Do Streamlined Bucket Portfolios Make the Grade?

A Bucket Portfolio that was worth $1.5 million in 2000 would have soared to nearly $2.1 million at the end of 2017. It would have also supplied more than $1.4 million in cash flows over that same period, according to my recent "stress test."

That sounds good, but is it, really? True, a bucket approach would have delivered on a main goal for many retirees: stability and growth of principal as well as real cash flow to meet in-retirement living expenses. From that standpoint, our hypothetical retirees are winning the game.

But would alternative strategies have performed even better? My 10-fund Bucket Portfolio, and the maintenance strategy I used to rebalance and extract cash flow, are admittedly a bit unwieldy; would a radically simpler approach have worked, too? In a rising market, how much higher would returns have been on a fully invested portfolio--one without a cash "bucket"?

Those are the questions I sought to answer with a second round of stress tests over the past week. I looked at how an ultrasimple portfolio would have fared as well as a fully invested portfolio with no cash component.

A Radically Simpler Portfolio Retirees have better things to do than manage their portfolios. While I'd argue that the portfolio featured last week would require just one good checkup a year--to rebalance and extract cash flows for the next year's living expenses--it does have 10 separate holdings. Would a single-fund option deliver similar results with less hassle?

I aimed to answer that question, employing a single fund--

To facilitate an apples-to-apples comparison with my earlier bucket stress test, I ran the simulation from 2000 through 2017. I combined the Balanced Index fund with a small cash bucket consisting of two years' worth of living expenses for our hypothetical retirees. Assuming a $1.5 million portfolio, they'd extract $60,000 in the first year of retirement, which translates into a $120,000 cash hoard at the outset of the portfolio simulation, in 2000. I also applied the same maintenance regimen as I did with last week's stress test. You can view the simulation in this spreadsheet (Microsoft Excel required).

Long story short: Not having discrete holdings to snip for living expenses put the simplified portfolio at a disadvantage relative to the multifund portfolio, where there was almost always something performing reasonably well and in need of pruning. At the end of the 18-year period, the simple portfolio was more than $400,000 smaller than the multifund portfolio.

The first years of simplified portfolio simulation illustrate the issue. As seasoned investors well remember, March 2000 marked the start of a massive sell-off in technology and Internet-related stocks, which had enjoyed an enormous runup in the second half of the 1990s. The Balanced Index lost money--although not catastrophic amounts--in 2000, 2001, and 2002, meaning that our hypothetical retirees quickly blew through their cash reserves in order to leave the depreciated positions undisturbed. The fact that the portfolio declined in value in those years also meant that our hypothetical retirees went without an inflation adjustment on their withdrawals in those years.

In contrast, the multifund portfolio fared better during that period. Whereas

The Takeaway: In part because of that poor start, the simple portfolio lagged the multifund portfolio by a meaningful margin over the time frame examined. It did so despite maintaining a higher allocation to equities over the 18-year period, and equities outperformed. In a nutshell, the portfolio encountered sequence-of-return risk; its particular structure left it with fewer tools to recover from that initial stressful market as robustly as was the case with the multifund portfolio.

Yet there are ways that an investor in search of a simplified portfolio could mitigate that risk. One would be to simply hold a larger cash bucket alongside the single fund, but that could drag on long-term returns. Alternatively, a retiree could simplify but not as radically as a one-fund holding--holding single discrete bond, domestic equity, and international equity fund positions, for example. That would give the retiree leeway to trim appreciated long-term positions while leaving others intact; for example, from 2000-2002 a retiree could draw from the bond fund as well as cash.

A Fully Invested Portfolio As I noted in last week's article, the cash bucket in my multifund portfolio didn't come in handy very often. Instead, it weighed on returns in a period in which stocks and bonds both performed quite well. I noted that a fully invested portfolio would have performed better, at least on the basis of returns alone.

But how much better? To see, I ran another stress test, jettisoning Bucket 1 and instead deploying the money into short- and intermediate-term bonds instead. Apart from its lack of a cash stake, the portfolio was the same as the multifund portfolio that I examined last week, and the maintenance regimen was largely the same, too. The major difference was that because the fully invested portfolio includes no cash, I pulled from the short-term bond fund when rebalancing proceeds were insufficient to meet cash-flow needs. You can view the simulation in this spreadsheet (Microsoft Excel required).

As I expected, the fully invested portfolio performed better than the portfolio with the cash bucket. By the end of 2017, its value stood at $2.2 million, more than $100,000 bigger than the portfolio with the cash component. While there would have been a few years in which our hypothetical retirees would have needed to raid the short-term bond fund to cover living expenses and top up depressed long-term positions (2001, 2002, 2008, 2011), this portfolio--like the portfolio with the cash component--consistently had a surplus of assets in the short-term bond fund. And because bonds generally outperformed cash over the 18-year period in our observation, that gave returns a boost relative to the portfolio with cash holdings.

The Takeaway: Over long periods of time, stocks and bonds have outperformed cash, so it shouldn't be surprising to see that a fully invested portfolio would have outperformed the portfolio with the dedicated cash component. However, it's worth noting that the 2000-2017 period featured a declining interest-rate environment, giving bonds a strong tailwind that may not be readily repeated. (Indeed, interest rates have begun to trend upward so far in 2018.) Thus, a fully invested retiree could confront the possibility of having to withdraw from bond holdings that have declined in value. Moreover, our simulation measured returns alone, and didn't capture the peace of mind that would likely have accompanied the portfolio with the higher cash cushion.

At the same time, the return lag of the bucket portfolio underscores the importance of not holding more cash in retirement than is necessary. While I sometimes shorthand the cash bucket as holding "two years' worth of living expenses," it's really "two years' worth of portfolio expenditures." By extension, retirees who are receiving a healthy share of their living expenses from pensions and Social Security won't need to maintain a significant cash cushion. This article discusses how to use your own portfolio spending to determine your allocation to cash and other assets.