Back in February, I spent an hour interviewing FPA's Robert Rodriguez, who made no effort to disguise his contempt for advisors, pension consultants and others who had dropped his fund because he could not find enough stocks selling at reasonable prices in early 2008 and had let cash build up in his portfolio. Rodriguez was so disgusted with everything-from asset management industry orthodoxy, to American consumers' penchant for living beyond their means, to politicians in both parties spending money they didn't have-that a month later he announced he was taking a sabbatical in 2010.

After two equity market corrections of nearly 50% in one decade, a growing number of advisors are starting to rethink the conventional buy-and-hold wisdom that contends one should set an asset allocation policy for a client-say 75% equities and 25% bonds for pre-retirees and 60%/40% for retirees-and stick with it through thick and thin. Next week, our sister publication, Journal of Indexes, will publish an article by Rob Arnott claiming that over the last 40 years ending February 2009, investors who bought 20-year Treasury bonds and simply rolled them over would have marginally beaten equities over the same time frame.

Arnott ultimately concludes that stocks do beat bonds, but by a far smaller margin than many advisors and investors thinks, like 2.5%, not 5%. Of equal significance, he challenges the beliefs underlying conventional wisdom about market behavior. Specifically, the notion that long-term equity investors typically enjoy long periods of outperformance, occasionally marred by sharp but often brief bear markets a la 1987, is not supported by the evidence. Indeed, Arnott's evidence reveals the opposite pattern may be more accurate: that long-term investors are more likely to experience lengthy "periods of disappointment, interrupted by some wonderful gains."

Needless to say, the 40-year period Arnott chooses from 1968 to 2008 was an era characterized by two bear markets for equities, the first running from 1966 to 1982 and the next running from 1999 to the present. That translates into 26 years or so of equity bear markets, which set up fixed-income securities for decades of equity-like gains, and only 14 years of bull markets for stocks.

The time period Arnott selected naturally favors bonds, and the past performance of the previous decade could set stocks up for some relatively easy gains. But 40 years is half of many clients' lifetimes and accounts for a much longer chunk of their investment lives.

Arnott doesn't argue that the next 40 years will be a replay of the last 40, but he does think that many investors and some advisors need to re-examine their expectations.