SBI Funds is going public | Plus: Why it’s so hard to gather land in India

SBI Funds is going public

The IPO of SBI Funds Management opens today. It is entering the market as India’s largest asset manager by AUM, a position it has held since 2021, with about ₹29.5 lakh crore under management, according to its prospectus.
However, there is an important nuance.
Despite managing more assets than ICICI Prudential AMC or HDFC AMC, SBI generates lower revenue than these rivals. This doesn’t imply it is a weaker business, but it shows that “largest” can be misleading in asset management. SBI’s business model is more complex than the headline AUM figure suggests.
When ICICI Prudential AMC listed last year, its prospectus helped explain how AMCs earn their fees. Rather than revisiting those basics, this discussion focuses on what makes SBI’s model distinct.

Not all AUM is equal

Consider the ₹29.5 lakh crore AUM.
This is not one homogeneous pool. Only around ₹12.5 lakh crore is in mutual funds. The larger portion, roughly ₹16.9 lakh crore, is in portfolio management services (PMS) and advisory mandates. SBI is actually the biggest PMS manager in India, with close to 40% market share.
On the surface, this looks like a well-diversified franchise. But of SBI’s approximately ₹4,400 crore operating revenue, about 96% comes from mutual funds. The huge PMS and advisory book contributes only around ₹155 crore of profit, a very small share relative to its size.
The key reason is that most of this PMS/advisory money is low-margin institutional capital — such as EPFO-type statutory and provident fund mandates — which come in very large tickets but pay extremely low fees. Rough estimates suggest SBI earns about 35 paise annually for every ₹100 managed in mutual funds, but only around 1 paisa per ₹100 on the rest.
So, AUM quality matters, even within mutual funds. Active equity schemes, where fund managers select stocks and charge higher fees, account for about 42.5% of mutual fund AUM but generate roughly three-fourths of fee income. Passive products — index funds and ETFs that simply track benchmarks at low cost — form about one-third of AUM but contribute only about 5% of fees. One rupee in an active equity fund is worth roughly ten times a rupee in a passive fund from a revenue standpoint.
This explains why SBI’s competitors appear stronger on revenue metrics. SBI earns around 35 basis points on its mutual fund AUM, versus about 52 bps for ICICI and 44 bps for HDFC. The difference is largely due to mix: ICICI and HDFC have a higher proportion of active equity assets than SBI.
This doesn’t make SBI an inferior business; it makes it a different one. SBI has become the largest by accumulating large, low-fee passive and institutional books. The trade-off is that this depresses revenue per rupee of AUM. In effect, SBI is a volume-focused rather than value-focused franchise.

The key metric moving in SBI’s favour

If this were the full picture, SBI would look like a giant being out-earned by leaner peers. But its asset mix has been shifting in a positive direction.
From FY24 to FY26, SBI’s equity-oriented AUM grew at about 22% annually. Over the same period, passive AUM grew around 13%, and low-fee liquid funds grew at under 6%. Equity’s share of the mutual fund book rose from 39% to 42.5%, while passive’s share declined. As a result, fee yield per rupee of AUM improved.
The change may appear modest, but it explains why revenue grew at nearly 28% per year in that period, even though overall AUM grew only about 17%. SBI not only gathered more assets; the average rupee it managed became more profitable.
What the prospectus does not clarify is how much of this improvement came from fresh inflows versus market appreciation in a bull phase. Rising markets can boost both mix and fee yield without necessarily indicating market share gains.

Where SBI really stands out

In asset management, pricing is a weak lever. Fee caps are largely set by regulation, and competition does the rest, so AMCs have limited pricing power. What they can control is cost, and this is where SBI clearly leads.
SBI’s operating expenses are about 8 basis points of AUM, the lowest among the top ten AMCs, which typically operate in the 10–25 bps range. Its core operating margin is around 79%. Once the research, compliance, technology and distribution infrastructure is in place, adding another ₹10,000 crore of AUM costs almost nothing extra. It is a high fixed-cost platform, and SBI runs the largest and most efficient one.
The top three AMCs have collectively held around 40% market share for years, while smaller, agile players have been gaining. Mid-sized AMCs in the middle have been squeezed, with their combined share falling from about 41% to 36% over five years. The industry is polarising into large-scale players at one end and niche specialists at the other, with a shrinking middle. In a world of steadily declining fees, the least efficient managers feel the pressure first, and SBI’s low-cost structure allows it to withstand that pressure longer than most.
Importantly, SBI’s moat is not superior investment performance. Its own disclosures show only about a quarter of its equity schemes in the top performance quartile.
This indicates that SBI became India’s largest AMC without consistently outperforming peers on returns. It underlines that in the Indian AMC business, distribution strength and investor trust can be as critical as performance.

