Money, Examined / No. 10 / Liquidity
Sleep well at night. Keep tomorrow in view.
Cash can buy room to breathe.
Building it can cost room to grow.
A dollar of liquidity does two jobs: it funds a disruption and lets you leave other money alone. The question is not whether cash is good. It is how much, for what, and in what sequence.
Here, SWAN means sleep well at night: a personal liquidity constraint, not a scientifically unique optimum. Separate the evidence for holding a buffer from the decision about its size.
Buffer-stock saving models connect precautionary wealth to income uncertainty, borrowing constraints, and preferences. Under the relevant conditions, Carroll's target is a wealth-to-permanent-income ratio, not a universal bank balance. Deaton shows why the income process and inability to borrow matter.[1][2]
The boundary: these models do not prescribe six months, $100,000, or this lab's formula.
Kaplan, Violante, and Weidner document households with substantial illiquid wealth but very little liquid wealth. Their analysis also challenges a simplistic “cash is always better” story: higher returns and access costs can make low-liquid-wealth portfolios understandable.[3]
The boundary: home equity, retirement assets and a stable-looking job are not spendable cash.
In a study of 585 UK bank customers, larger liquid balances were associated with better perceived financial well-being and indirectly with life satisfaction, after multiple controls.[4]
The boundary: this is not causal evidence. It does not show that moving money out of investments will make you happier, or locate a happiness threshold.
When invested money earns more than after-tax cash, postponing exposure can reduce later wealth. When markets fall or spending forces a sale, early cash can look better. Those are conditional model outcomes, not a guaranteed cost or a forecast.
The test: hold starting assets and saving constant. Change the funding order in Lab 02.
01 / Name the disruption
Start with expenses and access, not a percentage of the house. Compare a chosen full-expense buffer with a specific income disruption, then add known obligations and a separate one-off shock. The larger buffer is used to avoid double-counting the same months.
Include housing, food, care, insurance and minimum debt payments.
Spendable cash today, before the earmarks below. Not credit or investments.
A personal constraint, not a research-derived optimum.
USD; range 0 to 40,000
USD; range 0 to 40,000
%; range 0 to 100
%; range 0 to 100
months; range 0 to 36
Already committed spending, separate from monthly essentials.
A chosen stress amount; not a probability or forecast.
Taxable, saleable assets; volatile and not counted as cash runway.
Home equity + restricted/retirement assets - debts not already netted out. Exclude the cash and accessible investments above.
Cash covers this chosen disruption, leaving $2,000. That does not guarantee coverage of a larger or longer shock.
Earmarks and shock are charged at month 0. Only the monthly deficit draws cash; surplus income is not added to this conservative runway. Below zero means unfunded need, not a loan.
Investable wealth here means cash plus accessible market investments. Other net wealth changes the second denominator, never the runway. Percentages are undefined when the denominator is zero or negative.
$8,000 + $10,000 + max($30,000, $15,000) = $48,000
Surviving income: $2,500/month. Monthly deficit: $2,500. Cash after earmarks covers 6.8 months of that deficit. The larger of the two buffers is used because they protect overlapping spending, not two separate emergencies.
| Month | Cash remaining | Unfunded need |
|---|---|---|
| 0 | $17,000 | $0 |
| 1 | $14,500 | $0 |
| 2 | $12,000 | $0 |
| 3 | $9,500 | $0 |
| 4 | $7,000 | $0 |
| 5 | $4,500 | $0 |
| 6 | $2,000 | $0 |
The denominator is not the decision
A fixed $100,000 is 20% of $500,000, 10% of $1 million, and 2% of $5 million in investable wealth. That identity says nothing about how many months it covers. Spending, commitments, inflation and access still matter.
Near-term taxes, a planned purchase and emergency spending compete for the same cash. Do not promise one balance to three different jobs.
Liquid investments can be sold but may be down when needed. Illiquid net wealth can increase while cash stays scarce.[3]
A rising target needs funding. The next lab moves money within one budget; it never adds the reserve outside the portfolio.
02 / Fund the same future three ways
A minimum buffer plus parallel investing is a sequence to examine, not a universal recommendation. Move between wealth, cash share and cash balance; then add a market decline and a spending shock. Lab 02 uses independent inputs so changing your household runway does not silently alter this comparison.
Direct all new saving to the fixed goal. Invest new saving only after it is filled.
Fill the minimum buffer. Then split saving between cash and investing until the fixed goal is filled.
Use B's split, but raise the cash goal each year. Every extra dollar must still be funded.
