Cash back vs travel rewards — which earns you more, by the math

The cash back versus travel rewards debate is misframed in nearly every piece of mainstream personal finance content that addresses it. The dominant framing presents the choice as a binary preference question — “do you prefer cash or do you like to travel” — and then either recommends one or recommends both based on the writer’s own travel volume. This framing produces no actionable answer because the question is not preference; the question is mathematical. The right card for a given household depends on the household’s annual spending volume, the distribution of that spending across categories, the household’s willingness to redeem points on travel rather than cash, and the household’s actual ability to extract the higher cents-per-point valuations the travel cards promise. With those four inputs, the comparison resolves into a number — typically a difference of several hundred dollars per year, sometimes much more — and the right card is the card with the higher number.

This guide walks through the math behind the four categories of US credit card rewards (flat-rate cash back, tiered cash back, transferable travel points, brand-loyalty points), the conservative valuation framework we use to compare them on a single dollars-per-year basis, the spending and behavior thresholds that determine which category wins for a given household, the hidden costs that erode the headline earn rates of the higher-value categories, and a worked example that walks a representative household through the actual decision. The framework is structured so that a reader can apply it to their own annual spending in roughly twenty minutes with a download of last year’s credit card statement.

The four reward currency categories

US credit card rewards split into four broadly distinct currency categories, each with different conversion mechanics, different redemption ceilings, and different volatility risk.

Flat-rate cash back. The simplest category. The card returns a fixed percentage — typically 1.5%, 2%, or in a small number of cases 2.5% — of every dollar spent, paid as a statement credit or a deposit into a linked bank account, with no category restrictions. The Citi Double Cash, the Wells Fargo Active Cash, the SoFi Credit Card (2% to deposit holders), the Apple Card (3% on Apple, 2% via Apple Pay), and several others sit in this category. The reward currency is US dollars; there is no ambiguity in valuation; there is no expiration risk on accumulated balances on most products.

Tiered cash back. A category-specific elevated rate (typically 3% to 6%) on a defined set of categories, with the standard 1% rate on everything else. The categories vary by card: the American Express Blue Cash Preferred pays 6% on groceries up to $6,000 in spending per year; the Capital One SavorOne pays 3% on dining, entertainment, and groceries; the U.S. Bank Cash+ pays 5% on two categories the cardholder selects from a published list. The reward currency is still US dollars; the complication is the category-cap structure (the elevated rate only applies up to a stated spending ceiling, after which the rate drops to 1%) and the category restriction (only spending within the defined categories earns the elevated rate).

Transferable travel points. Points that can be either redeemed directly through the issuer’s travel portal at a fixed cents-per-point rate (typically 1.0 cent per point, with issuer-selected exceptions above that) or transferred to one of the issuer’s partner airline or hotel programs at a 1:1 (sometimes 1:2 or 1:0.5) ratio, and then redeemed for award flights or hotel nights. The transfer-and-redeem pathway is where the higher headline valuations come from: a Chase Ultimate Rewards point that redeems for a flat 1 cent in Chase Travel (more only on ‘Points Boost’ bookings Chase selects) can frequently be transferred to Hyatt and redeemed for 3 to 4 cents per point on a hotel night from a Sapphire Reserve — a quarter less from a Sapphire Preferred, which transfers to Hyatt at 4:3 for cards opened since June 15, 2026 — or to United and redeemed for 2 to 3 cents per point on a domestic flight, or to a partner program for premium-cabin international flights at 5 to 10 cents per point. The Chase Sapphire family, the American Express Membership Rewards family, the Capital One Venture family, and the Citi ThankYou family are the four major transferable-points ecosystems in the US market.

Brand-loyalty points. Points denominated in a specific airline or hotel program (Delta SkyMiles, United MileagePlus, Hilton Honors, Marriott Bonvoy, IHG One Rewards, Hyatt). The currency is not transferable to other programs; the redemption value is what the issuing brand sets in its specific award chart. Brand-loyalty cards typically earn at higher multiples on spending with the brand (5x to 10x on hotel stays with the brand, 3x to 5x on airfare with the brand) and at 1x to 2x on other spending. The reward valuation is the most volatile of the four categories — airlines and hotels can and do devalue their programs unilaterally on short notice — and the redemption optionality is the narrowest.

One feature that almost every travel-oriented card in the second and third categories shares — and that almost every flat-rate or tiered cash-back card lacks — is the absence of a foreign transaction fee on non-US purchases. For households that travel internationally, even modestly, the avoided 3% fee on every overseas purchase frequently exceeds the explicit rewards earned on the trip, which can flip the comparison entirely in the travel card’s favor. The full mechanics of the fee, the related dynamic currency conversion trap, and which cards waive each are in the foreign transaction fees and DCC guide.

