Chapter 6: The cash bucket and the dry-powder myth
Money market funds pay roughly 3.5–4% in June 2026, by the Cash & Bonds Guide's running count of the shelf. That sounds like good news for cash, and mostly it is. It is also why this chapter exists: when cash earned nothing, waiting in it announced its own cost every month, and now that cash earns 3.8% or so, waiting feels like a strategy. The dry-powder temptation got stronger the day cash got better, not weaker.
Chapters 4 and 5 split the stock sleeve and gave the bond sleeve its jobs. This chapter handles the smallest allocation, and the one carrying the most popular wrong job in retail investing.
The two jobs of cash
Job one is dates. A dollar owed to a calendar inside roughly the next two years, the first stretch of the Cash & Bonds Guide's parking bands, belongs in cash because its worst case is intolerable: the tuition bill and the moving truck do not accept a market recovery schedule. Chapter 1 already sorted these dollars into their buckets; cash is simply where the nearest bucket parks.
Job two is nerves. Some people hold a deliberate buffer, a few months of spending beyond the emergency fund, because the buffer is what lets them leave the 70/30 alone in a red October. Priced honestly, the buffer trails stocks in most years by design, and what it buys is behavior. The behavior is worth real money: Morningstar's Mind the Gap study (2025 edition) found investors earned 7.0% a year in funds that returned 8.2% over the ten years to 2024, and the missing 1.2 points a year came mostly from badly timed entries and exits, concentrated in the most volatile funds. A buffer that keeps you invested through the storm earns its keep several times over. Chapter 9 sizes yours against your sell line.
There is no third job. The pile labeled "waiting for the crash" claims one, and the rest of this chapter is the audit.
The third job, the one cash cannot do
Dry powder is cash held back to buy stocks after the crash you expect. It feels prudent, and it photographs well in a market that just hit a high. Strip the costume and it is a forecast: the pile has no date it must pay and no nerve it calms, and its entire job is to be smarter than the market about timing. The evidence on that job is unkind. SPIVA's 2025 scorecard found 79% of active large-cap funds trailed their benchmark, which means the professionals paid to time and select, with terminals and staff, mostly failed to. And Mind the Gap's 1.2-point annual gap is what amateur timing actually cost real account holders over the ten years to 2024. Neither group lacked conviction. Both lacked a schedule the market would honor.
The arithmetic of waiting
Waiting has a price meter running, so write the race down honestly. The waiter's money compounds at cash rates while it waits. For waiting to beat investing today, the market's total return between today and the entry day must come in below cash's, and then the entry must actually happen.
Put Mara's $40,000 dip fund through it, using 3.8% as a round middle of today's money fund shelf. Two years of waiting grows the pile to about $43,100, a 7.7% head start. Over those same two years, Vanguard's VCMM (March 31, 2026 run, model assumptions as always) compounds US equities 10.0–14.3% across its 4.9–6.9% band, and the 1926–2024 historical average of 10.5% a year compounds about 22.1%, turning the same $40,000 into roughly $48,800. So if no crash arrives within two years, the market she eventually rejoins must be 2–12% cheaper than today just for her to break even, and the hurdle grows every month the wait continues.
That is only the first win the dip-waiter needs. The second is the entry itself: standing in the window where prices sit below the cash line, on the week the headlines explain why prices are going lower, and buying. The Mind the Gap numbers are the record of how that week usually goes. People who sold to raise the powder face the same two wins in reverse order, out and back in, each on schedule.
What if the crash really does come?
Sometimes it will, and your plan already employs a buyer for it who needs no forecast. Chapter 10's rebalancing rule purchases stocks after they fall, on a written trigger, paid from the bond sleeve whose third job (Chapter 5) is exactly this ammunition. Your regular contributions keep buying straight through the decline as well, every payday, without an opinion about the bottom. Both mechanisms caught 2008 and 2022 for the people who had them, and neither required predicting either year. A dry-powder pile has to beat both of those disciplined buyers after paying its waiting costs, and it has to do so by winning the two decisions in the figure above on the worst weeks of the cycle. The crash being real does not make the timing problem easier; the gap data was collected from people who lived through real ones.