After essential costs and ordinary taxes; fixed nominal amount. Negative means spending exceeds income.
Fixed for A/B; grows for C. Must be at least the minimum buffer.
years; range 0 to 40
USD; range 0 to 1,000,000
USD; range 0 to 5,000,000
A chosen constraint. New saving fills this before parallel investing.
After the minimum is filled; the rest is invested. Stops at the goal.
A higher goal is not new wealth. It redirects the same saving budget.
Annual effective, before investment taxes/fees. A scenario, not a forecast.
Annual effective yield, held constant for this illustration.
Applied to interest only; 0% switches this tax off.
%; range 0 to 10
0 disables it. A year beyond the horizon has no effect.
Additional to ordinary monthly cash flow; can stand in for a disruption budget.
Applied once to all investments before that month's contribution. Not applied to cash.
A separate haircut on sale proceeds, not the market decline or a tax model.
This is a funding-order comparison, not an optimal allocation. Earlier market exposure wins under these inputs; it also exposes money to loss earlier.
year; range 0 to 25
| Measure | A / Cash first | B / Parallel | C / Rising goal |
|---|---|---|---|
| Cash | $158,090 | $117,724 | $146,449 |
| Investments | $757,277 | $887,421 | $854,465 |
| Nominal wealth | $915,367 | $1,005,145 | $1,000,914 |
| Real wealth, today's dollars | $493,740 | $542,166 | $539,884 |
| Cash / investable wealth | 17.3% | 11.7% | 14.6% |
| Cash goal / unfilled | $100,000 / $0 | $100,000 / $0 | $209,378 / $62,929 |
| Forced sales, gross to date | $0 | $0 | $0 |
| Sale friction to date | $0 | $0 | $0 |
| Unfunded needs to date | $0 | $0 | $0 |
No stress event occurs within this horizon. Financial wealth excludes illiquid assets and assumes no outstanding debt; it is not after-tax liquidation value.
Change only the gross investment return. These three paths are not confidence intervals, probabilities, or historical percentiles. All other assumptions, including any stress, stay the same.
| Annual investment return | B minus A |
|---|---|
| 3% | $6,444 |
| 6% | $48,425 |
| 9% | $133,335 |
| Year | A / Cash first | B / Parallel | C / Rising goal |
|---|---|---|---|
| 0 | $25,000 | $25,000 | $25,000 |
| 1 | $44,122 | $44,366 | $44,366 |
| 2 | $63,698 | $64,746 | $64,746 |
| 3 | $83,737 | $86,195 | $86,195 |
| 4 | $104,253 | $108,776 | $108,776 |
| 5 | $125,353 | $132,552 | $132,552 |
| 6 | $147,565 | $157,592 | $157,592 |
| 7 | $171,024 | $183,968 | $183,968 |
| 8 | $195,803 | $211,756 | $211,756 |
| 9 | $221,980 | $241,037 | $241,037 |
| 10 | $249,636 | $271,897 | $271,897 |
| 11 | $278,858 | $304,426 | $304,426 |
| 12 | $309,737 | $338,721 | $338,721 |
| 13 | $342,372 | $374,883 | $374,883 |
| 14 | $376,865 | $413,021 | $413,021 |
| 15 | $413,326 | $453,247 | $453,247 |
| 16 | $451,869 | $495,684 | $495,684 |
| 17 | $492,619 | $540,459 | $540,459 |
| 18 | $535,704 | $587,713 | $587,707 |
| 19 | $581,263 | $637,688 | $637,573 |
| 20 | $629,441 | $690,576 | $690,208 |
| 21 | $680,393 | $746,550 | $745,774 |
| 22 | $734,283 | $805,795 | $804,441 |
| 23 | $791,284 | $868,502 | $866,391 |
| 24 | $851,580 | $934,880 | $931,814 |
| 25 | $915,367 | $1,005,145 | $1,000,914 |
Model card / Illustrative, not predictive
These are deterministic bookkeeping models authored for this publication. They are not replications of the academic models, Monte Carlo forecasts, asset-allocation optimizers, or estimates of the probability of financial distress.
All policies begin with the same cash and investments. Existing investments are not sold to fill a target, and excess starting cash is not automatically reinvested. Only new positive saving is allocated. Cash interest stays in cash, so the balance can exceed its target. C starts with B's goal and raises it at the chosen annual rate.
Investable financial wealth = cash + market investments
No illiquid wealth or debt is modeled in Lab 02. Its “wealth” is therefore net financial wealth only under the explicit no-debt assumption, not total household net worth or after-tax liquidation value.
