Request an Invite to the Inveniam AI alpha Release →

February 29, 2024

Flawed Valuations Threaten $1.7 Trillion Private Credit Boom

Flawed Valuations Threaten $1.7 Trillion Private Credit Boom

This article is evidence of what happens when $7 trillion leaves the banking system and is redeployed via private credit and capital markets as high yield debt. The next version of this is DeFi, or decentralized finance. Everything will be traded digitally. There will a series of ledgers that everyone (as appropriate, public v private) can reference to see ownership, these ledger entries (tokens) need to reference real world asset data. The solution to this is not the worlds largest centralized data base to reference.

24 months ago all the debt was in the risk books of banks, with the OCC overseeing it. Each bank held it differently, and reserved against it differently. The OCC/FDIC/Fed monitored. This works, except when they fine banks for over reserving against the loans (for those who remember this was Suntrust who was fined for managing earnings by over reserving against loans ahead of the GFC).

This decentralized new reality, needs new tools to manage and report risk to lenders, shareholders, borrowers...the market broadly. There is an explosion of data and finding the signal in the noise is difficult and why compute has not pushed into private markets the way they have in Public Markets (high frequency trading, algrithmic trading, etc).

In Private Markets the lack of quality data is an obstacle. AI will solve this, but not on its own. There needs to be a mechanism by which AI is constrained to point at only verifiable data and is forced to deliver attribution to avoid hallucination. This is what Inveniam does working with folks at Amazon Web Services (AWS), Ibm Watson Center, and G42 we are making it possible to deliver the promise of AI at enterprise scale, because we force attribution, using Zero Knowledge proofs.

This mechanism will allow real time processing of edge data, where asset owners retain their data and push it to monthly third party marks, and these valuation experts are out of data collection and data entry and can focus on high value correlation and valuation work. This unlocks the ability to have AVM's valuing assets daily, which act as the collateral for this private credit.

Distributed ledger will help us understand ownership, distributed compute will help us understand the assets and collateral, distributed data is how asset owners will understand and control their own data, and monetize it for their own benefit and then monetize via DeFi.

Large asset owners will not turn it over to AI companies with out compensation, or they need a manner in which there is a zero-knowledge transfer. AI needs to understand the query and pull it via API from an encrypted vector data base on the client side. This is the infrastructure for AI, which is how we will process data at the edge, and then deliver compute to the private markets.


Apple PodcastsApple Podcasts