Skip to content
Reach Out to Us

Articles

rubyworks-image-36

Decision Velocity: How the Finance Function Is Being Rebuilt

0

For most of the last decade, finance transformation meant a faster close and cleaner data. In 2026 the question has changed. Boards are no longer asking how quickly finance can report what happened. They are asking how quickly the business can decide what to do next, and whether finance is speeding that up or slowing it down.

The Short Version

  • AI use is common, value is not. Gartner finds 59% of finance teams use AI, barely changed in a year, and Deloitte finds only 21% of full deployers see tangible value so far.
  • The bottleneck has moved. Data and the close are getting faster. The slow part now is the path from insight to a committed decision.
  • The team is being reshaped, not just shrunk. CFOs expect fewer junior roles and more mid-level analysts, turning the finance pyramid into a diamond.

What Decision Velocity Means

Decision velocity is the time between a signal appearing in the business and a committed, funded action in response. A price move, a supplier failure, a demand shift or a currency swing all start the clock. The clock stops when someone with authority commits money or changes a plan.

It is different from reporting speed. A company can close its books in three days and still take six weeks to reallocate a budget. Four measures capture it better than any single close metric:

  • Days to close: how quickly actuals are trusted. The average is 6 to 7 business days, and top performers close in 4 or fewer (Ventana Research; PwC, via Eagle Rock CFO).
  • Forecast refresh cadence: how often the view of the future is updated, and whether it rolls forward or resets once a year.
  • Scenario readiness: whether the likely shocks are already modelled before they arrive, so a response is chosen rather than built.
  • Signal-to-action time: how long it takes from a flagged variance to an approved change. It is rarely tracked as a formal metric, which in our view is often why it is the slowest step.

Why Now

Three forces have turned speed from a nice-to-have into the CFO’s main brief.

  • Volatility has become the base case. In Gartner’s early-2026 survey of 100 CFOs, responding to unpredictable market conditions ranked among the two hardest priorities for the next six months, alongside building AI talent. In India alone, September 2026 brought the Nifty’s worst September since 2018, heavy foreign selling and oil above $100, the kind of month where last quarter’s plan stops being useful.
  • CFOs are more confident, and expected to act on it. Deloitte’s CFO confidence score hit 6.6 in Q4 2025, its highest since late 2021, and half of large-company CFOs named digital transformation of finance as their top 2026 priority.
  • AI has made analysis cheap, so judgement is now the constraint. When a model can produce ten scenarios in an hour, the scarce resource is no longer the analysis. It is the decision process that has to consume it. 87% of CFOs expect AI to be very or extremely important to finance in 2026 (Deloitte Q4 2025 CFO Signals).

The Evidence: What the Latest Data Shows

AI adoption has plateaued; the value hunt has started.

  • Gartner puts finance AI adoption at 59% in 2025, against 58% in 2024 and 37% in 2023. The easy first wave is done.
  • 63% of finance organisations said AI implementation was slower than expected in 2025 (Gartner, May 2026).
  • Of companies that have fully deployed AI in finance, only 21% report tangible value so far, and just 14% have fully integrated AI agents (Deloitte Finance Trends 2026).
  • Only 8% of CFOs have deployed AI tools or agents at scale; 70% are still planning or piloting (Oliver Wyman Forum and NYSE CFO Survey 2026, 494 CFOs).

Quick wins are productivity wins.

  • Data extraction, payables and receivables automation, and report creation typically pay back in nine to ten months (Gartner, September 2026). Useful, but none of them changes a decision on its own.
  • Gartner now names low AI literacy, not hiring, as the biggest barrier to finance AI value.

Agents are the next priority.

  • 54% of large-company CFOs say embedding AI agents in finance is a top 2026 transformation priority (Deloitte Q4 2025 CFO Signals).
  • 44% already use AI for planning and budgeting and 41% for analysing financial data (Deloitte Q2 2026 CFO Signals).

Planning is still slower than the rhetoric.

  • Even among self-rated mature FP&A teams, only 11% use agile methods such as rolling forecasts, and many struggle to turn insight into swift action (Vena 2026 FP&A Impact Report, 400+ leaders).

The team shape is changing.

  • 64% of CFOs expect to shift away from junior finance roles; 41% expect more mid-level roles and 23% more senior roles. 61% expect headcount to stay broadly flat (Oliver Wyman Forum and NYSE).

The Real Shift: Five Insights

1. The bottleneck has moved from data to decision rights. A decade of spend went into the close, the data layer and analytics. Those steps are getting fast. The slow step now is the decision itself: approval chains, unclear ownership and budget rules built for an annual cycle.

Decision Velocity chart

Rubyworks analysis · the decision loop in a finance function

2. AI creates value when it is attached to a decision, not a report. The fastest-payback use cases (report creation, payables, data extraction) save hours but rarely change what the business does. That helps explain why adoption is near 60% while tangible value sits near 21%. Our view: the teams that capture value pick a recurring decision (pricing, working capital, hiring pace) and rebuild the loop around it end to end.

3. The biggest lever is deciding in advance. The fastest finance teams do not decide faster in the moment; they pre-agree responses. A scenario playbook that says “if input costs rise 10%, these three actions are pre-approved” turns weeks of debate into a same-day move. It costs little and needs no new software, which is why it is so often overlooked.

4. The diamond-shaped team has a hidden cost. Fewer junior roles means fewer people learning the business through reconciliations and variance work, the traditional path to controller and CFO. Firms that cut the base without redesigning how juniors learn judgement may create a leadership gap in five to seven years.

5. Faster decisions need faster controls. Speed without governance simply produces faster mistakes. CFOs already cite cost uncertainty, transparency and risk as their top AI governance worries (Deloitte Q2 2026 CFO Signals). Expect decision logs, model audit trails and clear agent authority limits to become core finance controls, much as reconciliations are today.

What It Means for Each Player

  • CFOs: the job shifts from owning the numbers to owning the decision process. Practical first steps are to map who can approve what and how fast, set signal-to-action targets for three or four recurring decisions, and pre-approve responses to the most likely shocks.
  • Finance teams: the valued skill is moving from preparing numbers to framing choices. The Vena survey and Gartner both point to the same gap: insight is produced, but not turned into action. Analysts who can say “here are the options and what each costs” will be in demand.
  • Boards and investors: decision velocity is becoming a quality-of-management signal. Useful questions for any company: how often is the forecast refreshed, how quickly was capital reallocated in the last shock, and which decisions have pre-agreed triggers?
  • Software vendors: the market is moving from tools that report to tools that act, with agents that draft entries, flag variances and propose actions. The winners will be those that fit inside approval workflows and audit trails, not just dashboards.
  • India’s finance centres: India hosts more than 1,700 global capability centres employing about 1.9 million people, and finance and accounting centres of excellence are often a company’s first India unit. As AI absorbs transaction work, these centres face a choice: stay processing hubs and shrink, or become the analytics and decision-support engine of the global finance function.

