Research

Much of our time is spent running the programs we sell so that we are certain we completely understand the products. We also use these products to provide research to our customers. In instances where the research is of interest to a general audience we publish the papers here.

Many of the following studies are in Adobe PDF format. Please visit www.adobe.com if you need to download the Adobe Acrobat Reader (free of charge).

Variations on The Trinity Study

The first group of studies are variations on the Trinity Study published in the February 1998 issue of The AAII Journal entitled, "Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable." The Trinity study was written by Philip Cooley, Daniel Walz and Carl Hubbard, professors of economics from Trinity University in San Antonio, Texas. Whereas the Trinity Study uses pre-defined asset allocations, our studies find optimal asset allocations using a computational technique known as simulated annealing (available in WATS).

Study 1 - Maximum Sustainable Withdrawal Rates With Varying Historical Success Rates Using Large Cap Stocks, Corporate Bonds and US T-Bills. Ending Value Goal: Above Zero (Don’t Go Broke). Data from 1946 to 2000.

Study 2 - Maximum Sustainable Withdrawal Rates With Varying Historical Success Rates Using Large Cap Stocks, Corporate Bonds and US T-Bills. Ending Value Goal: Equal or Greater than Original Corpus. Data from 1946 to 2000.

Study 3 - Maximum Sustainable Withdrawal Rates With Varying Historical Success Rates Using Large Cap Stocks, Corporate Bonds and US T-Bills. Ending Value Goal: Equal or Greater than Inflation-Adjusted Corpus. Data from 1946 to 2000.

Other Studies

A Choice of Risks When Spending in Retirement, Optimizing the Plan for each Retiree - This paper compares two spending strategies, one in which withdrawals are made to maintain one's standard of living, the other in which withdrawals are a fixed percentage of the value of the account. In the first case, the risk is insolvency; in the second, a reduced standard of living. 

Saving for College, A Case Study - Five fictitious investors are studied to see how asset allocation and investor mindset affect outcomes. A short investment horizon of eight years is used to determine how much variance there can be between very conservative and very aggressive investors.

The Dangerous Retirement Calculator Assumption - Most calculators on the major financial web sites assume that you will, somehow, magically earn the same rate of return every year in the future. Wouldn't that be nice?

Think Small (Small Cap, That Is) - A recent study proclaimed that, if you omit certain historical data, Large Cap Stocks actually outperform Small Cap Stocks. This study shows that Small Cap Stocks clearly and consistently outperform Large Cap Stocks.

Links to Other Research

The two predominant methods of modeling retirement strategies are historical simulations (used by WATS) and Monte Carlo Simulation (MCS). Both methods have shortcomings. Dr David Nawrocki, a professor of finance at Villanova University, explores many of the problems of MCS in his paper, The Problems with Monte Carlo Simulation, available on-line at http://www.journalfp.net/fpajournal/jfp1101-art12.cfm.

The Trinity Study, with minor editing, is available at http://www.aaii.com/promo/mstar/feature.shtml. Phil Cooley, PhD, one of the three professors who wrote this paper, will send you an un-edited copy of the study if you request.

Jaye C. Jarrett is an expert on asset allocation and spending policies. His web site contains several research papers. It is available at http://jjarrett.home.texas.net.

Copyright © 2000, 2004  Zunna, Inc.

Last updated on 02/12/2004