This article also appeared in:
By J.R. Robinson
Retirement income planning apps have proliferated in recent years, as people have sought out new tools to help them investigate how various investing and spending strategies may affect the values of their nest eggs upon retirement and sustainability thereafter.
Some of these apps boast a bevy of features, implying that more features mean a better, more accurate tool. This is not always the case. Often, these features are not only superfluous but may even damage reliability.
It is, of course, also true that calculators that are too simplistic in their design may be likely to produce unrealistic results. For instance, overlooking inflation, ignoring advisory fees and investment expenses, and failing to account for the possibility of negative returns all may render the output from many common calculators worthless to the user.
However, at a certain point, additional complexity may become the enemy of reliability. To the extent that certain inputs in the retirement software may cause the user to make flawed assumptions, output reliability may again suffer. The following are four input features that are commonly regarded as desirable, but which may actually lead to “garbage in, garbage out” analysis.
On the surface, accounting for the impact of income taxes in portfolio spending calculations seems like a fairly basic and important input consideration. The problem is that many users, including financial advisors, overestimate this number, most commonly by simply entering the client’s marginal tax rate. While this factor may be accurate if 100% of the withdrawals are coming from qualified accounts, such as pre-tax employer 401(k)s or traditional IRAs, financial planning best practices in retirement spending dictate balancing withdrawals from after-tax savings, tax-free Roth accounts and other retirement accounts so as to minimize the client’s overall tax liability. (See Vanguard’s publication, Spending from a Portfolio: Implications of Withdrawal Order for Taxable Investors.)
Through careful planning, it may be possible for many retirees to pay little or no tax on their nest-egg withdrawals throughout much of their retirement. Because of the propensity for user error, tax calculations are a factor that may be best left off the calculator feature list.
Users who wish to account for the impact of taxes and who can accurately estimate their tax liability may instead include this figure in their spending needs along with other living expenses, such as food, clothing, housing, travel, insurance and property taxes.
Incorporating a budgeting feature into a retirement planning calculator is a fast track to killing the user experience. Most families have no idea what their itemized living expenses are each year, and forcing them to estimate them in the input section of a retirement calculator creates complexity and the potential for inaccurate guesses. Simply put, budgeting should be an entirely separate exercise.
Additionally, while most retirees have no idea how much they spend on individual expense categories each year, it has been my experience that most do have a pretty good idea of the total amount they need to maintain their standard of living over and above what they receive from Social Security, pensions, and other guaranteed or predictable sources of income.
Social Security integration and optimization
Like budgeting, calculating a client’s optimal Social Security claiming strategy is a complex process that should be undertaken separately from and prior to retirement income planning analysis. As evidenced by the recent changes to the “file and suspend” and restricted application filing rules, making optimal decisions regarding claiming strategies often requires solid knowledge of arcane rules. Educated guesses or outright errors tend to lead to flawed and unreliable results.
Instead of adding to the input complexity of a retirement spending calculator, a simpler and better means of incorporating Social Security in the portfolio sustainability analysis is to derive the client’s portfolio withdrawal need by subtracting non-portfolio income sources (Social Security, pensions, rental income, etc.) for the client’s stated income need.
Some of the most popular retirement planning tools for financial advisors enable the user to import their real world portfolios into the application’s Monte Carlo simulations, a retirement planning technique that provides probabilities of different outcomes occurring based on certain assumptions.
The obvious benefit of this capability is that it enables the user’s real-world portfolio to much more closely mirror the model used in the application’s algorithm. In some cases, the applications allow for the accounting of asset classes as specific as international bonds, emerging markets, commodities and real estate investment trusts.
The closer the match, the more realistic the results, right? Maybe, maybe not. The Monte Carlo simulations in these calculators are based on assumptions about the expected or mean returns and standard deviations for each asset class. The problem is that mean returns and standard deviations, particularly for more esoteric asset classes, tend to be highly variable over time. If the expected return and volatility assumptions are wrong, the results may also be flawed. This is not to suggest that Monte Carlo simulations are useless, only that detailed portfolio integration may not be helpful.
In conclusion, when it comes to retirement planning software features, sometimes less is more. In my view, the best retirement planning applications include just enough inputs to cover the important factors that are within the user’s control and that may be accurately ascertained. If the input process takes more than a few minutes or requires the user to make educated guesses, it may be wise to question the reliability of the results.
At the end of the day, assessing the real-world applicability of any retirement planning calculator requires critical thinking and a clear understanding of the underlying assumptions in the app.
TO LEARN MORE VISIT NESTEGG.GURU