Broad reach, shallow wallet share

The second pillar of SBI’s moat is distribution, particularly strong where the industry is relatively underpenetrated: beyond the top 30 cities.
The RBI labels everything outside the top 30 cities as “B30”, and this is where SBI is strongest. B30 locations account for about 22.8% of its AUM, compared with an industry average of around 18%. Roughly two-thirds of its SIPs originate from these smaller centres. In such markets, brand trust carries more weight, making it harder for competitors to replicate SBI’s reach.
However, this reach is more about the number of relationships than their size. SBI has about 1.6 crore live SIPs, collecting around ₹4,000 crore per month. It holds about 15.5% of the industry’s SIP accounts but only 11.4% of SIP inflows. The average SBI SIP ticket is roughly 25% smaller than the industry average. Its share of SIP inflows has also declined from 12.9% in FY24 to 11.4% in FY26.
So SBI is adding SIP accounts faster than it is capturing SIP money. This aligns with its higher share in B30 markets, where investors typically start with smaller amounts. But it also highlights the need to increase ticket sizes over time, promote SIP step-ups, and deepen product penetration per client.

Growth drivers

An AMC essentially has two growth levers: increase AUM, or increase revenue per unit of AUM. SBI already leads on sheer AUM, so further asset growth is more of a defensive, incremental play. The more powerful lever is improving yield, which depends on (a) upgrading low-fee assets into higher-fee active equity, and (b) scaling higher-margin products for affluent investors.
This context explains the emphasis in SBI’s prospectus on newer segments. It highlights leadership in specialised investment funds (a new SEBI category), rapid growth in alternative investment funds (AIFs), and offshore products including GIFT City offerings.
These are genuinely higher-fee businesses, but currently very small. SBI’s 28% share of the SIF segment is in an industry of only about ₹10,000 crore, and its AIF book is just about 0.2% of total AUM. These are future options rather than current earnings drivers. For now, SBI remains primarily a scale-and-cost story, not a premium-fee one.
On the other side, there is a regulatory headwind that is not yet visible in the reported numbers. FY26, the last year covered in the prospectus, is also its best year on record in terms of fee yield, margins and revenue growth. But this period is entirely before SEBI’s new Base Expense Ratio (BER) framework, effective 1 April 2026, which will restrict expense levers for funds. None of that impact is reflected in FY26. At SBI’s size, even a 1 basis point reduction in fee yield translates to roughly ₹125 crore of lost revenue.

Interpreting the financials

Some headline metrics in the prospectus need closer scrutiny.
First, the profit figure. SBI reported PAT of about ₹3,067 crore, but this includes income from its own treasury and investment book. Excluding that, the core fund management business earns closer to ₹2,606 crore — a better indicator of operating profitability.
Second, the reported return on equity (ROE) of 43%, up from about 34% the previous year. Part of this jump is mechanical. Ahead of the IPO, SBI Funds paid a dividend of around ₹5,515 crore — about 1.8 times its annual profit — financed partly by selling investments. This reduced its equity base from roughly ₹8,300 crore to under ₹6,000 crore. With a smaller equity base, ROE rises even if underlying profitability is unchanged.
Stepping back from these details, SBI Funds is a very specific type of AMC. It is not positioned as the best stock picker or a pure technology platform. Instead, it is a scale-driven, trust-based, low-cost operator.
The key question for investors in the IPO is whether SBI can upgrade its vast base of small, low-fee relationships into higher-value ones quickly enough, while managing the active–passive mix that ultimately determines its earnings per rupee of AUM — and doing all this under tightening regulatory constraints on expenses.

Why land is so difficult in India

Land-related issues in India are notoriously complex.
Recently, about a thousand tractors converged on Gandhinagar as farmers from across Gujarat protested high-tension transmission lines being laid across their fields to evacuate power from solar and wind projects.
In 2015, after Andhra Pradesh was bifurcated, the state attempted to build a new capital at Amravati without resorting to forced acquisition. Farmers contributed around 34,000 acres under a pooling scheme, expecting land values to rise. The project then stalled for years, triggering prolonged protests. When it was revived, the state still had to compulsorily acquire about 1,800 acres from farmers who had lost faith in the arrangement.
In other cases, land is acquired but left idle. Karnataka, for example, acquired land for an infrastructure corridor in 2008, but it remained unused for 17 years until the High Court ordered its release.
Where land is not vacant, it is often entangled in irregularities. In Greater Noida, the national auditor found that nearly half of about 2,580 acquired plots had no functioning unit, and the rest were riddled with issues such as payment defaults and unauthorised transfers.
For economic development, land must move into productive use. Yet acquisition processes frequently break down at multiple stages. There is no comprehensive public database showing how land is owned, used and regulated. Information is fragmented across different sectoral systems that are not integrated.
Two broad challenges stand out: assembling land in usable parcels, and then converting that land into something that can legally and physically be developed. The discussion that follows focuses on these hurdles.

