Midbench: Tokenized Lending Pools with Disintermediated Value Capture

White Paper v2.2 | February 2026

Executive Summary

Midbench deploys on-chain lending pools for consumer credit, starting with U.S. auto loan refinancing. By replacing traditional financial intermediaries with smart contracts, the platform eliminates ~1% in annual fees that typically sit between borrowers and investors. The target segment shifts with interest rates: prime borrowers in low-rate environments (where 1% savings = 25% interest reduction), subprime in high-rate environments.

Target Market: U.S. auto borrowers seeking to refinance existing loans. The optimal segment shifts with the rate environment:

• Low-rate environment (Treasury ≤ 3%): Prime borrowers (FICO 720+) where our ~1% cost advantage represents a 20-25% interest savings
• High-rate environment (Treasury > 4%): Subprime borrowers (FICO 580-659) where rate sensitivity is lower and specialty finance overcharges

Credit Tiers: The platform uses a 5-tier credit segmentation: Super Prime (780+), Prime (720-779), Near Prime (660-719), Subprime (580-659), Deep Subprime (<580). Minimum eligibility is FICO ≥ 620. See Credit Model Documentation for details.

The core arrangement: Midbench funds all origination costs (marketing, underwriting, compliance) from its own capital, charging zero fees to Pool investors. In exchange, Midbench seeds each Pool as the founding investor, capturing value when loan returns exceed investor required returns.

The instrument: Each Pool is a Perpetual ABS—a continuously managed, tokenized lending vehicle with no stated maturity. Unlike traditional ABS trusts that amortize and wind down, a Midbench Pool can originate new loans, raise additional capital, buy back tokens, and distribute dividends indefinitely—all administered by smart contracts at near-zero marginal cost. This is a new instrument class enabled by blockchain infrastructure (see §3.6).

Conservative by design: The ~1% margin estimate is anchored to observable intermediary costs—not optimistic projections about market capture. Traditional securitizers (banks, captive finance companies) currently operate profitably while paying these costs, demonstrating that additional value exists in the lending chain. By claiming only the disintermediation savings, Midbench's projections represent a defensible floor, not a ceiling.

Three stakeholder groups:

• Borrowers: Refinance auto loans through the platform at competitive rates
• Pool Token Investors: Earn yield by investing in Pool tokens
• Midbench Equity Investors: Own shares in Midbench Inc., which profits from its founding positions

Borrower Highlights

Traditional auto loan payments traverse multiple intermediaries—payment processors, loan servicers, custodians—each extracting fees that get embedded in borrower rates. Midbench eliminates this chain. Each loan is deployed on Base with a dedicated payment address; borrowers pay in USDC directly to the loan contract.

Payments are flexible rather than fixed-schedule: borrowers pay any amount, at any time, provided cumulative payments remain above the minimum threshold. Interest accrues per-minute, so each payment immediately reduces the outstanding balance.

All loan state is on-chain and publicly verifiable in real time—balances, payment history, and terms are immutable and require no servicer reporting or reconciliation. A sample loan report illustrates this borrower-facing interface.

1. The Problem

When you make a car payment today, your money passes through a surprising number of hands—payment processors, loan servicers, banks—each one takes a cut. These costs get baked into your interest rate, and borrowers end up paying more than they should.

Traditional Flow:
Borrower → Payment Processor → Servicer → Bank → Lender
           ↓                  ↓          ↓
         (fees)            (fees)     (fees)

Each arrow represents fees extracted from borrower payments:

These costs compound, creating a persistent ~1% annual spread between what borrowers pay and what investors receive. On a $35,000 auto loan at 7$350/year that flows to intermediaries rather than reducing borrower costs or increasing investor returns.

2. The Midbench Solution

Midbench removes these intermediaries. Each loan is deployed on Base with a dedicated payment address. Borrowers pay in USDC, and payments flow directly to the loan contract.