What to Watch: The Next 12 to 24 Months

These are our signposts, not forecasts.

  • Whether the value gap closes. If the share of AI adopters reporting tangible value moves well above 21%, the rebuild is working. If adoption keeps rising while value stays flat, expect budget cuts to AI pilots.
  • Agents with authority. The real step change comes when agents are allowed to execute within limits (post a routine accrual, adjust a reorder, release a payment) rather than only recommend. Watch for companies publishing agent authority policies.
  • The death of the annual budget, in practice. Rolling forecasts are widely discussed but rarely used. Wider adoption would be the clearest sign that finance is organising around speed.
  • Decision velocity in disclosures. Expect more companies to cite faster reallocation and scenario readiness in earnings calls, and more investors to ask about it.
  • The junior pipeline. Watch graduate hiring in finance and whether firms redesign entry roles around exceptions and judgement, or simply stop hiring at the base.
  • India’s finance centres moving up the value chain, from transaction processing into FP&A, controllership and decision support.

The Bottom Line

The finance function spent ten years making the numbers faster. The next ten will be spent making decisions faster. AI is removing the analysis bottleneck, which exposes the real one: slow approvals, unclear decision rights and plans built for a calmer world. The CFOs who win this cycle will not be the ones with the most AI tools. They will be the ones who can show, with a number, how quickly their company turns a signal into an action, and who have built the controls to do it safely.

rubyworks-image-45

Data Center Consolidation: How AI, Power and Capital Are Reshaping Digital Infrastructure

0

Ownership of the world’s data centers is concentrating quickly, and the capital behind it is getting larger. Synergy Research Group counted 575 closed data center M&A deals worth $151 billion between January 2024 and early 2026, with private equity funding 84% of that value (Data Center Knowledge, Feb 2026). In July 2026 the largest deal in the sector’s history closed: the acquisition of Aligned Data Centers at an enterprise value of about $40 billion (Aligned press release).

Consolidation in this sector is happening on three levels at once:

  • Ownership. Independent and listed operators are being absorbed into a small group of large platforms backed by infrastructure funds, pension plans and sovereign capital.
  • Capacity. Hyperscale cloud and AI companies now control 48% of global capacity, while the share held in enterprise on-premise facilities has fallen from 56% in 2018 to 32% (Synergy, Apr 2026).
  • The value chain. Buyers are reaching into power generation, electrical equipment and cooling, because access to energy has become the binding constraint on growth.

This insight reviews the 2021 to 2026 record, the forces behind it, who is affected, and what the next two to three years could bring. Reported figures are attributed to their sources; forecasts are labelled as forecasts.

The 5-Year Trend (2021 To 2026)

Closed data center M&A value set back-to-back records in 2024 and 2025, after a sharp dip in 2023, while control of capacity shifted steadily toward hyperscalers and private platforms.

Data Center Consolidation chart

Source: Synergy Research Group (closed deals). 2021 approximate. 2025 derived: $151B closed Jan 2024 to early 2026, less $73B for 2024. S&P counts $69B for 2025 on its own basis.

Synergy recorded about $50 billion of closed deals in 2021 and $52 billion in 2022, then $26 billion in 2023 and a record $73 billion in 2024 (Synergy, DCD). For 2025, S&P Global counts 113 completed transactions worth more than $69 billion, about $8 billion above the prior record on its basis (DCK).

PeriodWhat defined it
2026 to dateThe $40B Aligned deal closed in July; KKR and Singtel agreed to take full control of STT GDC; Constellation closed Calpine. No full-year tally yet.
2025Record announcements, led by Aligned. Power-linked and financing deals (Hyperion JV) grew in importance.
2024Record $73B, lifted by the ~$16B AirTrunk deal that closed in late December and large equity injections into Vantage and EdgeConneX.
2023A relative lull: value fell by half to $26B as interest rates rose sharply.
2021 to 2022Take-private wave: QTS, CoreSite, CyrusOne and Switch, each valued at $10B or more.

The ownership of capacity has shifted in parallel. Enterprise on-premise facilities held 56% of global capacity in 2018 (Synergy). Today hyperscale operators hold the largest share, and Synergy projects two-thirds by 2031.

Share of global capacity202520262031 (forecast)
Hyperscale operators44%48%67%
Non-hyperscale colocation22%20%~14% (derived)
Enterprise on-premise34%32%19%

Sources: Synergy, 2025; Synergy, Apr 2026. The 2031 colocation share is the remainder of Synergy’s stated figures.

Concentration is also visible inside the hyperscale segment. The number of large hyperscale data centers doubled in five years to 1,136 at the end of 2024 (Synergy), and the three largest cloud providers now account for 57% of all hyperscale capacity (Synergy).

Key Drivers

Five forces are raising the minimum scale needed to compete, and each one favours owners with deep capital and secured power.

DriverKey evidenceWhy it favours consolidation
AI workloadsJLL expects AI to be about half of all data center capacity by 2030; AI training sites need roughly 10x the power density of traditional facilities (JLL). McKinsey sees AI demand reaching 156 GW of a 219 GW total by 2030 (McKinsey).Campuses of 100 MW to multiple GW require balance sheets and power access that few independents have.
Hyperscale capexAlphabet, Amazon, Meta and Microsoft entered 2026 guiding to close to $700 billion of combined capex, more than 60% above 2025 (CNBC).A few very large tenants now set demand, so operators need scale to win and serve their leases.
Power constraintsData centers used about 415 TWh in 2024 (about 1.5% of global electricity); the IEA base case is about 945 TWh by 2030 (IEA). JLL cites grid connection delays of up to four years (Data Centre Magazine).Energised sites command higher valuations (S&P via DCK); buying a powered platform is faster than waiting in a grid queue.
Capital requirementsAverage global build cost rose from $7.7 million per MW in 2020 to $10.7 million in 2025, with $11.3 million forecast for 2026 (JLL).Synergy notes that specialist operators have been unable or unwilling to fund growth alone, so private capital has stepped in (Synergy).
Scarce ready capacityNorth American primary-market vacancy hit a record low of 1.4% in H1 2026, and 80.4% of the 7,481 MW under construction was already preleased (CBRE via DCF).With little space to lease, acquiring existing platforms and pipelines is the quickest route to growth.

The common thread is that the scarce input has shifted from buildings to energised land and capital. That shift rewards large owners and pushes smaller ones toward a sale, a merger or a joint venture.

Market Activity

Deal sizes have stepped up sharply: the four largest deals before 2024 were each about $10 to $15 billion, AirTrunk set a new record at $16.1 billion in 2024, and Aligned more than doubled it at about $40 billion. The table lists the landmark transactions of the period, newest first.

