August 17, 2010
I will explain the calculations using a zero goal-risk example. At zero goal risk, TIPS are better-performing than any combination of stock and bond funds until the 23rd year. ATI for those 23 years using TIPS is $18.90. After the TIPS segment is determined, the software can analyze various asset allocations and the ATI under each allocation to achieve the goal of $1 withdrawal from year 24 to 30 at zero goal risk.
The results from the software-based optimization at 0% and 20% goal risk based on monthly sequence of returns analysis are presented below. You can view other combinations at the freely accessible online demo (no registration required) of the advice engine here. The withdrawals are pre-tax amounts and taxes are ignored for all calculations.
Figure 5: Recommendations for a retirement distribution portfolio at different risk levels
The withdrawal rate for zero goal risk is 4.6% (calculated as $100K/ $2.2 mm), which is much higher than the SWR of 4% widely quoted in the industry. When we increase the goal risk to 20%, the withdrawal rate rises to 4.9%. I am restricting my analysis to large-cap stocks; small-cap stocks would lead to higher withdrawal rates as well.
Other retirement distribution decision points
With goal risk-based portfolio selection, we have a base framework to make optimum decisions for other distribution decision points outlined by Joel Bruckenstein in his chapter titled “The Search for Software” in the book Retirement Income Redesigned. The recommendation is made on the combination of decision points that minimizes the ATI or equivalently maximize the withdrawal rate for a given goal risk constraint. Here is a brief outline of how this framework can be used for those decisions:
- Order of drawdown: The choice of an effective drawdown-order strategy across taxable, tax-deferred and tax-free accounts can be made by choosing the strategy that sustains the maximum withdrawal rate at a given goal risk constraint. All strategies, such as partial IRA withdrawals to take advantage of lower tax brackets in early years, can be analyzed in the framework.
- Annuities: Since the ATI for a systematic withdrawal portfolio and incremental ATI for each additional year one might live can be determined, the advisor has the metrics to make decisions regarding annuities. The software can optimize the mix of timing and quantity of annuities in an overall asset allocation.
- Timing and choice of method of Social Security payments: Various alternatives for the start of Social Security and the appropriate method for one’s spouse can be included in the optimization process to minimize the ATI.
- Reverse mortgages: Various options in using a reverse mortgage, such as immediate or use in case of shortfall, can be modeled to see their impact on the cumulative ATI.
Benefits and applications of the new framework
Pre-retirees can develop their retirement plan using this framework. The advantages of the framework for the retirement distribution problem are briefly outlined below:
- Better planning for uneven income needs
- Support for mental accounting by planning for different goals at different risk levels. For example, one can plan to cover basic expenses at a low goal risk and discretionary expenses at higher goal risk.
- Support for flexible planning horizons – any length horizon can be supported
- Joint optimization of all interacting decision points, for example the drawdown order across different types of accounts impacts the asset allocation in each account.
Although the analysis in this article is based on historical sequence-of-returns, my framework does not depend on them. The framework will work with scenarios generated using any means.
Manish Malhotra is President of Fiducioso Advisors Inc., a firm serving RIAs and other advisors. Fiducioso specializes in creation and management of private retirement portfolios for retail clients of the advisors. This article has been distributed for informational purposes only and should not be considered as financial advice.
Feedback on this article can be provided via email to , or by posting comments on twitter mentioning @fiduciosoAdv.
The author would like to thank Michael Kolfman, my colleague at Fiducioso, for implementing the software-based optimization for multiple withdrawals.
Display article as PDF for printing.
Would you like to send this article to a friend?
Remember, if you have a question or comment, send it to .
