AI Just Became Capital Risk

On 24 June, the Reserve Bank of India dropped a draft that does not look like an AI regulation at first read. It is titled Guidance on Regulatory Principles for Model Risk Management, 2026, and the language is pure prudential supervision: board-approved frameworks, three lines of defence, independent validation, risk tiering, model inventories retained for ten years after decommissioning. Spend twenty minutes with it and a sharper thought lands. India's central bank has just pulled AI into the family of supervisory tools normally reserved for capital and liquidity. The principle that a regulated entity remains fully accountable for the outcomes of every model it uses, including those sourced from third-party vendors, quietly closes a loophole that the entire fintech-banking AI stack had been leaning on.

What the draft actually says

The reach is wide. Eleven categories of regulated entities — commercial banks, small finance and payments banks, urban and rural co-operatives, NBFCs across all layers, AIFIs, ARCs, credit information companies — are inside the perimeter. The definition of “model” is wider still. It covers not just neural networks and generative AI but also algorithms, analytics tools, decision rules, even spreadsheet-based tools where they materially influence credit, pricing or risk decisions. Every such model must sit under a Board-approved Model Risk Management Framework. Every high-risk model must be cleared by the Risk Management Committee of the Board before deployment. Every customer-facing AI interface must disclose that it is AI, list its limitations, and offer the user a way to switch to a human. And every deployed AI system needs a working kill switch — a one-button deactivation if the outputs go wrong. Comments are invited until 24 July.

The deep idea: models as a new line on the risk ledger

This is the consequential part, and most early commentary is missing it. When a regulator says you cannot offload accountability to a vendor's API, it is doing for AI what Basel did for credit risk in the early 1990s. Models become a class of exposure. They have to be inventoried, tiered by materiality, validated independently, monitored for drift, and signed off at the highest level. That is the grammar of capital, not the grammar of code.

It is also the only grammar that scales. The RBI's own FREE-AI Committee survey last year found that nearly 21% of regulated entities were already deploying AI in production — across credit underwriting, cybersecurity, customer support, sales — and 67% wanted to go deeper. At that level of penetration, telling boards “your tech team has this” is not serious supervision. Every credit cycle eventually meets its model failure mode. A regulator that waits to discover the failure is a regulator already too late.

The kill switch rewrites the vendor market

Look at the supply side and the picture sharpens. A small set of global tech firms quietly supplies a disproportionate share of the AI models running inside Indian financial services. The draft flags this concentration explicitly as a systemic supply-chain risk. Combine it with the third-party-is-no-defence clause and the direction of travel is unmistakable: a serious push toward indigenous, auditable, swap-out-able model stacks. The new market is not the bank's market. It is a RegTech market — model validation firms, bias-testing labs, explainability auditors, red-teaming shops. A compliance officer reading this draft on Monday morning is realising that the next critical hire is not another data scientist; it is an independent validator who can sign off on someone else's data scientist.

Why this logic will travel beyond banking

I have spent enough time inside a national administration deploying AI on citizen-facing systems to read this draft as a template, not an end-point. Tax, customs, social security, urban service delivery, public health records — every government body running models on millions of files faces the same accountability question the RBI is now putting to banks. Who signs off on the model that decides a refund? A scrutiny notice? An eligibility threshold? The answer cannot be “the algorithm”. It has to be a named human, with a department behind them, with an institutional framework above them.

Three things from this draft will, I think, become the standard public-sector discipline within two budget cycles:

• the inventory rule — every active and decommissioned model on a register, with a ten-year tail;
• the explainability threshold — outputs interpretable to the extent the business process actually requires;
• the kill switch as a non-negotiable product feature, not a nice-to-have.

The trade-off, and the right side of it

The cost is real. Industry analysts already estimate a 50 to 100 basis point rise in IT spending for serious adopters. Smaller NBFCs and co-operative banks will feel it the most. Time-to-market for AI-driven products will lengthen. None of this is free, and the smaller institutions will need a transition window the draft does not yet promise.

But picture the alternative. A black-box model, bought from a vendor, denying loans to a cluster of small borrowers in a particular district, operating under no one's clear authority. Someone eventually discovers it — they always do. The regulator pays, the bank pays, the customer pays, and trust in AI itself pays the heaviest tax of all. The cost of the RBI's draft is the cost of preventing that discovery.

The cleaner way to say it: in a regulated industry, AI stops being a technology decision and becomes a capital decision. Capital decisions are made in boardrooms, not in model repositories. The window closes on 24 July. The more interesting question is not whether this framework is right for finance — it broadly is — but how quickly the rest of the public-private edge of AI adoption catches up. The RBI has, almost without saying so, written the first chapter of an Indian AI accountability code. Other regulators will write the rest, or they will inherit the failures of not having done so.

#RBI #AIGovernance #ModelRisk #BankingRegulation #IndianEconomy #AIAccountability #PublicFinance #RegTech

Government Isn't A Hyperscaler

On Monday, Alphabet fell about 5%, dragging the communication services sector down with it. Memory chip stocks plunged in Asia overnight and the selling crossed the Pacific by Tuesday morning. Beneath the price action, the cause was specific and revealing: investors are starting to ask whether the enormous sums being poured into artificial intelligence will earn their keep, and reports of senior talent leaving Alphabet's AI teams added to the unease.

That question — will the AI capex pay off — is the right question for shareholders. It is the wrong question for a government.

What the market wobble actually says

The sell-off was narrow, not broad. The Russell 2000 closed above 3,000 for the first time even as the big tech names tumbled. That tells you the market is rotating, not collapsing. The doubt is concentrated in a handful of names whose valuations had assumed the AI trade would compound for several more years without a pause.

The question being asked is simple. Amazon, Microsoft, Alphabet and Meta have collectively guided to roughly 25 billion in capital expenditure for 2026. Their combined free cash flow is forecast to fall to about billion in the third quarter — a decade low. At that ratio, the cash being spent is no longer comfortably backed by the cash being earned. That is what the nervous side of the market is pricing.

Why this question doesn't translate to government

Public administration does not sell tokens. It does not have shareholders. Its “revenue” is taxes and its “product” is rights, services, and the slow accretion of public trust. The whole logic of equity-style ROI — free cash flow divided by capex — is structurally absent.

And yet, watch any AI tender written by any government this year and you will see private-sector language smuggled in. Pilots. Use cases. Productivity gains. Cost savings per FTE. We are evaluating public AI investments in the only vocabulary the consultants brought to the room, and the result is that everyone ends up arguing over a metric nobody in the citizen-service chain actually cares about.

From inside a national tax administration, having worked on conversational AI for citizens at scale, I can say something specific. The right number was never “queries handled per rupee of compute.” It was something closer to: did the next confused taxpayer get an accurate answer, in the language she actually speaks, in time to act on it, without having to travel to a counter? That is an outcome question. It is not a capex question.

An outcomes-first lens

If the market is pricing AI on cash flow, the government should price AI on three different things:

• Service latency. How much faster does a citizen get an answer, a refund, or a decision after the system is deployed?
• Service reach. How many more citizens, in how many more languages and pin codes, can the same service touch without adding counters?
• Officer leverage. How much higher-order work can the same officer do because the system has absorbed the routine?

None of these are radical. They are simply restatements of what the public sector exists to do. The discipline is to write them into the procurement document before the vendor walks in, not after the dashboard is built.

A proposal: separate the two stacks

Here is a concrete proposal. A government should split its AI estate into two stacks and evaluate them by two completely different rules.

The internal stack — drafting, summarisation, case retrieval, file review, internal search across decades of orders — can be evaluated commercially. Hours saved, error rates, contractor-substitution ratios. That is fair, because the work being replaced was already priced.

The citizen-facing stack — multilingual chatbots, eligibility navigation, guidance on a new statute, status tracking on a refund — must be evaluated as service delivery. The metric is not unit economics. The metric is whether a person previously locked out of a service is now inside it.

Confuse the two stacks and the second one will always lose to the first on a finance committee's spreadsheet. It will be quietly defunded the moment an AI valuation cycle turns. And, as Monday's tape made plain, the AI valuation cycle will turn.

The public sector organisations point

Professor Michael Ting's course on the Analysis of Public Sector Organizations at SIPA made an argument that has aged into something close to a law. Public agencies serve multiple principals — legislatures, ministers, courts, auditors, citizens — and the metrics they adopt silently decide which principal they end up serving. Adopt private-sector ROI metrics for public AI and the system will end up optimising for the auditor's spreadsheet, not the citizen at the window. That is not a hypothetical risk. It is the default trajectory.