DateTargetBuyer(s)Value (US$)Why it matters
Pending, H2 2026DigitalBridgeSoftBank Group$16.00 per share in cash (reported at about $4B)Consolidation of the asset managers themselves; DigitalBridge also agreed to buy power investor ArcLight for up to $1.05B
Jul 2026 (closed)Aligned Data CentersAIP, MGX, BlackRock’s GIP~$40B EV, plus $5B growth capitalLargest data center deal ever; about 5 GW of operating and planned capacity across the Americas
Feb 2026STT GDC (remaining 82%)KKR, SingtelS$6.6B cash; S$13.8B (~$10.9B) EV2.3 GW platform across 12 markets in Asia Pacific, the UK and Europe
Feb 2026atNorthEquinix, CPP Investments$4.2BEquinix’s largest acquisition; adds a Nordic pipeline of over 1 GW
Jan 2026 (closed)CalpineConstellation Energy~$21.8B purchase pricePower-sector consolidation aimed at data center demand; 55 GW combined fleet
Oct 2025Hyperion campus JVBlue Owl (80%), Meta (20%)~$27B developmentHyperscaler moves a single campus off balance sheet using private credit
Oct 2025 (terminated)Core ScientificCoreWeave$9B all-stockRejected by Core Scientific shareholders; shows limits of neocloud roll-ups
Dec 2024 (closed)AirTrunkBlackstone, CPP InvestmentsA$24B (~$16.1B)Then the largest deal ever; lifted 2024 to a record
2023Compass DatacentersBrookfield, Ontario Teachers’~$5.5B (reported)Infrastructure funds buying scaled US developers
2022SwitchDigitalBridge, IFM$11BTake-private of a listed operator
2021 to 2022CyrusOneKKR, GIP$15BTake-private; closed Q2 2022
2021CoreSiteAmerican Tower$10.1BTower owner diversifies into data centers
2021QTSBlackstone$10BTake-private of a listed data center REIT

Four patterns stand out across these deals.

  • Private capital is the dominant buyer. Private equity accounted for 54% of closed deal value in 2020 and 65% in 2021, and has stayed in the 80% to 90% range since (Data Center Frontier).
  • Listed operators went private, then private platforms changed hands. The 2021 to 2022 wave took QTS, CyrusOne, CoreSite and Switch off public markets. The 2024 to 2026 wave is largely fund-to-fund, with Macquarie selling both AirTrunk and Aligned.
  • The managers are consolidating too. BlackRock bought GIP in 2024, Blue Owl bought IPI Partners for about $1 billion (Data Center Frontier), and SoftBank has agreed to buy DigitalBridge.
  • M&A is moving into power and equipment. Beyond Constellation and Calpine, Vertiv agreed to buy microgrid specialist Utility Innovation Holdings for $1.45 billion and Flex agreed to buy EPC Power for $4.4 billion (Data Center Richness).

Industry & Economic Impact

Consolidation is shifting bargaining power toward owners of powered capacity and pulling energy, finance and local governments into the data center story.

GroupWhat is changingEvidence
Data center operatorsScale and hyperscale tenancy drive value; enterprise-focused and sub-scale operators face pressure to sell or partner.S&P finds operators serving hyperscalers command higher valuations than enterprise-focused peers (DCK). Synergy says the fastest-growing colocation players are nearly all private.
Hyperscalers and AI firmsMoving from self-funded builds to external capital and leasing.Incremental debt rose from 9% of hyperscaler capex in FY24 to 32% by mid-2026, and FY26 free cash flow is expected near zero or negative for all but Alphabet and Microsoft (FactSet).
NeocloudsGPU-as-a-service providers lose their chip-access edge as supply improves.S&P expects consolidation within the segment and possible acquisitions by larger operators (DCK).
Private equity and infrastructure fundsNow the main owners of third-party capacity, with larger and longer-dated vehicles.JLL expects global data center core fund formation could top $50 billion in 2026, targeting returns of 10% or more (JLL). MGX closed its first fund at $49 billion (MGX).
Energy and utilitiesData center load is reshaping generation M&A and long-term contracting.Constellation’s Calpine deal was paired with 380 MW power agreements for CyrusOne campuses in Texas (Constellation 10-Q). The IEA expects natural gas generation for data centers to rise by about 175 TWh by 2030 (IEA).
Equipment and supply chainCooling, power conversion and grid equipment makers are acquiring to capture the build-out.Vertiv, Flex and others have made billion-dollar power acquisitions (see Market Activity). McKinsey allocates about $1.3 trillion of its $6.7 trillion estimate to power, electrical and network infrastructure (McKinsey).

Capital markets. Data centers have become a mainstream fixed-income asset. Combined data center ABS and CMBS issuance reached about $26 billion in 2025, more than 10 times the 2020 level (Impax). Barclays puts outstanding data center securitizations at $61 billion in 2026 year to date, about 12% of the esoteric ABS market, up from 3% in 2020 (SFA). The Meta and Blue Owl Hyperion financing was described by bookrunner Morgan Stanley as the largest private debt securities offering ever completed (Nonbillable).

The wider economy. McKinsey estimates cumulative data center capex through 2030 at roughly 1% of global GDP each year (McKinsey). Local economies feel both sides: data centers supplied $1.2 billion, or 39%, of Loudoun County’s fiscal 2026 budget (Washington Times), yet the county voted in September 2026 to pause new data center applications for 12 months (DCD).

Emerging risks. Concentration creates exposure to a small set of tenants whose free cash flow is being absorbed by capex. Securitization investors flag obsolescence risk as chip generations turn over quickly (IMF ABS monitor). Community and grid pushback is also rising as a constraint on new supply.

Regional Trends

Capacity and deal value remain concentrated in the United States, but growth and new consolidation platforms are spreading to wherever power is available. Synergy finds that just 20 state or metro markets hold 60% of global hyperscale capacity (Synergy).

