Tokenized private credit analysis: why falling TVL can hide a growing loan book.

What does TVL measure in tokenized private credit?

A dashboard’s total value locked, or TVL, is a starting point for research. Its meaning depends on how the product and data source define it. In a lending product, understanding the relationship between supplied funds and outstanding borrowing changes how you read the headline.

For Figure Markets Democratized Prime, the DefiLlama measure retained in this research covers unmatched lending supply across four pools: Home Equity, Automobile, SMB and Crypto-Backed Loan. This is supply still waiting to be borrowed.^[1]

That question turns a headline percentage into a useful investigation. It is the starting point for this worked example of tokenized private credit research.

Figure’s TVL fell while outstanding loans grew.

Between DefiLlama’s 18 and 25 September 2026 daily labels, unmatched supply fell $6,447,960. Borrowed value increased $4,225,281. Adding the two measures shows that total lender supply fell a much smaller $2,222,679, or 0.2769%.^[1]

USD values use DefiLlama’s pricing basis. Total lender supply and changes are Alkaron calculations. The source gives daily date labels without specifying exact measurement instants.

The result is more informative than “TVL fell.” Idle supply declined, outstanding borrowing rose, and the overall lender book contracted modestly. The full private credit report follows the four pools individually to show the activity behind those combined balances.

Why did Figure’s Automobile lending rate increase?

Lending rates also need an explanation. A rate can move because the share of supplied funds being borrowed changes, because the contract’s rate settings change, or because both move together.

In Figure’s Automobile pool, the annualized lending rate increased from 8.0117% to 9.8796% over the pinned weekly endpoints. Alkaron’s calculation separates the effects of the rate parameters and utilization.^[2]

• Contract parameter effect+2.3091 percentage points
• Utilization effect−0.4412 percentage points
• Net lending-rate change+1.8679 percentage points

The contract recorded a parameter update on 23 September. Its contribution exceeded the total rate increase, while lower utilization offset part of it. This explains the recorded rate mechanics; the evidence does not establish the reason for the parameter decision.

For anyone assessing private credit lending yields, that distinction matters. A higher quoted rate can reflect a change in the rate model even when utilization falls.

Apollo ACRED: higher NAV, unchanged endpoint supply.

Apollo’s ACRED offers a different analytical problem. To understand its token valuation, the report combines supply on all seven issuing networks with the retained RedStone NAV inputs.

At the historical midnight targets for 18 and 25 September, calculated token value rose from $95,484,543.68 to $95,538,803.54. Opening and closing token supplies were equal on each network, so the entire $54,259.86 endpoint increase came from higher NAV per token.^[3]

This is a token supply × NAV calculation. Equal endpoint supply establishes no net supply change; it does not describe every transaction between those endpoints. NAV excludes distributions, so the NAV change is distinct from an investor’s total return.

The Figure and ACRED examples answer different questions: how a lending book changed, and how token quantity and NAV affected valuation. Both belong in a careful tokenized private credit analysis, with their measurement bases stated clearly.

Sources and reporting dates

This article draws from Alkaron Private Credit Research: Figure & Apollo ACRED, 18–25 September 2026, corrected edition 1. It is a public selection of findings from the historical report.

1. 1. Figure balances and the TVL definitionDefiLlama: Figure Markets Democratized Prime ↗
   Retained daily USD series for the 18 and 25 September labels. The report derives combined supply and changes from those points. The live dashboard may display later observations.
2. 2. Figure lending-rate mechanicsProvenance blockchain API ↗ and Figure’s pool disclosures ↗
   Pinned contract states and recorded updates underpin the rate decomposition. The disclosures provide product context.
3. 3. ACRED supply and NAVThe report’s retained seven-network supply evidence and RedStone ACRED_FUNDAMENTAL NAV feed. Solana’s historical endpoint supplies were reconstructed from a later reading and intervening transactions returned by the queried node. The full report documents the reconstruction and calculation.