Skip to content
Menu
Finabase Learn

Portfolio and risk · Long-run planning

The bucket approach: dividing retirement money by when it is needed

How a three-bucket retirement plan works, tested against a fixed 60/32/8 mix from 2008 to 2026, and what the buckets do and do not change about risk.

On this page
  1. The idea
  2. How it works
  3. Using it
  4. Where it fails or misleads
  5. Compared with its neighbours
  6. How Finabase does it
  7. Frequently asked questions
  8. Summary
  9. Check your understanding

The bucket approach divides retirement savings into parts according to when the money will be spent. Money for the next year or two sits in cash. Money for the years after that is in bonds. Money that will not be needed for a decade or more is in stocks. Spending comes out of the cash bucket, which is refilled from the others from time to time.

The appeal is easy to see. A retiree with two years of spending in cash can watch the stock market fall without having to sell anything, and can think of the stock bucket as money for their eighties. The approach is popular with financial planners for that reason. Whether it changes the risk of the plan, as opposed to the way the plan feels, is a separate question, and the answer from testing is mostly no. This lesson explains the method, runs it on real data and separates what it does from what it only appears to do.

The idea

Sequence of returns risk: why the order of returns matters comes from selling assets after they have fallen in order to pay for living costs. The bucket approach attacks that directly. If the next two years of spending are already in cash, a bear market does not force any sale. The stock bucket can be left alone to recover, and by the time it is needed it probably has.

The three buckets, in a typical arrangement:

Bucket Holds Covers Purpose
1 Cash, money-market funds, Treasury bills Years 1 to 2 of spending Pay the bills whatever markets do
2 High-quality bonds, bond funds Roughly years 3 to 10 Income and stability; refill bucket 1
3 Stocks Year 10 and beyond Growth to outpace inflation

For a retiree with $1,000,000 who spends $40,000 a year, two years in bucket 1 is $80,000 (8%), eight more years in bucket 2 is $320,000 (32%), and the rest, $600,000 (60%), is bucket 3. Seen another way, that is a portfolio of 60% stocks, 32% bonds and 8% cash.

That second view is the important one. The buckets do not create safety that the allocation does not already have. What differs between a bucket plan and an ordinary balanced portfolio is the rule for where withdrawals come from and how the parts are topped up.

Retiring in January 2008: buckets against a fixed mix $600k $1.0m $1.4m $1.8m $2.2m Buckets Fixed 60/32/8, rebalanced yearly Two years in cash, the rest rebalanced yearly 2010 2015 2020 2025
Three plans for $1,000,000 from January 2008 to October 2026: 60% SPY, 32% AGG, 8% Treasury bills, with $40,000 a year (raised with CPI-U each January) paid monthly. Nominal dollars.

The figure tests three versions through the 2008 crisis and the years after. In the crash they are nearly indistinguishable. Afterwards they separate, and the reason has nothing to do with the cash bucket.

How it works

Everything depends on the refill rule. Three common ones:

  • Refill from winners. Each year, top up cash from whichever of stocks or bonds has done well. In a bad year for stocks, use bonds; if both are down, let the cash run down.
  • Mechanical waterfall. Bonds refill cash; stocks refill bonds; on a fixed schedule.
  • Rebalance. Set the cash bucket to its target and rebalance everything else to a fixed ratio. This is a balanced portfolio with a cash reserve.

A worked example: retiring in January 2008

The data: monthly total returns of SPY and AGG and the Treasury bill rate, January 2008 to October 2026. Each plan starts with $1,000,000 split 60% SPY, 32% AGG and 8% bills, and pays out $40,000 a year in monthly instalments, raised each January with consumer prices (to $61,677 by 2026). Figures are in nominal dollars; taxes and fees are not included.

  • Buckets, refilled from winners. Spending comes from cash. Each January, if SPY rose over the past twelve months, cash is refilled to two years of spending from stocks. If SPY fell, cash is topped up to one year of spending from bonds. Nothing else is rebalanced.
  • Cash reserve, rest rebalanced. Each January cash is set to two years of spending and the remainder is rebalanced to 65% stocks and 35% bonds (the same 60:32 ratio).
  • Fixed mix. Rebalanced to 60/32/8 each January; spending comes from cash first.
Plan Lowest balance (February 2009) Deepest fall December 2013 December 2022 October 2026 Stock share: lowest to highest
Buckets, refilled from winners $679,735 −32.0% $980,718 $1,113,766 $1,500,117 42% to 60%
Cash reserve, rest rebalanced $670,737 −32.9% $1,074,132 $1,418,843 $2,080,640 50% to 69%
Fixed mix, rebalanced yearly $668,323 −33.2% $1,086,597 $1,443,700 $2,115,797 50% to 70%

In the crash the bucket plan helped a little. Its lowest balance was $679,735 against $668,323, a difference of about $11,000 on a million, and its deepest fall was 32.0% against 33.2%. Its holder sold no stocks during 2008 or 2009: spending came from cash, and the January 2009 top-up came from bonds. That is the promise of the method, and it was kept.