Midbench Flow:
Borrower → Loan Contract → Pool → Investors
              (direct)

Every cent goes to principal reduction or investor returns.

Side-by-side comparison:

The ~1% annual savings comes from eliminating:

• Loan servicing — Smart contracts track balances automatically
• Payment processing — USDC transfers settle instantly with no processing fees
• Custodian/trustee fees — Collateral documentation lives on-chain
• Traditional compliance overhead — KYC verified once at origination

This ~1% becomes the Margin in loan pricing, creating the spread between Net Yield to Pool ($R$) and investor required returns ($r$). Instead of flowing to intermediaries, it flows to the Pool—captured by the founding investor (Midbench).

3. How It Works

3.1 Architecture

┌─────────────────────────────────────────────┐
│              MIDBENCH INC.                  │
│           (Delaware C-Corp)                 │
│                                             │
│  • Raises equity capital                    │
│  • Funds loan origination                   │
│  • Seeds each Pool as founding investor     │
└─────────────────────────────────────────────┘
                     │
                     ▼
┌─────────────────────────────────────────────┐
│           FACTORY CONTRACT                  │
│             (Base L2)                       │
│                                             │
│  • Deploys new Pool contracts               │
│  • Owned by Midbench multisig               │
└─────────────────────────────────────────────┘
                     │
        ┌────────────┼────────────┐
        ▼            ▼            ▼
   ┌─────────┐  ┌─────────┐  ┌─────────┐
   │ POOL A  │  │ POOL B  │  │ POOL C  │
   │         │  │         │  │         │
   │ ERC20   │  │ ERC20   │  │ ERC20   │
   │ Tokens  │  │ Tokens  │  │ Tokens  │
   └────┬────┘  └────┬────┘  └────┬────┘
        │            │            │
        ▼            ▼            ▼
   ┌─────────────────────────────────────┐
   │           LOAN CONTRACTS            │
   │                                     │
   │  Borrowers pay directly → Pool      │
   │  distributes to token holders       │
   └─────────────────────────────────────┘

3.2 Capital Flow

1. Midbench seeds Pool → Receives founding tokens
2. External investors buy tokens → Via sealed-bid auctions
3. Pool originates loans → USDC sent to borrowers
4. Borrowers repay → Payments flow to token holders pro-rata

3.3 The Fee-Free Model

Midbench charges zero fees to Pool investors—no management fee, no performance fee, no origination fee.

Instead, Midbench profits by being the founding investor. When loan returns ($R$) exceed investor required returns ($r$), the founding position captures value.

Incentive alignment: Since Midbench's returns depend on loan performance, it has direct incentive to underwrite conservatively and minimize defaults.

3.4 The Borrower Experience

Flexible Payments

Unlike traditional loans, there's no auto-debit pulling from the borrower's bank account on a fixed schedule. Borrowers pay when they want, as much as they want. As long as total payments stay above the cumulative minimum, the loan remains in good standing.

Interest accrues by the minute, so every payment immediately reduces what's owed—no waiting for end-of-month processing.

USDC-Native Payments

Borrowers pay in USDC on Base. When paying from a Coinbase account, gas is sponsor-covered, resulting in zero transaction fees for the borrower.

On-Chain Transparency

All loan state—balances, payment history, terms—is stored on-chain and publicly verifiable in real time. The record is immutable: neither Midbench nor any other party can alter payment history or modify balances post-deployment. This eliminates the reconciliation and reporting overhead inherent in traditional servicing.

See it in action: A sample loan report demonstrates the borrower experience—real-time balance, payment history, payment address, and original terms, all derived directly from on-chain state.

3.5 Midbench's Role

Before a loan goes on-chain, Midbench handles all off-chain processes:

1. Review application — Verify information, assess creditworthiness, determine rate
2. Fund refinance — Pay off existing loan and issue the new one
3. Deploy loan contract — Create dedicated payment address on Base

Once deployed, the loan record becomes immutable—Midbench cannot change terms, alter payment history, or modify balances. The loan exists independently, and every payment is recorded automatically.

