How Credit Card Rewards Actually Work — And Why Most People Leave Money on the Table

Understanding how this machine works is the difference between extracting a few tenths of a percent from your spending and extracting two or three percent from exactly the same spending. Same purchases, same month, same bank. Different outcome, because one person knows where the money comes from and the other does not.

Most people sign up for a rewards card, use it vaguely, and collect whatever falls out. That is like buying a gym membership and only using the sauna. The equipment is right there. You just need to know which machines do what — and, more importantly, who is paying for them.

Who Actually Pays for Your Rewards

Every card purchase moves money through a four-party chain. You pay the merchant. The merchant's bank — the acquirer — collects the money and passes it on, minus a merchant service charge. Inside that charge sits an interchange fee, which the acquirer hands to the bank that issued your card. The card network (Visa, Mastercard, Amex) sets the interchange schedule and takes a separate, much smaller, scheme fee.

The issuing bank keeps the interchange. Your rewards are carved out of it. That is the whole loop. Your 1% cashback exists because the merchant paid roughly 1% for the privilege of accepting your card, and your bank decided to hand part of it back to keep you spending on its plastic rather than someone else's.

This is not a theory you have to take on faith, because interchange is regulated and the numbers are published. In the European Economic Area, Regulation (EU) 2015/751 caps interchange at 0.2% of transaction value for consumer debit cards (Article 3) and 0.3% for consumer credit cards (Article 4). Those are hard legal ceilings on the pool of money that funds EEA consumer card rewards.

The United States regulates the same thing very differently, and this single asymmetry explains more about the global rewards market than anything else. The Federal Reserve's Regulation II caps debit interchange for covered issuers at 21 cents plus 0.05% of the transaction, with a further one-cent fraud-prevention adjustment for eligible issuers — but Regulation II applies only to debit cards. US consumer credit interchange is not capped at all.

So a US issuer funding a rewards card is drawing on an uncapped revenue stream, while an EEA issuer is working inside a 0.3% ceiling. That is why a headline American rewards card can plausibly advertise multipliers that no European consumer card can match on interchange alone, and why European card marketing leans so much harder on insurance, travel perks, and fee waivers instead. If you have ever wondered why the rewards landscape looks so much thinner outside North America, that is the answer, and it is written into law rather than into anyone's generosity.

The Interchange Ceiling — and Why "5% Cashback" Is Never What It Looks Like

Once you know rewards are funded out of interchange, the ceiling on honest flat-rate cashback becomes obvious. A card cannot durably pay you more than it collects. So when a card advertises 3%, 5%, or more, one of four things is happening, and it is worth knowing which:

• The rate is capped by spend. "5% on groceries, up to 3,000 per month" means 5% on the first 3,000 and the base rate on everything after. The blended rate is what matters, not the headline.
• The rate is capped by time. Rotating quarterly categories that require activation are designed around the fact that a meaningful share of cardholders will forget to activate.
• An annual fee is cross-subsidising it. This is legitimate, but it turns the card into an arithmetic problem rather than a free lunch. Our guide to when an annual fee is actually worth paying works the break-even.
• The merchant is funding it directly. Retailer-linked offers and co-brand deals sit outside interchange entirely; the shop is buying your visit.

None of these are scams. They are simply not what the headline number implies. The rule that survives all of them: convert every offer to a blended annual figure in cash before comparing anything.

The Three Types of Rewards, and What Each One Really Is

Cashback

A percentage of each purchase returned as a statement credit, bank transfer, or cheque. One unit of currency is one unit of currency, permanently. There is no redemption ambiguity, no award chart, and no devaluation risk. Cashback is the only reward type where the value you earn is the value you keep, and that certainty is worth more than most people price in.

Flat-rate cashback is the dominant retail-bank card format almost everywhere, from US issuers to European ones such as Komplett MasterCard or Nordea's Platinum tier. Check the current rate on the issuer's own page before comparing — cashback percentages and any monthly caps are revised regularly, and a rate quoted anywhere other than the issuer is a rate that may already be out of date.

Bank Points

A proprietary currency the issuer creates, values, and can revalue. This is the critical distinction: points are not money you hold, they are an unsecured promise the issuer can reprice. Some points redeem at a fixed rate through the issuer's own travel portal. The more interesting ones are transferable — they move into airline and hotel loyalty programmes at a set ratio, where their value is then determined by a third party.

American Express publishes on its Gold Card page that the Gold Card earns Membership Rewards points transferable to 20 airline and hotel travel partners, as published by American Express in July 2026. Chase publishes on its Sapphire Preferred page that points from that card transfer to travel programmes including United MileagePlus, British Airways Executive Club, Air Canada Aeroplan, World of Hyatt and Marriott Bonvoy, as published by Chase in July 2026. Neither issuer publishes a fixed cash value per point, and that is the point: transferability is what separates a points currency with upside from one that is just slow cashback, but the upside is only ever realised at redemption.