Afterwards it fell far behind. By October 2026 the unrebalanced bucket plan held $1,500,117 and the fixed mix $2,115,797.

Share in stocks: the bucket plan lets it drift 40% 50% 60% 70% 80% 90% Starting 60% Buckets Fixed mix Two years in cash, rest rebalanced 2010 2015 2020 2025
Share of each plan held in stocks at each month-end, January 2008 to October 2026.

The second figure shows why. At the bottom of the market in February 2009 the bucket plan was 46% in stocks and the fixed mix 56%: stocks had halved, the fixed mix had rebalanced into them that January, and the bucket plan had not. From then on the bucket plan sold stocks to fund every year's spending while its bonds sat untouched, and its stock share went as low as 42%. It went through the longest bull market on record with about half its money in stocks instead of 60% or more.

The version that rebalanced matched the fixed mix. With the same two-year cash reserve and the rest rebalanced yearly, the plan ended at $2,080,640, within 2% of the fixed mix. The cash bucket made almost no difference to the outcome. The rebalancing made all of it.

Other starting dates

Retired in Buckets, refilled from winners Cash reserve, rest rebalanced Fixed mix
January 2004 $1,515,252 $2,291,680 $2,339,389
January 2008 $1,500,117 $2,080,640 $2,115,797
January 2010 $3,360,537 $2,898,761 $2,834,370

Starting in 2010 the unrebalanced buckets came out ahead. With no crash at the start, stocks grew faster than they were spent, and the stock share drifted from 60% up to 84%. The plan earned more because it had quietly become a more aggressive portfolio, the same drift described in Rebalancing methods: calendar, threshold and what each one costs. Its deepest fall was 22.6% against 20.6% for the fixed mix.

In all three cases the result is explained by how much was in stocks along the way. Kitces called the bucket strategy an asset-allocation mirage for this reason.

Using it

  1. Size the cash bucket by spending, net of other income. If Social Security and a pension cover half of spending, two years of the remainder is what the bucket needs.
  2. Write down the refill rule. "Refill from stocks after a good year, from bonds after a bad one" is a rule. "I'll see how things look" is not, and tends to leave the cash bucket unfilled when markets are frightening.
  3. Rebalance the long-term buckets, or accept that the stock share will drift. Without it the plan becomes more conservative after a crash, which is the wrong direction, and more aggressive after a long rise.
  4. Check the implied allocation each year. Add up the buckets and compute the stock share. That figure describes the plan's risk; the labels do not.
  5. Do not oversize bucket 1. Five years of spending in cash is 20% of a portfolio at a 4% withdrawal rate, earning less than inflation in many periods. See Inflation and real returns: what an investment earns after prices.
  6. Use it for the reason it works. The evidence for buckets is behavioural. A retiree who can see two years of bills covered is less likely to sell stocks at the bottom, and avoiding that one mistake can matter more than any refinement of the allocation.

Where it fails or misleads

  • It does not reduce sequence risk beyond what its allocation does. In February 2009 all three plans were down by about a third. A larger cushion comes only from holding fewer stocks.
  • A cash reserve can run out. In this test the bucket plan's cash was down to nothing by the end of 2009. A bear market lasting longer than the reserve forces sales after all.
  • Refilling after a recovery means selling low anyway, just later. The January 2010 refill came from stocks that were still well below their 2007 level.
  • Cash drag. Woerheide and Nanigian's historical tests of buffer strategies found that, in most cases, holding a cash reserve did not raise the sustainable withdrawal rate, because of the return the cash gave up.
  • It hides drift. Because each bucket has a purpose, changes in the overall mix go unnoticed. The 2010 plan above reached 84% in stocks without anyone choosing that.
  • One period, one rule. A different refill rule, or a different eighteen years, would give different numbers. The general finding, that the refill rule and the implied allocation drive results, is what carries over.

Compared with its neighbours

Approach Where spending comes from Stock share over time Main benefit
Bucket approach A cash bucket, refilled by rule Drifts unless rebalanced Clarity; no forced stock sales in the short run
Fixed mix (total return) Whatever is overweight at each rebalance Held at the target Simplicity; buys low automatically
Glide path As fixed mix Follows a schedule De-risks around retirement
Income-only (live on yield) Dividends and interest Tilts towards high-yield assets Principal untouched
Bond ladder plus stocks Maturing bonds Rises as the ladder is spent Known cash flows for a set period

A fixed mix with spending taken from the overweight asset behaves much like a well-run bucket plan: after stocks fall, withdrawals come from bonds and cash; after they rise, from stocks.

How Finabase does it

Finabase does not offer bucket planning, model withdrawals or track a portfolio against spending needs.

A bucket plan can be watched with what it has. Separate portfolios can be created for each bucket, or one portfolio can hold everything: the portfolio page shows each position's market value and the cash balance, and its allocation bar shows the largest holdings and cash as shares of the total, which is the implied allocation that needs checking each year. The Treasury page gives current yields on bills and short-term notes, the return on a cash bucket. Price alerts on a stock fund can mark the level at which a refill from stocks would be due under a written rule.

