n my practice, I’ve found that the assumed rate of return represents the most consequential variable in retirement planning. Get this number wrong, and you either risk undersaving and running out of money or oversaving and unnecessarily sacrificing your current quality of life. After analyzing thousands of retirement scenarios and studying historical market data, I’ve developed a framework for selecting appropriate return assumptions based on your specific portfolio structure, risk tolerance, and time horizon. This guide will walk you through the evidence-based approach I use with clients to determine realistic return expectations.
Table of Contents
The Historical Context: What Markets Have Actually Delivered
Before projecting forward, we must understand historical returns. Since 1926, the S&P 500 has delivered an average annual return of approximately 10.1% before inflation and 6.9% after inflation. However, these long-term averages mask significant volatility and sequence risk.
Historical Annual Returns (1926-2023):
- Large-cap stocks: 10.2% nominal, 6.9% real
- Small-cap stocks: 11.9% nominal, 8.6% real
- Long-term government bonds: 5.4% nominal, 2.1% real
- Intermediate-term government bonds: 5.1% nominal, 1.8% real
- Inflation: 2.9% average annually
These numbers represent arithmetic means. The compound annual growth rates (CAGR) that actually determine wealth accumulation are approximately 0.5-1.0% lower due to volatility drag.
The Modern Realities: Why Historical Returns May Not Repeat
Several structural factors suggest future returns may be lower than historical averages:
Valuation Levels: With CAPE ratios near 30 compared to historical averages around 17, starting valuations suggest lower future returns. Research by Professor Robert Shiller indicates that high starting valuations predict below-average returns over the following decade.
Interest Rate Environment: While rates have risen from historic lows, the normalized risk-free rate remains below historical averages, pulling down expected returns across all asset classes.
Economic Growth: Demographic trends (aging populations) and productivity growth challenges suggest potentially lower economic growth, which correlates with market returns.
Globalization Effects: Increased correlation between global markets may reduce diversification benefits.
My Evidence-Based Framework for Return Assumptions
After synthesizing research from Dimson, Marsh, Staunton, Ilmanen, and others, I use the following real (after-inflation) return assumptions for retirement planning:
Conservative Portfolio (30% stocks/70% bonds): 2.5-3.0% real return
Moderate Portfolio (60% stocks/40% bonds): 3.5-4.0% real return
Aggressive Portfolio (80% stocks/20% bonds): 4.5-5.0% real return
These assumptions are approximately 1.0-1.5% below historical averages to account for current market conditions and the sequence of returns risk.
The Impact of Fees: The Silent Return Reducer
Many investors overlook the devastating impact of fees on long-term returns. A 1% annual fee doesn’t reduce your return by 1%—it reduces your ending wealth by approximately 20% over 30 years.
Fee Impact Calculation:
$1,000,000 portfolio over 30 years at 6% return:
No fees: 1,000,000 \times (1.06)^{30} = $5,743,491
With 1% fee: 1,000,000 \times (1.05)^{30} = $4,321,942
Reduction: $1,421,549 (24.8% less)
I always use net-of-fee return assumptions in retirement planning. If your portfolio costs 0.5% annually, reduce your expected return accordingly.
The Sequence of Returns Risk: Why Average Returns Mislead
The order in which returns occur matters tremendously during the distribution phase. Poor returns early in retirement can permanently impair portfolio survival, even if average returns eventually normalize.
Example: Two identical average returns, different sequences
Portfolio A: -15%, -5%, +25%, +18% (average = 5.75%)
Portfolio B: +18%, +25%, -5%, -15% (average = 5.75%)
For a retiree withdrawing 4% annually, Portfolio A might fail while Portfolio B succeeds, despite identical average returns.
I address this by using Monte Carlo simulations rather than straight-line return assumptions and by recommending more conservative return estimates for the first 10-15 years of retirement.