What to do this quarter

Three moves, for any department about to sign an AI contract:

• Write the outcome metrics before the price negotiation, not after.
• Build at least one citizen-facing metric whose headline a non-technical minister can defend in question hour.
• Keep funding the citizen stack even when the headline AI trade in the markets is going through one of its periodic doubts. Especially then.

The market is reminding everyone this week that the private bet on AI is a bet. Government is not making the same bet. It is deploying a tool that should be judged by whether the queue moved. The discipline is to keep saying that out loud while the news cycle is busy saying everything else.

#AIinGovernment #PublicSectorAI #DigitalGovernance #AICapex #IndiaAI #Governance #PublicFinance

The Filing India Doesn't Owe

On 30 June 2026, exactly a week from now, multinational groups across more than three dozen jurisdictions will lodge their first GloBE Information Return. It is the most ambitious cross-border tax filing the world has ever attempted. India is not in the queue.

That is not a footnote. It is the single most important fact about Indian international tax this year, and almost nobody is talking about it.

The week the global minimum tax goes live

The OECD's Pillar Two is now operational. The rules apply to multinational groups with consolidated revenues of at least €750 million. If their effective tax rate in any jurisdiction falls below 15%, someone, somewhere, collects a top-up. Roughly 140 jurisdictions joined the inclusive framework back in 2021. Thirty-seven have actually legislated a Qualified Income Inclusion Rule or a Qualified Domestic Minimum Top-up Tax that bites from the 2024 reporting fiscal year.

The OECD was still soldering plumbing in the last weeks before the deadline. On 18 May 2026 it issued a common understanding allowing groups to file centrally in one jurisdiction and avoid duplicate local returns, provided each domestic office gets a notification. That a structural feature of the regime had to be settled eight weeks before the first deadline tells you something about how unfinished this compact still is.

Three countries that stayed away

The United States, China and India have not implemented Pillar Two. The American position is now formal: in January 2026 the US Treasury announced that US-headquartered groups would be exempt, and the OECD's 5 January Side-by-Side package legalised that exit by carving out a safe harbour for groups parented in jurisdictions with ‘eligible’ tax regimes. The US is, as of today, the only jurisdiction on the OECD Central Record with a confirmed Eligible SbS Regime.

India's stance has been quieter, and to my mind more considered. We participated in the framework. We never legislated the rules. We watched.

Why I think the wait was right

Pillar Two is not really a tax. It is a coordination mechanism that exports one country's view of what another country's tax base should be. It is the first time in modern fiscal history that a domestic legislature has been asked to outsource the residual taxing right on profits earned at home to a residence jurisdiction abroad. That deserves more scepticism than it gets in polite international tax conversation.

For India, the cost-benefit was always thin. The corporate rate here is 22%, or 25.17% with surcharge and cess. The concessional rate for new manufacturers is 15% statutory, around 17.16% effective. We are not a low-tax jurisdiction. Pillar Two only matters when you are below the floor; we mostly are not.

The price of joining would have been real. A QDMTT regime to design. A GloBE return architecture to build. A workforce to retrain in jurisdictional ETR computation. Disputes to defend under accounting standards that are not ours. And acceptance that the rules will be rewritten by an OECD working party in which our vote is one among many. None of that grows the base. It expensively confirms what we already collect.

What we still need to claim

The harder side is data. The 2024 GIR is the first time multinational groups will publish, in a standardised XML schema, a jurisdiction-by-jurisdiction picture of where their profits arose and what tax those profits paid. Other administrations will receive that file automatically. India, outside the exchange relationships, will not, unless we sign on to receive it. That is a transparency dividend we should be claiming whether or not we ever impose a single rupee of top-up tax.

Soft power is the other piece. The Side-by-Side carve-out is, in effect, a US-only privilege today. The OECD has signalled other jurisdictions may be added. ‘May’ is doing the heavy lifting in that sentence. If a future investor's post-tax compliance burden is lighter under a US parent than under an Indian one, we have not lost any revenue, but we have lost a piece of the architecture of who matters in the next decade of international tax rule-writing.

A proposal

The smart move is not to copy Pillar Two. It is to ask for the data, build the analytical capacity, and use the next eighteen months to find out where Indian profit shifting is genuinely costing us revenue. Three concrete steps.

• Sign the GIR Multilateral Competent Authority Agreement. The cost is administrative. The value is a structured view of every in-scope group's worldwide tax footprint, delivered automatically.
• Commission a quiet domestic study of jurisdictional ETRs for India-headquartered MNEs. Two or three years of clean data will tell us whether a future QDMTT would collect ten thousand crore or ten lakh. Right now the number is asserted, not measured.
• Use the Income-Tax Act 2025 transition window to bake in a minimum-tax-compatible computational backbone. If we later choose to switch on a QDMTT, the systems already speak the schema. Building it after a political decision is far more painful than building it now, while the law is still warm.

Pillar Two will either be remembered as the most consequential multilateral tax instrument since the League of Nations drafted the first model treaties in the 1920s, or as a noble experiment that fractured the moment its largest economy walked away. We do not need to guess which today. We need to stay liquid: positioned to step in, positioned to step out, owning the data either way. That is the case for sitting out 30 June with intent.

#PillarTwo #GlobalMinimumTax #IncomeTax #IndianEconomy #TaxPolicy #OECD #InternationalTax

When Ships Paid In Renminbi

Somewhere in the European Central Bank's annual report on the international role of the euro, published a few days back, there is a sentence that ought to have set off more alarms than it did. Settlement activity on China's Cross-Border Interbank Payment System rose by more than a third in the days around the outbreak of the Middle East war. Quieter still, the report notes that some ships made payments in renminbi via CIPS, or in crypto-assets, to transit through the Strait of Hormuz during March and April. A handful of transactions, in absolute terms. But a tell.

The dollar is not dying. It does not need to die for the world's payment plumbing to start looking less like a single pipe and more like a switchboard.

The Hormuz Tell

For seventy years, the working assumption of treasurers and central bankers has been that when a regional crisis flares, dollars buy you out of trouble. They buy fuel, they buy insurance, they buy the wire that gets a cargo moving. The ECB's detail tells us that assumption is starting to fray at the edges. When a ship in a hurry could not, for whatever reason, settle a passage in dollars, the alternative chosen was renminbi or, more telling, crypto.

The size of the shift in flow is worth dwelling on. Customer-related cross-border payments by Chinese banks in renminbi reportedly hit USD 1.4 trillion in March 2026, roughly 30% higher than the previous month. CIPS itself had grown a meagre 3% in 2025, after years of 20%-plus expansion; the war reversed that deceleration in a single month. None of this dethrones the dollar. All of it builds optionality for whoever pays attention.

A Defensive Pile Of Gold, And Not Much Else

India sits in this picture awkwardly. We have, by ECB's reckoning, added about 130 tonnes of gold to the reserves stack since 2022, alongside Turkey, China and Poland. That is a sensible defensive instinct after watching Russia's reserves get frozen. It is also where the imagination seems to have stopped.

The position is genuinely odd. Of the major economies thinking about reducing dollar dependence, we are the only one that already runs a domestic retail payments network most of the world envies, settles more real-time transactions than the rest of the planet combined, and has spent a decade quietly exporting that stack to friendlier jurisdictions. From inside a national administration that has built and operated citizen-facing digital infrastructure at scale, I can say with some confidence that the hard engineering problem of cross-border instant payments was solved at home long before it became geopolitically interesting. What is missing is the political will to treat the stack as strategic infrastructure rather than a soft-power side project.

Wrong Question, Right Answer

Most Indian commentary on the de-dollarisation theme falls into a trap: which bloc should we join, the dollar one or the renminbi one. This is the wrong question, and it is asked by people who confuse a payment rail with a treaty. India's interest is not in joining a bloc. It is in being able to settle, invoice, hold and route in whichever instrument is cheapest, safest and least politically charged on the day a particular contract has to close. The word for that is optionality, and it is built, not declared.

Three moves are worth making.

Treat payment corridors as foreign policy

UPI acceptance in the Gulf, Southeast Asia, parts of Africa and small island economies is being built piecemeal, often as tourism convenience. Stop calling it that. A corridor that lets an Indian importer pay a Vietnamese supplier in rupees or dong, without a dollar leg, is strategic plumbing. Build it with line items in the budget and missions actively negotiating acceptance, the way other countries negotiate visa-on-arrival regimes.