RegionScale and growthConsolidation signals
North America and Latin AmericaThe Americas hold about 50% of global capacity and are forecast to grow fastest, at about 17% a year to 2030 (JLL). Atlanta overtook Northern Virginia in construction in H1 2026, and Northern Virginia vacancy fell to 0.2% (CBRE). Texas hyperscale capacity grew 71% in a year (Synergy).Aligned (US plus Mexico, Brazil, Chile, Colombia), Constellation and Calpine, and the Hyperion JV. Local limits are tightening, as the Loudoun pause shows.
EuropeLive capacity in the five FLAP-D hubs reached about 3.8 GW in H1 2026, and prime powered land costs rose 82% since 2021 to €2.26 million per MW (JLL EMEA). JLL expects EMEA to add a relatively modest 13 GW by 2030 (DCD).Growth and capital are moving to the Nordics, Iberia and Italy, where grid connections are faster (Rabobank). Equinix and CPP’s atNorth deal is the clearest example.
Asia PacificJLL forecasts APAC capacity rising from 32 GW to 57 GW by 2030, a 12% annual growth rate (JLL).Two of the largest deals of the period are regional platforms: AirTrunk (Blackstone, CPP) and STT GDC (KKR, Singtel). Bain sold its China data center business for $4 billion in 2025 (DCD).
IndiaCapacity reached about 1.64 GW in June 2026 and JLL expects about 6 GW by 2029, after 60% annual supply growth since 2021; vacancy hit a record low of 2.8% (JLL India, Construction World).Global hyperscalers have committed over $50 billion, with nearly 30% of new capacity to be self-built. JLL estimates the build-out needs about $110 billion of capital, which points to more joint ventures and platform sales.
Middle EastJLL expects Middle East capacity to quadruple by 2030 (JLL EMEA). Stargate UAE is a 1 GW cluster with a first 200 MW phase due in 2026, and Saudi Arabia’s Humain targets about 6 GW by 2034 (The National).Gulf sovereign capital is now a global buyer, not only a local builder: Abu Dhabi’s MGX co-led the Aligned acquisition.

For investors, the regional picture points the same way as the global one. The largest platforms are being assembled in the markets where power, land and permits can be secured fastest.

Outlook: 2027 To 2029

The published forecasts all point to continued concentration, but they are forecasts, and each depends on AI demand turning into durable revenue. The table separates what has been measured from what is projected.

IndicatorLatest reported factForecast (source)
Global data center capacityAbout 103 GW (JLL)About 200 GW by 2030, needing up to $3 trillion of investment (JLL)
Hyperscale share of capacity48% in 2026 (Synergy)67% by 2031; on-premise falls to 19% (Synergy)
Data center electricity use415 TWh in 2024 (IEA)About 945 TWh by 2030 in the base case (IEA)
Data center securitizationAbout $26B issued in 2025 (Impax)More than $40B possible in 2026 (Impax); $180B outstanding by end-2028 in a Barclays scenario (SFA)
US primary-market vacancy1.4% in H1 2026 (CBRE)CBRE expects power to hold US supply down through 2030 (W.Media)
India capacity1.64 GW in June 2026 (JLL India)About 6 GW by 2029 (JLL)

Our assessment for the next two to three years (analysis, not a forecast). Based on the evidence above, we expect five developments:

  • Exits by early investors. Funds that built platforms in 2018 to 2022 are selling, as Macquarie did twice. DCD has reported that Oaktree and Elliott are exploring sales of stakes in Pure and Ark, and that EQT is seeking a buyer for GlobalConnect (DCD). More fund-to-fund sales and continuation vehicles are likely.
  • Public markets reopening as an exit. Blackstone has reportedly considered a $2 billion IPO of a data center unit (DataCentres.com, citing Bloomberg). Listings would add a second exit route alongside private sales.
  • Power-first deal-making. Transactions pairing generation, grid equipment or behind-the-meter power with data center land should keep growing, because power access sets valuations.
  • Consolidation among neoclouds and mid-sized operators. S&P expects shake-out among GPU cloud providers, and sub-scale colocation operators face rising build costs and tenant concentration.
  • More scrutiny. Antitrust review has already required Constellation to divest plants to close Calpine, and local authorities such as Loudoun are slowing approvals. Regulation may lengthen timelines and raise the premium on already-permitted sites.

What could change the picture. The main downside risks are slower monetisation of AI, which would pressure hyperscaler budgets that are already absorbing most free cash flow; higher interest rates, which hurt a sector reliant on debt and securitization; and faster efficiency gains that reduce power and capacity needs. McKinsey’s own range for AI data center capex by 2030 runs from about $3 trillion to $8 trillion, which shows how wide the uncertainty is (McKinsey).

Key Takeaways

  • Consolidation has accelerated, not peaked. Closed deal value set records in 2024 and 2025, and the largest ever transaction, Aligned at about $40 billion, closed in July 2026.
  • Private capital owns the growth. Private equity has funded 80% to 90% of closed deal value since 2021, and the asset managers themselves are now merging.
  • Power is the new currency. Energised sites and secured grid capacity drive valuations, which is pulling utilities, generators and equipment makers into the deal flow.
  • Hyperscalers are concentrating demand and changing how they fund it. Close to $700 billion of 2026 capex is pushing them toward joint ventures, leases and debt.
  • Growth is spreading, ownership is not. New capacity is moving to Texas, the Nordics, Iberia, India and the Gulf, but it is largely being built by a small group of global platforms and their financial backers.
  • Separate the facts from the forecasts. Current data show tight supply and record deal flow. The 2030 projections are credible but conditional on AI revenue, financing costs and power delivery.

Sources

Industry research

Company filings and releases

News and trade press

Figures are as reported by each source and are not independently audited. Deal values are headline or enterprise values as announced. This article is for information only and is not investment advice.

rubyworks-image-28

The SIF Effect: India’s Next Product Shift

0

India’s markets are having a hard year. The Nifty 50 fell 5.67% in September 2026, its worst September since 2018, and foreign investors have pulled ₹2.45 lakh crore out of Indian equities this year. Yet one new product has kept growing right through the turbulence: the Specialised Investment Fund, or SIF.

The Short Version

  • SIFs went from zero to ₹31,175 crore in eleven months, with 33 strategies from 17 fund houses by August 2026. Assets grew 34% in August alone.
  • Investors are buying protection, not aggression. About 69% of the money sits in hybrid strategies that blend equity, debt and limited short positions.
  • The real engine is the tax wrapper. A SIF gives HNI-style strategies mutual fund taxation, which puts pressure on PMS and Category III AIFs.
  • The category has not yet been stress-tested. The current market fall is its first real exam, and how SIFs perform in it will decide whether this becomes a lasting product shift.

What a SIF Is, in One Minute

Think of a SIF as a mutual fund that is allowed to do a few things a normal mutual fund cannot. SEBI created the category in February 2025, and the rules took effect on April 1, 2025.

  • Entry ticket: ₹10 lakh per investor, counted at PAN level across one fund house’s SIF strategies. Accredited investors are exempt. That sits between a regular mutual fund and a PMS (₹50 lakh) or Category III AIF (₹1 crore).
  • The key freedom: up to 25% of net assets can be held in unhedged short positions through exchange-traded derivatives. The portfolio always stays net long.
  • Seven strategy types: equity long-short, equity ex-top 100 long-short, sector rotation long-short, debt long-short, sectoral debt long-short, active asset allocator long-short and hybrid long-short. Each fund house can run one strategy per type.
  • The quiet advantage: a SIF is still a mutual fund unit, so it is taxed like one. Gains are taxed only when the investor redeems, not on every trade (as in PMS) or at the fund level (as in Category III AIFs).

Why Now: A Market That Rewards Hedging

SIFs arrived at the moment long-only equity stopped feeling easy.

