We had an inquiry from a Money Magazine reporter doing a story on online retirement calculators. The basic question, "what is the best way to use these calculators?"
Let me start by saying that I am a big fan of online calculators. They have given the consumer far more information than they once had. Mortgage calculations used to be a total mystery to the average consumer. At least now, there’s a way for them to do the calculations themselves (not that anyone does).
That being said, I think retirement income calculators are HIGHLY flawed. They suffer the same problem as cash value life insurance illustrations. Namely, they assume linear returns. Let me give you an example using the three links that three different online retirement income calculators. Let’s say that you have a client who just turned 65 and is about to retire on nothing but their portfolio of $1,000,000. No Social Security, no pension. They need an income of $60,000. We will assume a diversified portfolio with an expected return of 9%.
Pretty reasonable, right? We're simplifying things by taking taxes out of the picture initially. It's just a matter of crunching the numbers, right? I actually entered this little experiment assuming that each of the calculators would produce similar results. I couldn't be more wrong.
If you plug those numbers in the MSN calculator, you get a portfolio that never runs out. In fact the graph is a straight line, never growing, never shrinking. See the image below. I guess is seems reasonable. You earn 9%, less 3% for inflation, and withdraw 6%. Pretty simple math. Unfortunately, way too simple/
Ask any financial planner if a 6% distribution with a 9% expected return is reasonable, and you will nearly always get a no. The problem is not the 6% withdrawal rate in itself (although some would say that it is too high), but the myth of the perpetual portfolio. What kills portfolios is not down years, but taking money out in down years. 9% annual returns does not actually mean 9% every year. We would love it if it did, but volatility is a reality in markets. I want to say "especially recently", but actually no, not especially recently. Always.
What happens when the market is down (that never happens, right?), and the portfolio is down 10% in a year. Then the client takes his routine 6% withdrawal. Now the $1,000,000 portfolio is only worth $840,000 ($100,000 in market loss, and $60,000 in withdrawals). Next year he still needs his $60,000, meaning that the remaining $780,000 needs to earn a return of 28% to get back to even. 9% the next year isn't going to cut it. In fact, 19% (9% is the expected, and 10% to make up for last year’s loss) is still way short.
Here’s where it gets even more nutty. Plug in the same numbers in the Bloomberg calculator, and you get a totally different result. At age 100, you will have $2,898,278. How is that possible? They calculate inflation differently. Probably more accurately, but you can see how small changes in the calculations make an enormous difference. MSN simply takes inflation off of your return. Bloomberg inflates your income by that amount each year. It's clear why that is the right way to do things. Inflation doesn't reduce your return, it increases your need. Here’s the Bloomberg graphic:
Additionally, the MSN and Bloomberg calculators did not touch taxes. When we enter the same assumptions in the FINRA calculator, which does account for taxes, we find that our client runs out of money at age 89. Here’s that one:
What a difference. Three calculators, same assumptions. The results:
- portfolio size is maintained forever, into perpetuity
- portfolio grows nearly three times its size over 35 years
- portfolio is depleted in 24 years
To answer the original question, “what's the best way to use these calculators?” I would say, cautiously and collaboratively. Be skeptical of the results generated by one calculator. Use all three, plus any others such as the one at bankrate.com.
Of course, the better way is to use a monte carlo simulation. Generally used by financial planners, this is also available online. Instead of telling you how much you will have or not have at some arbitrary age, it tells you the probability of success. Using many of the same numbers at Flexible Retirement Planner, the probability of success is 55.7%. Not very good.
Here’s that graphic:
What was different in these assumptions? Was is based on a lower return? Higher distribution? More taxes? None of the above. In fact, it was based on a higher average return of 10%, the distribution was exactly the same, and distributions were assumed to be strictly from a tax free (Roth) portfolio. The difference was volatility. And not that much of it. The standard deviation used was 14.3%, about the same as the S&P 500 over the last 10 years.
Comments