When to Use Cash, Individual Bonds, or a Short-Duration Bond ETF

Cash, an individual Treasury bill, a short-duration Treasury ETF, and an aggregate bond ETF are not interchangeable yield containers. They differ in the promise an investor owns, the date principal is expected, the path of market value, access before the planned date, and the risks hidden behind the quoted yield [1][2][3].

The decision begins with a liability: how much money, in what currency, on what date, with what tolerance for a temporary or permanent shortfall? Only then should the investor compare current yields, duration, credit exposure, bid-ask spread, fund fee, tax treatment, and account protections. The framework in stocks, bonds, and cash provides the allocation context.

What most investors get wrong: they ask which instrument will win after the next central-bank decision. That converts a funding problem into a rate forecast. The practical question is which instrument can do the required job even when the forecast is wrong.

Cash, an individual Treasury bill, a short-duration Treasury ETF, and an aggregate bond ETF are not interchangeable yield containers. They differ in the promise an investor owns, the date principal is expected, the path of market value, access before the planned date, and the risks hidden behind the quoted yield [1][2][3].

The decision begins with a liability: how much money, in what currency, on what date, with what tolerance for a temporary or permanent shortfall? Only then should the investor compare current yields, duration, credit exposure, bid-ask spread, fund fee, tax treatment, and account protections. The framework in stocks, bonds, and cash provides the allocation context.

What most investors get wrong: they ask which instrument will win after the next central-bank decision. That converts a funding problem into a rate forecast. The practical question is which instrument can do the required job even when the forecast is wrong.

Duration estimates a local price response; yield does not promise return

For a small parallel change in yield, the familiar approximation is ΔP/P ≈ −Dmod × Δy. If modified duration is 1.9, a one-percentage-point rise in relevant yields suggests roughly a 1.9% price decline before convexity, income, expenses, and other effects. It is a local sensitivity estimate, not a forecast and not a loss limit. Real yield curves move by different amounts at different maturities, and mortgage or callable cash flows can change as rates move [1][8].

Yield to maturity is also conditional. For an individual default-free bond held to maturity, it assumes promised cash flows and reinvestment conventions. For a fund, the portfolio continuously trades, receives flows, and replaces maturing holdings. The fund has no single date on which an investor is promised the original purchase price. Current SEC yield, distribution yield, and average yield to maturity answer different questions; none is a guaranteed holding-period return.

A dated Treasury bill is different. Treasury publishes the issue and maturity terms, and a bill held to maturity pays face value; selling before maturity exposes the owner to the market price [2]. An FDIC-insured deposit is a bank liability covered only when the bank, ownership category, and amount satisfy the applicable rules [3]. A security in the same brokerage screen does not become a deposit.

Scenario tests expose tradeoffs that one current yield conceals

A useful comparison holds the funding objective constant and changes one risk driver at a time. Test a roughly parallel 100-basis-point rise, a 100-basis-point fall, a steepening or flattening curve, wider credit spreads, a liquidity shock, and an earlier-than-planned withdrawal. These are stress tests, not probabilities. Federal Reserve H.15 data can anchor observable Treasury rates, but today's curve cannot reveal tomorrow's path [6].

Cash usually has little mark-to-market volatility for the depositor, yet its rate can reset quickly. A bill locks a dated cash flow if held to maturity but creates reinvestment risk after that date. A short Treasury ETF generally has limited duration and continually rolls exposure, so it can gain when yields fall or lose when they rise. A broader fund adds credit-spread, mortgage prepayment or extension, and greater duration effects [1][5].

Worked example: suppose a fund's modified duration is 2.0 and its relevant yields rise 0.75 percentage point immediately. The first-order price estimate is about −1.5%. That is not a prediction of total return: income accrues, the curve may twist, and the duration may change. The exercise is useful because it translates a vague phrase such as “low risk” into a loss budget that can be compared with the liability.

A short-duration ETF is not cash and does not automatically win during hikes

The phrase “short duration” describes rate sensitivity, not a guarantee. A Treasury ETF owns marketable securities and its shares trade through the day. Its NAV reflects current yields; its market price can differ from NAV; the fund charges expenses; and the portfolio rolls. The official fund page reports dated portfolio characteristics that change over time, so hard-coding one duration or yield into a long-lived guide is poor practice [4][10].

During a hiking cycle, an ETF can suffer an initial price decline while later reinvesting at higher yields. Cash rates may adjust faster or slower depending on the bank and product. A bill ladder can lock known maturity values but requires execution and reinvestment. Which produces the higher after-tax holding-period return depends on starting prices, the entire yield path, fees, spreads, taxes, cash-flow timing, and whether the investor sells early. “Rates rose” is not enough to identify a winner.

Deposit insurance also requires precision. FDIC coverage generally applies up to $250,000 per depositor, per insured bank, per ownership category, with aggregation rules; it covers qualifying deposits, not mutual funds or ETFs [3]. Brokered deposits can add operational and recordkeeping questions. Confirm the bank, ownership category, total aggregated balance, withdrawal rules, and whether a promoted rate is temporary.

The ETF structure guide and fund fact-sheet guide show how to verify the exact product rather than relying on a category label.

