In the wake of that "interesting aberration … we decided, or certainly I decided, that we shouldn't miss trying to capitalize on odd behavior," said Grantham.
Building its business on estimates of the long-term value of asset classes, GMO looked for opportunities to take advantage of a market predilection for going "periodically crazy" that's pretty much hardwired into humans, Grantham maintains.
"We had a few 100,000 years to work out what it was going to be like and it's pretty steady now," he said. "We are given to periods of euphoria" and much shorter periods of "blind panic," he said.
"My estimate is something like 85% of the time the market is approximately reasonable, approximately efficient. Close enough. And then 15% of the time, it's not. That divides something like 11% or 12% crazy optimism and 3% or 4% crazy pessimism. And that seems to be the model," Grantham said.
Until recently, however, the gains Grantham's GMO garnered by sticking to its valuation discipline when the broader market was pushing earnings multiples to dizzying heights only emerged after the firm first endured considerable pain.
An early test: Japan's stock market bubble of the late 1980s, which saw GMO slash its exposure to Tokyo stocks to zero a full two years ahead of the 1989 market peak that lifted Japan's weight in MSCI's Europe, Africa, Far East benchmark index to about 60%.
"We had some fairly dreadful underperformance but we still beat the S&P, so we were forgiven by those early adopters of foreign equity," Grantham said.
And then of course "it all blew up and we made it all back in profit," he said. Japan's high flying stock market fell off a cliff from the start of 1990, days after the Bank of Japan raised rates. GMO, with zero exposure to a Japanese market accounting for more than half of the EAFE index, more than made up for the performance pain it had endured owning little or nothing of that market during the bull run of the second half of the 1980s.
That experience had only a modest effect on GMO's approach to the next bubble – the internet craze. At the end of that period, the firm would once again come under pressure for making the right moves too soon.
The Japanese experience "gave us a passing interest in trying to hold our fire, a bit," Grantham said. "We did not take a position" against the internet bubble until the price-earnings ratio on the S&P was higher than it had ever been before – reaching a record 21 and a half times earnings around January 1998.
"When it did that, we started to move distinctly against the U.S. market and the U.S. market punished us by going from (a price-earnings multiple of) 21 to 35 – which is not an insignificant record breaking – (giving) us a horrific two and a half years of underperformance," Grantham said.
"Then it all blew up, and in asset allocation accounts we actually made money in 2000, 2001 and 2002. So, we had made nice money by the time the S&P was minus 50, which is pretty cool, you have to admit," Grantham said. And it was that performance that brought in a lot of assets to GMO between 2003 and 2006, he said.