Frequently asked questions

What is the bucket strategy for retirement?

A way of organizing savings into three parts by time: a cash bucket for the next year or two of spending, a bond bucket for the following several years, and a stock bucket for the long term. Spending is drawn from cash, which is refilled from the others.

How many years of cash should bucket 1 hold?

One to three years of spending not covered by other income is typical. More cash gives more comfort and lowers the portfolio's expected return.

Does the bucket approach protect against a market crash?

It prevents having to sell stocks for a year or two, which can help an investor stay the course. It does not reduce how far the portfolio's value falls: that depends on how much is in stocks.

Is a bucket plan different from a 60/40 portfolio?

In risk, hardly at all if the buckets add up to the same mix and are rebalanced. The difference is in presentation and in the rule for where withdrawals come from.

When should the cash bucket be refilled?

By a rule fixed in advance. A common one is yearly, from whichever asset is above its target share, which is rebalancing by another name.

Summary

The bucket approach holds near-term spending in cash, medium-term money in bonds and long-term money in stocks. Tested from January 2008, a bucket plan avoided selling stocks in the crash and bottomed about $11,000 higher than a fixed mix on a $1,000,000 portfolio. Its long-run result depended on rebalancing: without it the plan ended at $1,500,117 against $2,115,797 for the fixed mix; with it, at $2,080,640. A bucket plan is an asset allocation with a spending rule. Its value is in helping a retiree hold on through a fall, and it works best when the buckets are rebalanced and their combined stock share is checked each year.

Check your understanding

5 questions on this lesson. Each answer comes with the reason, and nothing is scored.

  1. A retiree has $800,000, spends $40,000 a year, and holds two years of spending in cash, eight years in bonds and the rest in stocks. What is the implied stock allocation?

    1. 40%
    2. 50%
    3. 60%
    Show the answer

    B. 50% Cash $80,000 and bonds $320,000 total $400,000. The remaining $400,000 in stocks is 50% of $800,000.

  2. In the 2008 test, how did the bucket plan and the fixed 60/32/8 mix compare at the bottom of the market?

    1. The bucket plan lost far less
    2. They were nearly the same, both down about a third
    3. The bucket plan lost far more
    Show the answer

    B. They were nearly the same, both down about a third Both started with the same allocation, so both fell about the same amount: lows of $679,735 and $668,323. The cash bucket changes what is sold, not how far the total falls.

  3. Why did the unrebalanced bucket plan end 2026 so far behind the fixed mix?

    1. Its cash bucket earned too little interest
    2. Its stock share fell in the crash and was never restored, so it held fewer stocks through the recovery
    3. It paid higher fees
    Show the answer

    B. Its stock share fell in the crash and was never restored, so it held fewer stocks through the recovery The fixed mix rebalanced into stocks at the low. The bucket plan did not, and then funded spending from stocks for years, staying underweight during a long rise.

  4. What is the main genuine benefit of a cash bucket?

    1. It raises the portfolio’s expected return
    2. It removes the need to sell stocks during a decline, which helps a retiree stay invested
    3. It eliminates sequence of returns risk
    Show the answer

    B. It removes the need to sell stocks during a decline, which helps a retiree stay invested The benefit is behavioural and practical: bills are covered without selling at low prices. Expected return is slightly lower, and sequence risk still depends on the allocation.

  5. Starting in 2010, the unrebalanced buckets ended ahead of the fixed mix. What does that show?

    1. That buckets always win in the long run
    2. That without rebalancing the stock share drifted up to 84%, so the plan earned more by taking more risk
    3. That cash outperformed stocks
    Show the answer

    B. That without rebalancing the stock share drifted up to 84%, so the plan earned more by taking more risk With no crash at the start, stocks outgrew spending and the plan became far more aggressive than intended. The result reflects the allocation, not the buckets.

Sources and further reading

  1. Harold Evensky, Stephen M. Horan and Thomas R. Robinson, The New Wealth Management: The Financial Advisor’s Guide to Managing and Investing Client Assets (Wiley, 2011) — the cash-flow reserve
  2. Walt Woerheide and David Nanigian, “Sustainable Withdrawal Rates: The Historical Evidence on Buffer Zone Strategies,” Journal of Financial Planning, vol. 25, no. 5 (2012)
  3. Journal of Financial Planning (September 2013), “The Benefits of a Cash Reserve Strategy in Retirement Distribution Planning” — summarizes Woerheide and Nanigian (2012)
  4. Michael E. Kitces, “Is a Retirement Bucket Strategy an Asset Allocation Mirage?” (Nerd’s Eye View, 2014)
  5. Hersh M. Shefrin and Richard H. Thaler, “The Behavioral Life-Cycle Hypothesis,” Economic Inquiry, vol. 26, no. 4 (1988) — mental accounting
  6. Finabase chart data — monthly bars and dividends of SPY and AGG, January 2004 to October 2026; 3-Month Treasury Bill rate (TB3MS) and CPI-U (CPIAUCSL) via FRED; simulations computed by Finabase