“Keep three to six months of expenses as an emergency fund.” You see it everywhere. So we went looking for the source of that number.
⚠️⚠️ We could not find a Korean public agency that states “three to six months.”
And ⭐⭐⭐ when we opened the US regulator's own research, the band it actually uses is not three to six months. It is one month.
This article does not set a new target for you. It shows where the agency material and the widely repeated advice part company, and what to check when choosing where to park the money — written for people dealing with Korean bank accounts.
⚠️⚠️ We could not source the "three to six months" rule
Starting in Korea: we could not find a sentence stating a target number of months at the Financial Supervisory Service, the Korea Deposit Insurance Corporation, or the Bank of Korea. Following the number back through Korean-language content leads to private banks and fintech blogs, and those posts cite no agency source either.
One widely quoted Korean private source uses a different split: “three months for a dual-income household, six months if you are the sole earner or live alone.” ⭐ Splitting on the number of earners in a household is reasonable on its face — two income streams mean one can stop without everything stopping. ⚠️ But that is the author's reasoning, not a finding, and the piece cites no institution. So this article does not promote that number into a recommendation.
⭐ “We could not find it” is itself information. We have hit this pattern repeatedly — see the interest-tax figure in our pay down debt vs save article, or the “35% payment history” claim in the credit score article.
⭐⭐⭐ The US regulator's actual dividing line is one month
The US Consumer Financial Protection Bureau (CFPB) published a report on emergency savings backed by survey and credit-panel data (March 2022). The bands it uses to group consumers are not three to six months.
| Emergency savings | Share of people | Debt 60+ days past due | Debt in collections | Could not pay all bills last month |
|---|---|---|---|---|
| ⚠️ None | 24% | ⚠️ 40% | ⚠️ 49% | ⚠️ 39% |
| Less than a month of income | 39% | 19% | 17% | 11% |
| ⭐ At least a month of income | 37% | ⭐ 5% | ⭐ 5% | ⭐ 1% |
⭐⭐ Note the unit: a month of income. The Korean rule of thumb is stated in months of spending. ⚠️ The denominators differ before you even argue about the number of months.
⭐⭐ The biggest change sits between zero and one month
The report does not print the gaps between bands, so we subtracted them ourselves.
| Measure | None → under a month | Under a month → a month or more | ⭐ First step, as a multiple |
|---|---|---|---|
| Debt in collections | ⭐ −32pp | −12pp | ⭐ 2.7× |
| Could not pay all bills | ⭐ −28pp | −10pp | ⭐ 2.8× |
| “Finances control my life” | −33pp | −21pp | 1.6× |
| Debt 60+ days past due | −21pp | −14pp | 1.5× |
| ⚠️ Financial well-being score | +10 points | ⚠️ +11 points | ⚠️ 0.9× |
⭐⭐⭐ On four of the five measures, the first step is the larger one.
“You need three months’ worth” is the kind of target that stops people from starting. What this data points at is that the first month — getting off zero — moves the most.
⚠️ The one exception is the well-being score (40 → 50 → 61), which keeps climbing. You could read that as peace of mind continuing to improve — ⚠️ but that reading is ours, not a sentence in the report.
⚠️⚠️ Two caveats that matter. ① This is US survey data and does not transfer directly to Korea. ② The report does not claim that savings cause fewer delinquencies — its own wording is that there are “striking differences in consumers' credit and debt profiles… by their level of emergency savings.” Causation could run the other way: people with steady incomes may both save more and fall behind less.
⭐ What does survive is the choice of bands itself. A financial regulator, cutting its own data, drew the line at one month rather than three to six. For reference, 63% of respondents (24% + 39%) held less than a month (our calculation).
Where to keep it — only two conditions
The requirements for an emergency fund are simple: ① you can get at it immediately, and ② the principal has not shrunk when you do. Yield is a distant third.
| Where | Immediate access? | Principal at risk? | Suitable? |
|---|---|---|---|
| Demand deposit (incl. "parking" accounts) | ⭐ Immediate | ⭐ No | ⭐ Yes |
| Term deposit | Yes, by breaking the term | Principal kept, contracted interest lost | Partly |
| ⚠️ CMA (Korean cash management account) | Usually fast | ⚠️ Depends on the type | ⚠️ Check the type |
| ⚠️ Funds, stocks, ETFs | Settlement delay after selling | ⚠️ Yes | ⚠️ No |
⚠️ Read the third row carefully. In Korean emergency-fund advice, CMAs are almost always listed alongside parking accounts — but a CMA behaves differently depending on its type. ⚠️⚠️ We could not open the deposit insurer's product list this time, so this article goes only as far as “check which type it is.” The deposit-protection marking appears on the sign-up screen; look at it.
⭐ The comparison with term deposits is in our deposit vs instalment savings article. For money that is not your emergency fund, see ETF basics and choosing a Korean ETF.
⭐ Deposit protection rose to ₩100 million on 1 September 2025
An emergency fund is money where “do I get it back if the bank fails?” genuinely matters. That ceiling changed recently.
| Item | Detail |
|---|---|
| Effective | 1 September 2025 |
| ⭐ Limit | from ₩50 million → ⭐ ₩100 million |
| Coverage | principal and interest included |
| ⚠️ Not covered | ⚠️ funds and other performance-linked products |
⚠️ A note on source tier. We verified this in a notice issued by the Korea Deposit Insurance Corporation and posted by the Korea Financial Investment Association (posted 26 August 2025). ⚠️⚠️ The deposit insurer's own pages did not load for us, so we did not read the agency's site directly.
