On Personal Finance | Better ways to crunch retirement numbers

You don't hear this one every day: Americans are saving

too much

for retirement!

Well, not all Americans, but some.

This is the view of a small group of academics who think the standard "static" financial calculators found on mutual fund and brokerage Web sites overestimate the nest eggs needed to fund a comfortable retirement. Newer "dynamic" calculators give a more realistic view of the odds that a retirement plan will meet the investor's goals, they say.

I don't believe that many people are saving too much. The low savings rate and small balances reported for retirement accounts prove otherwise.

But the current debate over the two types of calculators isn't just an academic exercise. An investor who thinks he's falling far short of his goal may give up - or go for broke with investments that are far too risky.

And imagine if you'd deprived yourself of good vacations, a nice home and better schooling for your children, only to find you'd saved more than you had to. How sad.

It could happen: If you overestimate your expenses in retirement by, say $1,000 a month, you'll need to add $300,000 to your nest egg. That will require a lot more saving in the years before you quit working.

I'll list a few of my favorite calculators in a moment, but first, a little on the debate.

All retirement calculators rely on assumptions, built into the calculator, or entered by the user, about things like inflation, tax rates and the rate of return the user will receive on his or her investments.

The problem is that the numbers you put in are sure to be wrong. Even if a stock-and-bond portfolio were to return 8 percent a year on average - a standard assumption - it will do better some years and worse in others.

Consider an example offered by T. Rowe Price Group Inc., the mutual fund company. Invest $100 with a 10 percent return for each of two years and you'll end up with $121. If, instead, you lost 5 percent the first year and made 25 percent the second, you still would seem to be averaging 10 percent a year. But because you'd start the second year with $95 instead of $110, you'd wind up with $118.75 even though you'd made a whopping 25 percent the second year.

Over 20, 30 or 40 years, this kind of volatility can leave you with a far different outcome than you'd expected if you assumed you'd make 8 percent every single year.

The newer dynamic calculators address the problem by running the figures hundreds of times, each time with different assumptions about returns, inflation and so forth. Afterward, they tell you what percentage of the calculations achieved the financial goal. Because of this odds-calculating feature, these are known as Monte Carlo simulations, for the famous casino town.

One of the best-known of these calculators is T. Rowe Price's free Retirement Income Calculator at http://www3. troweprice.com/ric/RIC/. I'd start with this because it's very easy to use. Don't do it just once. Play with the inputs to see how saving a little more each month or retiring a little later will affect the outcome.

Then move on to the much more elaborate Advanced FIRECalc at http://firecalc.com/index.php. I've looked at many free Web calculators, and this is my favorite because it gives the user vast control and the option of adding a number of variables, like accounting for an inheritance you expect someday.

And you can set it to match the mix of assets you actually own - big - and small-company stocks, foreign stocks, Treasury bonds, corporate bonds.

For an even better look at retirement investing, you may have to spend some money. One of the best-known services is offered by Financial Engines at https://www.financialengines. com. It costs $150 to $300 a year, but you may be able to get it for free through the mutual fund company that provides your 401(k) or similar plan. The Vanguard Group Inc. and some other financial-services companies offer the calculator free to investors who meet certain requirements.

And don't assume the old-fashioned static calculators are worthless. They can be useful if they give the user enough control and consider enough variables. I'm very impressed, for example, with the Retirement Planner in the inexpensive Quicken financial program found at any computer store.