Insist on settlement clauses in trade agreements

The next ten free trade agreements India signs should, at minimum, contain a clause allowing settlement in either party's currency for a defined share of trade, with central bank windows providing convertibility at agreed bands. Project mBridge, the multi-CBDC platform connecting China, Hong Kong, Thailand, the UAE and Saudi Arabia, shows what a serious version of this looks like. India is conspicuously absent from that table; that is a choice we keep making by default.

Stop confusing gold with a strategy

A 130-tonne pile is a backstop. It is not active monetary architecture. The state can hedge tail risks in metal and at the same time build the live system that determines who pays whom in normal times. The two are not substitutes.

What The Classroom Did Get Right

Years ago at Columbia, in Prof. Richard Robb's International Capital Markets class, the formative lesson was that the international monetary system runs on inertia. It does not change because a paper is published or a summit is held. It changes when, in some operationally tedious moment a ship in a hurry, a bank under sanction, a captain refusing to wait the cheaper, safer instrument turns out not to be the dollar. That moment does not need to come at scale to matter. It needs to come reliably enough that the next CFO writes the alternative into her treasury policy as a permanent option, not an emergency one.

Xi Jinping's 1 February call for the renminbi to become a global reserve currency, taken alongside CIPS opening up to multicurrency settlement from the same date, is best read in that light. It is not a slogan. It is a procurement plan for whichever country wants to underwrite the next system.

The Habit, Not The Asset

The dollar's strength has never been an asset class. It is a habit, a default keystroke on every treasurer's terminal. Habits are sticky. They are also vulnerable to small, repeated, observable counter-examples. A few ships at Hormuz paying in renminbi, or in stablecoins routed through some Gulf trading hub, will not by themselves break that habit. But anyone reading central bank reports for a living should treat what happened in March and April as the rehearsal it was.

India holds the world's best retail payments stack and a serious geopolitical hand. Both are pointed away from each other today. The most useful thing the Government of India could do this year is to draw a line between them: to convert a fintech achievement into an instrument of statecraft. The next time ships are in a hurry at a chokepoint, the question worth being able to answer is whether any of them are settling in rupees.

#InternationalFinance #DollarDominance #UPI #Renminbi #CIPS #Geoeconomics #IndianEconomy #PaymentSystems

Tokyo Just Turned Off The Tap

The number is small. The symbolism is enormous.

On Tuesday, the Bank of Japan raised its benchmark rate to 1.0%, the first reading at that level since 1995. For most central banks a quarter-point move is routine. For Tokyo, it closes a chapter that has lasted longer than many working economists have been working.

The decision passed 7-1, with one dissenter citing growth risks. Governor Ueda missed the meeting after being hospitalised; Deputy Governor Himino, who chaired it, told parliament three days later that the bank remained committed to further hikes. That the move passed without drama is itself the story. Cheap yen is no longer the world's default setting.

What Tokyo actually tightened

The obvious analysis writes itself. The yen is weak, hovering near 160 to the dollar despite roughly 11.7 trillion yen of intervention in May. Wholesale inflation hit 6.3% that month, the highest since 2023, lifted by the Iran-driven oil shock. Real rates remain deeply negative even at 1%. None of that is wrong, but it misses the larger point.

For three decades Japan has effectively been the world's liquidity tap. Near-zero rates plus a deep, freely-traded currency made the yen the borrowing leg of the most successful carry trade in modern finance: borrow yen at almost nothing, swap into something higher-yielding from US Treasuries to Brazilian sovereign bonds to Indian rupee debt, pocket the spread. That trade has financed a startling share of cross-border risk for a very long time. When Tokyo moves the funding rate, every position built on it gets re-priced.

The signal beyond the number

What Tuesday closed is the optionality of staying loose. The Bank of Japan has spent a decade signalling that normalisation was conditional on wages, on inflation expectations, on the global cycle. The April hold was framed as a pause to absorb the Iran shock. The June hike says, in effect, that even with that shock unresolved, the bank judges the risk of inaction worse than the risk of action. Himino's testimony three days later, flagging the possibility that underlying inflation may deviate upward from the 2% target, was a polite way of saying more hikes are coming.

That changes the global plumbing in a way much domestic commentary will miss. A 1% policy rate today, plus a credible signal of 1.25% by year-end, is enough to make a slice of yen-funded positions uneconomic at the margin. Past BOJ hikes since 2024 have each been followed by sharp drawdowns in high-beta global assets within weeks. The mechanism is not mystery — it is forced deleveraging.

Through the India lens

The reflex in emerging markets, including ours, is to read these moves through the dollar lens. That is half the picture. The other half is who funds the dollar.

India has benefited handsomely from a long stretch in which a slice of foreign portfolio money in our bonds and equities was, somewhere up the chain, financed in cheap yen. As that funding gets more expensive, two things shift. The marginal cost of foreign capital into India rises even if the Federal Reserve does nothing further. And the patience of that capital shortens; a carry that looks attractive at a 25 basis point Japanese funding rate looks distinctly less so at 100, and brittle at 125.

There is an old lesson from any honest course on international capital markets that the funding leg of a trade often matters more than the asset leg. Tokyo just moved the funding leg. The corollary, for a finance ministry, is that long-duration, sticky foreign capital should be treated very differently in tax and regulatory design from money that is essentially a leveraged carry position. From inside a national tax administration, I have repeatedly seen the same instrument behave very differently depending on who is funding it. The funding mix is now changing.

What I would actually do

If I were sketching responses, three would be ordered ahead of the rest.

Rebuild the muscle for scenario work. Not a war-game, a quiet quarterly exercise inside the relevant departments — including the direct tax administration, which sees cross-border income in unusual granularity — on what a sustained 150 basis point repricing of yen-funded capital does to specific sectoral flows.

Read the data we already have. Indian tax filings, treaty disclosures and inbound investor reports carry a much richer real-time picture of cross-border behaviour than is generally appreciated. The question is whether anyone is reading them that way.

Stop treating every additional inflow as a win. Some inflows are savings looking for a long home. Some are leverage looking for a quick exit. Treating them identically was tolerable when Tokyo was paying the world to borrow. It is less so now.

None of this is forecasting doom. A Japanese policy rate at 1% is, in the long run, healthy normalisation — a return to a world where the price of money is set by something other than insurance against deflation. The point is simply that the era in which a non-trivial share of global risk-taking was quietly subsidised from a single building in Tokyo is closing. We should plan for the world that comes next, not the one we got used to.

#BankOfJapan #YenCarryTrade #MonetaryPolicy #EmergingMarkets #IndianEconomy #CapitalFlows #PublicFinance

The Pause Is The Plan

At 10 a.m. this morning the Monetary Policy Committee kept the repo rate at 5.25%, unchanged for the third meeting in a row, with a neutral stance and a quietly composed press conference to follow. The headlines wrote themselves: RBI holds. Fixed-income desks shrugged. Equities opened a notch firmer. To anyone watching only the rate, today's decision looked like the absence of a decision.

It was not. The pause is the plan.

The rate is the top of a layered toolkit, not the toolkit

Think of Indian monetary policy as a stack. The policy rate sits at the top, visible, dramatic, easily understood. Below it lies a thicker, less photogenic layer: variable rate repo auctions, buy-sell forex swaps, open market operations, CRR adjustments, dollar liquidity windows for oil marketers, and the steady drumbeat of intervention in the spot and forward currency markets. Above 5.25% sits one lever. Beneath it sit a dozen.

The April hold was an early signal. Today's hold confirms the doctrine: until the data forces a move, the central bank will work the lower stack and leave the top untouched. DSP Mutual Fund's fixed-income team said it plainly in a pre-policy note, that the RBI rarely jumps straight to a rate move and follows a step-by-step sequence before pulling that trigger. Today's decision is a refusal to skip steps.

The shock is arriving through the exchange rate

Crude has been hovering near 96 dollars since the conflict in West Asia escalated. The rupee has slid to about 95 to the dollar, a level no one was forecasting a quarter ago. Wholesale inflation has crossed 8 percent. Retail inflation is still inside the band, but Icra's Aditi Nayar is right to flag that the second-round effects through fuel into transport, packaging and food are only beginning to show.

The temptation, particularly for analysts who default to a textbook reaction function, is for the RBI to hike in order to defend the currency. Standard Chartered already pencils in 50 basis points across FY27 and some expected the cycle to begin today. They were wrong, and I think rightly so. A defensive rate hike to prop up the rupee is the macroeconomic equivalent of treating a fever by raising the thermostat. It hurts the patient and does not fix the cause.