⭐ At emergency-fund sizes the ceiling is unlikely to bind — a month of income above ₩100 million is rare. ⚠️ The limit is widely described as applying per person, per financial institution, but we could not confirm that application rule in agency source text.
How much to start with
Taken at face value, the data above says the first target is not “three months” but “one month” — and that filling that first month is the step that moves the most.
- ☐ Decide your unit first — income or spending. ⭐ CFPB uses income; the Korean rule of thumb uses spending. Pick one and write it down.
- ☐ Use a separate account. Mixed in with day-to-day money, the purpose blurs.
- ☐ Automate a small transfer on payday. Getting off zero matters more than the amount.
- ☐ Compare rates at the moment you open the account. Demand-deposit rates move often. Korea's Financial Supervisory Service runs a comparison site with a demand deposit category.
- ☐ Once the first month is funded, the next question is whether to pay down debt first or use a tax-advantaged pension account.
Questions that remain
So do I need three months, or is one enough?
⚠️ This article recommends neither. Two things were verified: we could not source “three to six months” to any institution, and the US regulator drew its band at one month. ⭐ Set a larger goal if you want to — we simply could not find grounds for believing that anything short of three months is pointless.
Don't parking accounts pay less than term deposits?
Usually, yes. ⭐ But an emergency fund is money you pull out in a hurry, not money that earns, so the ordering is different. ⚠️ And as we found in the deposit article, there were real products where the headline rate with every bonus condition met came out below a plain unconditional rate — comparing advertised rates alone misleads.
What if I spend the whole fund?
That is what it is for. The complete habit is spending it and switching the transfer back on. ⭐ In the data above, the band that looks bad is “zero”, not “people who used it.”
Can I apply the US data directly in Korea?
⚠️⚠️ No. Unemployment insurance, medical costs and household-debt structure all differ. What this article takes from it is not a number of months but where a regulator drew its line, and the shape of the gaps between bands.
Why is the CMA row left vague?
⚠️ Because deposit protection differs by CMA type and we could not open the agency's product list. Not filling gaps with guesses is this site's standard, so we stopped at “check it on the sign-up screen.”
We could not source "three to six months" to any institution, and the line a US regulator actually drew was one month — with the largest gap sitting between zero and that first month.
Sources
- ⭐⭐ US Consumer Financial Protection Bureau — “Emergency Savings and Financial Security” (published 23 March 2022). Source for the bands (none / less than a month of income / at least a month), the 24%, 39%, 37% split, 60+ day delinquency 40/19/5%, collections 49/17/5%, unpaid bills 39/11/1%, well-being scores 40/50/61, and “finances control my life” 68/35/14%. ⚠️ A survey of US consumers.
- US Consumer Financial Protection Bureau — “Evidence-based strategies to build emergency savings” (17 July 2020). ⚠️⚠️ This report also gives no target number of months. It groups approaches into three categories instead: savings products, financial incentives, and behavioural approaches.
- Financial Supervisory Service, Korea — integrated financial product comparison (checked July 2026). Source for the fact that the savings section carries a demand deposit category alongside term deposits and instalment savings. ⚠️ We did not put specific rates in this article — they change too often.
- ⚠️ A Korea Deposit Insurance Corporation notice posted by the Korea Financial Investment Association (an association posting, not the agency's own site) — “From 1 September 2025, coverage rises from ₩50 million to ₩100 million, principal and interest included” (posted 26 August 2025). Source for the ₩100 million limit, effective date, principal-and-interest coverage, and the exclusion of performance-linked products.
- ⭐ Our own calculation. The gaps between bands (−32pp, −28pp, −33pp, −21pp, +10 points and so on), the multiples (2.7×, 2.8×, 1.6×, 1.5×, 0.9×) and the 63% holding less than a month are values we subtracted and divided from the report's figures, not numbers printed in it. ⭐ Check: 24 + 39 + 37 = 100.
What we could not verify
- ⚠️⚠️ A Korean public-agency source for “three to six months.” Not found. That is why this article does not state it as a recommendation.
- ⚠️⚠️ Whether a CMA is deposit-protected, by type. The deposit insurer's protected/unprotected product pages did not load. So the table says “depends on the type” and stops there.
- ⚠️ How the deposit-protection limit is applied. The “per person, per institution” description is widely used, but we could not confirm it in agency source text.
- ⚠️ The deposit insurer's own pages. Every limit-related page returned an error or an empty body, so we fell back on the association's posting of the notice.
- ⚠️ Causation in the CFPB data. The report shows differences across bands; it does not say savings reduce delinquency. We did not write it as causal.
- ⚠️ A Korean equivalent study. We found no Korean public survey linking emergency-fund size to delinquency or financial condition.
- Current rates on parking accounts and CMAs. They move constantly, so no figures appear here.
- How the financial well-being score (40/50/61) is constructed. We could not verify the scale's definition, so the body reports only the differences.
As of July 2026. The band-by-band figures come from the CFPB's 2022 report; the gaps and multiples are ours. ⚠️ This is general information, not a product recommendation or investment advice. Rates, conditions and protection status differ by product — check them on the sign-up screen before you open anything. Next steps are covered in pay down debt vs save and refinancing a loan.