Midbench remains available for customer support, but the accounting is handled by the loan contract itself.

3.6 The Perpetual ABS — A New Instrument

Midbench Pool tokens represent a structural innovation in asset-backed finance: a Perpetual ABS (a continuously managed, tokenized lending vehicle with no stated maturity).

Traditional asset-backed securities are born dying. An issuer pools loans into a trust, sells tranches, and the trust amortizes as borrowers repay—eventually winding down to zero. Every new lending cycle requires a new securitization: new legal documents, new trust formation, new rating agency engagement, new investor roadshows. The vehicle is disposable by design.

A Midbench Pool is the opposite. It is a living, perpetual vehicle that can:

• Originate new loans continuously as capital becomes available
• Raise new capital through iterative sealed-bid auctions into the same token
• Buy back tokens when market price falls below intrinsic value
• Distribute dividends to token holders from loan cash flows
• Reinvest principal as loans amortize, maintaining deployed capital

All of this happens programmatically through smart contracts, with no trust amendments, no bondholder consent votes, no new offering documents, and no incremental administrative cost.

Why This Couldn't Exist Before

Traditional finance attempted perpetual-like credit vehicles, but the cost structure made them impractical or unstable:

The ~1% annual intermediary cost isn't just a savings—it is the structural barrier that prevented perpetual ABS from existing. When ongoing administration costs ~1% of assets, a perpetual vehicle must either (a) charge fees that erode investor returns, or (b) be large enough to amortize fixed costs, creating barriers to entry. Smart contracts reduce marginal administration costs to near zero, making perpetual operation economically viable at any scale.

Closest Analogues (and Why They Differ)

Several existing instruments share individual characteristics with the Perpetual ABS, but none combine them all:

The closest structural analogue is a REIT—a perpetual vehicle that owns assets, pays dividends, and can issue new shares. But REITs are equity instruments backed by real estate with management companies that charge fees. Midbench Pools are direct credit passthrough vehicles with zero-fee smart contract administration.

CLOs have a reinvestment period (typically 3-5 years) during which the manager can originate new loans from principal repayments—but the reinvestment window closes, and the vehicle amortizes to maturity. The Perpetual ABS has no such constraint.

ABCP conduits were perpetual in practice, continuously rolling short-term commercial paper to fund longer-term assets. But the maturity mismatch created systemic fragility—when the commercial paper market froze in 2007-2008, conduits collapsed. Midbench Pools have no maturity mismatch: token holders own a perpetual claim on the underlying asset pool, not short-term paper that must be rolled.

The Innovation

The Perpetual ABS synthesizes five capabilities that have never coexisted in a single credit instrument:

1. Perpetual life — No stated maturity, no wind-down, no need to recreate the vehicle
2. Dynamic origination — Continuous lending as capital is available, not batch issuance
3. Programmatic capital formation — Sealed-bid auctions raise capital without investment banks or offering documents
4. Direct cash flow passthrough — Borrower payments flow to token holders pro-rata, with no intermediary handling
5. Real-time, symmetric transparency — Every stakeholder—founding investor, subsequent investor, borrower—sees the same on-chain state simultaneously

This combination is made possible by blockchain infrastructure that collapses the cost of trust, administration, and coordination to near zero. The Perpetual ABS is not an incremental improvement on traditional securitization—it is a new instrument class enabled by a new infrastructure layer.

4. Economic Model

4.1 Key Variables

Core requirement: $R > r$ for value creation.