Airline and Hotel Miles

Loyalty-programme currency earned directly, whether through a co-brand card such as an SAS EuroBonus Amex or through flying. Value is set entirely by the programme's award chart, which the programme can rewrite at will and frequently does. Excellent for people who genuinely fly one alliance and understand award availability. Poor for everyone else, who tend to accumulate a balance they eventually redeem for gift cards at a terrible rate.

Worked Example: Divide, Never Multiply

Here is the single arithmetic habit that separates people who win at this from people who think they are winning.

Suppose a card advertises two points per unit spent, and you spend 3,000 a month. That is 36,000 a year in spending and 72,000 points. Against a plain 1.5% cashback card, which returns 540, the points card sounds like an obvious victory — "2x" beats "1.5%" on any marketing slide.

It tells you nothing. To compare them you have to find the value of one point, and the only way to do that is to divide. Take the cash price of the thing the points actually buy and divide it by the points required.

If those 72,000 points, transferred to a partner programme, book a seat that genuinely sells for 900, each point is worth 0.0125 and you have earned 900 — a real 2.5% return, comfortably ahead of the cashback card. If the same 72,000 points buys a 360 gift card through the portal, each point is worth 0.005, you have earned a 1% return, and the 1.5% cashback card beat you while requiring no effort whatsoever.

The earn multiplier is the numerator. Almost all the variance lives in the denominator, and the denominator is the number the marketing never shows you. A card like the Chase Sapphire Preferred is interesting precisely because transfer partners give the denominator a chance to move in your favour — but only if you actually do the transfer and actually find the award seat. If you will not, treat the card as though its points redeem at the portal rate, because that is what will happen. We compare the two philosophies in detail in cashback versus travel points.

The Grace Period Is the Whole Game

Everything above is irrelevant if you pay interest, and the mechanism people misunderstand is the grace period.

The CFPB explains that a grace period is the window between the end of a billing cycle and the payment due date, that issuers must deliver the bill at least 21 days before payment is due, and — the part that costs people real money — that no law requires a grace period to exist at all. It is a courtesy conditional on paying in full.

Miss a full payment once and two things happen. Interest applies to the unpaid portion, which everyone expects. But you also lose the grace period on the next cycle, meaning interest starts accruing on each new purchase from the day you make it, not from the statement date. The true cost of the first missed full payment is not one month of interest on the shortfall. It is one month of interest on the shortfall plus the loss of interest-free float on everything you buy next month.

Put a number on that. The Federal Reserve's G.19 Consumer Credit release reported an average rate of 22.15% on credit card accounts assessed interest in May 2026 (preliminary) (20.94% across all accounts). At 22.15%, carrying 3,000 of new purchases for roughly thirty days with no grace period costs about 55 in interest that a full-payer would never have incurred — on spending that generated perhaps 45 in rewards at 1.5%.

The Break-Even: How Much Revolving Debt Cancels Your Rewards

The formula is simple enough to do in your head, and it is the most useful thing in this article.

Break-even balance = annual rewards ÷ APR.

Take a strong 2% cashback card and 3,000 a month of spending. That is 36,000 a year, earning 720. At the 22.15% average rate above, a balance revolving at an average of about 3,250 costs 720 a year in interest — because 3,250 × 0.2215 ≈ 720.

So roughly one month's spending, carried permanently, erases every unit of reward the card produced. Carry two months' spending and the card is costing you around 720 a year net while still sending you cheerful emails about how much cashback you have earned. This is not an edge case. It is the ordinary condition of a large share of cardholders, and it is why the CFPB's market reporting consistently observes that consumers who revolve pay far more in interest and fees than they receive in rewards.

Sign-Up Bonuses: Real Money, Real Distortion

Welcome bonuses are not marginal. The CFPB's May 2024 rewards issue spotlight found that nearly one in ten dollars earned by consumers in rewards is linked to sign-up bonuses. That is a large enough share that ignoring bonuses entirely is its own mistake.

The distortion is that a bonus is one-off and the card is forever. A generous bonus attached to a card whose ongoing earn rate is poor for your spending pattern is a single good year followed by many mediocre ones. Chase and Amex both structure bonuses around a minimum spend inside a fixed window, which is where people manufacture spending they did not need — and destroy more value than the bonus was worth. The test is whether you would have made the purchase anyway; if not, you are buying rewards at full retail. Hitting a minimum spend without burning cash covers the routes that survive that test, including large recurring bills that can legitimately be moved onto a card.

The application itself has a cost, though a smaller one than folklore suggests. FICO publishes that its score is composed of five weighted categories: payment history at 35%, amounts owed at 30%, length of credit history at 15%, and credit mix and new credit at 10% each. New credit — the category a hard inquiry touches — is the joint-smallest input, and FICO notes the risk signal is really about opening several accounts in a short period, especially for people without a long credit history. One application, well spaced, is a rounding error against a 35% payment-history weighting.

