A Bucket Portfolio Stress Test

Using so-called buckets to organize your retirement portfolio by time horizon has caught on among many Morningstar.com readers, in large part because of the psychological benefits. 

By maintaining a dedicated cash pool to draw upon for near-term living expenses--the linchpin of every bucket program--retirees can ride out fluctuations in the long-term portion of their portfolios. They can also switch on automatic withdrawals from their cash buckets to simulate a paycheck, and that reliable income stream can provide comfort in good markets and bad. Finally, a bucket strategy gets retirees away from what I consider to be an unhealthy form of mental accounting--focusing on income at the expense of total return. 

But does bucketing actually work in practice, meaning does it meet a retiree's cash needs while also generating a satisfactory level of return? 

I began creating sample bucket retirement portfolios--consisting of both traditional mutual funds and exchange-traded funds--just last year, so they don't have a sufficiently long track record to observe. But at the risk of being accused of data-mining, I took a look back at one of my bucket portfolios to see how it would have performed during the stress test of the 2008 financial crisis and in the subsequent equity market recovery. I tested the most aggressively positioned of the model bucket portfolios--the one geared toward retirees with 25-year time horizons. That portfolio's heavy equity weighting has been a big help recently as stocks have rallied, but its equity-heavy stance also left it with a big hole to claw its way out of in 2008. 

The exercise yielded some encouraging results during this admittedly arbitrary time period. The headline is that at the end of 2012, the value of the aggressive bucket portfolio was ahead of the starting value of the portfolio, even though our fictitious retiree was also withdrawing 4% of the portfolio, adjusted for inflation, per year. Of course, there's no guarantee that a bucket portfolio started today will fare as well, particularly given rock-bottom bond yields and the fact that equity market valuations are nowhere near as attractive as they were following the 2008 market crash. Specific categories, such as the Treasury Inflation-Protected Securities in portfolio holding  Harbor Real Return , appear to have limited upside potential today. 

The Starting Portfolio
The starting portfolio, featured in depth in this article, is geared toward young retirees with an anticipated time horizon of 25 years. They will use a 4% withdrawal rate, with an annual inflation adjustment, which translates into a $60,000 withdrawal from their $1.5 million portfolio in the first year of retirement. They have a high risk tolerance. 

The sample portfolio features the following holdings in the following dollar amounts. 

Bucket 1: $120,000
$120,000: Cash (In my original portfolio, I used cash plus  PIMCO Enhanced Short Maturity ETF (MINT) for bucket 1. But because the PIMCO fund has only been around since late 2009, I used Vanguard Prime Money Market as a proxy in the return simulation.) 

Bucket 2: $480,000
$130,000:  T. Rowe Price Short-Term Bond (PRWBX) 
$150,000:  Harbor Bond (HABDX) 
$100,000: Harbor Real Return  
$100,000:  Vanguard Wellesley Income (VWINX)     

Bucket 3: $900,000
$400,000:  Vanguard Dividend Growth (VDIGX) 
$200,000:  Harbor International  (HAINX) 
$100,000:  Vanguard Total Stock Market Index (VTSMX) 
$125,000:  Loomis Sayles Bond (LSBDX) 
$75,000:  Harbor Commodity Real Return   (Note that this fund wasn't launched until 2008, so I've used near-clone  PIMCO Commodity Real Return (PCRAX) in its place in the return simulation.) 

Cash Flow and Rebalancing Rules
One of the keys to making a bucketing strategy work is to have a plan for bucket maintenance: how you'll refill bucket 1 if it becomes depleted and how you'll manage rebalancing. 

For the purpose of back-testing our sample portfolio's returns, we used the following system: 

• Reinvest all dividends and capital gains. 
• Rebalance positions at the end of the calendar year if the position size exceeds 110% of its starting value in 2007. 
• Use rebalancing proceeds from appreciated positions to meet living expenses. 
• If rebalancing proceeds are insufficient to meet living expenses, withdraw cash from bucket 1 to meet planned living expenses. 
• If rebalancing proceeds exceed living-expense needs, use to refill bucket 1 if lower than starting level ($120,000). 
• If rebalancing proceeds meet living expenses and bucket 1 is full, use rebalancing proceeds to add to positions that have declined the most (or appreciated the least) since 2007. 

The Stress Test: Year by Year
For a detailed look at the year-by-year results of our bucket stress test, you can download this spreadsheet (Microsoft Excel required) or this PDF. What follows is a synopsis of the portfolio's performance in each calendar year as well as the steps I took to maintain the portfolio.

2007
Starting Portfolio Balance: $1,500,000
Ending Portfolio Balance: $1,648,498
Amount Withdrawn: $60,000 

Domestic-equity stocks deliver just so-so performance this year, but the portfolio's international and commodity positions generate very strong returns. Harvesting gains from those two positions as well as Harbor Commodity Real Return, all of which quickly appreciated 10% above their starting value, easily meets this year's living expenses of $60,000. The additional $12,408 of rebalancing proceeds gets moved into the year's lowest-returning position, cash.

