By
Mark Hulbert,
MarketWatch
A recent academic study is perhaps the most comprehensive analysis yet of whether bucket approaches are superior to other retirement funding strategies.
The idea behind bucket approaches is to divide your retirement portfolio into several different buckets, with transfers between those buckets carefully structured. The first bucket would contain enough cash and liquid assets to fund several years of retirement, with other buckets containing riskier assets. That first bucket would be replenished only when those riskier assets perform well. Bucket approaches have great intuitive appeal, since they insulate the retiree from concern about the impact a bear market might have on their retirement.
The traditional alternative to bucket strategies is to have just one retirement portfolio, or bucket, divided between cash and riskier assets according to a predetermined allocation—say 60% stocks, 30% bonds, 10% cash. A key part of these non-bucket portfolios, of course, is the regular rebalancing that takes place to bring actual allocations back in line with the intended allocation.
The study that provides the comprehensive analysis of bucket strategies was published on the Social Science Research Network in December, by Javier Estrada, a professor of finance at the University of Navarra’s business school in Barcelona: “The Bucket Approach for Retirement: A Suboptimal Behavioral Trick?” Professor Estrada constructed three different bucket strategies and compared them to 11 different non-bucket strategies.
To analyze each strategy, Estrada assumed a 4% withdrawal rate over a 30-year retirement. He studied all possible 30-year periods in the U.S. and 20 other countries between 1900 and 2014. In no case did he find that any of the bucket strategies came out on top. The best performer was the portfolio with 60% allocated to stocks and 40% to T-bills.
Consider the failure rate, which measures the number of years in which the strategy had insufficient assets with which to fund the required withdrawal. The 60% stock/40% T-bills portfolio had a 0% failure rate, indicating that in no 30-year period in the U.S. or foreign country was the approach unable to fund the withdrawal. The three bucket strategies’ failure rates ranged from 4% to 5%. (See accompanying chart.)
A 4% to 5% failure rate might not seem alarming, but it is. Such a rate means that the strategies can be expected to fail in one of every 20 to 25 years. Assuming a 30-year retirement, that means that each of the bucket strategies Estrada studied can be expected to fail at least once. That should certainly give retirees pause.
Why are bucket strategies more likely to fail than the non-bucket strategies? The answer has to do with the periodic rebalancing transactions periodically undertaken by the non-bucket strategies. Such rebalancing, of course, involves selling a portion of outperforming assets and purchasing more of underperforming ones, in order to bring the portfolio’s allocation back in line with its intended allocation. This rebalancing means that the approach constantly is buying low and selling high, which needless to say is a winning strategy.
The bucket portfolios, in contrast, at most engage in only half of these rebalancing transactions, since they do sell some of their outperforming assets to fund withdrawals. But they do not engage in the other half of rebalancing—purchasing more of underperforming assets. And that puts a significant damper on these portfolios’ longer-term returns.
So retirees are paying a high price for the comfort and ease of accounting that accompany a bucket strategy. Therefore, Estrada recommends that financial planners “attempt to convince retirees that however plausible, comforting, and easy to implement the bucket approach may be,” a portfolio with an appropriate asset allocation and which is periodically rebalanced “would be just as easy to implement and would ultimately make them better off.”
