Tuesday, December 21, 2010

Stress-Testing a Retirement Plan

It’s easy to make a plan for spending retirement savings that looks good on paper. But there are many unknowns when it comes to investing. A little bad luck in the first few years of retirement can sink your plan.

Proper stress-testing of a retirement plan is the main theme of Jim Otar’s book Unveiling the Retirement Myth. Instead of testing a plan with average return figures for stocks, bonds, and other asset classes, Otar simulates retirement plans using actual market data since 1900. These simulations check the results when a virtual copy of you retires in each year to see what happens.

Suppose you have a $1 million nest egg and plan to withdraw $50,000 rising with inflation each year. If you assume your investment returns will be 5% above inflation each year then the $1 million will last forever. But this is unrealistic. Returns vary unpredictably. This retirement plan that looks so great shrivels up when Otar’s analysis shows how often you’d run out of money before age 90.

As I explained in an earlier post, I have concerns about the way that Otar reduces dividends in actual historical returns before using them, but the basic idea of checking a retirement plan against historical data is useful.

According to Otar, those who create a single projection of a retirement savings balance aren’t necessarily naive: “most of these plans are produced for only one reason: to sell dreams. Many financial planners do that, many stockbrokers do that, mutual funds do that, hedge funds do that, many pension managers do that.” I guess it’s hard to sell a financial plan to a client if it shows a 25% chance of the client eating cat food.


  1. I can see why Otar would be fond of annuities if he is so concerned about stock market returns and longevity risk.

    What would be a ballpark withdrawal rate going forward to maintain principal using a 100% stock portfolio? I would guess 3% if income is derived mainly from dividends and capital gains. To me, that seems about right. If a focused investor is able to add 2% to market returns by focusing on quality companies, he/she could probably bump that up to 4%.

    I remember you had some very good posts about this, and came up with a 4% of current portfolio number?

  2. @Gene: Otar's focus was on choosing a fixed percentage of the starting portfolio value. I think it makes more sense to take a percentage of current portfolio size.

    I found two of my posts on this subject:

    Modifying the 4% Rule

    An Experiment with the Modified 4% Rule

    I didn't come up with the 4% figure, though. I just analyzed it.

  3. @Financial Cents: Thanks. Yes, I have done a stress-test of sorts on my finances. I figure I'm good right now until my early 70's with high probability. But that's not good enough. So, I continue to work. Even if I had enough I'd probably work at something anyway. I haven't made a final decision on what constitutes the correct stress-test. But I don't have to figure this out until I actually contemplate full retirement.