Open two credit card offers side by side and the American economy splits in half. One envelope promises a $750 sign-up bonus, airport lounge access, and a 0% introductory rate. The other demands a $200 security deposit, charges a $99 annual fee before the card even arrives, and carries an APR above 28%. Both envelopes are arriving more frequently. What is disappearing is the middle: the plain, decent card for the borrower who is neither wealthy nor in crisis.
That anecdotal contrast now has federal data behind it. The Consumer Financial Protection Bureau’s credit card origination tracker, which counts actual new accounts opened across six risk tiers, shows a clear divergence since 2021. The share of new credit card accounts going to superprime borrowers (scores of 780 and above) has climbed steadily. Originations for deep subprime and subprime borrowers (below 620) have also ticked upward. The near-prime and prime segments, roughly scores of 620 to 719, have flattened or lost ground. As of the bureau’s most recent release, reflecting activity through approximately the fourth quarter of 2025, the credit card market is splitting at the seams.
Economists recognized this shape early in the pandemic recovery. Moody’s Analytics chief economist Mark Zandi and others called it a “K-shaped economy,” a world where upper-income households surge ahead while lower-income households fall behind. By mid-2026, that dynamic has migrated squarely into the lending data, and the consequences reach well beyond a line on a chart.
What the federal data actually shows
The CFPB’s Consumer Credit Trends platform tracks originations across credit cards, auto loans, mortgages, and student debt using anonymized, account-level records from major lenders. For credit cards, the bureau segments new accounts into six risk tiers: deep subprime, subprime, near-prime, prime, prime-plus, and superprime. The definitions, laid out in the agency’s downloadable data dictionary, have remained stable for years, which means shifts in each tier’s share are not artifacts of changing methodology.
The pattern is not subtle. Superprime originations have grown as lenders compete aggressively for their safest customers, the borrowers who qualify for 0% intro APR offers, premium rewards cards, and credit limits that can stretch into six figures. At the other end, subprime originations have expanded as issuers push deeper into high-fee, high-interest products: secured cards, store-branded lines, and niche offerings built to tolerate elevated default risk. The middle tiers, where tens of millions of working-age Americans sit, have not attracted the same growth in new lending.
These are records of actual account openings, not survey responses or model projections. A federal regulator is counting how many new credit lines went to each slice of the score distribution, and the count shows a market pulling apart.
Why the split is widening
Several forces are reinforcing each other.
Asset appreciation locked in at the top. Homeowners who secured sub-4% mortgage rates during 2020 and 2021 still hold that advantage. Stock and retirement portfolios, despite periodic volatility, have grown substantially since the pandemic lows. According to the Federal Reserve’s Distributional Financial Accounts, the wealthiest 20% of households captured a disproportionate share of the approximately $40 trillion increase in aggregate household net worth recorded between early 2020 and late 2024. Higher wealth translates into lower utilization ratios, fewer missed payments, and credit scores that drift upward almost automatically.
Debt pressure is compounding at the bottom. The Federal Reserve Bank of New York’s Household Debt and Credit Report showed total credit card balances surpassing $1.2 trillion by late 2025, with delinquency transition rates for borrowers under 30 reaching levels not seen since 2010. When balances grow faster than incomes and payments slip, scores erode, pushing more consumers into subprime territory and making them targets for the high-cost products that dominate that tier.
Interest rates are squeezing the middle hardest. The Federal Reserve’s benchmark rate, while down from its 2023 peak, remains elevated by post-2008 standards. The average credit card APR sat above 20% through early 2026, according to the Fed’s G.19 consumer credit release. Superprime borrowers largely avoid that cost because they pay in full each month. Subprime borrowers are already priced for risk. But near-prime borrowers carrying revolving balances absorb the full weight of a 21% or 22% rate, making it harder to pay down debt and easier to slide backward on the score scale.
