Lookback: March 2025, when the Fed held rates but India's yield curve refused to stay flat

The 19 March 2025 FOMC was supposed to be a non-event. By the time the dust settled three weeks later, it had quietly reset the carry trade into Indian debt and forced rate-sensitives to reprice.

The script for that Wednesday was almost written before Jerome Powell walked to the podium. Consensus across the Mumbai sell-side had pencilled in a hold at 4.25 to 4.50%, a marginal trim to the SEP dot plot, and a careful re-statement of the data-dependent line. The Fed delivered exactly that. What nobody on the desks had quite priced was the reaction function inside the Indian rates complex, which moved with a conviction that the FOMC statement itself did not justify. The Indian 10-year benchmark, which had been hugging the 6.95 to 7.05% band for most of February, slipped roughly eight basis points intraday on 20 March (the morning after the US session) and kept drifting lower into month-end. Foreign portfolio investors, who had been net sellers in Indian debt through most of January and February, flipped to net buyers within the same week. The disconnect was the story.

I want to walk through what the tape actually did across those three weeks, because the temptation in hindsight is to compress a slow rotation into a single dramatic candle. It was not that. It was a methodical handover from one regime to another, and the equity market took its cue from rates with a delay of about four to five sessions.

The set-up: how the desks positioned into 19 March

The week running into the FOMC was characterised by a curious split. Equities, measured by the Nifty 50 and the Bank Nifty, traded in a tight 1.2% range through 12 to 18 March, with intraday volatility compressing to levels that India VIX had not seen since the August 2024 lows. Bank Nifty in particular looked coiled. The index had repeatedly failed at the 51,800 to 52,000 zone through early March, and the open interest profile on the weekly options strip showed heavy call writing at 52,000 and put writing at 50,500. That is the textbook shape of a market positioned for a non-event.

Cash flows told a different story. FII activity in the cash segment, per the NSE provisional data through 12 to 18 March, was a thin net negative, with the daily prints oscillating around the flat line. DIIs, mostly the domestic mutual fund complex absorbing SIP inflows, continued their mechanical bid. The interesting tell was in the derivatives book, where FII index futures net long had been bleeding lower through February and the first week of March, and where the put-call ratio on the Nifty options strip had crept up to levels that historically marked short-term bottoms. The street was not bullish. The street was hedged.

The bond desks had a different problem. The Indian 10-year had been sticky around 7.00% even as the US 10-year had drifted from a February peak above 4.55% down toward 4.25% by mid-March. That widening of the real spread, on a hedged basis, was sitting on the desks of every Singapore and London FPI rates trader, and the chatter on the brokerage calls in the week of the FOMC was explicit about it. The spread looked dislocated. Nobody wanted to be first.

USDINR was the other piece of the jigsaw. The pair had been parked just above 87.20 through much of mid-March, with the RBI's hand visible on the upside in the form of sporadic dollar sales around 87.40. The DXY, which had topped above 108 in January, had drifted toward 103.50 by 18 March, and that dollar weakness was the quiet enabler for everything that followed. Crude (Brent) sat in a 71 to 74 dollar range, which the macro desks correctly read as benign for India's current account math.

Caption: the weekly frame shows the compression into 19 March and the slow expansion through the back half of March and into April, with the FOMC week itself sitting as the inflection candle rather than the climax.

The weekly frame is worth pausing on, because it captures something that the daily chart obscures. The two candles before the FOMC were inside bars. The candle that followed was a wide-range expansion to the upside. In Indian equities, that pattern, after a multi-week consolidation, has historically marked the start of a directional leg, not its end. The skeptics who called the post-FOMC rally a relief bounce had to contend with the fact that the bounce did not roll over.

What actually happened on the day, and the day after

The Indian session on 19 March itself was a non-event in equities. The FOMC decision was scheduled for after the Indian close, so the Nifty and Sensex closed near flat with a defensive bias, and Bank Nifty did its now-familiar drift into 51,600. Currency and bond markets in Asia caught the Powell presser in their morning windows of 20 March, and the reaction was immediate.

The 10-year benchmark opened lower in yield. By the time the OIS desks had repriced, the one-year OIS had shed roughly six to seven basis points and the five-year had shed closer to ten. That move was not about the Fed. The Fed had held. The move was about what the Fed signalled it would not do, which was push back on the market's pre-existing dovish drift. Powell's tone on inflation was read as confirmation, not as hawkish surprise. The Indian curve, which had been waiting for permission, took permission.

