CD Ladder

A CD ladder is an arrangement of several certificates of deposit bought with different maturity dates so that one matures at regular intervals, commonly once a year. A borrower with $50,000 and a five-year ladder might hold $10,000 each in one-year, two-year, three-year, four-year and five-year CDs, so that $10,000 becomes available every twelve months without triggering an early withdrawal penalty.

The staggered maturities are only half of the structure, and they are the half that is visible from the outside. The other half is a reinvestment rule: as each rung matures it is renewed at the ladder's longest term rather than at its original one. On the five-year ladder above, the maturing one-year CD is replaced with a new five-year CD, then the two-year with a five-year, and so on. After five years every rung on the ladder is a five-year CD, one still matures every year, and the structure sustains itself indefinitely without further decisions.

The phrase is market vocabulary rather than a defined term. No bank regulator or standard-setting body defines a CD ladder, and it is not a distinct product with its own legal character, however banks and brokerages package the idea: it is a way of holding ordinary certificates of deposit. What a CD is legally, what the early withdrawal penalty is and is not, and why a brokered CD behaves differently are covered in the material on certificates of deposit.

What the structure actually buys, stated as the two problems it solves. The first is liquidity. A single long CD locks the whole sum until one date, and reaching it early costs whatever the deposit agreement says, which is set by contract rather than by any federal ceiling and can exceed the interest earned. A ladder converts that into a series of dates on which part of the money is free, so an unforeseen need can often be met by waiting for the next maturity instead of paying a penalty. The second is reinvestment-rate risk, meaning the risk of having to reinvest a large sum at whatever rate happens to prevail on one particular day. Spreading maturities across five years means no single date's rate determines the return on more than a fifth of the money, so the ladder earns a rolling average of the long rate over the cycle rather than the rate on one morning. That averaging is the point, and it cuts both ways by design: it also means a ladder never captures a peak.

The case in which a ladder is worse than the simplest alternative, and it is the case to check before building one. A ladder assumes the yield curve slopes upward, so that committing money for longer is paid for. When short maturities pay more than long ones, the assumption fails and every part of the structure works against the saver. The rungs at the long end pay less than the short end, so the blended yield is below what simply rolling the shortest CD would produce. Worse, the reinvestment rule then directs every maturing rung into the longest term, which is the lowest rate available, so the ladder systematically feeds money toward the worst price on the curve for as long as the inversion lasts. And it is doing this while giving up access to four fifths of the money. The saver has paid illiquidity for a negative yield premium.

The same problem appears in a milder and more common form. Where a savings account or a high-yield savings account pays more than the longest rung on the ladder, the ladder's entire yield benefit disappears while its costs, the administrative effort and the penalty exposure on any unplanned withdrawal, remain. Since a savings rate is variable and a CD rate is not, the honest way to frame that comparison is that the CD is buying rate certainty rather than yield, and the question is whether rate certainty is worth what it costs on the day.

Two alternative structures, because a page that names only its own subject is incomplete. A barbell holds very short and very long maturities with nothing in the middle, which produces more liquidity at the short end and more yield at the long end than an evenly spaced ladder, at the cost of holding nothing that matures in the intervening years. A no-penalty CD buys the liquidity directly: it allows withdrawal without the contractual penalty after a short initial period, and it pays a lower rate than a comparable standard CD in exchange. That initial period is not the bank being cautious: under 12 CFR 204.2(c)(1)(i) a deposit counts as a time deposit only if it carries a penalty of at least seven days' simple interest on money withdrawn in the first six days, so no CD can be penalty-free from the day it is opened. Where the reason for laddering is purely the fear of needing the money, a no-penalty CD may answer it at a known price rather than through a structure.

One thing a ladder does not do, and the confusion is worth heading off. Staggering maturities changes when money is available. It does not diversify credit risk, reduce exposure to any institution, or extend deposit insurance, because a ladder built at one bank is one bank's obligation across five certificates. Spreading deposits across institutions or across ownership categories in order to keep more than the insured limit covered is a separate exercise with separate rules, and the published material on deposit insurance covers it. A ladder and an insurance strategy can be run at the same time; they are not the same strategy, and describing a ladder as safer than a single CD confuses the two.

Divide the sum into equal parts, buy one CD at each maturity from one period out to the length of ladder you want, and record every maturity date. As each rung matures, renew it at the longest term on the ladder, or take the money if you need it. The structure needs one decision a year after that.

A hypothetical example of the arithmetic and of the steady state. Basia splits $50,000 into five parts of $10,000 and buys one-year, two-year, three-year, four-year and five-year CDs. Assume, purely to illustrate, an upward-sloping set of rates of 3.4%, 3.6%, 3.8%, 3.9% and 4.0%.

In the first year the blended yield is the average of the five, 3.74%, so the ladder earns roughly $1,870 ($50,000 × 0.0374). Holding the whole sum in one-year CDs would have earned about $1,700 and given up nothing in access. Holding it all in a single five-year CD would have earned about $2,000 and given up access to all of it for five years. The ladder sits between the two on both measures, which is what it is for.

When the one-year rung matures it is replaced by a new five-year CD. Repeat that each year and by the end of year five every rung is a five-year CD, one matures annually, and the ladder pays a rolling average of the five-year rates of the preceding five years. That is the steady state, and it is the structure's actual product: the long rate, averaged over the cycle, with a fifth of the money reachable each year.

Now invert the assumption, which is the case worth checking before building one. Suppose the rates run 4.6%, 4.4%, 4.2%, 4.0% and 3.9% from shortest to longest. The ladder's blended first-year yield is 4.22%, or about $2,110, while simply rolling one-year CDs would pay 4.6%, about $2,300. The ladder gives up roughly $190 in the first year, commits four fifths of the money for longer, and its reinvestment rule sends every maturing rung into the 3.9% five-year rate, the lowest on the curve. Nothing about the structure is malfunctioning. The structure is designed for a curve that slopes the other way.

Pros

• Part of the money reaches maturity at regular intervals, so an unplanned need can often be met without paying an early withdrawal penalty.
• No single date's rate determines the return on more than one rung, which removes the risk of committing everything at a low point.
• After one full cycle every rung earns the long-term rate while one still matures each period, so the structure sustains itself.
• It imposes an annual decision point, which is a natural moment to reconsider whether the money should stay in CDs at all.
• It is built from ordinary certificates of deposit, so it needs no special product and carries the same deposit insurance as any other deposit at the institution.

Cons

• It assumes long maturities pay more than short ones. Where they do not, the ladder locks in a lower yield and pays for illiquidity it is not being compensated for.
• In that situation the reinvestment rule makes it worse, because it directs each maturing rung into the lowest rate on the curve.
• Where a savings account pays more than the longest rung, the yield case disappears while the penalty exposure and the administration remain.
• It never captures the best available rate, since averaging is the whole mechanism.
• It is more work than one deposit, with several maturity dates, several renewal notices and several grace periods to track, and a missed grace period can roll money into a fresh term at an uncompetitive rate.
• It changes only the maturity profile. It does not reduce credit risk or extend deposit insurance.