Gita Gopinath made the better point earlier this week: some rupee adjustment is what should happen when the world's oil price changes. Trying to keep the currency frozen would only postpone the move and bleed reserves in the process. The professional discipline is to let the price absorb part of the shock and use the lower-stack tools to keep the adjustment orderly.

Why doing nothing is the hardest trade in the room

To anyone who has watched government respond to external shocks from inside the administrative system, the instinct to act visibly and audibly in a crisis is almost overwhelming. Holding is the harder trade because it offers nothing for the news cycle to consume. There is no announcement, no ribbon to cut, no graph to point at.

What today protects is not the rupee, which will move as it must, but the credibility of the easing cycle behind it. Between February and December 2025 the RBI cut rates by a cumulative 125 basis points. Those cuts have transmitted into home loans, MSME credit and personal borrowing. A panicked reversal of that work, on the back of a single quarter of oil-led inflation, would shake the very domestic-consumption story that gives India its growth premium. The MPC's third consecutive pause is, in effect, a statement that the easing of last year will be defended.

This is the lesson at the heart of Prof. Richard Robb's International Capital Markets course at Columbia: a small open economy hit by a real external shock should let the exchange rate absorb the blow while the central bank manages the volatility, not the level. That is the doctrine on display this morning, even if the press release did not say so.

The harder question is what fiscal does next

The blind spot in today's commentary is that monetary policy cannot carry this alone, and should not have to. If crude stays near 96 dollars, the second-round inflation will come through fuel cesses, GST on logistics, and the price line of every state-distributed commodity. The fiscal authority owns more of those switches than the RBI does.

A serious response over the next two quarters has to include targeted excise rebates on transport fuels rather than blanket cuts, faster GST input-tax refunds to small businesses caught in the cash-flow pinch, quicker direct-benefit transfers to insulate the bottom three deciles from food inflation, and a clear medium-term fiscal anchor so the bond market prices the borrowing programme without demanding a higher term premium. Tax and expenditure administration, for once, can be the country's first line of macroeconomic defence rather than its slowest.

The closing thought

Markets are trained to read central bank announcements for what changed. The lesson of today is to read them for what stayed the same, and to notice what is doing the work underneath. The repo rate at 5.25 percent is the surface of the water. The currents that matter run below it: forex tools, liquidity operations, and a quietly disciplined refusal to let a temporary oil shock undo a year of carefully transmitted easing.

If the next 90 days bring a calmer crude tape, today will look like skill. If oil heads higher, the hold will be tested, but the architecture for the test is visibly in place. Either way, I think the MPC made the right call. Doing less, well, is harder than doing more, badly.

#RBI #MonetaryPolicy #IndianEconomy #RepoRate #MPC #Rupee #PublicFinance #CentralBanking

The Code Goes Live

Somewhere in a finance department of a company, a junior accountant is staring at a payment portal that now asks her to select Tax Year 2026-27 instead of Assessment Year 2027-28. That single drop-down is where the new Income-tax Act, 2025, actually begins. Not on 1 April, when it commenced. On 15 June, when the first advance-tax instalment falls due and a real bank account moves money under a real new statute.

Tax law, like a building, is finished only when somebody walks in.

The first deadline is the real commencement

The Income-tax Act, 2025, has been on the books for months. The Income-tax Rules, 2026, replaced the 1962 Rules from 1 April. Hundreds of FAQs have circulated. None of that is the test. The test is the 15 June advance-tax instalment for Tax Year 2026-27, the first time a taxpayer estimates a year's liability under the new code and pays fifteen percent of it.

That is when the law moves from print to pay-in. Until then, every conversation about the new Act has been a rehearsal. From 15 June, the choreography is live.

Two statutes, one tax year

There is a peculiar transition reality that gets understated. The 1961 Act stands repealed from 1 April 2026, but the repeal does not disturb anything relating to tax years before that date. Assessments, appeals and proceedings for earlier years continue under the old Act. The advance-tax instalment due in March 2026 for FY 2025-26 was governed by the old law. The instalment due on 15 June 2026 is governed by the new one.

So the department, and every CFO of any scale, is now running two statutes simultaneously: one for past income, one for current income. From inside any large tax administration, this is the un-televised reality. A transition is not a date; it is a multi-year overlap, where the same officer handles a 2023-24 reassessment under the old code in the morning and a 2026-27 advance-tax compliance under the new code after lunch. The simplification on paper does not eliminate that doubling of cognitive load. Only time does.

The small print of a clean statute

The new Act is shorter, cleaner and more readable than the 1961 Act after six decades of grafts and provisos. But cleanness creates its own friction at the implementation layer. "Previous year" has become "tax year". Section numbers have shifted: advance tax now sits in Sections 403 to 410 rather than the familiar Section 208 onwards. Form 16 is now Form 130. Form 3CEK is now Form 173. Form 26AS is now Form 168.

None of this changes any substance. All of it changes muscle memory. Every payroll system, every ERP, every accountants-office macro, every internal departmental tutorial, every taxpayer's mental shorthand must be rewritten. The headline simplification will be felt only after that rewriting is done. The cost of clarity is paid up front, in transition friction; the dividend comes back later, in compliance ease. Reformers tend to celebrate the first. Only practitioners feel the second.

An administration learning in public

I think the more interesting question is what a new code does to the administration that runs it. A tax department is, in operational terms, a very large rule-execution machine. When the rulebook changes, every loop in that machine must be re-instructed. The portal must accept new minor-head codes. Notices must cite new sections. Officers must learn to draft orders in the new language without slipping back into the 1961 cadence they have known their whole careers. Helpline scripts must be rewritten. Internal training becomes, in effect, an ongoing exam.

This is where productivity either materialises or evaporates. If the first advance-tax cycle runs smoothly, if challans clear, refunds reconcile, mismatches are caught early, then taxpayer confidence in the new regime is established for the next decade. If it is messy, every news cycle of the next few months will be about glitches, and a reform that is substantively sound will be remembered as operationally rocky. The legislative work is over. The implementation game has just begun.

A small proposal for the transition window

One concrete suggestion. Through the first full tax year under the new Act, every taxpayer-facing communication, intimation, notice, demand, refund order, should carry the equivalent old-Act section in parentheses next to the new one. A demand under Section 405 of the 2025 Act should read "(corresponding to Section 234C of the 1961 Act)". This is not a dilution. It is a translation layer.

It would cost the department almost nothing. It would save tens of thousands of taxpayers from a private translation exercise each. It would also reduce the small but real risk of misdescription in appellate orders during cross-over years, where a citation gets stuck between two statutes. Translation layers age well. They quietly disappear once they are no longer needed, leaving a cleaner system behind. The concordance table mapping the old sections with the new one needs to get etched in the muscle memory.

For most taxpayers, 15 June 2026 will feel like any other deadline. For the new Income-tax Act, 2025, it is the day the simulation ends and the actual run begins. Everything written about the new code so far has been on a whiteboard. From that date, it is on the ledger.

#IncomeTaxAct2025 #AdvanceTax #TaxYear2026 #IndianTaxation #TaxAdministration #CBDT #PublicFinance

Buffett's Quiet Bet On Compute

The most quietly important capital markets story of yesterday was not the eighty billion dollars. It was the ten billion inside it.

Warren Buffett's Berkshire Hathaway put ten billion dollars into Alphabet's latest equity raise: five billion of Class A shares at $351.81 and five billion of Class C at $348.20. Buffett does not normally pay for compute. For most of his career he avoided technology companies on the grounds that their capital was consumed faster than it returned. That he is now anchoring an eighty-billion-dollar stock sale designed almost entirely to fund AI data centres is a more telling signal than the headline number.

An industrial capex cycle has begun

Alphabet expects to spend roughly $180 to $190 billion on capex in 2026. Together with Microsoft, Meta and Amazon, the four hyperscalers are projected to put close to $700 billion into AI infrastructure this year, and Wall Street estimates point north of a trillion in 2027. These are not software numbers. They are railway, electricity-grid, telecom-trunk numbers, the kind of capital intensity that defines an industrial cycle, not a product cycle.

The raise itself is structured to look like one: thirty billion in underwritten public offerings, of which fifteen billion is mandatory convertible preferred stock, a forty-billion at-the-market programme to be drawn through the year, and the Berkshire anchor on top. Mandatory convertibles are the financing instrument of a company that wants equity-like permanence without immediate dilution. ATM programmes let you tap the market the way a utility taps a balance sheet. This is not a tech firm raising opportunistically. It is a hyperscaler behaving the way a power utility did in the 1920s.