  • Indian equities are under pressure. The Nifty 50 closed at 22,716 on September 29, 2026, nearly 14% below its 52-week high of 26,373, as high US bond yields, a strong dollar, crude prices and foreign selling weighed on the market (IndiaIPO, Sept 2026).
  • Foreign money is leaving. FPIs have withdrawn a net ₹2.45 lakh crore from Indian equities in 2026, already more than the ₹1.66 lakh crore pulled out in all of 2025 (Angel One, Sept 2026).
  • Domestic money is not. Mutual fund assets reached a record ₹87.08 lakh crore in August 2026, SIP contributions hit a record ₹32,297 crore, and equity funds logged their 66th straight month of net inflows (AMFI).
  • Investors are already seeking balance. Hybrid mutual funds drew ₹10,045 crore in August, led by multi-asset and arbitrage funds (AMFI data via Business Remedies).

Our read: this is the affluent investor’s dilemma of 2026. Many built large mutual fund portfolios during the bull run and do not want to retreat to fixed deposits, but they are uncomfortable being fully long in a falling market. The SIF offers a middle path inside a wrapper they already trust. That is why its growth has accelerated as markets have weakened, not in spite of it.

The Early Scorecard

Eleven months in, the numbers say SIFs are real, concentrated and still very young.

The SIF Effect chart

Source: AMFI monthly data as reported by Business Standard, Angel One, Sahi, Whalesbook and DSIJ. Months without a published figure are omitted.

  • It is almost all new money. Cumulative net inflows have reached ₹30,028 crore against ₹31,175 crore of assets, so roughly 96% of the category is fresh capital, not market gains (our calculation from AMFI-based reports).
  • Hybrid is the bestseller. Hybrid strategies hold ₹21,390 crore, about 69% of assets, and hybrid long-short alone holds ₹19,670 crore. No fund house has launched a debt SIF yet.
  • The buyer is the mass-affluent investor. 1,25,539 folios hold ₹31,175 crore, an average of about ₹25 lakh each (our calculation). Hybrid long-short folios average about ₹40 lakh, well below the ₹50 lakh PMS threshold.
  • Three brands dominate. Edelweiss Altiva, SBI Magnum and ICICI Prudential iSIF held about 79% of category assets in April 2026 (SIF360). Distribution reach, not track record, decided the early winners.
  • Performance is early and uneven. Quant’s hybrid long-short SIF reportedly returned 34.78% over six months to August 28, 2026 (INDmoney), while earlier equity variants were flat to negative. Almost no strategy has a one-year record.
  • It is still tiny. ₹31,175 crore is about 0.36% of mutual fund industry assets (our calculation), so the runway is long.

The Real Shift: Five Insights

The headlines focus on growth. The more important story is what SIFs change underneath.

1. This is a wrapper migration, not a new asset class. Long-short equity already existed in India through Category III AIFs. What SIFs change is the packaging: a ₹10 lakh ticket and mutual fund taxation, against a ₹1 crore ticket and fund-level tax at the maximum marginal rate. The tell is supply-side behaviour. PMS and AIF houses such as ASK, Nuvama and Marcellus sought mutual fund licences to run SIFs (Business Standard, April 2025), and the AIF industry lobbied for tax parity. When producers change their licence to follow a tax wrapper, money follows too.

2. Investors are buying a better balanced fund, not a hedge fund. Two-thirds of the money sits in hybrid strategies. In the buyer’s mind, the benchmark is a balanced advantage fund or a fixed deposit, not a global hedge fund. Managers who run aggressive equity long-short books may be solving a problem their buyers do not have.

3. Distribution is beating track record. With almost no strategy older than a year, the early leaders are the fund houses with the widest bank and wealth distribution. That concentrates assets quickly, and it raises a suitability question: a ₹10 lakh minimum filters out small investors, but it does not ensure they understand derivatives and short positions (IMPRI, 2026).

4. The 25% short cap makes SIFs smarter long funds, not market-neutral funds. Because the book stays net long, SIFs will still lose money in a sharp fall, just less. The September 2026 sell-off is the category’s first real test. The metric that matters most is net equity exposure, and investors should ask for it.

5. The fee pool is moving to mutual fund houses. SIFs let AMCs capture affluent-client assets that used to go to PMS and AIF managers. For listed AMCs, that means a higher-value product line built on their existing distribution, an under-appreciated earnings lever in a year when equity markets are weak.

Who Gains, Who Must Adapt

  • Large, bank-linked AMCs gain first. They own the relationship managers who sell to ₹25 lakh to ₹50 lakh clients, which is exactly where SIF tickets are landing.
  • PMS managers face the sharpest squeeze below ₹1 crore. A client with ₹50 lakh can now buy a comparable strategy with better tax treatment. The response is either a mutual fund licence or a move toward bespoke, higher-ticket mandates that SIFs cannot replicate.
  • Category III AIFs are the most exposed. Their listed long-short strategies now compete with a cheaper, more tax-efficient cousin. Expect them to lean into what SIFs cannot do: higher leverage, unlisted assets and fully customised structures.
  • Wealth managers get a new mid-ticket product, and new duties. SIFs fill the gap between SIPs and PMS, but advisers will need stronger suitability checks for derivatives-based strategies.
  • Investors gain access and tax efficiency. In return, they need to watch net exposure, liquidity terms (some strategies can be interval or closed-ended) and costs, and remember that SEBI requires fund houses to monitor the ₹10 lakh threshold, with redemptions triggered if an investor falls below it for reasons other than market losses.

What to Watch: The Next 12 to 24 Months

These are our signposts, not forecasts.

  • How SIFs hold up in the current fall. If hybrid SIFs lose clearly less than balanced advantage funds through this sell-off, the category earns a place in affluent portfolios. If they do not, flows could stall as fast as they arrived.
  • The first debt SIF. No fund house has launched one yet. A credible debt long-short product would open SIFs to conservative investors who never touch derivatives.
  • Whether flows come from Category III AIFs and PMS. Watch for AIF redemptions and PMS asset shifts in affluent segments. That will show whether SIFs are cannibalising or expanding the market.
  • Concentration versus new entrants. Seventeen AMCs are live and more are awaiting approval. If the top three keep most of the assets, distribution will have won over strategy.
  • Tax and regulatory tweaks. Any move to give Category III AIFs tax parity, or to tighten SIF suitability rules, would change the economics quickly.
  • The 1% milestone. SIFs are about 0.36% of mutual fund assets today. Reaching 1% would imply roughly ₹87,000 crore at current industry size (our illustration, not a forecast).

