When a company that rose through "financial arbitrage efficiency" sees its profit model collapse under the dual pressures of regulation and economic cycles, all narratives about an "AI transformation" essentially answer the same question: can it redefine its own reason for existence? The most brutal aspect of capital markets is their indifference to sentiment, focusing solely on the sustainability of cash flows. When the old profit engine is severed, and a new story fails to provide genuine profit support, former valuation peaks can instantly become cliffs for a frantic exit.
Yiren Digital's current predicament is not an isolated case but rather a rite of passage the entire loan facilitation industry must face in an era of heightened regulation. The 45% single-day crash reflects not mere market emotion but a verdict on a business model receiving a "stay of execution."
The company's recent earnings report is not simply a case of missing performance expectations; it is a concentrated manifestation of a failing business model. Quarterly losses per share were -$1.44, a sharp reversal from a profit of $0.52 per share a year earlier, while revenue fell 31% year-over-year. Viewed in isolation, these figures might be dismissed as cyclical volatility. However, against the backdrop of new policies, they indicate a structural collapse. The core issue is not about earning more or less, but whether earnings are possible at all.
Over the past decade, Yiren Digital (formerly known as Yirendai) has essentially engaged in one activity: facilitating high-interest credit packaged with technology. During the P2P era, it profited from information asymmetry, surviving on the interest rate spread between fund providers and asset seekers. In the loan facilitation era, it capitalized on regulatory gray areas, hiding comprehensive financing costs within complex fee structures involving service charges, guarantee fees, and membership fees. This model thrived in a lenient regulatory cycle but became a fatal weakness during a compliance crackdown.
The implementation of the new "Regulation No. 9" effectively blocked both of these profit avenues. The regulation explicitly caps the maximum interest rate at 24% (on a comprehensive cost basis), mandates that all hidden fees be made transparent, prohibits platforms from charging borrowers directly, and imposes a "whitelist system" for bank partnerships. This signifies the institutional elimination of Yiren Digital's core profit source—those shadow charges once disguised as "technology service fees." Previously, a loan might have had a nominal rate of 18%, but with various fees added, the borrower's actual cost could exceed 30%, with the difference constituting the platform's profit margin. Now, this margin is forcibly compressed to within 24% and must be transparent. In other words, this is not a growth problem but a fundamental shift: "The money you earned in the past is no longer permissible."
The market's 45% plunge reflects precisely this realization—it is not a short-term negative but a long-term reconstruction of the valuation framework. Investors recognize that the old high-margin model is unsustainable, and a new one remains unproven. When arbitrage opportunities are leveled by regulation, fintech companies revert to being ordinary credit intermediaries, whose valuation logic is far below that of high-growth tech stocks.
Against this backdrop, Yiren Digital has aggressively reinforced a new label: an AI company. It launched the "Zhiyu large model" and the "Mofang AI Agent Platform," emphasizing full-process intelligentization from customer acquisition to risk control and collections. This sounds like a perfect "finance + AI" narrative. Management attempts to signal to the market: "We are no longer just a lender; we are AI-savvy."
However, the investment community is increasingly清醒: for financial institutions, AI's primary role is typically "cost reduction," not "revenue reinvention." A simple comparison illustrates this: For companies like OpenAI or Nvidia, AI *is* the product itself, with customers paying directly for computing power or model access. For banks, AI is an "efficiency tool" to reduce manual review costs. For loan facilitation platforms, AI merely "optimizes lending," aiming to improve approval rates or reduce bad debts. Yiren Digital's problem lies in attempting to present the third narrative as the first.
Real-world data does not support this vision. Research and development expenses surged significantly (+176% in 2024), while net profit declined (-26% for the first three quarters of 2025), and the revenue structure showed no fundamental change (credit-related activities still account for over 90%). This indicates that AI has not created new revenue streams but is merely optimizing an old business being compressed by regulation. High R&D investments have failed to translate into new growth drivers, instead exacerbating current profit pressures.