The compute concentration that follows

Capital flows are not neutral. A trillion dollars of capex absorbed annually into four American balance sheets, almost entirely deployed inside the United States with custom silicon and dedicated power, is a redirection of the world's marginal investable capital toward one geography and one stack. The natural counterpart is already visible. Foreign investors pulled more than twenty billion dollars out of Indian equities in the first four months of this year alone, eclipsing the full-year record of 2025, and the rupee has touched all-time lows against the dollar under the combined pressure of energy shocks and capital outflows.

It would be a mistake to credit all of this to the Iran war. The war is the proximate shock. The deeper current is that risk capital globally is being pulled toward an AI capex story that, for now, is built and owned almost entirely by a handful of American firms. In Prof. Richard Robb's International Capital Markets class at Columbia, the point worth remembering was that cross-border capital follows narrative as much as yield. The narrative now has one destination.

Where the sovereignty instinct misfires

The reflex in capitals from Delhi to Brasília will be to demand domestic equivalents. Indian compute sovereignty. Indian foundation models. Indian sovereign silicon. Some of this is genuinely necessary. Most of it, at the scale being discussed, is not feasible.

Consider the arithmetic. Alphabet alone will commit more capex to AI infrastructure in 2026 than the Union government's entire capital expenditure outlay for the year. A press release calling for a Bharat-class hyperscale buildout that can sit at parity with the four American giants is not a strategy. Even a credible five percent of that footprint would consume a meaningful share of domestic risk capital and crowd out infrastructure, housing and credit to small firms.

The harder question is the right one. If we cannot build at parity, and the world's compute will sit on a small number of foreign balance sheets, what is the negotiating position?

The smarter move

I think the correct frame is to treat compute the way India once thought about energy: as a strategic input that does not have to be domestically owned to be domestically governed. A few concrete moves follow from that.

• Sovereign compute partnerships, not sovereign compute factories. Long-duration capacity contracts with hyperscalers, partly funded by domestic institutional capital, locking in priced access to GPUs and inference for Indian public workloads over a decade. The mental model is closer to an LNG offtake agreement than to Make-in-India.
• Co-investment vehicles routed through sovereign and pension capital. When the most capital-hungry sector in the world is issuing equity in eighty-billion-dollar tranches with Berkshire on the cover page, the disciplined move is to be a calibrated buyer of those issues, not an admiring spectator.
• Domestic compute concentrated on what only India can do. Indian-language models, public-data fine-tuning, citizen-facing inference at low cost, audit and compliance tooling fitted to Indian law. Not foundation-model parity. Sovereignty at the layer that actually touches citizens.
• A tax architecture that recognises compute as infrastructure. Depreciation schedules, transfer pricing rules and withholding treatment for cross-border compute services were drafted for an earlier world. They need a careful refresh, because in five years a large share of value added in Indian sectors will be running on rented foreign GPUs, and the statute should not be ambiguous about where that value is taxed.

The Buffett tell, again

Return to Buffett. The investor who built a career on insisting that capital must compound, not be incinerated, has now decided that the compute build is the rare case where the spend itself is the moat. For an Indian observer that is the cue, not to argue with the cycle, but to find the cleanest seat inside it. The wrong response is national chest-thumping about sovereign models. The right one is the unglamorous work of contracts, capital allocation, and statute. The cycle is industrial now. Treat it as such.

#CapitalMarkets #AICompute #IndianEconomy #Alphabet #BerkshireHathaway #ComputeSovereignty #Hyperscalers #EmergingMarkets

Mind The Reinvestment Gap

In the first nine months of FY26, gross foreign direct investment into India hit roughly $73.7 billion. Over the same window, net FDI turned negative for six consecutive months through January 2026. That is not a paradox to paper over with a press release. It is the gap that should be on every policymaker's desk this week.

The Finance Ministry's Monthly Economic Review, released on 30 May, made the now familiar case that the macro is resilient. PMIs are expansionary, GST is healthy, rural demand is holding. All true. But the FDI signal is the one the headline does not catch, because it requires reading the following two numbers together: how much came in, and how much quietly walked out.

The metric that flatters us

For a decade, gross FDI has been India's favourite slide in every investment pitch. The figure has held up. What has changed is the denominator the world now looks at - net FDI, which subtracts repatriations by foreign investors and outward investment by Indian firms.

The trajectory is uncomfortable. RBI reported net FDI of $10.1 billion in FY24, then just $0.4 billion in FY25 - a 96 per cent collapse in a single year - even as gross inflows climbed from $71.3 billion to $81 billion. In the first nine months of FY26, net FDI inched up to about $3 billion. The inward door is fine. The outward door has been thrown open.

Where the gap really opens

Two channels drive the leak. Repatriation and disinvestment by foreign investors rose to $51.5 billion in FY25, up from $44.5 billion in FY24 and $29.3 billion in FY23 - a near doubling in three years. Outward FDI by Indian companies surged to $29.2 billion in FY25, a 75 per cent year-on-year increase. Singapore, the United States, the UAE, Mauritius and the Netherlands took the bulk of it.

Sectorally, the rotation is sharper than the totals suggest. FDI into banking fell from $898 million in FY23 to $115 million in FY25 - an 87 per cent drop. Software and hardware's share of inflows fell from 44 per cent in FY21 to 14 per cent in FY25. Renewable energy is the bright spot, with FDI up about 50 per cent in a year. The composition is telling foreign investors a story about where they think Indian returns will be earned, and where they will not.

A reinvestment problem, not an entry problem

The instinct in Delhi will be to read this as a confidence problem and respond with another round of FDI cap relaxations. I think the diagnosis is wrong. India does not have an entry problem. It has a reinvestment problem.

Listed multinationals are now doing what corporate finance textbooks tell them to do. After 2021, several foreign businesses listed their Indian subsidiaries on local exchanges; a large slice of the capital raised was promptly sent home. Indian equities at roughly 22 times forward earnings versus 13.6 times for the MSCI emerging markets index are a structural invitation to take chips off the table. India is now expensive enough that the rational move for a foreign owner, after a good run, is to sell some down.

Meanwhile, the chief economic adviser has publicly observed that Indian private firms are not stepping up capex in proportion to their profitability. The two trends - foreign owners harvesting, domestic owners deploying capital abroad - are not separate stories. They are the same story told twice. The marginal rupee of profit, foreign or Indian, is finding it more attractive to leave than to build the next plant here.

What might actually move the needle

From inside a national tax administration, a few things become obvious that do not always show up in market commentary.

Reinvested earnings deserve a separate, named regime. An MNC parent paying tax on dividend repatriation under treaty rates faces no real incentive structure that distinguishes "I am taking this money home" from "I am ploughing it back into a new line here." A modest tax credit, or a lower effective rate for verified reinvestment into greenfield capacity, would be fiscally cheap and signal-rich. Prof. Richard Robb's International Capital Markets course at Columbia drilled in a point that still travels well: capital is taxed at the margin where it can move, and small wedges decide whether it stays.

Certainty pays more than concession. Repeated assurances on tax stability matter only if assessment behaviour at the field level matches the rhetoric. The reinvestment call is made in a boardroom in Tokyo or Seoul by someone reading not just the statute but twenty years of dispute outcomes. A measurable improvement in dispute closure timelines is worth more to that boardroom than another headline rate cut.

India can keep celebrating gross inflows, or it can begin measuring what actually builds capacity. The honest scoreboard is net FDI plus retained earnings reinvested in the country. Until that number recovers, every "highest ever FDI" headline is, with respect, a vanity metric.

#FDI #IndianEconomy #PublicFinance #CapitalMarkets #Macroeconomics #ForeignInvestment #Reinvestment

After the Carve-Out, India's Window

Reporting in the New York Times on Friday put a hard number on something tax administrators around the world had been quietly calculating since January: roughly forty billion dollars in US corporate income tax has gone unpaid since the start of 2025 because the Trump administration walked out of the global minimum tax negotiation. Thermo Fisher Scientific alone reportedly shaved $3.5 billion off its tax bill via Malta. American Express, PayPal, Pepsi and others routed earnings through Cyprus, Bermuda, Switzerland and the Cayman Islands. The US Treasury took the position on day one of the second Trump term that the OECD's Pillar Two had "no force or effect" in the United States, and in January the OECD blinked.

That is the bare news. The harder question is what countries like India should do now.