The conservative valuation framework — what a point is actually worth

The largest single source of error in cash back versus travel rewards comparisons is the use of aspirational point valuations rather than realistic ones. A blogger who points out that a Chase Ultimate Rewards point can be worth 10 cents on a Lufthansa first-class redemption is technically correct, but the same valuation applied across all of an applicant’s annual spending dramatically overstates the real value the applicant will extract. The aspirational redemption is available a small number of times per year, requires booking 11 months ahead, requires flexibility on origin and destination, and consumes hundreds of thousands of points per redemption. A household earning 60,000 points a year cannot live in the aspirational valuation universe.

The conservative valuation we use at finbarrow is the cents-per-point a typical household actually extracts on average across a year of redemptions, factoring in the mix of high-value redemptions (transfers to high-value programs) and low-value redemptions (cash-equivalent statement credits) most cardholders use. For the four major transferable-points programs, the conservative valuations are:

Chase Ultimate Rewards: 1.6 cents per point as a household average. The flat 1-cent Chase Travel redemption sets the floor (the 1.25-cent Preferred and 1.5-cent Reserve portal multipliers ended for new points in October 2025, replaced by selective Points Boost uplifts); the Hyatt transfer at a typical 2.5 to 3.0 cents per point from a Reserve — 1.9 to 2.25 cents through the Preferred’s 4:3 ratio — pulls the average up. Households that consistently transfer to Hyatt or United for high-value redemptions can reach 2.0 cents per point sustained; households that primarily redeem through the portal stay at 1.0 cent.

American Express Membership Rewards: 1.5 cents per point as a household average. The Pay With Points portal floor at 1.0 cent and the high-value transfers (ANA, Air Canada, ANA business class redemptions) at 2.0 to 4.0 cents combine to a typical 1.5. Households that aggressively transfer can reach 1.8 to 2.0 sustained; the portal-dominant household stays close to 1.0.

Capital One Venture miles: 1.3 cents per mile as a household average. The Capital One Travel portal at 1.0 cent and the transfer partners (Turkish Airlines and Wyndham being two of the higher-value options) at 1.5 to 2.5 cents combine to a typical 1.3.

Citi ThankYou points: 1.4 cents per point as a household average. The Premier card portal floor and the transfer partners (Avianca LifeMiles, Turkish, Singapore) combine to a typical 1.4.

These valuations are deliberately conservative. They reflect what the median sophisticated cardholder extracts in a typical year, not what a points expert extracts in the best year, and they are the basis for the math throughout this guide. A reader who is confident they can extract higher valuations consistently should adjust upward; a reader who knows they will be a portal-dominant redeemer should adjust downward toward the portal floor.

The effective earn rate — translating multipliers into dollars

The headline earn rate on a credit card (3x on dining, 2x on travel, 1x on everything else) does not translate directly into dollars-per-year without two additional pieces of information: the household’s spending distribution by category, and the point valuation applied to the resulting points. The combination produces an effective earn rate, expressed in dollars-per-dollar-spent, that is comparable across cards and across reward currencies.

A worked calculation for a household spending $40,000 per year on credit cards, distributed as $7,000 on groceries, $4,500 on dining, $3,500 on travel, $1,500 on gas, $1,800 on streaming and household subscriptions, and $21,700 on other (general merchandise, online shopping, services, miscellaneous):

On the Chase Sapphire Preferred (3x on dining, 3x on online groceries up to $1,000/month, 2x on travel, 1x on other, $95 annual fee, valuation 1.6 cents per point):

• Dining: $4,500 × 3 points = 13,500 points × 1.6 = $216
• Online groceries (capped at $12,000/year, well above the household’s $7,000): $7,000 × 3 = 21,000 points × 1.6 = $336
• Travel: $3,500 × 2 = 7,000 points × 1.6 = $112
• Other: $24,000 × 1 = 24,000 points × 1.6 = $384
• Total rewards value: $1,048. Net of $95 annual fee: $953 per year.

On the Citi Double Cash (2% across the board, no annual fee):

• $40,000 × 2% = $800 per year, no annual fee, $800 net.

On the American Express Blue Cash Preferred (6% on groceries up to $6,000, 6% on streaming, 3% on transit/gas, 1% on other, $95 annual fee):

• Groceries: $6,000 × 6% = $360, remaining $1,000 at 1% = $10
• Streaming: $1,800 × 6% = $108 (assuming all streaming/sub qualifies; in practice partial)
• Gas: $1,500 × 3% = $45
• Other: $30,700 × 1% = $307
• Total rewards: $830. Net of $95 annual fee: $735.

