Such changes would have huge implications for taxpayers and public employees. A reduction in the investment projections would put more pressure on taxpayers and workers to support the two retirement systems, which already are significantly under-funded. The less the California Public Employees' Retirement System and the California State Teachers' Retirement System believe they'll earn from their investments, the more they have to depend on other sources.

For instance, CalSTRS long has assumed it would earn 8 percent a year on its investments – but is now being urged by consultants to scale back that figure. Board members are sweating the potential impact.

"If we go less than 8 percent, we're going to make it up where?" board member Kathy Brugger asked her colleagues at a meeting last month. The board plans to decide in September if it should lower the estimate.

Even if the two funds don't lower their investment forecasts, they're still looking to taxpayers to help them recover from recent investment losses.

CalPERS, which has the authority to set its contribution rates, is imposing increases of about 6 percent to 10 percent on the state as well as the local governments that use it for pensions. CalSTRS, which needs the Legislature's approval to raise rates, plans to petition lawmakers for an increase sometime next year. Employee contributions might go up as well.

In rethinking their investment forecasts, the two funds are responding to the combined $100 billion they lost in the fiscal year that ended last June. Although they've recouped some of their losses, the prospect of a tepid economic recovery is prompting pension funds across the country to lower their predictions. In CalPERS' case, some of the pressure is coming from the Schwarzenegger administration, which accused the fund of not facing up to its problems.

Nationwide, several public pension funds are cutting or are about to cut their investment forecasts, consultant Nick Collier told the CalSTRS board last month.

Collier, of Milliman Inc., said CalSTRS' current forecast of 8 percent is reasonable but "somewhat aggressive." Most of his clients assume annual returns of 7.75 percent, and several are cutting the forecast by a quarter point. While he said there's no one right answer, he suggested CalSTRS reduce its forecast.

That put board members in a bind. Several indicated they would be willing to reduce the forecast but openly fretted about the implications.

"We need to do something," said board member and state Controller John Chiang. "I also don't want to overreact."

Predicting investing results is difficult – but also vitally important. About 75 percent of CalSTRS' money comes from its investment returns, with the rest coming from school districts, the state and teachers. Although results in any given year can vary widely from what's predicted, the forecasts have an almost sacred quality to them. They represent a pension fund's fundamental long-term expectation of its financial health.

They change incrementally – maybe a half a percentage point or less – and hardly ever.

Although both California funds re-examine their assumptions every three years or so, CalSTRS hasn't altered its forecast since 1995. The CalPERS estimate has held steady at 7.75 percent since 2003.

Those forecasts were thrown seriously out of whack when the financial markets collapsed in late 2008. CalPERS lost 23 percent of its portfolio in the fiscal year that ended last June, and CalSTRS lost 25 percent.

"Negative 25 percent wasn't quite what we were thinking of," said Mark Olleman, another Milliman consultant, in remarks to the CalSTRS board.

With markets as volatile as they are, some experts say it's silly for pension funds to agonize over quarter-point or half-point changes in their forecasts.

"There's no way we can estimate expected rates of return with that degree of precision," said Brad Barber, a professor of finance and pension expert at the Graduate School of Management at UC Davis.

But even minor changes in investment assumptions could have enormous consequences for the two California funds, which control a combined $331 billion in assets and have billions of dollars in pension obligations stretching decades into the future. Both systems are significantly under-funded, which means their liabilities outweigh their assets by tens of billions of dollars. Cutting their investment predictions makes that outlook worse.

If the forecasts are reduced, "your liabilities are going to look bigger," said Alicia Munnell, director of Boston College's Center for Retirement Research. "It matters a lot."

While CalSTRS expects to decide its forecast this fall, CalPERS is just beginning to examine the issue and doesn't plan a board vote until next February.

Yet CalPERS already is facing some pressure to lower its estimate.

One of CalPERS' outside investment managers, the chief executive of New York's BlackRock Inc., jolted CalPERS officials last summer by telling them that investment returns will be "subpar for many years."

Gov. Arnold Schwarzenegger, who wants to reduce pension benefits for newly hired workers, has called CalPERS' financial outlook "unsustainable." His economic adviser David Crane has said the fund's investment assumptions are "ungrounded in reality."

For now, CalPERS says it's comfortable with its 7.75 percent forecast.

"Is our assumption reasonable and prudent? Yes," Chief Investment Officer Joseph Dear told the board in November. Even with the stock market crash of 2008, the fund has earned an average return of 7.9 percent a year the past two decades, he said.

But in comments to the board last month, Dear said the CalPERS investment staff will re-examine its forecast as part of a broader look at its investment strategy that it conducts every three years.

"Is the 7.75 percent return objective reasonable and attainable, and should we be considering other possible outcomes?" Dear said.