What the side-by-side actually did

On 5 January 2026 the OECD Inclusive Framework released the Side-by-Side package. It is not a withdrawal of Pillar Two. In the polite language of international tax it is a structural exemption: US-headquartered multinationals are deemed to satisfy the global minimum tax standard because they are subject to the rebadged GILTI regime, now called Net CFC Tested Income, or NCTI. The Income Inclusion Rule and the Undertaxed Profits Rule will not be applied to them by other jurisdictions for fiscal years beginning on or after 1 January 2026.

The problem, which technical readers know and policymakers are still digesting, is that NCTI does not work the way Pillar Two does. NCTI permits global averaging across jurisdictions. Pillar Two requires a country-by-country effective tax rate at the 15 per cent floor. The two regimes differ by design. Calling NCTI "robust enough" to substitute for Pillar Two is a polite legal fiction, and the empirical evidence is now in. Forty billion dollars in fifteen months is not a rounding error.

India's seat at the table, and the larger irony

India was not a passive bystander in any of this. The OECD's own implementation handbook on the global minimum tax was prepared under the Indian G20 presidency. The G20 finance ministers' communiqué from Rio in July 2024 explicitly urged all jurisdictions to implement the minimum tax. India quietly withdrew its 2 per cent equalisation levy in August 2024 to align with Pillar One and Pillar Two, sacrificing a domestic revenue stream as the price of multilateral consensus.

Eighteen months on, the multilateral consensus has a hole the size of US corporate income in it, and India has neither the equalisation levy nor a domestic minimum tax. That is a bad place to be.

The QDMTT question is now the only question

Within the Pillar Two architecture there is a quietly useful instrument: the Qualified Domestic Minimum Top-up Tax, the QDMTT. It allows a country to impose its own top-up tax on undertaxed profits earned by multinationals operating within its borders, so that the revenue accrues domestically rather than to the parent country. The OECD's own Side-by-Side text reinforces QDMTTs as the principal mechanism for protecting local tax bases, particularly in developing countries.

For India, the case for a QDMTT was always strong. With the US carve-out, it has become close to obvious. If a US-headquartered multinational operates a significant capability centre or manufacturing presence in India and pays an effective rate below 15 per cent because of incentives or transfer pricing arrangements, the choice is straightforward. Either India collects the top-up itself, or it leaves the difference on the table. Under the side-by-side, no other jurisdiction will collect that top-up from a US parent. The money simply disappears.

The Income-tax Act, 2025, which came into force on 1 April this year, is the cleanest vehicle in which to introduce a QDMTT. The Act was designed as a generational rewrite. Slotting in a domestic minimum tax now, while the architecture is still fresh and the rules are being shaped, is administratively far easier than retrofitting later.

A closing argument from inside administration

The instinct of every tax administration when global rules get watered down is to wait and see. That instinct is wrong here. The signal from January is not that Pillar Two is dying. It is that Pillar Two will survive in a fragmented form, and the countries that move first on domestic top-ups will keep their taxing rights intact. Those that wait will discover, two budget cycles from now, that revenues have leaked through a perfectly legal door.

There is a quieter argument as well. India's credibility in international tax forums was built on two decades of substantive contribution. Walking away from that posture because the United States walked away first would be the wrong lesson. The right lesson is that multilateralism, when it cracks, is best repaired by countries who take the rule and apply it at home, not by countries who wait for the rule to be restored from above.

A domestic minimum tax of fifteen per cent on large multinationals, implemented through the new Act with the standard €750 million revenue threshold, would do four things at once. It would protect the domestic tax base. It would signal continuity of India's commitment to the BEPS framework. It would give Indian-headquartered groups a level playing field with US-headquartered ones. And it would, frankly, put forty billion dollars back on the table as a question other countries have to answer for themselves.

The carve-out is done. The window for the response is open. The next budget is the place to take it.

#IncomeTax #PillarTwo #BEPS #GlobalMinimumTax #IndianTaxation #PublicFinance #InternationalTax #CBDT

Three Cents On The Dollar

Three cents.

That is roughly what one extra dollar of AI model spend is replacing, in routine knowledge work, at the firms most exposed to the shift. Among the most AI-exposed companies, each $1 drop in spending on online labour marketplaces corresponded to only about three cents in added AI model spend by the third quarter of 2025. The ratio comes from firm-level payments data analysed by Ryan Stevens of Ramp, and it has been doing the rounds in public-sector commentary this week.

If that number holds, the implication is not that AI is ‘cheaper’ in the comforting, marginal way most agency notes describe it. The implication is that the market price of a meaningful slice of first-pass, reviewable knowledge work has collapsed by roughly an order of magnitude. That is not a procurement story. It is a public finance story, and we should treat it as such.

What the 33-to-1 ratio actually says

A 33-to-1 substitution ratio is not, strictly, a substitution ratio. It is a ratio between two budget lines, captured at one moment, in one slice of the economy, by one provider’s payments data. Read with care, it tells you three things.

One: buyers have found a class of output where the first pass is now an AI deliverable rather than a human one, and the first pass is the part they used to outsource. Two: the unit cost of producing that first pass has fallen far enough that even a sceptical CFO is rewriting last year’s contractor budget. Three, and most important: the saved spend is not, mostly, being reinvested in a more expensive in-house model. It is leaving the books altogether.

For a government department, that last point matters most. In the private sector, ‘leaving the books’ eventually shows up as somebody’s lost job. In the public sector it shows up - or refuses to show up - as a backlog that does not grow, a clearance cycle that does not lengthen, a sanctioned post quietly left vacant. The change is invisible until you go looking for it.

Why government cannot just copy the ratio

It would be a mistake to read this number and rush to swap people for tokens inside a tax administration. Public-sector value is not what the private sector buys when it buys ‘first-pass output’. From inside a national tax administration, one learns quickly that the value of a routine notice is not in its first draft. It is in the fact that the draft is recorded, attributable, reviewable and defensible six years later under cross-examination.

None of that vanishes if AI does the drafting. But it changes who is accountable for what, and at which moment in the workflow. Records-management law, natural justice, statutory limitation, the discipline of audi alteram partem - these are not procurement constraints. They are the architecture of the trust the department holds with the citizen. Any honest AI strategy in this domain begins there and works backwards to the model, not the other way around.

The serious question for the head of any large administration is therefore not ‘where can I deploy a model?’ It is: where in my workflow is the first pass currently a bottleneck, who reviews it, and what does that review actually verify? If the review is genuine - a senior officer reading, weighing, signing - AI is a gift; the bottleneck moves to where it belongs, which is judgment. If the review is rubber-stamping, AI will reprice that work to zero and the institution will not notice until the courts do.

Where the repricing is already real

The categories where this repricing is real will be familiar to anyone who has worked the field. Drafting of routine intimations and standard show-cause notices. Cross-referencing returns against third-party information. Summarising voluminous assessment records before a hearing. Preparing first-pass replies to grievance petitions. Tagging and classifying taxpayer correspondence. Translating dense statutory text into plain-language guidance for citizens.

Each is a place where the marginal cost of a first pass has fallen sharply. Each is also a place where a human signature still ought to mean something. The design problem is precisely how to keep the signature serious while the draft becomes cheap.

Three operating shifts that would actually help

• Measure the bottleneck, not the licence. An honest dashboard tracks clearance cycle times, refund-issue latency, time-to-first-response on grievances. If those are not moving, the model is decorative.
• Promote the reviewer. Where AI does the first pass, the human downstream should be more senior, not less. The saving is bought by elevating the reviewer’s standard, and paying her for judgment rather than throughput.
• Budget for the audit trail before the model. Every AI-assisted output in a tax administration must carry a verifiable record of what the model saw, what it produced, what the officer changed, and why. The audit trail is the primary product. The draft is secondary.

The budget question we should be asking

Michael Ting’s course on public sector organisations at Columbia used to insist that the deepest puzzle in bureaucratic design is not how to motivate effort but how to allocate scarce attention. The Ramp number, read honestly, is a story about attention, not effort. Cheap first passes free attention; what an institution does with that freed attention is the whole game.

The danger a department most needs to guard against is the most boring one: cashing the saving without rebuilding the work. If sanctioned posts are quietly left vacant, if the freed hours are absorbed by more of the same routine work, and if the audit trail is patched on at the end, then the 33-to-1 ratio will become the public sector’s number too - and we will have repriced our own knowledge work without ever asking what we wanted to do with the difference.