4.2 Loan Pricing Formula

Each loan is priced using:

$\text{Gross APR} = r + \text{Margin} + \text{ECL}$

Where:

• $r$ = Market-observable investor expected return. Computed as 5-Year Treasury rate + Auto ABS spread (~75 bps). This is dynamic:
  • Low-rate environment (2021-2022): r ≈ 1.75-2.5% → Best APR ≈ 4%
  • High-rate environment (2024-2025): r ≈ 4.5-5% → Best APR ≈ 7%
  • Future: Will be discovered via Pool token trading price
• Margin = Disintermediation savings (~1%), at least initially. This is the cost traditional lenders pay for intermediary services that Midbench eliminates.
• ECL = Expected Credit Loss, computed as $\text{PD} \times \text{LGD} \times \text{EAD}$:
  • PD (Probability of Default) — XGBoost model trained on 17M loans, test AUC 0.775
  • LGD (Loss Given Default) — Segment-level empirical averages (~83%), derived from 35K+ historical chargeoffs
  • EAD (Exposure at Default) — Empirical lookup factors by term and credit tier, derived from 344K+ historical defaults

Why Margin = ~1%: Traditional lenders pay ~1% annually for intermediary services (servicing, custody, trustees). Midbench eliminates these costs, so this ~1% can be added to loan pricing without making rates uncompetitive. The Margin creates the $R - r$ spread that funds the platform.

Why ~1% is a Conservative Floor:

Traditional auto loan securitization is a profitable business. Investment banks and captive finance companies (Ally, Toyota Financial, Ford Credit) routinely:

1. Originate or purchase loans at par or slight premium
2. Pool loans into asset-backed securities
3. Sell tranches to investors at prices that generate profit

They accomplish this while paying the ~1% in annual intermediary costs (servicers, trustees, custodians). Their profitability comes from multiple sources:

The implication for Midbench: If traditional players generate returns after paying intermediaries, then eliminating those intermediaries provides Midbench with at minimum that ~1%. But Midbench also occupies positions in the value chain that securitizers currently monetize—origination, capital deployment, and residual ownership via founding tokens.

What we claim vs. what may exist:

The 1% figure represents the identifiable, auditable cost layer being removed—not the total value opportunity. By anchoring projections solely to disintermediation savings and ignoring potential capture of securitizer spreads, the economic model is deliberately conservative. The actual value capture could be higher, but we anchor projections to a defensible floor.

Net Yield to Pool: $R = r + \text{Margin}$

With dynamic r based on Treasury rates:

• High-rate scenario (Treasury 4.25%): $r = 5\%$, $R = 6\%$
• Low-rate scenario (Treasury 1.5%): $r = 2.25\%$, $R = 3.25\%$

4.2.1 Why Constant R Across Credit Tiers?

A natural question: shouldn't riskier loans require higher R to compensate for greater return volatility?

In finance theory: Higher variance should command a risk premium, even if expected returns are equal. However, this applies at the individual loan level.

In a diversified pool: Idiosyncratic default risk diversifies away. With 10,000+ loans:

• Individual loan outcomes are binary (default or not) → high variance
• Pool-level outcomes converge to expected values → low variance
• The "risk" investors bear is pool-level, not loan-level

Systematic risk is priced into r: Risk that doesn't diversify (recessions affect all borrowers) is reflected in the Pool Token Investor Required Return, which is discovered via auction or market trading. If the pool's credit composition shifts riskier, investors demand higher r.

This approximation works when:

1. Pool is well-diversified (10,000+ loans)
2. Pool composition is controlled via underwriting criteria
3. Investors price the whole pool, not individual loans
4. Auction/trading mechanisms adjust r if pool risk profile changes

4.3 Value Creation

Each dollar invested creates value when $R > r$:

$\text{NPV per dollar} = \frac{R - r}{r}$

With $R = 6\%$ and $r = 5\%$, each dollar creates $0.20 in NPV.

4.4 Value Capture

Value capture is how Midbench profits without charging fees. By eliminating intermediaries, the platform creates the Margin (~1%) that would otherwise go to servicers, custodians, and trustees. As the founding investor, Midbench captures this value through its token position—not through fee extraction.

In an efficient market with known future returns:

• Founding investor (Midbench): Captures value from all future growth via founding position
• Subsequent investors: Earn exactly $r$ (their required return)

Later investors pay fair value and receive fair returns. Midbench is compensated for building the platform and funding origination.