The ranking for this household: Sapphire Preferred at $953, Citi Double Cash at $800, Blue Cash Preferred at $735. The travel card wins by approximately $150 over the flat-rate cash back. But the win depends critically on the household actually extracting the 1.6 cent-per-point valuation. At the flat 1-cent Chase Travel valuation that applies to points earned since October 2025, the 65,500 points are worth $655, and net of the $95 fee the Sapphire Preferred number drops to $560 — well below both the Citi Double Cash and the Blue Cash Preferred.

This sensitivity is the core of the cash back versus travel rewards comparison. The math swings on the realized valuation of the points, which swings on the household’s redemption sophistication and willingness to plan travel around point availability. A household that is honest with itself about how it will actually redeem (portal-dominant, low effort) will frequently see the math favor cash back even when the headline numbers suggest travel rewards win.

The break-even — when does travel actually beat cash

The break-even between transferable travel points and flat-rate cash back depends on three variables: total annual credit card spending, distribution across bonus categories, and realized point valuation. The break-even patterns that emerge across thousands of household examples:

For a household spending under $25,000 per year on credit cards with the spending broadly distributed across non-bonus categories, flat-rate 2% cash back nearly always wins. The bonus categories on a travel card cover too small a fraction of the spending to overcome the lower base rate on the rest.

For a household spending $25,000 to $50,000 per year with at least 25% of spending in bonus categories (dining, travel, groceries on a card that bonuses groceries), the travel card and flat-rate cash back are close, and the realized point valuation tips the decision. At 1.5+ cents per point, travel cards typically win by $100 to $300 per year; at 1.25 cents per point or below, flat-rate cash back typically wins.

For a household spending above $50,000 per year with substantial dining or travel spending (over 30% of total), travel cards typically win by $500 to $1,500 per year, even at conservative point valuations. The compounding of the higher multipliers on the larger absolute dollar volumes overcomes the cash back simplicity advantage.

For a household at any spending level that genuinely does not travel (under three flights per year, no hotel stays) or genuinely is not willing to redeem points for travel, cash back wins regardless of the math. Points that sit in the account at 1.6 cents per point on paper are worth zero if they are never redeemed; cash back deposited as a statement credit is worth its face value the moment it lands.

The hidden costs that erode the headline numbers

The math above assumes the household actually captures the full headline rewards on each card. In practice, several frictions reduce the realized rewards substantially on the higher-rate cards, and the realized advantage of travel cards over cash back is frequently smaller than the calculation suggests.

Annual fee carrying costs that do not earn their keep. A card with a $95 annual fee that pays $200 in marginal rewards beyond the no-fee alternative is a $105 net positive — but a card with a $95 annual fee that pays $80 in marginal rewards is a $15 net negative. Households that hold multiple annual-fee cards because each one looks favorable in isolation frequently find the aggregate is below what a single no-fee 2% card would have produced.

Maintenance cost of multiple cards. Tracking three or four cards across different bonus categories, remembering which card to use for which purchase, and reconciling statement balances costs real time. For a household where the marginal dollar of rewards is worth more than ten dollars of additional cognitive load, the math is positive; for a household where additional cognitive load is genuinely costly, the simpler 2% card frequently wins on a quality-of-life basis even when it loses on raw dollars.

Bonus category misses. Tiered cash back cards and rotating cash back cards reward higher rates only on spending that fits the defined categories. The categories are not always clear in real time — is a meal at a restaurant inside a hotel “dining” or “travel” or “lodging”? The merchant code (MCC) the merchant submits determines the category, and the merchant has discretion over which MCC to use. A category-tiered card systematically misses spending the cardholder believed would qualify; the realized bonus capture is typically 80% to 90% of the theoretical maximum on category-restricted cards, not 100%.

Devaluations on brand-loyalty points. A travel card that earns Hilton Honors or Marriott Bonvoy points at high multiples is exposed to devaluation risk on the underlying loyalty program. Hilton and Marriott have both made unannounced devaluations several times in the past decade, dropping the redemption value of accumulated points by 20% or more overnight. Transferable points are less exposed because the holder can choose which transfer partner to use after the devaluation; brand-locked points are fully exposed.

Statement credit complexity on premium cards. The Chase Sapphire Reserve, American Express Platinum, and similar premium cards offset their $795 to $895 annual fees with assorted statement credits (Uber, dining, airline incidentals, lounge access, Global Entry). The credits are real but only valuable if the cardholder actually uses them on spending they would otherwise have made. A $200 airline incidental credit is worth $200 only if the household would have spent $200 on the qualifying airline anyway; for households that would not have, the credit is worth zero. The advertised “net cost” of a premium card after credits is consistently lower than the realized net cost across a representative cardholder population.