The better view, I think, is to treat this moment as a budgeting exercise, not a tooling exercise. Ask, for every freed officer-hour, what higher-value question we now want that officer to be asking. Then build the model around the answer. The cheapest first pass in history will be wasted on the same old second pass.

When The Rich Quietly Quit Dollars

UBS has just published its 2026 Global Family Office Report, and the number to underline is 60. Sixty percent of family offices are planning the biggest strategic changes to their portfolios in five years. Globally, North America is the only region they intend to cut.

The signal in family-office money

These are not retail flows. Family offices have no clients to redeem on a bad month, no benchmark to hug, no consultant grading them quarterly. They have the longest discretion in private finance. When that money rebalances, it is rarely noise.

The detail under the headline is sharper than the headline itself. Two-thirds expect confidence in the dollar’s reserve role to fall. Nearly half say they are already overexposed to the dollar. The Swiss franc and the euro are the preferred diversification currencies. Emerging-market equities, infrastructure and gold get a top-up. The first-ranked risk for both the next twelve months and the next five years is geopolitical uncertainty.

This is not a BRICS communique or a yuan-internationalisation press release. It is balance-sheet behaviour from the deepest patient pools in private wealth, and it deserves to be read as such.

The Hormuz crucible

The timing is not coincidental. We are in the third month of the Strait of Hormuz crisis. ORF’s input-output modelling places the structural CPI ceiling for India somewhere around 4.5 percent on this shock. MUFG’s adverse-scenario USD/INR sits above 95.

More interesting than the price is the plumbing. Roughly 60 million barrels a month are reportedly settling in yuan and dirhams under wartime arrangements. That is not large next to the global oil trade. But it is the first time in this cycle that non-dollar settlement rails have moved real volume under stress, and importers have, for the first time in years, seen the cost of dollar-denominated energy clearing exposed as a strategic vulnerability rather than a piece of cheap plumbing.

India: beneficiary and victim at once

There is a comforting reading of the UBS data for Delhi. Money rotating out of North America flows naturally into emerging-market equities, and India sits at the top of that allocation pile. The eighteen rupee-invoicing arrangements the Reserve Bank has quietly enabled with trading partners are exactly the rails the world is now testing under fire.

The discomfort is that the same regime is squeezing us simultaneously. Foreign portfolio investors pulled over a billion out of Indian equities in the first four months of 2026. The current-account arithmetic is being held together by intermittent yuan-priced crude cargoes and a slowly bleeding reserves stock. Anyone telling you that a multipolar monetary order is unambiguously good news for India is selling something. It is a structurally improved bargaining position with a near-term liquidity risk. Both things are true; the policy has to address both.

Four moves that would actually matter

Commentary at this point usually retreats into “deepen capital markets”. The list that moves the needle is shorter and more specific.

Operationalise rupee invoicing properly. The Gulf bottleneck is not policy; it is correspondent banking economics and Vostro account uptake. If an Indian refiner finds it cheaper to settle a UAE crude cargo in INR than in USD, the share moves on its own. Today the friction runs the other way, and no announcement fixes that.

Resize the strategic petroleum reserve. Roughly 45 days of cover is acceptable in normal times. It is thin in a Hormuz regime, where supply losses since February have already exceeded a billion barrels globally. We can either top it up with leased commercial space and term contracts, or this becomes a recurring crisis the RBI is quietly asked to fund out of forex reserves.

More sovereign gold. Emerging-market central banks have absorbed roughly 225 million ounces of gold since 2008, and still hold about half the physical gold of advanced economies. India is underweight by every reasonable benchmark. The argument that buying more is atavistic is the wrong frame: this is now the most consensus trade in central banking, and the data has stopped being shy about it.

Clean the rails for inbound capital. If we want the rotation into Indian equities and rupee bonds to be sticky, settlement, repatriation timelines and tax certainty for non-residents must be visibly frictionless. A family office choosing between Mumbai and Rio de Janeiro (where i did my Columbia Capstone) will not decide on yield alone. It will decide on execution friction. We control that variable entirely.

The takeaway

One framing from Kent Daniel’s capital-markets course at Columbia has stayed with me longer than the lecture notes: in any market, the patient money tells you where the cycle is going long before the loud money does. The patient money right now, very politely through a UBS survey and very impolitely through gold, is saying the post-1971 dollar order is being repriced. We have something like the next decade to position for that. The Hormuz crisis is the first real stress test. If the policy answer is reduced to “buy more Russian crude”, we will have wasted the lesson.

Frameworks Don't Move Rocks

The photograph from Hyderabad House this week is the kind diplomacy is good at producing: two foreign ministers, two folders, two signatures, one framework. The actual constraint on India's rare-earth ambition, though, is not the absence of a framework. It is the absence of plants.

India and the United States on Tuesday signed a bilateral framework aimed at securing the supply, mining, and processing of critical minerals and rare earth elements. The scope is wide: the framework seeks to deepen cooperation across the critical minerals and rare earths supply chain, including mining, processing, recycling and related investments. Read carefully, the document concedes that mining is the easy bit.

The Gap Between Reserve And Production

India is not poor in this geology. Government estimates put the country's monazite at 13.15 million tonnes, containing roughly 7.23 million tonnes of rare earth oxides. A serious endowment by global standards. And yet India currently produces only four critical minerals — copper, graphite, phosphorous and titanium — owing to limited exploration and a lack of proper infrastructure and processing technology. The gap between what we have under the soil and what we ship out of a factory is the entire story.

Which is why the comparison everyone reaches for — China — is more sobering than it first looks. The International Energy Agency estimates China accounted for about 91% of global separation and refining production in 2024 and 94% of sintered permanent magnet production. The challenge is not about geology. It is about industrial depth and policy consistency.

Why Processing Is The Real Chokepoint

Rare earths, despite the name, are not really rare. What is rare is the chemistry plant downstream. Three features of that plant make it unusually hard to finance the conventional way.

It is dirty. Processing costs are high, and mining involves heavy use of chemicals that generate toxic waste. Environmental approvals in India are slow, contested and political.

It is long-dated. A separation-and-refining line takes five to seven years from licence to first commercial yield. Banks dislike the gestation. Public equity markets do not pay for it. Strategic patience is not a line item on a term sheet.

It is exposed to a single buyer's price war. In 2022, to maintain its control over the rare earth market, Beijing increased REE processing by 25% to lower global market prices, causing foreign producers to limit or even halt production. That is the memory every private board has when an Indian processing proposal is put before it.

What The Framework Cannot Do By Itself

India's 2026-27 Union Budget introduced plans for ""rare earth corridors"" in Odisha, Kerala, Andhra Pradesh and Tamil Nadu to support mining, refining, research and magnet manufacturing. ""Corridor"" is a useful planning vocabulary. It is not, by itself, a financing instrument.

Under the Quad framework, governments and private companies are expected to mobilise up to $20 billion through loans, guarantees, subsidies and long-term purchase agreements. Twenty billion sounds large until you remember it is split across four countries and several links of a global chain. India's share, on its own, will not move a single tonne of separated neodymium to a port.

What Would Actually Work

The interesting question is not what the framework says. It is what the Indian state's response to it should look like. Three instruments, all sitting in tools we already own.

1. Use The Tax Code As Strategic Patience

The tax code is the cheapest and most flexible instrument the state owns for shaping long-gestation capex. A targeted package — accelerated depreciation on rare-earth processing assets, an investment-linked deduction with a long carry-forward, a concessional rate on income from notified critical-mineral output — would change the IRR of a separation line more than any subsidy headline. Having watched, from inside, how one statutory regime gives way to another, I am wary of overpromising what an incentive can do. But a calibrated incentive for an industry whose cash flows will not be linear is one of the few honest uses of the tool.

2. Sovereign Offtake At A Floor

The single most important thing the framework enables is the construction of guaranteed demand. The American model is instructive. The Pentagon took a $400 million equity stake in MP Materials last July, the first investment of its kind in Pentagon history, including a guaranteed floor price for some of the company's output and a ten-year commitment to purchase magnets from its planned Texas facility. A floor price from a sovereign buyer is worth more than a soft loan, because it survives a Chinese price war. India's defence, railways, renewables and an eventual critical-minerals reserve can together write that kind of contract. The framework gives us the scaffolding. Procurement has to use it.

3. A Separate, Time-Bound Clearance Track

Environmental clearances for notified processing units should sit on their own track with public timelines and named accountability. Uncomfortable to say in print. Also the difference between a list of corridors and a list of plants.