Mara's buckets, sized
Chapter 1 gave Mara's sabbatical its own bucket and date: four years out, six months away from work, $6,000 a month of spending, so $36,000. The Cash & Bonds Guide's bands place a four-year dollar on the short-duration shelf now, rolling into bills and money funds as the date crosses inside two years. That is a dates job, sized exactly, parked to a calendar.
Alongside her $310,000 of invested funds, though, sits a separate $40,000 in a money market fund, and it has a different story.
| Mara's cash, after the audit | Dollars |
|---|---|
| Sabbatical bucket (June 2030) | $36,000 |
| Added to the retirement mix, on schedule | $4,000 |
| Waiting for a crash | $0 |
| Total | $40,000 |
The $40,000 had waited since 2024, born from one persuasive article about a correction that was always six months away. At today's yields it earned about $1,500 a year, which felt like proof of wisdom. This week she ran the two-job test: she asked the pile for its date and found none, then for the person it calms and found her sleep already covered. Then the meter reading: two more years of waiting puts her near $43,100, while even the conservative end of Vanguard's model band compounds past $44,000 and history's average runs near $48,800, before the harder problem of actually buying on the worst week. The relabeling took ten minutes. $36,000 took the sabbatical's name and its June 2030 date, and $4,000 joined her retirement mix through the regular schedule. Nothing was sold in fear or bought in a hurry; every dollar simply received a job.
Every cash dollar carries either a date or a name: a payment it will make or a person it keeps calm, written into the plan. A dollar waiting for a market level is an equity decision wearing a money-fund yield, and it goes back into the plan it left.
Where people go wrong
Letting the buffer grow with the headlines. A nerves bucket sized once, in Chapter 9's calm, is a tool. A buffer that gains a month of spending after every scary Tuesday is dry powder assembling itself, and it never announces the change.
Waiting to get back in "once things settle." Settled and cheap do not arrive together; by the time the news reads calm, the entry window in Figure 6.2 has usually closed above the cash line. Sellers-turned-waiters must win the round trip twice, out and back in, each on schedule.
Counting the emergency fund as powder. That money already holds the first job money can have, and spending it on a dip converts a safety net into a stock position at the exact moment the paycheck is least secure. The Cash & Bonds Guide built that fund; leave its job alone.
Mistaking the yield for a strategy. Earning 3.8% feels like winning while you wait. The VCMM's own cash assumption is 2.9–3.9% against 4.9–6.9% for US equities (March 31, 2026 run, model assumptions); even the cautious model expects patience in cash to cost about 2 points a year. The yield is a parking fee refund, never a return engine.
1. Every goal, with its date and its bucket (Chapter 1).
2. Target mix: __% stocks / __% bonds, chosen from the worst-year column (Chapter 2).
3. The storm list: what protects me in each kind of crash (Chapter 3).
4. Inside stocks: __% US / __% international (Chapter 4).
5. Bonds: __%, duration matched to __ (Chapter 5).
6. Cash: $__ for __ (dates), nothing for dips. New this chapter. Mara's line reads "$36,000 for the sabbatical (June 2030), nothing for dips."
Key takeaways
- Cash holds exactly two jobs: dollars with dates inside roughly two years, and a nerves buffer that buys behavior. Money funds paying roughly 3.5–4% (June 2026) make both jobs cheaper to hold and create no third one.
- Dry powder is a forecast wearing a yield. SPIVA 2025: 79% of active large-cap funds trailed their benchmark; Mind the Gap 2025: real investors earned 7.0% in funds returning 8.2% over the ten years to 2024, with the gap living in mistimed entries and exits.
- The waiting arithmetic runs uphill: $40,000 at 3.8% reaches about $43,100 in two years, while Vanguard's VCMM band (4.9–6.9%, March 31, 2026 run, model assumptions) compounds 10.0–14.3% and the 1926–2024 average of 10.5% a year compounds about 22.1%. The dip must beat the meter, and then the entry must actually happen.
- The plan already owns a licensed dip buyer: Chapter 10's rebalancing rule, funded by the bond sleeve, plus contributions that buy through every decline without a forecast.
- Mara's audit: $36,000 + $4,000 + $0 = $40,000, every dollar holding a date or a job, and the crash fund retired without a sale in fear.
Sources: Morningstar Mind the Gap · SPIVA U.S. scorecard · Vanguard return forecasts (VCMM) · Vanguard model allocation data · Finvest Cash & Bonds Guide