Rekenthaler Report

# When Vanguard's Retirement Income Calculator Stopped Making Sense

How to break an investment tool--but gain insights from it, anyway.

Try Harder!
The president of the American Finance Association, Dr. David Hirshleifer, noticed something peculiar about Vanguard's Retirement Income Calculator. By its reckoning, a hypothetical 55-year-old investor who saved at the tool's highest possible rate, earning the highest possible return, holding the highest possible current retirement assets, and willing to settle for the lowest possible income-replacement rate … would fail. Her projected monthly income would fall short of the projected goal.

I doubted that Vanguard had botched the math. The second-order calculations for retirement-income forecasts can be tricky--for example, if and how to model portfolio volatility--but the basics are straightforward. The investor owns this much money, is saving at that rate, and therefore will possess the following estimated amount on his 65th birthday. Annuitize that total and divide by 12 to make the payments monthly. It's a spreadsheet exercise.

Vanguard did not botch the math. These are the tool's minimum and maximum settings:

• Income \$400,000
• Savings Rate 15%
• Current Assets \$2 million
• Expected Annual Return 10%
• Replacement Rate 60%

Enter those figures, and the tool projects \$15,789 in monthly retirement income, which sounds about right. But the goal is \$20,000 monthly, which is exactly right (\$240,000 annually, that being 60% of the investor's working salary). Even the most generous Social Security estimate won't square that circle.

The Wrong Settings
The problem is the limits. The calculator's cap on current assets is too low. In the grand scheme of things, a \$2 million nest egg is terrific. (It has been surprisingly achievable, too; higher-income workers who have participated in a 401(k) plan for 30 years, have invested the legal maximum, and have held mostly equities will be near the mark.) When trying to replace even a 60% slice of \$400,000, though, that cool \$2 million is insufficient, if the goal lies only 10 years away.

One interpretation of this seemingly obvious mistake is that Vanguard wishes to fail its tool's users. Its business, after all, is collecting investment assets. Vanguard doesn't generate sales from people who are satisfied with their retirement plans and who spend their spare change instead on second homes. Financial-services firms sometimes exaggerate their customers' retirement needs. Perhaps Vanguard has done the same.

Perhaps. On the other hand, Vanguard isn't known for its irresponsibility, and, when errors occur, follies are the more common cause than conspiracies. Also, I can see why Vanguard would wish to set a relatively low limit on current assets. Permit the slider to show \$10 million, and there's the user who may immediately quit, deciding that the tool wasn't made for people like him. So, I can understand how the decision might have been made. But it was a poor one, because a calculator that fails the apparent best case looks pretty silly. ("Ludicrous," writes Hirshleifer.)

The caps on the settings sharply reduce the options available to our hypothetical investor, but they do not eliminate them entirely. We can't give her more assets, further reduce her replacement rate, or push her savings rate past 15%. We can, however, enter her expected Social Security payments and have her retire later.

The Social Security Administration's quick calculator projects monthly receipts of \$2,668, expressed in today's terms. Such forecasts embed many assumptions, but there's no point in fussing those details. Adding the Social Security checks boosts the forecast monthly income to \$18,457, putting our investor \$1,543 short of her goal. Having her retire one year later, at age 66, nearly eliminates the gap. Extending until age 67 closes the deal.

Time heals many wounds. That said, the apparent certainty is artificial. Who is to say that the investor's original retirement date was not year-end 2006, and that by bumping it two years, to December 2008, she inadvertently slashed the assets available to her upon retirement? Even calculators that run simulations show midpoint results that imply more inevitability than exists.

(Overly) High Expectations
What's more, we projected a 10% annual return. That has indeed been the norm for several decades now, albeit with volatility, but it would be exceedingly brave to expect The Great Bull Market to continue indefinitely. Also, that figure implies an all-stock portfolio until retirement, at which point presumably Vanguard's program waves its electronic hands, and translates the lump sum into an annuity rate. (That's what I did when I worked on investment calculators.)

That 10% forecast must be reduced, substantially. I suggest halving it. With 10-year Treasury notes yielding less than 2%, the investor will need to put half her portfolio into equities, and have those equities thrive, to achieve the seemingly modest average of 5%. One can hope for more, but should not expect it.

Halving the investment return destroys the plan. Time no longer meets the need. To reach the targeted level of retirement income, the hypothetical investor must work until age 81, all the while receiving her full salary. Such an event is more likely to occur than pigs flying, but less than the Cubs winning the World Series. That 60% replacement rate won't be happening.

Exceptions to the Rules
As problems go, that's not a bad one. The calculator won't permit a lower replacement rate because 60% is already regarded as borderline insufficient. But trying to survive on \$24,000 annually when accustomed to \$40,000 is quite different than doing so when each figure carries another decimal point. Assuming that she does not face high debt, either as mortgage payments or through other obligations, our investor will live quite nicely on \$200k per year, thanks much.

The moral: Retirement-income calculators serve middle-tier workers well. Their settings are calibrated accordingly, and their rules of thumb, such as the appropriate range of replacement rates, are sound. They are unsuited for those on the outside, though. The analysis that is used for the middle fractures when applied to either the wealthy or the poor.

John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.