First accrue cash and investment returns using equivalent monthly rates. Next apply the one-time market decline, if enabled. Then allocate that month's positive saving, or fund its negative cash flow. Finally pay the stress expense at that month's end. Costs are funded from cash first, then market sales net of the sale haircut.
Any remaining shortfall is recorded as cumulative unmet spending, not a loan. A later recovery in assets does not erase the earlier unmet need. Ending wealth alone is not a fair welfare comparison once a plan fails to fund spending.
Annual effective returns, nominal monthly saving and nominal spending shocks are held constant except for the optional single decline. There is no automatic rebound. Cash interest uses the entered tax rate; investment taxes, investment fees, capital-gains timing and tax-account constraints are omitted. The sale haircut is a separate stress assumption, not a capital-gains calculation.
Real wealth divides nominal wealth by accumulated inflation. A higher gross market-return assumption may overstate the advantage of investing if taxes and costs would narrow the spread.
For US bank deposits, FDIC coverage is generally $250,000 per depositor, per insured bank, per ownership category. Coverage is conditional on the institution and account structure; it does not cover non-deposit investments or an uninsured intermediary's failure.[5]
Check the actual account, withdrawal conditions and aggregation. An investment called a money market fund is not an FDIC-insured bank deposit. This publication does not equate deposit protection with protection from inflation, or prescribe a product.
The source ledger
Access and editorial review date: . A linked paper supports only the claim identified here, not an endorsement of this publication or its scenarios.
Christopher D. Carroll. The Quarterly Journal of Economics 112(1), 1-55 (1997). Working paper: NBER 5788 (1996). DOI: 10.1162/003355397555109
Supports: Under the paper's conditions, uncertainty and precautionary motives interact with impatience to produce a target wealth-to-permanent-income ratio.
Boundary: Not a recommendation for a fixed dollar reserve, a separate savings-account balance, or this calculator's chosen number of months. Model assumptions and income processes matter.
Primary NBER record; journal title, author, year and DOI verified against Crossref. Journal publisher page restricted automated access. Access/review: 2026-09-05.
Angus Deaton. Econometrica 59(5), 1221-1248 (1991). DOI: 10.2307/2938366
Supports: Borrowing constraints and the process governing income change the role of assets as a consumption buffer.
Boundary: The results depend on the persistence and growth of income and other model assumptions. No unique emergency-fund size follows for a real household.
Author-hosted paper link accessible; bibliographic identity and DOI verified against Crossref. The primary paper is a PDF. Access/review: 2026-09-05.
Greg Kaplan, Giovanni L. Violante, and Justin Weidner. Brookings Papers on Economic Activity, Spring 2014, 77-138. Linked DOI is the NBER Working Paper 20073 version (2014). DOI: 10.3386/w20073
Supports: Households can hold substantial illiquid wealth yet very little liquid wealth. Access costs and higher illiquid returns help explain that portfolio pattern.
Boundary: Descriptive patterns and a model, not a trial of emergency-fund policies. Low liquidity can reflect a return/access-cost tradeoff, not simply a household mistake. Historical survey shares are not presented here as current prevalence.
Primary Brookings article and linked paper; NBER version and authors verified against the NBER record and Crossref. Access/review: 2026-09-05.
Peter M. Ruberton, Joe Gladstone, and Sonja Lyubomirsky. Emotion 16(5), 575-580 (2016). DOI: 10.1037/emo0000184
Supports: Among 585 UK bank customers, larger liquid balances were associated with more positive perceived financial well-being and, indirectly, life satisfaction, with multiple controls.
Boundary: Not randomized. Controls do not establish causality or eliminate omitted variables and reverse causation. A selected UK sample does not establish a universal threshold or prove that moving investments to cash makes someone happier.
UCL institutional repository record and abstract, with publisher DOI and open-access manuscript link. Access/review: 2026-09-05.
Federal Deposit Insurance Corporation. Official FDIC consumer guidance; living web page.
Supports: FDIC insurance protects qualifying deposits at insured banks, generally $250,000 per depositor, per insured bank, per ownership category. Non-deposit investments are not covered.
Boundary: Coverage depends on ownership, aggregation and the institution. It does not insure market returns, inflation-adjusted purchasing power, or failure of an uninsured intermediary.
Official FDIC guidance accessed directly. Verify your institution and account structure rather than treating this page as a coverage determination. Access/review: 2026-09-05.