4.5 Numerical Example

Subprime auto refinance market (based on ABS-EE data analysis):

How the numbers connect (subprime example):

Gross APR       = r + Margin + ECL
    14%         = 7% + 1% + 6%

Net Yield (R)   = r + Margin
     8%         = 7% + 1%

Value Created   = Margin = R - r = 1%

Why All Credit Tiers? Competitive analysis using historical Treasury rates shows Midbench pricing is competitive across all segments: near_prime (+44 bps vs market median), prime (+104 bps), super_prime (+181 bps), and subprime/deep_subprime (within market IQR). The ~100-200 bps spread above market represents intentional margin (Base Margin + Uncertainty Buffer for launch phase), not model error.

Model parameters: $R = 6\%$, $r = 5\%$, Midbench seeds $100K, 100 subsequent$1M raises

Key insight: External investors earn their required return. Midbench earns outsized returns for being first and funding origination.

5. Stakeholder Analysis

5.1 Pool Token Investors

What you get:

• Expected return of $r$ (risk-adjusted required return)
• Real-time on-chain transparency
• ERC20 tokens (transferable)
• Zero fees
• Aligned manager (Midbench holds same tokens)

Risk factors:

• Realized $R$ may differ from expected $R$
• Credit losses could exceed projections
• Smart contract risk

Comparison to traditional credit funds:

5.2 Midbench Equity Investors

Value capture: Midbench captures value created by disintermediation through its founding positions across multiple Pools.

Returns depend on:

• Maintaining $R > r$ spread
• Scaling loan origination volume
• Managing credit risk effectively

5.3 Borrowers

• Access to refinancing at competitive rates
• Direct payments to smart contracts
• No intermediary handling of funds

6. Risk Factors

Critical threshold: If $R \leq r$, no value is created. However, the ~1% margin estimate is anchored to observable intermediary costs, not optimistic projections. Traditional ABS markets validate that spreads above this floor exist—the question is how much Midbench captures, not whether value exists.

6.1 Why the Model is Conservative

The economic projections in this white paper are deliberately anchored to a defensible floor rather than optimistic estimates:

1. Only disintermediation savings are modeled — The ~1% margin comes from eliminating intermediary costs. No credit is taken for origination margins, excess spread, or other value that traditional securitizers capture.

2. Traditional players validate the opportunity — Banks and captive finance companies are profitable after paying intermediary costs. Their continued profitability demonstrates that additional value exists in the lending chain beyond what we claim.

3. Upside is unmodeled — If Midbench captures any portion of what securitizers currently extract (origination spreads, excess spread), returns exceed projections. The model assumes zero capture of these sources.

4. Conservative credit assumptions — PD model over-predicts defaults (providing safety margin), LGD uses segment averages (not optimistic recovery assumptions), and pricing includes Uncertainty Buffers for launch phase.

This approach means:

• Downside scenario: Model is approximately correct; Midbench earns the projected ~1% margin
• Base case: Midbench captures some securitizer value; returns exceed projections
• Upside scenario: Midbench captures significant securitizer value; founding positions appreciate substantially

By not claiming upside in projections, the model provides a realistic floor that investors can underwrite with confidence.

6.2 Market Validation: Completing What Figure Started

Figure Technologies pioneered on-chain lending and proved it works. Founded in 2018, Figure built the Provenance blockchain and has originated over $15 billion in loans with ownership, custody, and trading occurring on-chain. Their success established a crucial foundation:

• On-chain lending is market-accepted — Rated securitizations, institutional investors, regulatory comfort
• Traditional finance will engage — Banks and asset managers participate in blockchain-based transactions
• The economics are proven — Billions in originations demonstrate viability

Midbench completes the vision.

