“I’ll have a great retirement— provided I die by Friday.”

For those who have little or no savings, this grim witticism pretty much sums it up. For the rest of us, retirement spending is a delicate dance. Uncertainties here (What kind of return will I get on my savings? Will inflation be greater than I expect?). And unknowns there (When will I die? Will I end up in a nursing home?).

So how do we puzzle through it?

There’s no definitive answer, but the tools for thinking about it are improving. You can explore with me— and do your own exploring— by visiting the website firecalc.com. The site has the calculating engine to do what you and I can’t do “in our spare time at home”— figure the odds of going broke in retirement.

Let’s start from the probabilistic wisdom financial planner William Bengen introduced 20 years ago. Bengen was seeking a safe withdrawal rate for a 30-year retirement. He found that a typical retirement portfolio could survive for 30 years if you withdrew 4 percent at the start. After that, you adjusted the withdrawal upward for inflation.

If we assume such a portfolio on the firecalc website, that’s the result we get because it uses the same method the Bengen used. (Ten Nerd points if you shouted, “That’s stochastic!)

On firecalc, a balanced portfolio (60 percent stocks, 40 percent 5-year Treasurys) has a 97.4 percent chance of survival for 30 years. It will have those survival odds if you take \$40,000 in the first year from a \$1 million portfolio and then adjust the \$40,000 withdrawal upward by the rate of inflation each year.

So you’ll be pretty secure if you retire at 65. The problem: when your portfolio survives you 97.4 percent of the time, you’ll leave behind money you could have spent.

How much? As much as five times what you retired with. The average amount left behind will be 1.5 times what you start with.

We can fix that by spending more. Raise your initial withdrawal amount to 6 percent. But the odds of portfolio survival drop to 50.9 percent. Not good. The average amount left behind will be about 20 percent of what you started with. The largest will be a hefty 3.5 times starting value.

It would be great to spend 50 percent more in retirement. But the risk of going broke will be more than most want to face. So we spend less. Retirement will be secure, but less comfortable.

Is there a work-around that will make things better?

Yes, there are two easy methods. The first is to follow the required minimum distributions dictated by the Internal Revenue Service. Yes, it's that simple. As I pointed out earlier columns (here and here), this method is likely to provide a rising income for a long retirement. You’ll never run out of money, but if you make it to 95, your income may be shrinking.

The second method is also simple. Be flexible. This means you spend a little less after your portfolio has had a bum year. Withdraw the greater of 90 percent of your previous year’s withdrawal or 6 percent of your current portfolio and your long-term spending may be more, or less, than the inflation adjusted initial 6 percent. But you’ll never run out of money.

Yes, there can be a lot of handwringing about the uncertainty here. But don't reflexively spend less so you can enrich your heirs more. Why? Because there are several other safety valves on future spending:

• in a couple, the odds are against even one person living 30 years. Using the Vanguard longevity tool, for instance, there is only an 18 percent chance that either person in a 65-year-old couple will live 30 years.
• living expenses decline when a member of a couple dies.
• living expenses decline as we age. For most people the declines in regular consumption are greater than the increases in medical spending.
• shelter is our biggest expense. Many older couples happily reduce their shelter spending as part of a retirement plan or as part of widowhood.

Flexibility is forced on some. Or we exercise it happily. Either way, it can be as important as the return on our savings.