Equities followed within an hour of the open. Bank Nifty broke through the 52,000 call wall by 10:15 IST, and that single break dragged a great deal of trapped short delta into the cash market. Public-sector banks, which had been the laggard cohort through February, led the move. The half-hour bars on 20 March show the textbook shape: a gap up, a brief retracement to fill, and then a steady step-higher pattern with each retracement holding above the prior swing low. By the close on 20 March, Bank Nifty had added in the region of 1.4 to 1.6% off the prior day's close, and the breadth on the NSE 500 was the strongest single-session reading since the December 2024 highs.

Caption: the half-hour frame shows how the move was paced, with the gap-up on 20 March holding through the European open and into the US session, instead of fading by lunch as so many FOMC-day reactions in 2024 had done.

The character of the 20 March session is what mattered. Fade-the-gap had been the dominant intraday pattern through late 2024 and early 2025, where any FOMC-driven gap up in India tended to give back two-thirds of its move by 14:30 IST. The 20 March bars did not do that. The retracement was shallow, the volume on the up-bars was visibly heavier than on the down-bars, and the close was in the upper third of the daily range. Tape readers know what that means. It means the move had committed buyers, not just short-covering tourists.

USDINR cracked through 87.00 by 21 March and traded down to the 86.65 to 86.75 zone by the following Monday. That was not a small move in the context of February, when the pair had been bid on every dip. It was the FPI dollar selling, on the back of the resumed debt inflows, that did the work.

The sector tape: who actually owned the rally

If the move had been pure beta, every basket would have moved in line. It was not pure beta. The sectoral breakdown across the 19 March to 2 April window had a clear architecture, and it is worth reading as a positioning map.

PSU banks were the standout. Even before the FOMC, the Nifty PSU Bank index had been forming higher lows through March, and the 20 March session put it into a clean breakout. By the first week of April, the cohort had outperformed the broader Nifty by a meaningful margin, with names like SBI, Bank of Baroda, and Canara Bank doing most of the heavy lifting. The driver was not Fed-specific. The driver was the read-through from a softer USDINR and a friendlier rates regime into the carry on government bond portfolios, which sit fat on PSU bank balance sheets.

Private banks lagged on a relative basis through the first week. HDFC Bank, ICICI Bank, and Axis Bank participated, but the leadership baton clearly sat with the PSU cohort for the first five sessions. By the second week, with the broader rotation maturing, private banks caught up.

Realty and NBFCs (capital-market sensitive ones) led the second-derivative trade. The thesis was simple. A softer rates regime, combined with renewed FPI appetite, was a tailwind for housing finance and a tailwind for the wholesale-funded NBFC book. Names with high duration on their liability books outperformed, and the Nifty Realty index posted one of its strongest fortnights of the 2025 calendar through the 19 March to 2 April window.

IT was the puzzle. The sector should, on paper, have done well from a softer DXY and friendlier global rates. It did not. The tape on IT through that window was anaemic, and the underperformance versus the Nifty was visible by the first week of April. The reason was idiosyncratic. The US recession chatter that had been bubbling in February had not gone away, and the buy-side was reluctant to add to large-cap IT into the April earnings prints without first seeing what the FY26 guidance looked like. The Fed hold did not solve the demand question, and the tape respected that distinction.

FMCG and pharma behaved as the defensives they were positioned as. They participated mildly on the up-days and held up reasonably on the few intraday pullbacks, but they were not the trade. The capex names, the infrastructure cohort, and the cement basket all did better, with the latter benefiting from a parallel narrative around an early monsoon and a softer commodity input bill.

Metals were the contrarian sleeve. Despite a softer dollar (which should have helped) the metals basket lagged through the first week because the China demand chatter remained mixed. By the second week, with the dollar holding lower and copper finding a bid, the cohort caught up.

The cleanest read on the tape was this: anything geared to domestic rate sensitivity outperformed, anything geared to a US growth call underperformed.

Flows, the five-session follow-through, and the ten-session test

The FII cash flow flipped from net negative to net positive within the week of 19 March, and stayed net positive through the back half of March. The cumulative net buy through 19 March to 2 April, per the NSE provisional aggregates, was the strongest fortnight of net FPI cash buying since the September 2024 window. That is the data point I would anchor the whole analysis on. The Fed itself did not change. The flow regime did.

