
A certificate of deposit asks one real question: how long. The APY you get, the penalty you would pay to get out, whether the account renews itself at maturity — all of it follows from the term you pick. Most term advice tries to settle that question by guessing where rates go next.
This settles it differently. You get the current shape of the CD curve from FDIC national average data, the arithmetic that shows exactly what renewal rate a shorter term needs in order to beat a longer one, what an early-withdrawal penalty costs in dollars, and a way to choose a maturity from a calendar date instead of a forecast.
One ground rule for the numbers. Every rate below is an FDIC national average as of August 17, 2026. A national average is not an offer and not a rate you can go open today. Use it as a reference line for judging whatever a bank puts in front of you.
The two things a term decides
A term decides two things.
First, it fixes your APY for the full length of the CD. If rates move after you fund it, your CD does not move. That is the point of the product, and it cuts both ways.
Second, it sets the date your money comes back without a penalty. Before that date, the deposit agreement governs. CDs generally charge an early-withdrawal penalty, stated in the disclosure and often quoted as a number of months of interest.
What the term does not change is your insurance. FDIC standard coverage is $250,000 per depositor, per insured bank, per ownership category — the same for a three-month CD as for a five-year one. Credit union share certificates carry the same standard limit through the NCUA. Ownership categories (single, joint, certain retirement accounts, revocable trust) are insured separately from one another. FDIC trust-account rules were revised effective April 1, 2024, so if a trust holds the deposit, run it through the FDIC’s EDIE estimator rather than any formula you remember.
One more distinction that matters when you compare quotes: APY includes the effect of compounding, a stated interest rate does not. Compare APY to APY.
The national average term curve
FDIC national averages, as of August 17, 2026:
| Term | National average APY |
|---|---|
| 1 month | 0.22% |
| 3 months | 1.14% |
| 6 months | 1.41% |
| 12 months | 1.71% |
| 24 months | 1.57% |
| 36 months | 1.34% |
| 48 months | 1.27% |
| 60 months | 1.36% |
For context, the savings national average is 0.38% and the money market average is 0.63%, both as of the same date.
Two features are worth reading off that table. The peak sits at 12 months, not at the far end. And only the one-month average, 0.22%, falls below the savings average of 0.38% — every term from three months out averages above both savings and money market.
A curve where shorter terms yield more than longer ones is called inverted. When the averages look like this, “lock it up longer for a better rate” no longer describes the average, and the case for a long term has to come from something other than yield.
Break-even math for the shorter term
Here is the test that replaces a forecast. Put $10,000 to work for 24 months, two ways, using the national averages above.
Path A — one 24-month CD at 1.57% APY. Year one: $10,000 × 1.0157 = $10,157.00. Year two: $10,157.00 × 1.0157 = $10,316.46. Interest over two years: $316.46.
Path B — a 12-month CD at 1.71% APY, then a renewal for another 12 months. Year one: $10,000 × 1.0171 = $10,171.00. Year two is the unknown. For Path B to tie Path A, the renewal has to turn $10,171.00 into $10,316.46. That works out to a renewal APY of about 1.43%: $10,171.00 × 1.0143 = $10,316.45, one cent short of Path A.
So the shorter term wins unless the renewal a year from now comes in below roughly 1.43% APY — 0.14 points under today’s 24-month average and 0.28 points under today’s 12-month average.
That figure is the whole decision, and you can recompute it for any pair of offers with two lines of multiplication. You are not predicting anything. You are naming the rate that would have to appear for the long lock to have been worth it, then deciding whether you would be comfortable if it did.
Run the same test at the long end and it gets starker. The 60-month average of 1.36% sits 0.35 points below the 12-month average of 1.71%, so a five-year lock at the average starts you behind on yield and asks you to stay there for four more years. That can still be the right call if you have a date five years out and want the rate settled now. It is not a call the yield makes for you.
The cost of breaking early
The break-even above assumes you hold to maturity. Price the other case before you sign.
Take that same $10,000 in a 24-month CD at 1.57% APY, with a penalty of six months of interest, and suppose you need the cash at month 12. Using simple monthly proration — annual interest divided by twelve, no compounding inside the year, the convention used throughout this example:
- Interest earned in 12 months: $10,000 × 1.57% = $157.00
- One month of interest: $157.00 ÷ 12 = $13.083333
- Penalty of six months: $13.083333 × 6 = $78.50
- Net interest kept: $157.00 − $78.50 = $78.50
You keep $78.50 on $10,000, an effective 0.785% for the year.
Now compare it two ways. Money left in savings at the 0.38% national average would have earned $38.00, so the broken CD still came out $40.50 ahead. A 12-month CD at the 1.71% average would have paid $171.00, so the term mismatch cost you $92.50 against the CD you should have picked.
Run the same penalty at month six and the arithmetic turns ugly: six months of interest earned is $78.50, six months of penalty is $78.50, and you net $0.00. Some agreements take the shortfall out of principal when the penalty exceeds interest earned. Your disclosure states whether yours does.
Maturity dates and spending dates
The rule that beats a forecast is a calendar rule.
Write down the date you need the money. Not “sometime next year” — a date. Then pick a term that matures on or before it. A CD that matures two months after the tuition bill is due is the wrong CD no matter what it pays.
If you cannot name a date, the money probably needs to stay reachable, which is what savings and money market accounts are for: 0.38% and 0.63% national average respectively, lower yields with no maturity to work around. Regulation D’s six-transfer limit was suspended in 2020, though many banks still set their own transfer limits, so check the account agreement before assuming unlimited access.
Between those two cases sits money with a soft date — a car sometime this year, a move maybe next spring. That is where a short term does its best work. It pays a real yield, and it hands the money back often enough that a slipping plan does not cost you a penalty.
Ladders and split deposits
Splitting one deposit across several terms buys a maturity on a regular schedule instead of one date at the end.
Put $10,000 into four equal $2,500 pieces at 3, 6, 12, and 24 months. At national averages that is 1.14%, 1.41%, 1.71%, and 1.57%. Because the pieces are the same size, the simple average is also the deposit-weighted average:
(1.14 + 1.41 + 1.71 + 1.57) ÷ 4 = 5.83 ÷ 4 = 1.4575%
Against putting the whole $10,000 in the 12-month at 1.71%, the ladder gives up 0.2525 percentage points of blended yield. What it buys is a maturity in three months, another in six, another in twelve — four decision points instead of one, each of which can roll into a longer term or come out as cash.
That trade is worth making when you are genuinely unsure of your own timeline. It is not worth making when you know the date, because then you are paying for flexibility you already have.
The short version
- Name the date first. The term follows from the date, not from a rate view.
- No date usually means it is not CD money.
- Longer does not mean higher. As of August 17, 2026 the national average peaks at 12 months (1.71%) and falls to 1.27% at 48 months.
- Before accepting a longer lock, compute the renewal rate that would make the shorter path lose. If that rate looks unremarkable, the long lock is not buying much.
- Read the penalty line before funding, and check whether it can reach into principal.