The Bottom Line

SIFs are India’s first serious attempt to bring hedged, strategy-led investing into the mutual fund wrapper, and they have landed at exactly the moment investors want protection without leaving equities. The early money is real, mostly new and heavily tilted to hybrids. The product shift will be judged not by how fast assets grow in a rising market, but by how SIFs behave in this falling one. For investors, the question to ask is simple: what is the fund’s net exposure, and what did it do in September 2026?

rubyworks-image-42

The Listed Asset Manager Effect: What Changes When the Investment Business Itself Goes Public

0

When an investment firm lists its own equity, it gains a second client with different goals. Limited partners (LPs) want net returns. Public shareholders want predictable, growing fee income. Most of what is changing in private markets in 2025 and 2026 (bigger platforms, the push into private credit, insurance and retail wealth, and a wave of manager-on-manager acquisitions) follows from how firms balance those two clients.

The scale is now material. Blackstone manages about $1.35 trillion, and Apollo and Brookfield each manage more than $1 trillion (company reports, June 2026). Blackstone, KKR and Apollo sit in the S&P 500, so millions of passive investors own a slice of the private markets industry without owning a single private fund.

This insight looks past the headline growth. It asks what actually changes inside a firm when the business itself trades daily: what it builds, what it buys, how its risks travel, and what that means for LPs, portfolio companies and markets. Reported facts are attributed to sources. Our own analysis and views are labelled as such.

Where Things Stand: 2025 To 2026

The past 18 months delivered record scale and the first sustained test of the listed model: liquidity stress in retail private funds, priced every day in the managers’ own shares.

  • Sep 2026: the credit discount persists. Private credit manager shares slid again even as redemption requests eased, with HPS redemptions down to 11.5% from 13.3% (Benzinga). The discount is outlasting the flow data.
  • Jun 2026: evergreen stress spreads to Europe. Partners Group limited withdrawals from an evergreen PE fund; its shares fell 17%, Bridgepoint 9.4% and ICG 5.4% on the day (Proactive Investors). Liquidity became a sector-wide equity risk.
  • Apr 2026: gates hit market value. Blue Owl capped two credit funds at 5% redemptions after $5.4 billion of requests, and its shares hit a record low of $7.95, 68% below the January 2025 peak (Zacks via Yahoo Finance).
  • Feb 2026: contagion through sentiment. Blue Owl permanently halted redemptions at a retail credit fund, and listed peers fell 5% to 10% intraday even without the same exposure (Morningstar).
  • H1 2026: scale keeps building. Blackstone reached about $1.35 trillion of AUM, Apollo crossed $1 trillion and Brookfield stayed above $1 trillion; Brookfield raised $67 billion by early May with trailing FRE up 18% to $3.1 billion.
  • Dec 2025: strong demand in India. ICICI Prudential AMC raised about ₹10,600 crore in an issue subscribed 39 times, and listed about 20% above its price.
  • 2025: record earnings on scale and credit. CVC grew management fees 9% to €1.5 billion on €23 billion of inflows, and BlackRock completed its purchase of HPS in July.

What the market is paying for now. Public investors pay a premium for recurring management fees: PwC estimates fee-related earnings (FRE) attract on average about three times the multiple of performance earnings. The 2026 sell-off added a second pricing factor. Investors now also discount how much of a manager’s FRE comes from vehicles that promise periodic liquidity, because that is where a gate can appear.

Key Drivers

Firms list for five reasons, and each benefit carries a cost that is rarely discussed in listing prospectuses.

Why firms list2025 to 2026 evidenceThe less visible cost (our view)
A currency for acquisitionsListed managers closed or agreed deals for HPS, Oaktree, Arctos, Coller Capital and ArcLight within 18 months.Buying other managers lifts FRE quickly but adds integration and key-person risk that LPs feel first.
Liquidity and succession for partnersBlue Owl’s shares fell 68% from their January 2025 peak by April 2026, cutting the listed wealth of its partners and staff.When partner wealth sits in the stock, short-term price moves can shape long-term investment behaviour.
Index membership and a lower cost of capitalBlackstone, KKR and Apollo are all S&P 500 members, so index funds are permanent holders.Index holders judge the firm every quarter, whatever the fund cycle.
A balance sheet for permanent capitalApollo crossed $1 trillion of AUM in 2026, with its Athene retirement business the largest component (Praxis Rock, Aug 2026).The manager starts to resemble a financial conglomerate, with insurance and credit risk inside the equity story.
Trust with retail wealth clientsBlackstone is reported to hold about half of the private wealth alternatives market (Praxis Rock, Aug 2026); CVC’s wealth channel grew about 40% (PE Insights, 2026).Wealth vehicles offer periodic liquidity, which links fund redemptions to the share price.

The common thread is that the public market pays most for fee streams that do not depend on selling assets well. That single pricing signal explains the strategic direction of almost every large listed manager.

Market Activity

The defining activity of 2025 and 2026 is manager-on-manager M&A: listed firms using their equity and balance sheets to buy fee streams they cannot build fast enough.

DateBuyerTargetReported priceTarget AUMPrice as % of AUM (our calculation)
May 2026 (closed)KKRArctos Partners~$1.4B~$16B~9%
2026 (agreed Oct 2025)BrookfieldRemaining 26% of Oaktree~$3B~$334B~3.5% (implied for 100%)
Jul 2025 (closed)BlackRockHPS Investment Partners~$12B~$150B~8%

Other 2025 to 2026 moves: SoftBank agreed to buy DigitalBridge at $16 per share (closing expected H2 2026), DigitalBridge agreed to buy power investor ArcLight for up to $1.05 billion, and EQT bought secondaries leader Coller Capital.

Our reading of the 2025 to 2026 deal flow:

  • Buyers pay for access, not assets. Arctos at about 9% of AUM priced sports league approvals and GP relationships that would take a decade to build. The same logic applies to any franchise with a licence, a regulator or a relationship moat.
  • Credit is still the centre of gravity, despite the stress. HPS and Oaktree are both credit platforms, and BlackRock’s private credit AUM reached $203 billion by the end of 2025 (Alt Goes Mainstream, Apr 2026). The 2026 sell-off has made buyers selective about retail credit, not about credit.
  • Power and infrastructure managers are the next targets. The SoftBank, DigitalBridge and ArcLight chain links AI capital, digital infrastructure and electricity in one listed group.

Secondaries are the other hot segment: EQT bought Coller Capital, Goldman bought Industry Ventures and Bridgepoint bought Newbury. Secondaries suit a listed owner because they earn fees from other managers’ need for liquidity, which rises when exits are slow.

What Changes When the Manager Is Public

Listing turns a private partnership into a machine that compounds recurring fees, and 2026 showed the same machine can run backwards.

The Listed Asset Manager Effect chart

Our analysis · the listed-manager fee flywheel and its 2026 reverse gear

The loop explains most strategic choices of 2025 to 2026. Its weak point is the wealth channel: the vehicles that add the most fee AUM also promise liquidity, and when those promises are tested the damage reaches the share price and the deal currency at once.