More critically, the barriers to entry for AI in finance are not high and are rapidly dissipating. Banks can develop their own risk control models, with major state-owned and joint-stock banks establishing fintech subsidiaries and building their own large models. Core model capabilities can be procured externally, as general foundational models grow more powerful and fine-tuning for specific sectors becomes easier. Data advantage does not belong to loan facilitation platforms; core credit data resides with credit reporting systems and banks, while the scenario data platforms possess is seeing reduced utility under strict privacy protection laws. As AI becomes "infrastructure," Yiren Digital's purported technological edge can be easily eroded.
This creates a sharp paradox: the more it emphasizes AI, the more it exposes its lack of a new business model. Companies with genuine AI moats would see changes reflected in their revenue structure first, not just in technical parameters on presentation slides. The market's concern is not the absence of AI, but that its AI cannot generate independent revenue and remains dependent on a shrinking credit business.
If it continues down the path of "loan facilitation + AI," Yiren Digital's ceiling is clear: shrinking scale, compressed profits, and marginalization in valuations. Squeezed by the 24% interest rate cap and the trend of banks handling more business in-house, the market space for loan facilitation platforms is being physically constrained.
However, from another perspective, the company is not entirely without opportunity. The key lies in a pivot: from "profiting from interest spreads" to "profiting from services." In other words, stop being a "quasi-financial institution" and become a "fintech service provider." This path has been validated globally: the US's Upstart is essentially an AI risk control service provider, bearing no credit risk itself and only offering a decision engine; India's Lendingkart is gradually shifting towards technology output, providing digital solutions for small and medium-sized banks; China's Ant Group is also de-emphasizing proprietary finance in favor of technology services, aiming to survive compliance by "exporting technology" rather than "facilitating funds."
If Yiren Digital is serious about transformation, the logic should follow a clear three-step process: 1. Completely abandon reliance on high-risk loan facilitation. Accept scale reduction and short-term profit deterioration as the cost of "deleveraging." It must cease the extensive model of covering high risk with high interest, even if it means revenue halving. 2. Modularize and commercialize its capabilities. Package its risk control, customer acquisition, anti-fraud, and collection systems into standardized APIs or SaaS offerings. Shift from "I will help you lend" to "I will sell you the tools to lend yourself." This would change the revenue recognition model from "gain on sale" or "service fee sharing" to stable "technology subscription fees." 3. Partner with banks as a "technology supplier," rather than acting as a "parasite on banks." Transition from being a "traffic distributor" to becoming a bank's "technology vendor." Previously, the platform controlled users while banks provided funds, giving the platform a strong position. In the future, banks will control funds and core data, with the platform providing technical supplements, placing banks in the dominant position. This redefined relationship implies reduced bargaining power but also lower compliance risk.
This path is difficult but offers a logically coherent future. In the AI era, true scarcity lies not in the models themselves, but in accumulated financial data, risk control experience, and scenario-specific capabilities. These are precisely the assets Yiren Digital has accumulated over the past decade. If it can encapsulate this experience into code rather than loans, it might find a new space for survival. However, this requires significant strategic resolve, as the scalability of technology service revenue is much lower than that of credit business, making financial statements appear very unattractive in the short term.
In conclusion, as its market capitalization falls to single digits, profits turn negative, and regulatory tightening continues, the market has voted with its feet: Yiren Digital is no longer viewed as a "growth company." But not all undervalued companies have a future, and not all suppressed industries lack opportunity. The crucial factor is whether the company is willing to admit one thing: its past path to success has completely failed.
If it persists with the "lending + AI" narrative, it risks becoming just another financial platform phased out by the times, gradually withering in the cracks of regulation. However, if it genuinely completes the role shift from capital intermediary to technology service provider, it might survive—albeit less profitably, but more sustainably. For investors, this company is no longer a "high-growth target" but rather a case study: an example of how a company is forced to rewrite its business model at the intersection of intense regulation and the AI wave. The lessons here extend far beyond stock price fluctuations—in an era where compliance is the baseline and technology is a tool, any attempt to repackage old arbitrage with new concepts will ultimately face a reckoning of its true value. Genuine rebirth involves not just telling a new story but having the courage for radical reform, rebuilding authentic technological barriers from the ground up.
Comments