 1998; however, in his latest book, The Four Pillars of Investing, published in May 2002 Bernstein does not increase his prediction of future real returns and in fact uses 3.5% vice 4% (the 4% may just be 3.5% rounded off) for future returns on US stocks. In The Intelligent Asset Allocator he also claims that it is folly to expect returns that exceed the "natural" real growth rate of 2% plus dividends.

I think that it is important to udnerstand that Bernstein is not being all that precise in his estimates because they are rough estimates. It is also important to note that Bernstein believes that historical returns were so much higher because the US market was so risky and that future returns will be significantly lower becasue the risk is greatly reduced.

Regards,
Prometheuss

telegraph: So if we are down from the peak by 40-50%, then we have to assume there is no more down, that 'risk' has been removed? Bear market happened (worst case) and it is over? or another bear market in XXX years, where XXX is to be determined by divining?

If we assume Bernstein was (1) being precise and (2) that his 4% estimate was from the S&P 500 peak forward and that (3) he means 20 years when he says long term (I doubt 1, but read him to mean 2 & 3 though long term could be 20 to 40 years because he is not being precise), then the estimate of future real returns looks something like this:

1485.46  August 2000 index value
 879.39  December 2002 index value

 40.8%  loss to date
  7.7%  estimate of future average annual real returns
        from January 2003 through July 2020.

This does not make a lot of sense if you believe what Bernstein writes, though. The current PE10 still exceeds 20 which means that the market is either over valued and will fall further or that the market is less risky and prices will remain high and returns will be lower.

Bernstein gives us no hint on divining bear markets.

Regards,
Prometheuss
(Can someone check my math?)

William Bernstein provides the first part in The Intelligent Asset Allocator in Table 2.3 (reproduced below), copyright 2001.

Unfortunately, he's wrong:

Expected inflation adjusted return

0 to 3% T-Bills
3% All other high-quality bonds

These are the inflation-adjusted return of NEW bonds. The return of old bonds can diverge rather substantially from these numbers.

Worse bear market loss

None T-Bills
None (short duration) All other high-quality bonds
10% (long duration)

Again he's wrong, except on T-bills. Increases in expected inflation (among other things) cause increases in the prime rate, causing existing fixed-rate bonds en masse to decline in market value - a bear market.

And then factor in inflation.

Invest $1000 in a long-term bond index, and $1000 in the S&P500, and hold them both for three years; you are more likely to have lost purchasing power in the bond investment, than in the stock investment.

Again he's wrong, except on T-bills. Increases in expected inflation (among other things) cause increases in the prime rate, causing existing fixed-rate bonds en masse to decline in market value - a bear market


What if he is talking about holding the bond to maturity and not selling it? Then you lose no value except to inflation right?