Figure moved loan records on-chain but left loan payments on legacy rails. Borrowers still pay via ACH bank debits, meaning Figure incurs all the costs of traditional payment infrastructure—processing fees, reconciliation, failed payment handling, and multi-day settlement times.

Midbench takes the final step: fully on-chain payments via USDC.

The result: ~1% additional margin.

Figure proved that on-chain lending works. Midbench captures the remaining ~1% by eliminating the last legacy dependency—payment infrastructure. This isn't incremental improvement; it's the completion of a fully disintermediated lending stack.

Figure's success at scale validates the foundation. Midbench builds on that foundation to capture the full disintermediation opportunity.

7. Model Validation & Backtesting

Validation run: January 24, 2026

7.1 Data Foundation

The credit model is trained on 16.7 million loans from SEC ABS-EE filings (2021-2025), representing the complete universe of U.S. auto loan securitizations:

7.2 PD Model Discrimination

The XGBoost probability-of-default model was validated on a held-out test set:

The model correctly rank-orders borrowers by risk (top decile captures 3.85× lift), though predicted PDs exceed actual defaults. This conservative bias is acceptable for launch—overprediction leads to higher APR quotes, providing a safety margin.

7.3 Backtest Results

Historical backtest applying the pricing engine to 500K sampled loans with known outcomes:

Note: Market APR uses median (robust to outliers), excludes rate outliers >30% (data errors) and rate-subvented loans (captive 0% APR promotions).

By credit tier (approved loans):

APR Comparison (Model vs Market Median):

7.4 Market Positioning

Analysis of model APRs vs. actual market rates (sample: 598,671 non-subvented loans):

Note: Rate-subvented loans (31% of ABS pool) are excluded from this analysis. Captive finance 0% APR promotions are manufacturer subsidies, not market-comparable rates.

Estimated market capture: At current pricing with historical Treasury rates, our model is competitive across all segments. Near_prime shows the tightest spreads (+44 bps), followed by prime (+104 bps) and super_prime (+181 bps).

Strategic implication: With dynamic rate-based pricing using historical Treasury rates, Midbench is competitive across all credit segments. The ~200 bps average spread above market represents intentional margin (Base Margin + Uncertainty Buffer) rather than model miscalibration.

7.5 Stress Testing

Profitability assessed under adverse economic scenarios:

Key findings:

• PD breakeven: ~3.5× base assumptions (portfolio remains profitable through 3× PD severe recession)
• LGD resilience: Can absorb LGD increase from 83% to 100% and remain profitable at base PD
• Capital adequacy: 99th percentile (severe recession) requires ~11% capital; 99.9th percentile ~18%

7.6 Model Limitations (Honest Assessment)

Launch phase design: Pricing is intentionally conservative. Uncertainty Buffer (100 bps), DeFi Channel Premium, and Segment Risk Margins stack to provide cushion against model error. Early production exists to generate calibration data, not to prove the model is correct.

8. Technical Implementation

8.1 Smart Contracts

Pool Contract: ERC20 token + sealed-bid auctions + loan management + distributions

Loan Contract: A deliberately "dumb" contract—minimal container that holds borrower payments and can be accessed only by Pool

8.2 Security

• UUPS upgradeable proxies with timelock
• Multisig ownership
• Reentrancy protection
• KYC whitelist and transfer restrictions
• Emergency pause capability

8.3 Deployment

Built on Base (Coinbase L2):

Zero-fee borrower payments: Coinbase sponsors gas for USDC transfers from Coinbase accounts. This means:

$\text{\$1.00 paid by borrower} = \text{\$1.00 received by pool token investors}$

Every cent borrowers send reaches investors—no transaction fees, no payment processing costs.

Negligible protocol gas costs: ~$0.15/loan/year, covered by Midbench. On a$35,000 loan saving ~$350/year from disintermediation, gas is over 2,000× smaller than the savings.