Asset Class Specific Assumptions
Based on current market conditions, these are my recommended nominal return assumptions for major asset classes:
US Large-Cap Stocks: 6.5-7.5% nominal (4.0-5.0% real)
International Developed Markets Stocks: 7.0-8.0% nominal (4.5-5.5% real)
Emerging Markets Stocks: 8.0-9.0% nominal (5.5-6.5% real)
US Aggregate Bonds: 4.5-5.0% nominal (2.0-2.5% real)
TIPS: 2.0-2.5% real (already inflation-adjusted)
These estimates incorporate current yield levels, valuation metrics, and economic outlook.
The Withdrawal Rate Connection
Your assumed rate of return directly impacts sustainable withdrawal rates. Using the historical 4% rule with reduced return assumptions may be overly optimistic.
Revised Safe Withdrawal Rates Based on Current Expectations:
- 3.0-3.5% for portfolios with lower expected returns
- 3.5-4.0% for average expected returns
- 4.0-4.5% for above-average expected returns (rare in current environment)
I typically use a 3.25-3.75% initial withdrawal rate for clients with moderate portfolios retiring today.
Personalization Factors: Adjusting for Your Situation
Your specific circumstances should influence your return assumptions:
Investment Costs: Reduce expected return by your total expense ratio + advisory fees + trading costs.
Tax Status: Taxable accounts should use after-tax return assumptions. A 7% pre-tax return becomes approximately 5.5% after taxes for many investors.
Behavioral Factors: Investors who panic-sell during downturns may experience returns 1-2% below market averages.
Time Horizon: Longer time horizons can justify slightly higher return assumptions due to increased risk capacity.
The Retirement Phase Glidepath
I don’t use a single return assumption throughout retirement. Instead, I implement a decreasing return assumption glidepath:
Years 1-10 of retirement: Use conservative assumptions (0.5-1.0% below long-term expectations)
Years 11-20: Use moderate assumptions
Years 21+: Use long-term expectations
This approach better accounts for sequence risk in early retirement when portfolio damage is most impactful.
Practical Application: Case Study
Let’s examine a 60-year-old couple with a $1,500,000 moderate portfolio (60/40 stocks/bonds) planning to retire at 65:
Return Assumptions:
- Stocks: 7.0% nominal (4.5% real)
- Bonds: 4.5% nominal (2.0% real)
- Portfolio: (0.6 \times 7.0) + (0.4 \times 4.5) = 6.0% nominal
- After 2.5% inflation: 3.5% real return
- After 0.4% fees: 3.1% real return
Withdrawal Rate:
Initial withdrawal: $60,000 (4.0% of $1,500,000)
Adjusted for conservative assumptions: $52,500 (3.5%)
Using Monte Carlo simulation, this withdrawal rate has an 85% success probability over 30 years.
The Behavioral Aspect: Managing Expectations
Perhaps most importantly, realistic return assumptions prevent destructive investor behavior. When actual returns exceed expectations, investors feel confident. When returns fall short of optimistic assumptions, investors may abandon their strategy at the worst possible time.
I find clients who use conservative assumptions are more likely to stay the course during market downturns, ultimately achieving better results than those with optimistic assumptions who panic-sell.
The Verdict: Recommended Rates for Current Conditions
Based on current market conditions, economic outlook, and historical evidence, I recommend these real (after-inflation) return assumptions for retirement planning:
Pre-Retirement Accumulation Phase (10+ years from retirement):
- Conservative: 3.0-3.5% real
- Moderate: 4.0-4.5% real
- Aggressive: 5.0-5.5% real
Early Retirement (Years 1-10):
- Conservative: 2.5-3.0% real
- Moderate: 3.5-4.0% real
- Aggressive: 4.0-4.5% real
Late Retirement (Years 20+):
- Use long-term historical averages minus 0.5%
Remember that these are planning assumptions, not predictions. The actual returns you experience will vary, often significantly. The purpose of conservative assumptions isn’t to predict the future but to create a robust plan that can withstand less-than-ideal outcomes.
I revise these assumptions annually based on market valuations, interest rate environments, and economic conditions. The most successful retirement plans are those built to survive disappointing returns, not those requiring optimal outcomes to succeed.