Points Are a Liability, Not an Asset

The same CFPB spotlight documents what the fine print permits: issuers increasing the points required for a redemption after you have earned them, points vanishing when an account closes, expiry applied without prior communication, and conditions buried deep enough that the marketing and the terms describe different products.

Treat that as the design of the system rather than a malfunction. A points balance is an unsecured, unindexed, revocable claim on a company that can unilaterally change the exchange rate and has a commercial incentive to do so. Cash does not have that property. This is the honest argument for cashback over points for most people, and it has nothing to do with which one earns a bigger headline number.

The practical rule: do not hoard. A balance you can redeem this year at today's rate is worth more than a larger balance redeemed in three years at whatever rate exists then.

Spending Abroad: Where Rewards Quietly Go Negative

Foreign transaction fees are the fastest way to turn a rewards card into a losing proposition, because they are charged on the full transaction while rewards are a small percentage of it.

The arithmetic is brutal and takes one line. A card charging a 3% foreign transaction fee while paying 2% back returns negative 1% on every purchase made in another currency. Spend 4,000 abroad on a trip and you have paid 120 in fees to earn 80 in rewards — a 40 loss on a card you believed was making you money. Dynamic currency conversion at the terminal, where the merchant offers to bill you in your home currency at their exchange rate, layers a further markup on top; we take that apart in dynamic currency conversion and hidden FX fees.

The structural fix is to separate the two jobs. Keep the rewards card for domestic spending where interchange funds a real return, and hold a genuinely low-cost multi-currency card such as Wise for anything denominated in another currency, so the FX leg is priced at something close to the mid-market rate instead of quietly eating the rewards you just earned.

Who Rewards Cards Are Wrong For

This is the section most comparison sites skip, so it deserves stating plainly. A rewards card is the wrong product if any of the following describes you.

• You carry a balance, even sometimes. See the break-even above. The rewards are not close to covering the interest, and a low-rate or balance-transfer card is the correct product instead.
• You are near the edge of your budget. Rewards cards are engineered to increase spending. If a 2% return is causing you to spend more than you otherwise would, you are paying 98 to receive 2.
• You have a thin or damaged credit file. Approval odds are poor and the applications cost you inquiries for nothing. Build the file first.
• Your income is irregular. The grace period requires paying in full every single month. A card that punishes one bad month with a full cycle of daily-accrued interest is a poor fit for lumpy income.

The regulatory record is unusually blunt on the underlying dynamic here. The FCA's credit card market study found that competition in the market "is focused on certain product features such as introductory promotional offers and rewards" with "less competitive pressure on interest rates outside of promotional offers and on other fees and charges" — and identified around 650,000 people who had been in persistent debt for three years or more, plus a further 750,000 making systematic minimum repayments over the same period.

Read that carefully. The regulator's finding is that rewards are the competitive battleground precisely because interest rates are not. The rewards are the shop window; the interest is the business.

The Practical Framework

Audit your spending. Pull three months of statements and categorise: groceries, fuel, dining, online, travel, everything else. You need percentages, not impressions — almost everyone misjudges at least one category badly.

Match cards to your actual categories. Pair one category card covering your largest bonus-eligible spend with one strong flat-rate card for everything else. Two cards capture the large majority of available value; a fourth and fifth card add complexity faster than they add return.

Set up autopay for the full statement balance. Not the minimum, not a fixed amount — the full statement balance. This is the single control that keeps the grace period intact and the whole exercise profitable.

Use the right card at the point of sale. A category card left in a drawer earns the base rate on nothing. Set it as the default payment method for the merchants where it applies.

Separate the FX leg. Domestic spending on the rewards card, foreign-currency spending on a low-FX card.

Review once a year. Issuers change earn rates, categories, and caps, and your spending changes too. Fifteen minutes a year through our rewards calculator is enough.

The Red-Flag Checklist

• "Up to X% cashback" — the "up to" is carrying the sentence. Find the base rate.
• Tier structures needing a flowchart — complexity is a margin strategy, and it is not your margin.
• High annual fee against modest rewards — the fee must be cleared by realistic value, not maximum theoretical value.
• No cash redemption option — if you cannot convert to cash, the issuer controls the exchange rate permanently.
• Caps on bonus categories — always compute the blended rate across your real monthly spend.

The Bottom Line

Credit card rewards are a share of interchange returned to you to keep you spending, bounded by law in the EEA at 0.3% for consumer credit cards and bounded by nothing at all on US credit interchange. Within that structure the bank expects to win on average, and it wins mostly through interest at rates the Federal Reserve puts north of 22% on accounts assessed interest.

Your edge is narrow but completely real: pay in full every month so the grace period holds, value points by dividing rather than multiplying, keep foreign-currency spending off the rewards card, and re-run the numbers once a year. Do those four things and you are extracting two to three percent from a system designed to give you as little as it can. Skip any one of them — particularly the first — and you become the customer the issuer likes best: the one who is quite sure they are winning.