Sarabesh Muralidharan

MONETARY ECONOMICS

Irving Fisher’s equation of exchange, MV = PQ, is one of the oldest identities in monetary economics. Velocity is, in effect, a measure of how hard each rupee in the economy is working. A rupee that sits in a locker does no economic work; a rupee that moves through ten transactions in a month does a great deal.

On the face of it, UPI looks purpose-built to raise velocity. It replaced a payment process that could take a working day, or require standing in a queue, with one that settles in seconds, at any hour, for free. The scale of adoption has been extraordinary: from around 1.2 billion transactions in FY 2017-18, UPI volume grew to roughly 241.6 billion transactions in FY 2025-26 — a nearly 12,000-fold increase in eight years — with the platform now processing more than 660 million transactions every single day.

Each of those transactions represents a rupee changing hands. If we take Fisher’s definition literally — velocity as the frequency with which money is exchanged — the transactional velocity of money in India’s digital economy has almost certainly risen sharply since 2016, simply because the friction that used to slow a rupee down between one holder and the next has been largely removed.

A rise in the number of UPI transactions does not necessarily mean the same rise in velocity measured against broad money (M3) and nominal GDP, for a specific reason worth taking seriously: a large share of UPI activity is very small in ticket size. Eighty-six percent of person-to-merchant UPI payments are below ₹500.

Some of this growth reflects genuinely new economic activity moving into the formal, trackable digital system for the first time — which would indeed raise velocity. But some of it may simply reflect the same underlying spending being split into more, smaller, more frequent digital transactions than before. A single cash withdrawal that once funded a week of small purchases now shows up as a dozen separate UPI payments — inflating the transaction count without necessarily raising the true velocity of the underlying money stock.

This distinction matters for anyone tempted to cite UPI’s transaction growth as direct proof of higher monetary velocity in the Fisher sense. The honest position is a conditional one. UPI has, with very high confidence, increased the frequency and speed with which money changes hands at the transaction level, and has likely pulled a meaningful share of previously cash-based, undocumented activity into the visible digital economy. Both point toward higher macro velocity. But confirming the size of that effect on M3 velocity specifically would require comparing RBI money-supply and nominal-GDP data before and after UPI’s scale-up, holding other drivers of velocity constant — an empirical exercise this article has not attempted, and one worth flagging clearly rather than assuming away.

My tentative answer: very probably yes, but the transaction-count evidence alone is suggestive rather than conclusive. UPI is exactly the kind of infrastructure change monetary economists would expect to raise velocity — and precisely how much remains a genuinely open, and answerable, empirical question.