DII flows did what they always did. The mutual fund SIP machine continued to absorb whatever incremental supply came to the market, with the daily DII cash buy holding in a steady band. The interesting feature, however, was on the derivatives side, where the FII net long in index futures climbed off the February lows through the back half of March, and where the put writing in the Nifty weekly options strip became visible at strikes that the index had only just broken above. That is positioning language for confidence.

Caption: the daily frame shows the flow inflection lining up with the price inflection on 20 March, and the five and ten-session follow-through that confirmed the move was not a one-day relief bounce.

The five-session test, which is the first filter every event-driven setup has to clear, was passed with room. By 26 March, the Nifty was holding well above the pre-FOMC close, the Bank Nifty had consolidated above 52,500, and breadth had not deteriorated. The ten-session test, which clears by 2 April, was where the rotation matured. PSU banks held their leadership, private banks caught up, realty and NBFCs extended, IT continued to lag, and the broader market made fresh highs versus the early March base.

There were two intraday wobbles worth flagging. The first was 24 March, a Monday, when the Nifty opened soft on the back of weekend headlines around US tariff geometry. The dip was bought within the morning session, which was the tell. The second was 28 March, around the F&O expiry, when option-led volatility briefly took the index down to test the 22,800 zone on the Nifty. Same outcome. The dip was bought. In both cases, the half-hour bars showed accumulation rather than distribution.

The verdict the desks reached, and the one they should have

The brokerage notes that came out 24 to 72 hours after the FOMC mostly read the move correctly in direction and underweighted it in magnitude. The dominant call across the major Indian sell-side desks was constructive on rate-sensitives, neutral to mildly negative on IT, and constructive on PSU banks. That was, in retrospect, the right tape call. The note that landed best in my reading at the time was the one that flagged the FPI debt re-entry as the load-bearing variable, because it was that single variable that made the rest of the rotation work. Without the FPI bid on debt, the rupee does not strengthen. Without the rupee strengthening, the equity beta does not get a second leg. Without the second leg, the breakouts in PSU banks and realty do not hold their character.

The cleanest mistake in the post-event commentary was the one that read 20 March as a sentiment trade. It was not. It was a positioning trade. The shorts in PSU banks, the underweights in realty, the cautious FPI debt books, all of those positions were on the wrong side of a regime that the FOMC simply did not push back against. The market did not need the Fed to cut. The market needed the Fed to not surprise hawkish. That is what it got.

The broader question, the one that mattered most for the months that followed, was whether 19 March marked the start of a trend or whether it was a tactical window. My read at the time, and the read I would still defend, was that it was both. It was tactical in the sense that the immediate re-rating in rate-sensitives happened over a compressed three to four week window. It was structural in the sense that the FPI behaviour through April and May 2025 continued the pattern that 20 March had inaugurated, with debt inflows running well above the run-rate of late 2024.

VERDICT

Stance: BULLISH (in retrospect, for the 19 March to 2 April window and the five to ten session follow-through) Horizon: 1mo Rationale: the Fed hold itself was the non-event the desks had priced; the load-bearing variable was the FPI debt re-entry and the consequent rupee strength, which lifted PSU banks, realty, and rate-sensitive NBFCs with a clean five and ten-session follow-through, while IT lagged on idiosyncratic US-demand caution.

Looking back from May 2026, the March 2025 FOMC sits in a particular category of Indian-market events. It was not a market-moving decision in its own content. It was a permission slip. The tape had been waiting for the Fed to not push back, the FPI desks had been waiting for the rupee to find a floor, and the domestic rate complex had been waiting for the green light to compress the spread to US Treasuries. All three got what they were waiting for inside the same 24-hour window, and the rotation that followed had the shape of a positioning unwind rather than a sentiment shift. That is a useful distinction. Sentiment shifts fade. Positioning unwinds persist until the positioning is genuinely cleared, which in this case took the better part of a month.

The piece of the tape I keep returning to is the 20 March half-hour series. The gap that did not fill, the retracement that held above the prior swing, the close in the upper third of the range. Those are the signatures of a market that has decided. The Fed did not decide for it. The Fed simply got out of the way.