Is a credit card annual fee worth it?

At a glance#

The basic calculation#

An annual fee is a fixed price for a bundle of benefits. What the bundle costs matters less than what you used from it, and only what you would have paid for without the card counts. Everything else is an offer that cost you nothing and gave you nothing.

Brochure value and personal use are two very different figures. The advertised value of a package almost always sits far above the benefit you took, and the gap between them is the gap between an honest calculation and a promotional one.

At the end there is one number. If the value used exceeds the fee, the card carries itself.

The calculation runs over a year, because the fee falls annually and the benefits reset annually. A year is also long enough to even out outliers, and across several years one exceptional redemption or one unused certificate distorts the average far less. Holding a card longer simply means carrying on with the sums: three lines per year are enough, and after three years the trend is unmistakable.

What belongs in the calculation#

Benefits fall into three groups: awards with a fixed purpose, credits in points, and benefits that mainly save time. Each is valued differently.

Awards with a fixed purpose#

Free nights, certificates and credits are the easiest to value. They have a cash value, and that value is the price of what you would otherwise have bought. For a free night that means the night you would have booked anyway, as free night certificates describes, and not the most expensive property in the category.

Purpose-bound credits count only insofar as they were used. A monthly credit for a service you do not use has no value to you. Credits split by month are particularly tricky, because an amount per month lapses monthly: using it three times a year means counting three months, not twelve.

Points and miles#

Earning on spending is redeemable currency and enters at its redemption value, which what points are worth works through in detail.

Valuing it at a brochure rate flatters the card. Valuing it at zero because no redemption has happened yet understates it. A conservative value that matches the way you redeem works best, stays constant across years and keeps the calculations comparable.

Bonus earning on particular categories belongs here too. A higher rate on travel counts only for the part above the base rate, because the base would also come from a card without a fee.

Benefits that save time#

Lounge access, shorter queues and earlier boarding are hard to quantify. One workable measure is the price a single entry would have cost, multiplied by the number of actual visits. Two visits a year are two entries, and an unlimited membership does not change that.

Valuing time above money is allowed, provided the valuation stays the same across years; consistency matters more than the level. Collecting several currencies calls for the same approach for each, otherwise the card with the most generous valuation always looks best, which says nothing about its use.

Status benefits sit in the same group and count only when they worked: a bag checked without a fee, a seat chosen without a charge, earlier boarding that made a difference on a tight connection. A tier that never came into play over the year contributes nothing, however good it looks on the card carrier.

Worked example: two years, one card#

The same card is held for two years, with a welcome bonus in the first year and none in the second.

Two years like these make up the most common pattern. The first year almost always carries itself; the second decides the card.

Most bad decisions start with the free night. A benefit used in the first year and not in the second still gets counted from memory, and the card survives on a night that never happened.

Insurance is the hardest item. It only pays out in a claim, and a year without one does not mean it was worthless. A workable approach is the premium of a comparable standalone policy: somebody who would have bought such a policy anyway enters its price, and somebody who would not enters zero.

Fee reimbursements, for instance for expedited screening applications, arise rarely. When they do, they count in full, but only in the year of use.

“Check and decide” is where the real work lies. Checking does not mean closing automatically; it means looking at the alternatives from the same issuer. A cheaper card in the same family, a retention gesture or a closure are all possible outcomes, and only the second-year figures show which one fits.

When it is not worth keeping the card#

The fee is not worth paying in these situations:

• When the benefits go unused. A bundle without use is a subscription without value.
• When travel frequency falls. Expensive cards bundle benefits that assume a certain number of trips, and below that level the cheaper card wins.
• When the same benefits exist elsewhere. Duplicate insurance and duplicate lounge access count once.
• When the points are never redeemed. A currency without a redemption has no value, however large the balance.
• When the card changes how you spend. Spending that happens only because of the card means paying more for points than they are worth.

Putting two years side by side shows the trend immediately. A card that sits just below the fee in the second year is a candidate for a downgrade before an outright closure.

The issuer’s no-fee card belongs in the comparison as well. When that free alternative delivers eighty percent of the benefit, the expensive card carries only the remainder, and that remainder has to justify the whole fee.

What to check before closing#

Three points decide whether closing is the right answer.

First, where the points sit. Bank points hang on the account and frequently lapse on closure; programme currencies sit in the programme and stay. Holding several cards from one issuer adds a twist, because the points accounts often pool, and then the shared account decides what happens to the balance, whichever card is closed.

Second, the alternative within the same issuer. Moving to a card without a fee keeps the account, and cancel the card or downgrade it explains how that runs.

Third, the timing. The natural moment for the whole check is when the fee posts, because the decision is due then anyway and all of last year’s numbers exist. After posting some issuers allow a refund window, and once it closes the fee is gone.

Signalling an intention to cancel sometimes brings a gesture from the issuer, such as a credit or bonus points. There is no entitlement to it, and the answer depends on your history with the bank. An offer of that kind goes into the calculation like any other credit, at the value you will use.

Spending deserves a look too. A card that barely gets used loses not only its benefits but its earning as well, and the two together frequently decide against keeping it.

A short list is enough: fee, awards used, credits used, lounge visits, insurance claims, points earned, and the total below. The first run takes half an hour and last year’s statements. From the second year on, the old list with new numbers does the job and the decision takes minutes.

Why the first year is its own calculation#

In the first year the welcome bonus sits on the credit side. It is a one-off and therefore no argument for the second year.

Carrying it forward means holding a card that never pays for itself again, and that is how stacks of cards nobody oversees come about. Two separate calculations keep it clean. The first answers whether opening the card paid, the second whether keeping it pays, and what a welcome bonus is worth covers the first half.

A waived first-year fee changes nothing about the logic. The decision simply moves twelve months, and the calculation stays the same.

Supplementary cards sometimes carry their own fees and sometimes their own benefits, and both belong in the same calculation as the main card. Ongoing discounts with partners count only where you shop anyway.

Noting each use during the year works far better than reconstructing it at the end. One line per use is enough, and memory stops deceiving you.

What applies with several cards#

Several cards mean several separate calculations, followed by a check of the overlaps. A benefit present in two cards counts in one.

Insurance packages and lounge access overlap most often. Both are expensive in the fee and worthless when duplicated.

Start with the most expensive card, because that is where the largest amount is at stake, and put all the checks for a household into the same month so that overlaps show up in one pass.

For the second-year decision one question helps. Would you apply for this card today if there were no bonus? A no is a clear result.

Documenting several years reveals your own patterns. In heavy travel years expensive cards carry themselves easily; in quiet years they almost never do. Accepting that usually leaves two or three cards where there were six, and the survivors are the ones whose benefits arise in your own travel year. The annual fee is the only item in this field that is certain to arrive every year. Everything else is an offer, used or not.

The most common mistakes#

• Calculating with brochure values. Only what you would otherwise have bought belongs in the sum.
• Carrying the bonus forward. The welcome bonus counts in the first year and never again.
• Counting unused credits. A credit is worth the part you used, and the rest lapsed.
• Counting duplicate benefits twice. Lounge access held on two cards is one benefit, so check across all of them.
• Closing after the fee posts. Once the refund window has passed the fee is spent, which is why the decision belongs before the posting date.
• Valuing points without a redemption practice. Set a conservative value and keep it from year to year.

Related topics#

• What a welcome bonus is worth
• Cancel the card or downgrade it
• What points are actually worth
• Free night certificates
• Who gets into the lounge
• Points after closing a card