The Mathematics of Pawn Loans: Why Monthly Compounding Breaks Standard Banking Software
Standard banking software models 365-day APRs and monthly EMIs. Indian pawnbrokers compute day-accurate simple interest and 30-day compounding cycles on fluctuating gold collateral. Here is how we modeled it in Kotlin.
In India's informal lending and pawnbroking markets (Girvi and Bandhak), the arithmetic of money lending is fundamentally different from commercial bank loans or consumer credit cards.
Commercial loans assume fixed amortization schedules: an equated monthly installment (EMI) where each payment is mathematically partitioned between principal reduction and reducing-balance interest over a 365-day annual basis.
A traditional pawnbroker operates on an entirely different financial framework built around immediate physical collateral, 30-day lunar or calendar settlement periods, and compounding interest cycles (Chakravriddhi Byaj).
1. Day-Accurate Pro-Rata vs. Whole-Month Rounding: When a borrower pledges gold ornaments to meet an emergency expense and returns 42 days later, how is interest computed? Traditional ledgers often round to two full months, causing friction and customer disputes. Modern pawnbrokers require day-accurate pro-rata computation: one full month's agreed percentage plus exact daily fractional accrual (Monthly Rate / 30 * 12 days).
2. Partial Repayments & Principal Reallocation: When a customer pays ₹5,000 toward an outstanding pledge balance of ₹35,000, how is that capital allocated? In informal lending practice, accrued interest must be satisfied first to prevent balance deterioration, with residual funds credited directly to principal reduction. Hand-written ledgers frequently make errors during multi-cycle partial payments.
3. Dynamic Gold Loan-to-Value (LTV) Ratios: While institutional NBFCs operate under an RBI-mandated 75% LTV cap against market spot prices, independent pawnbrokers adjust valuation based on net gold purity after stone deduction, hallmark karatage (22K, 18K, 14K), and melting loss allowance. When gold prices fluctuate rapidly, calculating collateral security margin by hand is prone to catastrophic error.
4. Avoiding Floating-Point Drift in Financial Code: In building Girvi, standard IEEE 754 floating-point numbers (Double or Float in Kotlin) were strictly avoided. When computing compounding interest across thousands of historical pledge transactions, floating-point rounding errors accumulate into rupee discrepancies. We modeled all financial pipelines using fixed-point integer cents (Paise) and BigDecimal to guarantee mathematical precision down to the exact Paisa.
By translating centuries-old counter lending conventions into rigorous Kotlin domain models, we gave pawnbrokers an app that matches their authentic accounting principles while eliminating manual mathematical error.
Explore how we translate these principles into production software.