An aggregate bond ETF is a diversified bond allocation, not a cash substitute

A broad U.S. aggregate fund may own Treasuries, agency securities, agency mortgage-backed securities, and investment-grade corporate bonds. That diversifies issuers and sectors, but it also combines interest-rate risk, credit spreads, and mortgage prepayment or extension behavior. Its index weights reflect the market's debt outstanding, not a household's next liability date [1][5].

Longer duration can help when relevant yields decline, but that is a sensitivity, not a promise that bonds will offset equities. Inflation shocks can pressure both stock valuations and nominal bonds. Credit spreads can widen during recessions even while Treasury yields fall. Mortgage durations can extend when refinancing slows. Diversification is valuable because outcomes are uncertain; it is not a guarantee of negative correlation in every drawdown.

The fund can still be appropriate for a strategic fixed-income sleeve whose horizon and risk budget tolerate its drawdowns. It is usually a poor match for a known payment due in months when the minimum proceeds cannot fall. Separate the liquidity reserve from the strategic allocation, then measure the total household duration and correlated loss paths. See risk measurement, risk-adjusted metrics, and the 60/40 portfolio.

A $100,000 twelve-month reserve starts with the obligation, not the forecast

Worked example: a household needs at least $100,000 in U.S. dollars twelve months from now for a property closing. The payment date is firm, the minimum is not negotiable, and early access is unlikely. The first gate is therefore principal availability on a known date—not whether analysts expect cuts. Keeping extra cash for timing uncertainty and matching the remainder with Treasury maturities near the payment date may fit the obligation more directly than buying a rolling ETF [2].

A single twelve-month bill may not mature on the exact closing date. A small ladder or a bill maturing earlier can reduce settlement risk, while the final days remain in a qualifying insured deposit. The household must account for purchase price, auction and settlement dates, brokerage or TreasuryDirect procedures, maturity instructions, taxes, and any chance the payment moves. A Treasury bill sold before maturity has a market price; “government backed” does not remove early-sale risk [8].

If the spending date is flexible and a modest temporary shortfall is acceptable, a short Treasury ETF may reduce ladder maintenance and offer daily exchange liquidity. That convenience comes with NAV, spread, fee, and execution risk. An aggregate fund asks the household to tolerate more risk that has no obvious connection to this liability. The answer changes if the $100,000 is a rolling emergency reserve rather than a fixed payment.

Practical takeaway: write the liability specification before opening a quote screen: date range, minimum dollars, early-access probability, acceptable temporary loss, acceptable permanent loss, guarantee requirement, tax account, and person responsible for execution.

Rate-cycle history explains outcomes after the fact, not the next trade

The Federal Open Market Committee sets a target range for the federal funds rate, while market yields across maturities embed expectations, term premiums, inflation uncertainty, and supply and demand [6][7]. By the time a hike or cut is announced, Treasury prices may already reflect a different expected path. A rule that buys short duration in “hikes” and long duration in “cuts” therefore needs an investable signal date, execution price, benchmark, costs, tax assumptions, and out-of-sample evidence—not a chart labeled after the cycle ended.

Duration math still matters. If relevant yields fall unexpectedly, greater duration generally creates a larger positive price response; if they rise, it generally creates a larger negative response [8]. But a curve twist can make one maturity behave differently from another. Credit spreads and mortgage cash flows can overwhelm the Treasury move. Carry and rolldown depend on the starting curve. Rebalancing time changes realized results.

The uncomfortable implication: there is no instrument that is simultaneously principal-stable, perfectly liquid, high-yielding, inflation-proof, tax-free, and certain to outperform after every policy move. Backtests that select cycle boundaries with hindsight are especially vulnerable to regime labels and timing leakage. The guides to regime detection and allocation regimes explain that limitation.

Use history to size plausible sensitivities and failure modes, not to turn a policy-cycle label into a guaranteed trade.

A decision tree should reject mismatches before ranking expected return

Decision tree: first ask whether a minimum nominal amount is required on a fixed date. If yes, prefer instruments whose maturity or insured account terms match that date, and keep an operational buffer. If no, ask whether the reserve is rolling and whether small NAV losses are acceptable. A short Treasury ETF may then be eligible. Only use credit or intermediate duration when the horizon, portfolio role, and loss budget deliberately permit those risks.

Second, verify the product. For deposits, verify the insured bank, ownership category, aggregate balances, rate conditions, and withdrawal restrictions. For a bill, verify CUSIP, auction or market price, maturity, settlement, reinvestment, and custody. For an ETF, verify index and mandate, effective duration as of a dated fact sheet, holdings, credit quality, spread, premium or discount, volume, fee, distribution policy, and tax treatment [3][4][5][10].

Third, compare consistent scenarios and after-tax proceeds. Record what happens if the money is needed early, yields move one percentage point either way, spreads widen, the bank changes its rate, or the payment date moves. Then document the choice and review trigger in an investment policy statement. Rebalancing belongs to the strategic portfolio; a required near-term payment should not depend on a favorable rebalance. See tax-aware rebalancing.

Checklist: liability documented; legal promise understood; coverage or maturity verified; duration and spread risks measured; early-sale scenario tested; costs and taxes compared; operational owner assigned; review trigger written. If any required gate fails, reject the instrument before debating a few basis points of yield.

Related guides

• Related guide 1