Prediction: The Best-Performing Big Bank Stock of the Next 2 Years Is Not JPMorgan Chase

SoFi added 1.1 million customers (+35% year over year) and 2.2 million products in Q2, while adjusted net revenue rose 40% and EPS increased to $0.12 from $0.08. Nu Holdings grew Q2 revenue 39% year over year and net income 49% to more than $1 billion, with average revenue per active customer rising to $17 from $13. The article argues that both digital banks offer materially faster growth potential than JPMorgan, supported by cross-selling, product innovation, AI capabilities and geographic expansion.

Analysis

The relevant differentiation is not user growth but monetization quality and funding economics. SOFI's product-per-member inflection can raise lifetime value and lower acquisition payback, but its valuation remains highly exposed to credit normalization and the cost of maintaining deposits as rates fall. The key 1-3 month catalyst is whether incremental products translate into sustained contribution-margin expansion rather than higher promotional spend; absent that proof, the market should treat new feature launches and the Composer acquisition as narrative, not earnings drivers.

NU has the cleaner operating leverage setup: increasing revenue per active customer while using a low-cost, data-rich deposit base can support earnings growth ahead of customer growth. The less appreciated risk is that moving upmarket and expanding Mexican credit increases exposure to more competitive pricing and potentially weaker loss visibility; a faster mix shift into unsecured lending could make the current profitability multiple vulnerable. Mexico is a meaningful 6-18 month optionality source, but bank-charter economics require evidence of deposit gathering and risk-adjusted loan yields, not merely account additions.

JPM is the natural quality hedge rather than a direct growth competitor. If consumer delinquencies reaccelerate or fintech funding costs remain sticky, JPM's diversified deposit franchise and underwriting scale should attract relative flows; conversely, a benign credit/rate-cut cycle favors duration-sensitive SOFI and higher-beta NU. Consensus likely underweights that AI underwriting is only economically valuable if it reduces loss rates or servicing expense without creating model-risk, fair-lending, or regulatory costs—none of which is independently established here.

Key Decisions for Investors

• Prefer a 6-12 month long NU / short SOFI pair, sized beta-neutral: NU has better demonstrated operating leverage and geographic runway, while SOFI carries greater U.S. consumer-credit and funding-cost sensitivity. Target 15-25% relative return; exit if NU's revenue per active customer stalls for two quarters or Mexico credit losses rise materially.
• Maintain JPM as a 3-6 month defensive hedge against a consumer-credit scare: long JPM versus a basket of SOFI and fintech ETF ARKF. The thesis is falsified by sustained deposit outflows or a material deterioration in JPM's card/net charge-off guidance.
• Do not add directional SOFI solely on product-growth claims. Set an alert for next earnings: consider a long only if contribution margin and adjusted EBITDA guidance rise alongside stable personal-loan delinquency trends; otherwise, failed monetization conversion is a credible 10-15% downside catalyst.
• For NU, accumulate on broad EM/Brazil risk-off weakness rather than chase momentum. Monitor Brazilian policy rates, FX, and quarterly cost of risk; a sharp BRL selloff or credit-cost increase would overwhelm company-specific ARPAC gains in the near term.