Hundi vs SWIFT: Ancient System, Modern Parallels

What Global Finance Learned (and Didn't Learn) from Indian Merchants

The SWIFT network processes $5 trillion daily across 11,000+ institutions, the backbone of global finance. Yet its architecture echoes principles Indian merchants discovered centuries ago: correspondent banking, trust hierarchies, and standardized messaging. This lesson compares Hundi networks to SWIFT, revealing what modern finance retained, what it improved, and what it might have lost.

Two Systems, Five Centuries Apart

Brussels, 2024. A banker at HSBC London sends a payment instruction to Standard Chartered Singapore. The message travels through SWIFT, the Society for Worldwide Interbank Financial Telecommunication. Within seconds, $10 million moves.

Surat, 1750. A Shroff draws a Hundi on his correspondent in Muscat. A courier carries the paper across the Arabian Sea. Within weeks, the equivalent of that fortune changes hands.

The technology differs radically, fiber optics versus paper, seconds versus weeks. But the underlying architecture? Surprisingly similar.

Correspondent Networks: Then and Now

Neither system moved physical value. Both moved instructions between trusted parties who then settled locally.

Hundi Correspondent Network:

Merchant A (Surat) → Shroff B (Surat) → Hundi → Correspondent C (Muscat) → Recipient D (Muscat)

SWIFT Correspondent Network:

Company A (London) → Bank B (London) → SWIFT Message → Correspondent C (Singapore) → Recipient D (Singapore)

The parallel is exact. In both cases:

1. The sender deposits value with their local institution
2. The institution sends instructions to a correspondent elsewhere
3. The correspondent pays out locally
4. Settlement happens between institutions separately

Trust Hierarchies: Tiered Architecture

Not all banks connect to SWIFT equally. Major global banks connect directly. Smaller banks access SWIFT through correspondents. This creates a trust hierarchy.

The Hundi system had identical structure:

SWIFT Tiered Architecture:

• Tier 1: Major global banks (JP Morgan, HSBC)
• Tier 2: National banks using Tier 1 as correspondents
• Tier 3: Regional banks using Tier 2 as correspondents

Hundi Tiered Architecture:

• Tier 1: Major Shroff houses (Jagat Seth, Chettiars)
• Tier 2: Regional merchants using Tier 1 as correspondents
• Tier 3: Local traders using Tier 2 connections

The economic logic is identical: trust is expensive to establish. Rather than each party verifying every counterparty, the system creates verified nodes that others connect through.

Message Standardization

SWIFT's power comes from standardization. An MT103 payment message from Tokyo to Toronto uses identical fields.

Hundi communities developed similar standardization:

SWIFT MT103 Fields: Sender BIC, Receiver BIC, Currency/amount, Ordering customer, Beneficiary

Hundi Standard Elements: Drawer identification, Drawee identification, Currency/amount, Payee details, Time terms, Authentication

What took SWIFT decades to standardize, Hundi communities developed over centuries through practice.

What SWIFT Improved

Fair assessment requires acknowledging SWIFT's genuine advances:

Speed: From weeks to seconds. This isn't incremental, it's transformational.

Scale: 11,000+ institutions in 200+ countries. The Hundi network never achieved such comprehensive coverage.

Standardization: One global standard versus many community standards.

Regulatory Compliance: SWIFT enables anti-money laundering and sanctions enforcement.

Resilience: Distributed architecture with multiple data centers.

What SWIFT May Have Lost

Yet the comparison also reveals potential losses:

Flexibility: Hundis could be customized to any transaction need. SWIFT messages are standardized but rigid.

Cost Efficiency: SWIFT transactions carry fees at each correspondent hop. Hundi correspondents often settled through ongoing trade relationships.

Integration: Hundi settlement integrated with trade flows. SWIFT separates payments from trade.

Financial Inclusion: SWIFT requires bank accounts and formal identification. Hundis worked for anyone trusted within the community.

The Hawala Parallel

The closest living relative of Hundis is the hawala system, still operating for remittances:

1. Worker in Dubai gives cash to hawaladar
2. Hawaladar calls counterpart in Kerala
3. Kerala counterpart pays worker's family
4. Settlement happens through ongoing trade

This is exactly how Darshani Hundis worked, but with phones instead of paper. Hawala remains popular because it's faster and cheaper than bank transfers for many corridors.

Data Comparison

Lessons for Modern Finance

1. Technology Changes Faster Than Architecture

SWIFT's correspondent banking model is essentially the Hundi model with electronic transmission. The fundamental architecture hasn't changed in 500 years.

2. Trust Infrastructure Is Fundamental

Both systems work because participants trust each other. SWIFT backs trust with legal contracts; Hundis backed trust with community and reputation.

3. Integration Beats Separation

Hundi settlement integrated with trade flows, reducing friction. SWIFT separates payments from trade, requiring separate processing.

Your SWIFT Moment

Every international transfer you make relies on SWIFT or similar networks. The technology is invisible but essential.

As you use modern payment systems, remember that you're participating in an ancient human challenge. The merchants of Surat would recognize what you're doing, even if they wouldn't recognize how.

Global Perspectives on Payment Infrastructure

The question of how to move value across distances and institutions has occupied thinkers worldwide. How do contemporary scholars view the fundamental challenges that both Hundi and SWIFT address?

Perry Mehrling (born 1959), the economist known for his 'money view' framework, argues that all payment systems are fundamentally about managing the 'hierarchy of money', where some promises are more acceptable than others. The Hundi system created its own hierarchy: Jagat Seth's word outranked a local trader's, just as JP Morgan's SWIFT messages carry more weight than a regional bank's. Mehrling's insight: payment systems are always about trust hierarchies, whether traditional or modern.

Charles Goodhart (born 1936), the British economist who served on the Bank of England's Monetary Policy Committee, emphasized that payment systems require public goods, shared infrastructure that no single participant would build alone. The mahajan assemblies provided this for Hundi networks; SWIFT provides it today. Goodhart's lesson: successful payment systems blend cooperation with competition.

Barry Eichengreen (born 1952), the UC Berkeley economist who studies international monetary systems, documents how dominant payment networks persist long after their original advantages fade. Sterling-based clearing survived decades after British economic dominance ended; SWIFT may persist similarly. His parallel: Hundi networks persisted into the 20th century even after colonial banking arrived, established trust networks resist displacement.

These scholars reveal a striking continuity: the problems of payment infrastructure are universal, and the solutions, trust hierarchies, cooperative infrastructure, network persistence, emerge regardless of technology. The Hundi merchants discovered through practice what economists theorize today.

In our final lesson, we'll bring the Hundi story to the present: what does this heritage mean for India's financial future in 2026 and beyond?

Economists recognize that some solutions are 'fundamental', they'll be reinvented whenever the problem arises.

Indian merchants discovered correspondent banking independently and early. Understanding this indigenous innovation helps appreciate that modern finance builds on universal solutions.

SWIFT's architecture documents from 1973 describe correspondent banking as if newly invented. Yet the Jagat Seth network operated identically 200+ years earlier.

Ronald Coase's transaction cost theory explains that integration reduces friction between related activities.

Hundi correspondents were often trading partners. This integration reduced settlement friction, SWIFT's separation creates friction.

30-40% of Hundi obligations settled through trade offsets rather than cash transfers.