There's no one-size-fits-all recipe for maintaining a bucket portfolio. Rather, the "right" bucket-maintenance regimen depends on the investor's preferences, the desired complexion of the portfolio throughout retirement, as well as the end goals for that portfolio. Figuring out where you come down on the following three decisions will help you devise a bucket-maintenance approach that fits with your own situation.

Decision 1: How will you extract assets from your long-term portfolio?

One of the key decisions--important for anyone maintaining an in-retirement portfolio, not just those employing the bucket strategy--is this: How will you extract cash from your portfolio on an ongoing basis? While many retirees like the idea of subsisting off of the income stream from their portfolios, in practice generating a livable income stream has gotten tricky as yields have slunk lower.

Option 1: Reinvest income and capital-gains distributions back into the long-term portfolio; rely exclusively on rebalancing.

This is what I've called the "strict constructionist total return" approach to portfolio management, in that the retiree doesn't rely on any of the portfolio's "organic" dividend and capital-gain distributions for cash flow, but instead reinvests them and uses rebalancing proceeds for cash-flow needs. This approach tends to have the most support in the academic community, in that a portfolio that prioritizes total return over income will tend to be more efficient, and rebalancing allows tight control over a portfolio's asset allocations on an ongoing basis.

Yet there will be years in which rebalancing doesn't yield sufficient cash flows to meet living expenses, pointing to the value of holding a cash bucket for those instances. My bucket portfolio stress tests, which employed a pure total return and rebalancing regimen, also illustrate the virtue of rebalancing individual positions when they exceed target levels, not just when the whole portfolio's asset allocation is out of line. The latter types of rebalancing opportunities are too rare to supply regular cash flows.

Option 2: Use income distributions to help meet annual cash flow needs; use rebalancing to supply additional cash needs. Reinvest capital-gains distributions.

This hybrid tack is one employed by many financial advisors. The retiree spends stock dividends and income payouts from bonds, while relying on rebalancing to meet additional cash flow needs.

The income distributions supply a baseline of income into the household and are somewhat predictable, whereas mutual fund capital-gain distributions aren't predictable at all. On the downside, spending income distributions may not be desirable in depressed markets; from the standpoint of maximizing returns, the retiree may be better off reinvesting them into the portfolio and using the cash bucket instead.

Option 3: Reinvest income and/or capital-gains distributions when the market is low and spend cash instead; spend distributions when valuations are lofty.

This is a more opportunistic, market-sensitive approach. When stocks or bonds are perceived to be cheap, the income and capital-gains distributions are plowed back into the portfolio. Because rebalancing proceeds may be scant in such environments, too, the retiree would tap cash in such years to meet living expenses. Yet in more richly valued market environments, the retiree spends those distributions (or at least the portfolio's income distributions). This approach holds a lot of appeal at first glance, but it's also a version of market-timing: Arriving at appropriate valuation levels for whole asset classes is notoriously difficult to do.

Decision 2: What type of "glide path" will you employ?

Identifying a desired in-retirement asset-allocation mix--or glide path--is crucial when establishing a bucket-maintenance regimen, as it will dictate where to pull cash for living expenses and where to deploy excess profits. In my model portfolio stress tests, for example, I plowed excess profits into cash and short-term bonds, which led to an ever-higher weighting in safe securities during periods of market appreciation. Such a setup might not be desirable for every retiree.

Option 1: Glide toward a more conservative portfolio stance.

This is the standard approach to asset allocation before and during retirement, with bonds taking up an increasingly large share of the portfolio as the years go by. This approach addresses the fact that as the portfolio's value diminishes during retirement, a higher percentage of it will need to be held in safe securities to meet income needs during retirement. Moreover, rising health-care outlays later in retirement may necessitate a higher level of liquidity than was needed in the middle retirement years. From a practical standpoint, retirees can achieve a more conservative glide path by rebalancing out of equities on a regular basis. Yet there are a couple of key drawbacks. First, this approach might not suit retirees with very large portfolios, for whom leaving money to heirs or charity is a major goal; such retirees may desire a consistently high level of equity exposure in order to grow those assets they won't use during their lifetimes. Second, the investor who enters retirement with an aggressively positioned portfolio is most vulnerable to sequence-of-return risk.

Option 2: Maintain a static stock/bond allocation throughout retirement.
Under this approach, the retiree would regularly rebalance back to a static allocation--say, 50% stocks, 50% cash and bonds. In normal, appreciating market environments, maintaining such an allocation would generally entail spending or rebalancing out of equities on an ongoing basis, as stocks will typically generate higher gains than bonds. From a practical standpoint, maintaining a static asset allocation across a shrinking portfolio may mean that a retiree doesn't have sufficient assets in safe securities. For example, an 80-year-old retiree with a 10-year time horizon who's targeting a 50% stock/50% bond and cash mix may not have sufficient portfolio liquidity.

Option 3: Glide toward a more aggressive portfolio stance.

This approach--more cash- and bond-heavy at the outset, more aggressive later in retirement--may help avoid sequence-of-return risk and in turn the viability of a portfolio plan. On the downside, such an approach may be psychologically unappealing and also tees up the possibility that a retiree will need to draw from stocks if they happen to be at a low ebb later in retirement.

Decision 3: When will you tap bucket 1?

The linchpin of the bucket system is bucket 1, which holds enough cash to fund one to two years' worth of living expenses. Having those liquid assets helps a retiree avoid tapping an asset class--whether stocks or bonds--while it's at a low ebb, and helps ensure that even in a catastrophic market environment, the retiree's standard of living needn't change radically. Plus, those liquid assets provide a valuable intangible in rocky markets: peace of mind. That bucket 1 shouldn't be allowed to run dry isn't up for debate, but when it should be tapped is a matter of personal preference.

Option 1: Spend bucket 1 on an ongoing basis and periodically refill.

Under this approach, a retiree uses bucket 1 to fund ongoing living expenses, then refills it, either throughout the year or as part of a once-yearly portfolio checkup. This is the "sleep-easy" approach, in that if a source of cash flow from the portfolio becomes disrupted—for example, if a reliable dividend-paying company cuts it payout—the retiree doesn't have to make up for the shortfall on the fly. She can be systematic in deciding how to make up for the shortfall when she does her "portfolio maintenance" at year-end. On the other hand, this approach makes bucket maintenance a bit more labor-intensive than tapping bucket 1 only in catastrophic market environments.

Option 2: Spend bucket 1 only in catastrophic market environments.

This is the approach that Harold Evensky, the originator of the bucket approach, says he uses with clients in his practice. The retiree relies on income, rebalancing proceeds, or a combination of the two to fund ongoing living expenses. Evensky uses rebalancing as the main source of cash flow for his clients. He or she taps bucket 1 only in extreme market environments, when the usual sources of cash flow--income, rebalancing proceeds, or both--are insufficient to provide living expenses. The big advantage of this approach is that it requires less oversight; if bucket 1 isn't being spent on an ongoing basis, the retiree won't need to worry about regularly refilling it.