Derek Tharp: An Alternative Approach to Calculating In-Retirement Withdrawals

Episode Summary

Executive Summary: Morningstar hosts retirement researcher Derek Tharp on dynamic retirement-income planning and related investment choices. Tharp favors predefined, risk-based spending guardrails over static Monte Carlo targets, arguing retirees can adjust spending as conditions change. He also discusses sequence risk, Social Security timing, portfolio diversification, annuities, tax-loss harvesting, ESG investing, robo-advisors, and the behavioral effects of planning tools.

Main Topics: Dynamic retirement planning and guardrails: Tharp argues that retirees should use ongoing plans with predefined spending adjustments rather than treating a one-time probability-of-success estimate as a fixed prescription. Risk-based guardrails can make responses to portfolio changes clearer and provide peace of mind. Probability of success and underspending: For retirees who update plans and adjust spending over time, a probability-of-success target has a different meaning than in a static plan. Very conservative targets may favor legacy accumulation over current spending, potentially leading to underspending. Sequence risk, Social Security, and early-retirement spending: Front-loaded withdrawals—such as spending more before delayed Social Security begins—can increase sequence risk. Tharp suggests holding several years of near-term spending in stable assets and retaining the option to claim Social Security earlier if circumstances warrant. Portfolio construction, bonds, and TIPS: Tharp considers claims about the demise of the 60/40 portfolio overstated and favors stocks plus high-quality bonds for many retirees. He sees TIPS as especially useful for clearly defined future expenses, while emphasizing real returns and accepting bond-price fluctuations. Annuities and retirement income: Simple income annuities may be underused because they can provide mortality-credit benefits, though exchanging a lump sum for income presents psychological barriers. Their suitability depends on household income needs, taxes, and strategies such as Roth conversions. Tax-loss harvesting and portfolio trade-offs: Tax-loss harvesting can help, especially for higher-bracket investors and during downturns, but its value is often overstated when future capital-gains taxes and portfolio drift are ignored. Tharp is particularly cautious about direct indexing letting tax considerations drive investment choices. Behavior, ESG, and automated advice: Tharp says planning tools and how they frame risk can affect confidence and spending decisions. ESG may help investors feel comfortable investing, even if its real-world impact is uncertain; robo-advisors can automate tax management and provide continuity if an investor experiences cognitive decline.

Key Arguments: A static Monte Carlo probability-of-success figure often fails to explain when or how a retiree should adjust. Predefined guardrails make potential spending changes concrete and can reduce uncertainty. In Tharp's analysis, retirees who continually update plans and adjust spending had surprisingly similar maximum and minimum spending outcomes when targeting a 20% versus a 95% probability of success over a 30-year retirement. In this dynamic setting, the target more strongly tilts toward current income or legacy than it signals a simple chance of depletion. A bad market year alone does not establish a damaging sequence of returns; the risk becomes clearer after sustained adverse outcomes. Front-loading withdrawals can heighten that risk, so cash and bonds for near-term needs and flexibility around Social Security claiming can help. Higher nominal yields do not necessarily make bonds more attractive in real terms when inflation is also higher. Tharp remains comfortable with a stock-and-high-quality-bond portfolio and views the criticism of 60/40 as overdone. TIPS are particularly appropriate when matched to specific future liabilities, such as a planned expense. For general retirement spending, a diversified bond portfolio may be sufficient. Simple income annuities can strengthen retirement income through mortality credits, but consumer reluctance to give up a lump sum limits uptake. Annuities should be considered alongside portfolio withdrawals and tax planning. Tax-loss harvesting has legitimate but conditional value. Assessments should account for eventual taxes, investor tax brackets, market conditions, and the danger that tax goals distort portfolio construction. Financial-planning tools can unintentionally encourage underspending, while ESG preferences or robo-advisor automation may help some people invest consistently and maintain a plan despite behavioral or cognitive challenges.

Data Points: Ongoing probability-of-success targets compared: 20% and 95% - Tharp said a comparison over a simulated 30-year retirement found the differences in maximum and minimum spending were far smaller than expected when retirees adjusted plans continuously. Retirement period in the spending comparison: 30 years - The period over which the ongoing 20% and 95% probability-of-success targets were compared. Illustrative distribution-rate guardrails: 5% starting rate; adjust at 6% or 4% - Tharp's simplified example: reduce spending if the distribution rate rises to 6%, or consider increasing it if the rate falls to 4%, returning toward 5%. Near-term reserves suggested in a bucket approach: 5 to 7 years of anticipated spending - Tharp described targeting this amount in stable investments such as cash and bonds to help cover early retirement spending and manage sequence risk. Social Security delay example: Age 65, defer for 5 years - A hypothetical retiree could reserve assets for the period before Social Security begins, while retaining the option to claim earlier if markets perform poorly. Longer pre-Social Security period: 15 to 20 years - Tharp noted that managing sequence risk may be more challenging for someone retiring at 50 or 55. Capital losses deductible against ordinary income: $3,000 - Tharp noted the annual amount of losses that may be used to offset ordinary income, with unused losses carried forward. Capital-gains tax bracket: 0% - Tharp said tax-loss harvesting can be harmful for investors in the 0% capital-gains bracket, for whom harvesting gains may be more appropriate. Potential savings-rate change associated with ESG comfort: 1% to 2% - Tharp offered this as an example of how an investor who feels better about a portfolio might increase their savings rate.

Pivotal Quotes: "But what you really get with the guardrails is predefining: okay, here's where my lower guardrail is. Here's where I'm going to cut back if needed. Here's where my upper guardrail is and where I'm going to increase my spending if it's appropriate." — Derek Tharp: His central case for setting spending adjustments in advance rather than relying on a probability-of-success result alone. "And once you do that, it's just a totally different situation. The metrics have really different meaning in a way that I think might not be well appreciated." — Derek Tharp: Tharp explains why conventional Monte Carlo metrics can be misleading when retirees plan to make ongoing adjustments. "I just don't think it should be what's driving the portfolio, and you want to make sure it's right for you as an individual." — Derek Tharp: His conclusion on tax-loss harvesting: consider its potential benefit, but do not let the tax strategy dictate investment choices.

Implications: Retirement plans are more useful when they specify how spending can change as markets and income evolve. Investors should weigh real returns, taxes, behavior, and personal goals—not rely on static success scores or tax tactics alone—and review annuities and automated tools as parts of an individualized plan.