Real wage growth has been uneven. Bureau of Labor Statistics data show that while nominal wages have risen, inflation-adjusted gains have been modest for middle-income workers, particularly in sectors like retail, food service, and administrative support. That sluggish real income growth limits the ability of near-prime households to build the financial cushion that separates a 660 score from a 720.
What it feels like on each side of the K
For a borrower with a 790 credit score, the current market is a buyer’s paradise. Card issuers are locked in a rewards arms race, offering sign-up bonuses worth $500 or more, lounge access, travel credits, and cash-back rates that effectively subsidize everyday spending. Credit limits are generous, and balance transfer offers at 0% for 15 to 21 months remain widely available. These consumers are not just surviving the rate environment. They are profiting from it.
For someone with a 560 score, the inbox fills up too, but with very different offers. Secured cards requiring a $200 or $300 deposit. Unsecured subprime cards with annual fees of $75 to $125 and APRs north of 28%. Credit limits of $300 or $500 that leave almost no room for a stumble before utilization spikes and the score drops further. The products exist, and their proliferation shows up in the CFPB origination data, but they are designed to extract revenue from risk rather than reward loyalty.
The near-prime borrower with a 660 score occupies the most frustrating position. Too risky for premium products, not distressed enough to be courted by subprime specialists, this group faces a narrow set of plain-vanilla cards with middling limits and few perks. Lender marketing budgets flow to the extremes because that is where the return on investment is clearest: low losses at the top, high fees at the bottom.
The gaps the data does not fill
The CFPB origination files are strong evidence for credit cards, but they do not settle every question.
Geography and demographics are missing. The public data does not break originations down by metro area, race, or income bracket. Researchers cannot confirm, using these files alone, whether superprime growth clusters in high-cost coastal cities while subprime expansion concentrates in rural or lower-income regions. That hypothesis aligns with broader economic geography, but it remains unproven in this specific dataset.
Cross-product comparisons are limited. News outlets and credit bureau reports have described similar polarization in auto lending and mortgage originations, but those claims typically rely on proprietary data from Equifax, Experian, or TransUnion rather than regulator-curated public files with transparent methodology. The patterns may well be consistent, but the evidentiary standard is not identical.
There is a lag. The CFPB’s most recent credit card origination data reflects conditions through roughly the fourth quarter of 2025. Any shifts in lender strategy during the first half of 2026, whether driven by policy changes, economic slowdown, or competitive pressure, would not yet appear in the numbers.
Migration vs. composition is unclear. The origination data shows which tiers are receiving more new accounts, but it does not directly reveal whether individual borrowers are moving between tiers or whether the shifts reflect new entrants to the credit market (younger consumers, recent immigrants) arriving disproportionately at the extremes. Credit bureau migration studies could answer that question, but they are not part of the CFPB’s public release.
What the shrinking middle means for borrowers right now
The verified takeaway is narrow but significant: in the credit card market, new lending has become measurably more polarized by credit score. That is not a talking point or a think-tank projection. It is a structural shift visible in federal data spanning multiple years.
For households stuck in the middle, the practical risk is stagnation. Without targeted products or clear pathways to move up the score ladder, near-prime borrowers can spend years in a credit no-man’s-land, paying higher rates than superprime consumers but receiving none of the fee-heavy “second chance” marketing aimed at subprime segments. Financial advisors often recommend these borrowers focus on utilization (keeping balances below 30% of limits), set up autopay to avoid late marks, and periodically request credit limit increases. Those steps can nudge a score upward over 12 to 18 months, but they require disposable income that the rate environment is actively consuming.
For regulators, the data raises a question the CFPB has not yet answered publicly: does the growth of subprime originations represent expanded access to credit, or does it signal a new cycle of high-cost lending that will generate defaults and consumer complaints down the road? The bureau has the data to investigate. Whether it chooses to do so, particularly amid ongoing political debate over the agency’s scope and leadership, remains an open question heading into the second half of 2026.
In the meantime, the two arms of the K keep spreading. The rewards keep getting richer at the top. The fees keep getting steeper at the bottom. And the borrowers in between keep waiting for an offer that never quite arrives.