The 80% rule — for most US households, cash back wins

A defensible rule of thumb that emerges from the math: for roughly 80% of US households who carry credit cards, a flat-rate 2% cash back card is the right primary card. The 20% who benefit meaningfully from a travel rewards card are characterized by some combination of high total credit card spending (over $40,000 per year), substantial spending concentration in bonus categories (over 30% in dining-and-travel for example), genuine willingness to redeem points for travel, and the time and inclination to optimize point transfers.

For the 80%, the cash back math is straightforward, the cards are simple, the rewards arrive as cash that fully fungible with any other money, and the time cost of optimization is essentially zero. The marginal $150 a year a travel card might pay over the cash back card, in expected value, is consistently lower than the time cost of the optimization required to capture it.

For the 20%, the travel rewards math compounds. A household at $80,000 in annual credit card spending with a Sapphire Preferred plus a Freedom Unlimited plus a strategic American Express or Capital One layer can plausibly capture $1,800 to $2,500 in rewards value per year, against $900 to $1,200 on a flat-rate 2% card. The marginal $1,000 a year is worth the optimization time for a household that values travel and spends the volume. The sign-up bonus math guide covers the year-one yield calculation that frequently dominates the year-one decision regardless of the ongoing-rewards comparison.

The mistake to avoid in both directions: a low-spending household applying for premium travel cards because the math looks favorable in marketing collateral, and then under-extracting the value; a high-spending traveler defaulting to a 2% cash back card out of simplicity, and leaving substantial money on the table that a moderate-effort travel strategy would capture. The math is the discipline that prevents both errors.

A worked example — three households, three different answers

Consider three households, each with different spending profiles and different preferences. The right card for each is dictated by the math, not by a one-size-fits-all preference.

Household A — single adult in a low-cost city, $22,000 annual credit card spending, $3,500 on groceries, $2,000 on dining, $0 on travel. No travel ambition, prefers simplicity. The math: flat-rate 2% card pays $440 per year, no annual fee. The Blue Cash Everyday (3% on groceries) pays $105 on groceries + $370 on other (1%) + $40 on dining (1%) = $515; but the realized capture is closer to $480 after category-miss friction. The simple 2% card is within $40 of the optimized cash back card and saves cognitive overhead. The right answer for Household A: a flat-rate 2% cash back card.

Household B — couple in a high-cost city, $48,000 annual credit card spending, $9,000 on groceries, $7,500 on dining, $4,000 on travel, $4,000 streaming and entertainment. Travel two to three times a year, willing to redeem points on hotels and airfare. The math: Sapphire Preferred earning 21,000 points on groceries, 22,500 on dining, 8,000 on travel, 23,500 on other (with category caps factored in) at 1.6 cents = $1,200 net of $95 annual fee. Flat-rate 2% pays $960. Travel card wins by $240. The right answer for Household B: the travel card, with the awareness that the win depends on actually redeeming for travel at the conservative valuation.

Household C — frequent business traveler, $75,000 annual credit card spending, $4,000 on groceries, $14,000 on dining, $15,000 on travel, $42,000 on other. Travel monthly, redeems for premium-cabin international flights using transfer partners. The math: Sapphire Reserve (3x on dining, 3x on travel, 1x on other, $795 annual fee, $300 travel credit, $400 dining credit) at 1.8 cents per point realized valuation = $14,000 dining × 3 = 42,000 points + $15,000 × 3 = 45,000 + $42,000 × 1 = 42,000 + $4,000 × 1 = 4,000 = 133,000 points × 1.8 = $2,394; net of $795 fee, plus $700 in usable statement credits, = $2,299. Flat-rate 2%: $1,500. The travel card wins by roughly $800 per year. The right answer for Household C: the premium travel card, with the awareness that the win compounds at high spending volumes and high redemption sophistication.

The pattern across the three: the right card scales with spending volume, redemption sophistication, and bonus-category concentration. There is no single answer; there is a calculation per household that returns one.

Sources

• US credit card market structure and issuer ecosystem: Federal Reserve — Consumer Credit (G.19) and CFPB — Credit card market report.
• Point valuation methodology and historical valuations: The Points Guy — Monthly Valuations (we use roughly 70% of TPG’s valuations as our conservative baseline because TPG’s are aspirational; the conservative haircut produces the household-average valuations cited above).
• Reward card issuer terms and bonus structures: each card’s specific application page on the issuer site (Chase.com, AmericanExpress.com, CapitalOne.com, Citi.com).
• Merchant category code (MCC) mechanics: Visa — Merchant Data Standards Manual.
• Loyalty program devaluation history: tracked by personal finance communities and aggregated at Frequent Miler.

If a valuation on this page looks off against current realized valuations, the personal finance communities update faster than this article; let us know via contact and we will reconcile.