2008
Starting Portfolio Balance: 1,588,498
Ending Portfolio Balance: $1,288,729
Amount Withdrawn: $60,000 

Owing to routs in the equity and commodities markets, the portfolio sheds $300,000 of its value during the worst year of the financial crisis. High-quality bonds are a bright spot, however, and rebalancing proceeds from Harbor Bond help feed the desired $60,000 distribution. (We forgo the inflation adjustment this year.) Withdrawals from bucket 1--cash--provide the bulk of the funds needed for living expenses, however. 

2009
Starting Portfolio Balance: $1,228,729
Ending Portfolio Balance: 1,480,868
Amount Withdrawn: $61,800 

Stocks begin clawing their way back in 2009, and bonds also generate strong gains as the Federal Reserve Bank embarks upon its quantitative easing campaign. (The only casualty? The cash position, whose yield shrivels to next to nothing.) The portfolio's pure equity holdings are still well lower than their initial position sizes, but trimming the bond holdings and Vanguard Wellesley Income allows us to meet living expenses of $61,800 (our original $60,000 distribution plus a 3% inflation boost), top up the cash stake, and add a little to the position that has fallen the furthest since the portfolio's inception, Harbor International. 

2010
Starting Portfolio Balance: $1,419,068
Ending Portfolio Balance: $1,562,901
Amount Withdrawn: $63,654 

Owing to strong returns across the board, our portfolio ends the year at a higher level than when we started. Lightening up holdings in Vanguard Wellesley Income and Loomis Sayles Bond, which have appreciated more than 10% from their initial values, help meet part of our desired living expenses; a smaller distribution from the cash pool (bucket 1) gets the total distribution up to our $63,654 target ($61,800 adjusted for inflation). Our equity holdings enjoy another year of strong gains but are untouched because they're not yet 110% of their original value. 

2011
Starting Portfolio Balance: $1,499,247
Ending Portfolio Balance: $1,549,334
Amount Withdrawn: $65,564 

Although bond returns were muted and international stocks posted losses owing to European debt woes, high-quality U.S. stocks had a solid year, and so did our Vanguard Dividend Growth and Wellesley Income funds. Pruning gains from strong performers such as Dividend Growth, Harbor Bond, and Harbor Real Return, all of which hit our rebalancing target, provide ample income for living expenses. Rebalancing proceeds also enable us to top up our cash bucket and add to our stake in Harbor International, which is still well below our original position size. 

2012
Starting Portfolio Balance: $1,483,770
Ending Portfolio Balance: $1,637,996
Amount Withdrawn: $67,531 

All of the major asset classes enjoy robust gains, save for our cash holdings, which return next to nothing. Several of our holdings--including Vanguard Dividend Growth, Loomis Sayles Bond, and Vanguard Wellesley Income--hit our rebalancing target. With the rebalancing proceeds, we meet living expenses and add to our worst-performing positions since inception, Harbor International and Harbor Commodity Real Return.

Takeaway 1: Diversification Proves Its Mettle
Among the key conclusions from the exercise is that, even more than the bucket framework, holding and rebalancing a diversified portfolio in retirement can help provide decent performance and deliver a steady cash flow under a variety of market conditions. During the depths of the bear market, for example, rebalancing out of the portfolio's appreciated stakes in high-quality bonds helped meet planned withdrawal amounts. Those rebalancing proceeds, plus cash withdrawals from bucket 1, obviated the need to withdraw from equity assets while they were down. Maintaining a sizable equity stake, meanwhile, allowed the portfolio to rebound more than it would have if it were focused strictly on bonds and other income-producing assets, particularly as yields plummeted coming out of the financial crisis.

Takeaway 2: Maintenance Strategy Matters--a Lot
The back-testing of our bucket strategy also highlights the role that bucket maintenance plays in the portfolio's performance--and how many different variations there are to consider. The maintenance strategy featured here--which used rebalancing proceeds to top up the portfolio's cash stake before adding to depressed positions--arguably reduced the portfolio's total-return potential in exchange for keeping cash in the range of two years' worth of living expenses. (In essence, it values safety and peace of mind more than growth potential.) An investor who's more interested in generating high long-term total returns and less concerned about safety and liquidity, on the other hand, might take the opposite tack, plowing rebalancing proceeds into depressed positions and then topping up the cash stake. Alternatively, one might rebalance based on percentage variations relative to asset-allocation targets, rather than variations relative to initial position sizes, as featured here. (Note that if a retiree wishes to make his portfolio more conservative as it is drawn down, the rebalancing strategy would need to change to allow for larger cash and bond components.) In future articles, we'll test alternative bucket-maintenance and rebalancing strategies.

Takeaway 3: It's More Complicated Than It Looks
Our model portfolio is less complicated than most retirees' situations, in that most people come into retirement with multiple accounts, both taxable and tax-sheltered; couples' planning adds even more wrinkles. That means real-life bucket setup and maintenance is going to be more complex than is the case for the single portfolio featured here, owing to asset-location and withdrawal-sequencing issues and the need to take required minimum distributions from tax-deferred accounts, among other considerations. That shouldn't deter you from employing buckets in your own retirement distribution plan, but our sample portfolio arguably oversimplifies what it takes to implement such a system. We'll be discussing additional aspects of bucket setup and maintenance in future articles.

See More Articles by Christine Benz