The global credit market is undergoing a fundamental shift from the traditional vertically integrated banking model to a horizontally modular architecture. Who can become the choke point that controls the new global economic system: artificial intelligence or blockchain? This excerpt and summary is taken from ARTEMIS BIG FUNDAMENTALS.
The loan system shifts from vertical integration to a horizontal modular architecture
The global credit market, with a size of 348 trillion dollars, is undergoing a structural reorganization. In the past, traditional banks adopted a one-stop, vertically integrated model, completing everything—from deposit gathering to lending—within a single system. However, modern finance is evolving into a five-layer horizontal architecture: Origination, Distribution, Underwriting & Risk, Capital, and Infrastructure. In this new architecture, credit is no longer confined to physical banks; instead, it is embedded via APIs into e-commerce platforms or software. Private credit is expected to reach 5 trillion dollars by 2029. The “Capital” layer, which simply provides funding, is increasingly becoming commoditized, and market value is rapidly moving toward the layer that holds core technology.
Artificial intelligence underwriting becomes the primary choke point for risk control
“Underwriting & Risk” is seen as the primary key choke point. Traditional credit scoring models (such as FICO) rely on only about 20 to 30 variables and are often viewed as opaque “black boxes” in terms of information. By contrast, AI-driven underwriting models can evaluate more than 1,600 variables. They not only increase approval rates by 44%, but also reduce default rates by 53%. Take Upstart as an example: 92% of its loan decisions are fully automated, cutting processing time from days under traditional committee reviews to minutes. Companies that control these proprietary scoring engines effectively control the entire credit market’s economic returns, and can use that to charge service fees to other institutions that lack technical capabilities.
Blockchain infrastructure redefines settlement and costs
“Infrastructure” is the main choke point in blockchain. Blockchain technology is replacing old settlement systems that have been used for decades. The origination cost of a traditional mortgage is about 11,000 dollars per loan, but through blockchain rails (such as Provenance Blockchain), the cost can be lowered to below 1,000 dollars, and settlement time can be reduced from an average of 28 hours to just seconds. This technological advantage not only eliminates a large number of intermediaries, but also establishes very high barriers to entry. Companies that have blockchain settlement layers or combine them with banking licenses (such as SoFi or Figure) can act as platforms to support the entire industry, turning compliance and technology into a “regulatory moat” and exporting infrastructure services to the outside.
In the future credit market, the “winners” with competitive advantages will be those platforms that control the AI risk layer or the blockchain settlement layer, or “compounding players” that can vertically integrate multiple key layers. Conversely, traditional lenders that rely solely on balance-sheet scale and lack proprietary technology will face the problem of marginal profits falling to zero.
This article, which first appeared on ABMedia, is titled: “The traditional credit market has flipped—who is the choke point in the new world, AI or blockchain?”