WhoWhat changes in practice (our analysis)
LPsThe manager’s incentives now favour asset growth. LPs gain leverage on fund-size caps, fee offsets, and change-of-control and key-person clauses as managers buy each other.
Portfolio companiesOwnership shifts toward longer-hold vehicles, continuation funds and credit. Owners have less need to sell on a fixed fund clock and more interest in steady yield.
Partners and staffMore pay arrives in listed stock. A falling share price becomes a retention problem at the moment the firm most needs its investment teams.
Wealth clientsLiquidity terms are now the product. Gates, once an edge case, are disclosed, studied and priced by the equity market.
Public investorsOwning an S&P 500 index fund now means indirect exposure to private market fundraising and to the liquidity of retail private funds.

Regional Trends

The listed-manager effect is showing up differently in each region in 2025 and 2026, shaped by what each market’s listed managers actually earn from.

RegionCurrent picture (2025 to 2026)What to watch (our view)
United StatesHome of the largest listed platforms: Blackstone ($1.35T AUM), Apollo (>$1T), KKR (>$750B) and Ares ($670B) as of mid-2026. The 2026 private credit sell-off was concentrated here, led by Blue Owl.Credit-heavy and retail-heavy managers carry a liquidity discount until redemption data stabilise for several quarters.
EuropeEQT manages about €341 billion after buying Coller Capital; CVC reported record 2025 results and 40% growth in its wealth channel; Bridgepoint returned a record €16.6 billion to investors in H1 2026. Partners Group’s June gate hit the whole listed group.European managers lean heavily on evergreen and wealth products, so their share prices are unusually sensitive to redemption headlines.
IndiaListed managers are mostly mutual fund houses. ICICI Prudential AMC, India’s largest by active fund assets with ₹10.1 trillion, listed in December 2025; Prudential kept about 35% and plans to return about $1.4 billion of proceeds to its shareholders.Alternatives are likely to reach Indian public markets through the AIF arms of listed AMCs and wealth firms, rather than through standalone private equity IPOs.
Japan and the GulfSoftBank’s agreed purchase of DigitalBridge brings a US listed manager into a Japanese listed group. Gulf sovereign platforms such as MGX raised large dedicated funds in 2026.Strategic and sovereign buyers are becoming acquirers of managers, not only LPs in their funds.

Outlook: 2027 To 2028

Management guidance still points to growth, but the equity market will judge the next two years on liquidity as much as on fees.

What companies have said (targets and guidance, not our forecasts)

  • Brookfield expects 2026 growth to exceed its long-term targets, with its infrastructure and private equity flagships on track to be its largest ever (Q1 2026 results).
  • Partners Group raised $16 billion of new commitments in H1 2026 and confirmed guidance of $26 to $32 billion for the year.
  • CVC says it is highly confident in its growth path to 2028 (2025 full-year results).
  • KKR expects its new Solutions business, built around Arctos, to reach about $100 billion of AUM over time.

Our view: five things likely to define 2027 to 2028

  • Liquidity disclosure becomes a valuation input. Managers that publish evergreen redemption rates next to FRE, and show they are stable, should earn back the 2026 discount first.
  • Manager M&A continues, priced for access. Secondaries, specialist franchises and power-linked infrastructure managers remain the most sought-after targets. Credit platforms may change hands at lower prices after the sell-off.
  • The listing window favours platforms, not pure buyout shops. The next manager IPOs are more likely to come from firms with a mix of credit, insurance and wealth fees than from single-strategy funds.
  • LP terms tighten. Expect more negotiation over key-person, change-of-control and fund-size provisions as ownership of managers keeps changing.
  • India’s listed AMCs move into alternatives. Growth in their AIF arms is the most likely route for Indian public investors to gain exposure to private markets fees.

Signals to watch: quarterly redemption requests at the largest retail credit and evergreen funds; the share of FRE coming from vehicles with periodic liquidity; the price paid per dollar of AUM in manager acquisitions; and the closing of SoftBank’s DigitalBridge purchase.

Key Takeaways

  • A listed manager serves two clients. LPs want returns; shareholders want recurring fees. Most strategy in 2025 and 2026 follows from that tension.
  • The market pays for fees, so firms build fees. Credit, insurance, wealth vehicles and manager acquisitions all add FRE, which carries the highest valuation.
  • Deal prices reward access. Buyers are paying around 8% to 9% of AUM for hard-to-build franchises such as HPS and Arctos.
  • Liquidity is now priced in the share. The Blue Owl and Partners Group gates of 2026 showed that retail fund stress becomes equity stress within a day, across the whole listed group.
  • The next re-rating will go to transparency. Managers that disclose and stabilise evergreen redemptions are best placed to recover the 2026 discount.
rubyworks-image-50

The New Import Substitution Map: India’s Next Domestic Manufacturing Plays

0

$112 billion. That was India’s trade deficit with China in 2025-26, a record. Electronics imports alone crossed $116 billion. And in roughly 636 product lines, China now supplies at least 80% of what India buys from abroad.

Those numbers are a warning. They are also a shopping list. Every large, concentrated import bill is a market that a domestic manufacturer could take, if the economics, technology and policy line up. This insight maps where they are lining up now, where they could in three to five years, and where the prize is real but the path is still long.

The Map in 30 Seconds

  • Ready now: electronics components and PCBs, solar cells, defence. Policy is live, capacity is being built and demand is already local.
  • Building: semiconductor packaging, lithium-ion cells, pharmaceutical ingredients. Large prizes, early plants running, but scale and cost parity are years away.
  • Long shots: rare earth magnets and high-end medical devices. Strategically vital, but India is starting close to zero.

Why It Is Back, and Why This Time Is Different

Import substitution has a bad name in India. The old version relied on high tariffs and licences and protected inefficiency for decades. The new version uses different tools:

  • Pay for output, not for existence. Production-linked incentives reward units actually made and sold.
  • Mandate quality domestic supply where it matters. Solar’s approved-manufacturer list and defence’s positive indigenisation lists create guaranteed local demand.
  • Link it to exports. Defence exports hit a record ₹38,424 crore in 2025-26, up 63% in a year (Ministry of Defence). The goal is to compete globally, not just to replace imports.

Our read: the easy layer is done, the hard layers are next. India substituted the final assembly of phones and electronics, but dependence on the inputs has deepened at the same time. The number of electronics product lines where China supplies at least 80% of India’s imports rose from 44 in 2018-19 to 71 in 2025-26 (Koan Advisory and Institute of Chinese Studies). In pharma, China’s share of India’s API and intermediate imports rose from 68.2% to 73.7% between FY21 and FY25, even as incentive schemes drew more investment than targeted.

Each layer further upstream has smaller volumes, harder chemistry and thinner margins for error. That is where the next manufacturing plays sit, and it is why they need patience as much as capital.