The Honest Measure

Rare earths are the small, dense node where industrial policy, foreign policy and tax policy meet. The Indian instinct — sign a framework, announce a corridor, wait for FDI — is unequal to the problem. Every successful rare-earth processor today exists because a patient combination of state capital, state demand and state tolerance for environmental cost held the project together long enough for unit economics to mature. Japan did it after 2010. The United States is doing it now. China did it for forty years.

The agreement signed in Delhi gives India a partner, a forum and a public commitment. What it does not give us is a single tonne of separated neodymium. That has to be built — and the building is mostly a domestic exercise of fiscal patience, regulatory courage and procurement discipline. The honest measure of this deal two years on will not be the count of MoUs that followed. It will be the tonnes of magnet-grade output the country ships in 2028.

The Three-Cent Government Hour

A small number in a payments-data paper has been bothering me. Among firms most exposed to generative AI, every $1 fall in spending on online labour marketplaces by the third quarter of 2025 was matched by roughly three cents of added AI model spending. Three cents where a dollar used to sit. That is not a productivity nudge. It is a repricing of a whole category of work.

What the ratio actually says

The number comes from a Ramp analysis picked up in commentary on US federal AI policy. The exposed firms were not abandoning the work. They were buying the same first-pass output — drafts, summaries, light code, document review — at a fraction of the prior price. A slice of knowledge work has moved from a labour line item to a software line item, with the ratio between the two collapsing by more than an order of magnitude.

For private firms, this plays out through hiring and margins. For governments, it plays out through almost everything: workforce composition, training pipelines, procurement rules, and the implicit deal that lets young officers grow into senior ones.

Why tax administrations sit in the bullseye

Walk through any direct-tax office and ask what the work actually is. Reading. Drafting. Summarising. Comparing one provision against another. Translating dense statute into plainer language. Spotting the inconsistencies between a return, a third-party report and a bank statement. Almost the entire stack is language-heavy. That is exactly where the three-cent ratio bites hardest.

I have watched this from the inside. Mapping an old direct-tax statute against a new one — tracking which old section migrates where, what is dropped, what is reorganised — is genuinely difficult professional work. A few years ago that effort would consume dozens of officers for months. A current-generation model, given the right corpus and a careful prompt, will now produce a competent first draft of much of it in an afternoon. Not the final word. But a credible first pass.

The same is true of taxpayer-facing communication. Building an assistant that can field lakhs of routine questions on a new law — what the slabs are, how to elect a regime, what to fill where — was, until recently, a serious capital and talent project. Today the floor for that capability has fallen sharply. The hard part is no longer building the bot. The hard part is governance: what it is allowed to say, how its mistakes are caught, how a taxpayer appeals an answer that turned out to be wrong.

The junior officer problem

A Stanford working paper this year found a 16 per cent relative employment decline for workers aged 22 to 25 in the occupations most exposed to generative AI. Read that and a managerial instinct should fire. In any large department, the junior cadre is not just there to do work. It is there to learn. The years spent reading scrutiny files, drafting orders, sitting through hearings — those years are how a tax officer becomes a tax officer.

If the model does the first draft, what does the junior do? The wrong answer is: nothing, until they are senior enough to “supervise” the model. A supervisor who has never written the draft cannot meaningfully review one. Within a decade we would be running an administration whose middle ranks know how to prompt AI for a note but cannot tell when the note is quietly wrong on a point of law.

What the junior role should become

I think the redesign is closer to this. The junior officer is no longer the drafter. They are the evaluator, the contester, the case-builder. From day one they are taught to interrogate a machine-generated draft: where is the citation, is the authority current, does the inference survive cross-examination by a contrary view. They are taught to construct the hard cases the model gets wrong, and to document them. They become the institution’s quality control function rather than its typing pool. That is more demanding work, not less. It suits the calibre of people the service actually recruits.

A proposal worth piloting

If a tax administration wants a concrete way in, here is one. Pick a single, well-bounded workflow — say, drafting routine rectification orders, or first-level responses to grievance petitions. Build a model-assisted pipeline with three deliberate seams: a machine first draft, a structured human evaluation against a checklist, and a logged audit trail of every override. Track three numbers — time saved per case, override rate, and downstream litigation outcomes for the cases that went out. Run it for a year. Publish the numbers.

This is unfashionable advice in a moment that prefers headline pilots and grand strategies. The three-cent ratio is not going to wait for a strategy document. It is already changing what a dollar of knowledge work buys in the market. A serious department asks what that means for its own internal economics, builds the governance to capture the gain safely, and quietly redesigns its junior roles before the redesign happens to it.

The departments that get this right will not be the ones that bought the most expensive tools. They will be the ones whose officers learned, early, to argue with a machine and win.

After Aadhaar, the Productivity Test

On 28 April 2026, NITI Aayog and the Frontier Technology Hub released a roadmap called DPI@2047. The Chief Economic Adviser described India's next phase of digital public infrastructure as a total factor productivity engine. That phrase is doing more work than it looks. It marks the end of one era and the beginning of a harder one.

Reach was the easy part

The first decade of DPI was a reach problem. Could we give 1.4 billion people a verifiable identity? Could we route real-time payments at population scale? Could we plug welfare into a single pipe? JAM, UPI, DigiLocker and GeM answered yes. DPI@2047 quietly concedes that the welfare-delivery question is settled, and sets a different one. The new task is not to reach the citizen; it is to lift what the citizen, the small firm and the small farm can actually produce.

That is a different test, and most departments are not yet measuring themselves against it. The roadmap proposes eight sectoral transformations across MSMEs, agriculture, education, health, credit, energy and social protection, with a state-led, district-executed model and pilots from 2026-27. Notice the language: district-executed. The unit of accountability is shifting downwards. Productivity, unlike inclusion, cannot be claimed in a press release; it has to show up in somebody's actual output.

Why most departments will misread this

The instinctive reading of DPI 2.0 inside government will be: more APIs, more dashboards, more apps. That misreads the brief.

Phase one of DPI worked because it disentangled the rails, identity, payments and data, from the application layer. Phase two is being asked to do something subtler. It must disentangle existing work from existing process. A welfare benefit can move through a new pipe without changing what the benefit is. Productivity gains demand the opposite. The pipe is uninteresting; what changes is the work itself. Deloitte's Government Trends 2026 puts it crisply: the biggest gains come not from automating old processes but from redesigning the work itself. A UK trial of over 20,000 civil servants using generative AI for three months saved an average of 26 minutes a day per person, nearly two working weeks a year. That is not because the AI replaced anybody; it is because the work was finally allowed to be done differently.

Inside Indian government, I have watched well-meaning officers turn a transformative tool into a faster version of the form it was meant to replace. The form persists, the discretion persists, the file persists. Faster, but unchanged. DPI 2.0 will succeed or fail on whether departments are willing to give that comfort up.

Three moves a department should make immediately

Stop digitising forms; redesign the file. The unit of bureaucratic work in India is the file. Every project I have seen that put a digital wrapper around an unchanged file reproduced the same delays in colour. Pick three high-volume work-streams, write down what an ideal file looks like with AI-assisted drafting embedded in it, and re-engineer backwards. The form is downstream; the file is upstream.

Build agents that draft, not bots that retrieve. Most public-sector AI in India today is a chatbot that finds a circular. That was the right starting point. It is now the ceiling. The gain lies in agents that draft an order, prepare a notice, summarise a representation, and present a ready-to-sign output. The officer reviews and decides; the typing is gone. We are using a five-times leverage tool as a 1.2-times search tool, and calling it transformation.

Measure officer-minutes, not transactions. If DPI 2.0 is about productivity, the metric must be productivity. Most dashboards still count transactions: files moved, returns filed, calls answered. None of that tells you whether the work got lighter. The metric that should matter, and that nobody is asked to report, is officer-minutes saved per case. A department that reports this number will, within two quarters, look very different from one that does not.

The harder test is institutional will

The CEA said something else at the launch that has not been quoted enough: India has strong design capabilities, but success will depend on sustained institutional will to move from strategy to execution. That is the polite version of the real problem. Indian bureaucracy is excellent at announcing platforms and indifferent at reorganising work around them. DPI 1.0 succeeded partly because it was built outside the line department, on rails the rest of government had to either ride or be left behind by. DPI 2.0 is being handed to the line departments themselves.

So the productivity test is, in the end, a leadership test. Which Secretaries will decide that their teams write fewer pages, sign fewer files and answer fewer queries by the end of next year? Those are the departments where DPI 2.0 will arrive. The rest will get a new portal.