9. Conclusion

Midbench eliminates ~1% in annual intermediary costs from consumer lending. This creates value that flows to:

1. Pool Token Investors: Higher net yields, zero fees, real-time transparency
2. Midbench: Founding investor returns that fund platform development
3. Borrowers: Potential for better rates as platform scales

The model works when loan returns exceed investor required returns ($R > r$). Midbench's founding position creates alignment—both Midbench and Pool investors benefit from strong loan performance.

10. Iterative Capital Raises & Token Pricing

This section details the capital formation mechanics of the Perpetual ABS instrument introduced in §3.6.

10.1 The Iterative Appreciation Model

Because each Midbench Pool is a Perpetual ABS—not a fixed-life trust—it can raise capital through multiple sequential auctions into a single token, indefinitely. This creates iterative appreciation for existing holders:

Round 0: Midbench seeds $100K at BV = $1.00     → 100,000 tokens
Round 1: Investor A buys at $1.20/token         → Premium flows to Pool
         Pool now: $220K assets, 200K tokens    → BV = $1.10
Round 2: Investor B buys at $1.32/token         → Premium flows to Pool
         Pool now: $352K assets, 300K tokens    → BV = $1.173

Each new capital raise at a premium increases book value for all existing holders.

10.2 Fair Value Pricing

The fair value of a pool token is determined by discounting expected cash flows at the investor's required return:

$P_{fair} = \frac{R}{r} \times BV$

Where:

• $R$ = Net Yield to Pool (expected return on deployed capital)
• $r$ = Investor Required Return (discount rate)
• $BV$ = Book Value per token (total assets ÷ total supply)

At fair value, the investor earns exactly their required return $r$.

10.3 Reserve Price Rationale

Each auction includes a reserve price—a minimum clearing price below which the auction does not settle. This serves three purposes:

1. Protect existing holders: New capital should be accretive, not dilutive
2. Enforce fair value: The reserve price is typically set at fair value ($\frac{R}{r} \times BV$)
3. Sustainable business model: Midbench's founding position must appreciate to fund operations

The premium is dynamic, not fixed:

The fair value premium ($\frac{R}{r} - 1$) depends on current market conditions:

The 20% figure is illustrative based on current market conditions (R ≈ 6%, r ≈ 5%), not a guarantee. Actual premiums will vary with:

• Treasury rates: As r changes, the R/r ratio shifts
• Credit spreads: As R changes with loan performance and market pricing, so does fair value
• Relative investment sizes: The dilution/appreciation impact on BV depends on new capital vs. existing supply

10.4 Fair Value by Round

As book value compounds through successive raises, the fair value price increases. Note: The premium percentage is not fixed—it depends on R and r at the time of each raise:

The premium floats with market conditions; the absolute price reflects both BV growth and the current R/r ratio.

10.5 What You're Buying

When bidding in a Pool token auction, investors are purchasing:

1. A pro-rata claim on future loan cash flows that yield $R$ annually
2. NOT a claim on current USDC (which will be deployed into loans)
3. A perpetual instrument with no maturity date

Your expected return depends on your entry price:

$\text{Expected Return} = R \times \frac{BV}{P_{bid}}$

10.6 Comparison to Traditional ABS Pricing

The reserve price mechanism is similar to Dutch auction IPOs (e.g., Google 2004), where the issuer sets a floor and lets the market determine the clearing price above it.

10.7 Auction Failure

If an auction cannot clear at or above the reserve price:

1. The auction does not settle
2. All bidder commitments are refunded
3. No new tokens are minted
4. Existing holders are unaffected

This is a feature, not a bug. A failed auction indicates the market does not value the opportunity at fair value—Midbench can then reassess loan pipeline, credit performance, or market conditions.

10.8 Reserve Price Calculation

For each auction, the reserve price is computed as:

$P_{reserve} = \frac{R}{r} \times BV_{current}$

Using observable inputs:

Appendix: Model Equations

© 2026 Midbench Inc. All rights reserved.

This white paper is for informational purposes only and does not constitute an offer to sell or a solicitation of an offer to buy any securities.