The trigger was supply risk, not just cost. When China restricted rare earth magnet shipments in 2025, Indian automakers scrambled for supply (Reuters). That episode moved import substitution from an industrial-policy idea to a board-level risk question.

The Map

We plotted eight industries on two questions: how big is the import prize, and how ready is India to take it?

The New Import Substitution Map chart

Rubyworks assessment, September 2026. Positions are judgement-based, drawn from import data, approved capacity and scheme progress cited in this note.

How to read it: the top-right is where capital can be deployed with the least execution risk today. The top-left holds the largest prizes, but they need time, technology partners and sustained policy. The bottom-left is strategically important but still at pilot scale.

Ready Now: The Scale Plays

1. Electronics Components and PCBs

  • The prize: electronics imports reached $116 billion in 2025-26 (Commerce Ministry). China’s dominance sits in motors, cables, switching equipment and other core parts, not finished goods.
  • What is happening: the Electronics Component Manufacturing Scheme pipeline covers 75 projects with ₹61,671 crore of proposed investment. Once running, approved projects are expected to meet about half of India’s PCB demand.
  • The catch: components are high-volume and thin-margin, and Chinese suppliers have a decade of scale. Design depth in India is still limited.
  • Our call: the most investable near-term play, because the customers (phone, auto and appliance assemblers) are already in India. Multilayer PCBs, connectors and display modules look strongest.

2. Solar Cells

  • The prize: India localised solar modules first but kept importing the cells inside them.
  • What is happening: from June 1, 2026, approved solar projects must use domestically made cells. About 30 GW of cell capacity was on the approved list by May 2026.
  • The catch: the next layers, wafers and polysilicon, are still imported. Mandated demand also invites overbuilding once the initial rush ends.
  • Our call: cells are a play now. Wafers and polysilicon are the next, larger and riskier opportunity, and whoever integrates backward first will set the cost curve.

3. Defence

  • The prize: defence production reached a record ₹1.78 lakh crore in 2025-26, up 15.6% in a year, against a government target of ₹3 lakh crore by 2029 (Ministry of Defence).
  • What is happening: ten positive indigenisation lists now name 5,521 items for domestic sourcing, and the private sector’s share of production has reached a record 24%.
  • The catch: orders are large but lumpy, cycles are long and state-owned firms still dominate production.
  • Our call: the demand is the most secure on the map. The underappreciated opportunity is in sub-systems and component suppliers to the big platforms, where the indigenisation lists bite first.

Building: Big Prizes, Early Plants

4. Semiconductors

  • Where it stands: twelve units approved with over ₹1.64 lakh crore of investment; Micron, CG Semi and Kaynes are in commercial production, all in assembly and testing. The first silicon fab is scheduled for 2028. China still supplied 48.9% of India’s semiconductor imports in 2025-26.
  • The catch: packaging is the easier half; front-end fabrication needs process know-how India does not yet have at scale.
  • Our call: the near-term play is packaging, testing and the equipment and materials suppliers that cluster around them, not the fab itself.

5. Lithium-Ion Cells

  • Where it stands: imports from China more than doubled since 2021-22 to $3.9 billion, and China’s share reached 83.6%. Only 1.4 GWh of the 50 GWh targeted under the battery cell incentive scheme was commissioned by October 2025 (IEEFA).
  • The catch: the scheme’s slow start shows how hard cell chemistry, equipment and mineral supply are to localise at once.
  • Our call: cells for electronics and energy storage will localise before EV-grade cells. Recycling, as a domestic source of lithium and cobalt, is an underrated adjacent play.

6. Pharmaceutical Ingredients (APIs and KSMs)

  • Where it stands: 39 bulk-drug projects covering 28 APIs are commissioned, including Penicillin G and clavulanic acid, and 191 of 726 targeted APIs, starting materials and intermediates are now made in India for the first time. Yet China’s share of these imports still rose to 73.7% in FY25.
  • The catch: China’s fermentation scale sets prices India cannot yet match for high-volume antibiotics.
  • Our call: the winners will be molecules where buyers pay a premium for supply security, and makers who pair domestic capacity with export contract manufacturing.

Long Shots: Strategic, But Starting From Near Zero

7. Rare Earth Magnets

  • Where it stands: a ₹7,280 crore scheme targets 6,000 tonnes a year of capacity and drew 20 bids. India holds the world’s third-largest rare earth reserves, yet imports almost all its magnets, and China controls about 90% of processing.
  • Our call: small tonnage but outsized strategic value. Automakers and defence buyers may pay a premium for secure supply, which could make the economics work sooner than volumes suggest.

8. High-End Medical Devices

  • Where it stands: India’s device market is 70% to 80% import-dependent by value, a figure that has barely moved in a decade, with imaging and implants the most import-heavy. Consumables are already competitive.
  • Our call: substitution will climb from consumables into mid-tech devices first. High-end imaging will need global partners, not stand-alone Indian plants.

The Traps: Where Substitution Fails

  • Assembly without depth. Making the final product locally can raise imports of the parts inside it. Electronics exports and imports both grew strongly in 2025-26, which shows how easily assembly and dependence rise together.
  • The mandate hangover. Guaranteed domestic demand draws a wave of capacity. When the wave peaks, prices fall and weaker players get squeezed. Solar cells are the category to watch.
  • One layer’s protection is the next layer’s cost. Duties and mandates that help cell makers raise costs for module makers and power developers. Every substitution decision moves margin along the chain.
  • Incentive-dependent economics. If a plant only makes money while incentives last, it is a subsidy, not a business. Investors should test unit economics with incentives stripped out.
  • Hidden upstream imports. Equipment, specialty chemicals and minerals for chips, batteries and magnets are still largely imported. A plant can be domestic while its inputs are not.

What to Watch

  • Commissioning, not approvals. The number of component, cell and chip projects actually producing, against the number approved.
  • The China-concentration count. Whether the 71 electronics lines with 80%-plus China dependence start to fall. That is the cleanest test of real substitution.
  • The first domestic magnets. India aims to start producing rare earth magnets by the end of 2026. On-time output would move magnets up the map fast.
  • Solar’s next layer. Announcements of wafer and polysilicon capacity will signal who is integrating backward.
  • Battery cell ramp-up. Whether commissioned cell capacity moves meaningfully beyond pilot scale in the next 12 months.
  • API data for FY26. A falling China share would be the first proof that the bulk-drug scheme is changing the mix, not just adding capacity.

The Bottom Line

India’s first import substitution wave replaced finished goods. The second has to replace what goes inside them, and that is harder, slower and more selective. The near-term money is in electronics components, solar cells and defence supply chains, where demand is local and policy is working. The larger prizes in chips, battery cells and pharmaceutical ingredients will reward investors who can wait three to five years and who judge projects on unit economics, not incentive announcements. The one number to track is simple: how many product lines still depend on a single foreign supplier.