Hurdle 1: Assembling land

Among all inputs a growing economy needs, land is arguably the hardest to organise. It cannot be created or relocated. Infrastructure and industry require continuous tracts of land, often in specific locations.
But putting such tracts together is extremely difficult. Much of India is divided into countless small plots. Establishing who owns a given parcel is often unclear. There is no single, authoritative record that captures every inheritance, sale or family partition. Even official land records do not conclusively prove ownership. In many areas, maps and textual records contradict each other.
Even when the owner is identifiable, acquisition is not guaranteed.
There is the classic “hold-out” problem, common wherever land is privately owned. Suppose a factory needs 200 contiguous plots. You might secure 199, but if one owner in the middle refuses to sell, the entire project can be blocked. The more fragmented the land, the worse this becomes. In India, where plots are frequently co-owned by multiple family members, many people effectively have veto power over any project.
On top of this, state-level regulations can obstruct voluntary purchases. Some states restrict the sale of agricultural land to non-farmers; others impose landholding ceilings. There is no assurance that a buyer can assemble the required parcel even with willing sellers.
The usual workaround is to rely on the state to acquire land.
This approach dates back to an 1894 colonial law that empowered the British government to compulsorily take land. Instead of negotiating separately with each owner, the state could run a single acquisition process for a defined area, then take possession. Existing claims on the land would be extinguished, and disputes would shift to compensation rather than title. The state could then use the land for public works, industrial estates, townships or even projects ultimately used by private companies.
In 2013, this colonial law was replaced by a new statute that made acquisition more consultative, costly and stringent. It requires governments to minimise the land they take and to prefer leasing over outright purchase where possible. But the core power — compulsory acquisition without consent — remains.
This is not unique to India. Many countries allow compulsory acquisition as a last resort. In England, for instance, authorities are expected to try to buy land by agreement, but they can simultaneously prepare a compulsory purchase order if negotiations fail.
The Indian peculiarity is that this “backstop” has become the default method of assembling land.

Hurdle 2: Ownership is not enough to build

Even if you legally own a large, dispute-free parcel — say, 100 acres for a factory — your challenges are not over. Ownership is only one dimension of land rights; you also need the right to use the land for your intended purpose.
If the land is classified as agricultural, it must be formally converted to industrial or residential use in the revenue records. Separately, your proposed use must comply with the area’s master plan. These two processes are handled by different agencies. A parcel may pass one test and fail the other. You might convert farmland to non-agricultural use but find the master plan does not permit a factory there; or the master plan may zone the area for industry while the revenue record still lists it as farmland.
Some states have begun to address this mismatch. Late last year, Karnataka amended its revenue rules so that if a master plan designates an area for industry, agricultural land there can be converted to industrial use without special approval from the deputy commissioner. The need for such a reform shows how cumbersome the earlier system was.
Around the same time, the Karnataka High Court highlighted another systemic issue: land records, forest boundaries and planning zones were all based on different mapping systems that did not align. The court directed the state to integrate these into a unified record.
In practice, you can buy land from the correct owner at a fair price and still lack the legal right to develop it as planned.
Beyond this, there are multiple other government interfaces. You need road access, power and water connectivity. Depending on the project, you may also require environmental clearance, pollution control consent, forest clearance or permission to extract water. Different agencies control each of these, and many approvals are discretionary. A clearance from one authority does not guarantee approval from another. There is no single map or system that clearly indicates what can be done on a given parcel.

Hurdle 3: Incentives to hoard rather than develop

To navigate these complexities, India created specialised bodies — often industrial development authorities such as KIADB, MIDC or the Noida Authority.
These agencies acquire and consolidate land, earmark it for industrial use, build internal roads and utilities, subdivide it into plots, and then lease or sell these plots. They effectively offer “ready-to-use” land with much of the legal and physical groundwork completed, which is highly attractive to investors.
But this model also runs into problems.
Take KIADB as an example. The state auditor found it holding about 6,600 acres of developed but unused land and more than 30,000 acres of undeveloped land. Some parcels were awaiting infrastructure, some were stuck in litigation, and others were reserved for future phases. At each stage — possession, building approval, commencement of production — fewer plots progressed. The auditor also noted that KIADB did not maintain a proper inventory of its holdings.

Why do such projects stall?

Greater Noida offers a clue. Industrial plots there were deliberately priced low to attract industry. This meant that from day one, the market value of a plot often exceeded its acquisition cost. For many allottees, it was more profitable to hold the land idle than to build a factory. The plots became speculative assets. Constructing a plant involves demand risk, construction risk and long-term capital lock-in. Simply holding the plot and waiting for surrounding development to push up prices, then seeking a transfer, could be more attractive. Authorities often failed to cancel such non-compliant allotments and reclaim the land.
These agencies became adept at acquiring and allotting land — activities with clear political and administrative rewards. But the harder tasks of cancelling idle allotments, recovering land and fighting legal battles received less attention. As a result, they struggled to ensure that land was actually developed.

Alternative approaches

To break this deadlock, other models are needed.
One promising alternative is land pooling. Under this approach, multiple landowners contribute their plots to a common pool. The government then uses part of the pooled land for roads, drainage and public spaces, sells some parcels to raise funds, and returns the remaining land to the original owners. Although each owner receives a smaller plot, it now has infrastructure and services, making it more valuable.
Some Indian states already use variants of this model. Gujarat, for example, has long relied on town-planning schemes to reshape urban areas.
Pooling is less coercive than outright acquisition, but it depends heavily on trust. Japan, which has used land pooling for over a century, illustrates the prerequisites: accurate and reliable land records, a clear and binding development plan, transparent rules for valuation and compensation, and implementing agencies with a strong track record. Without such trust, pooling schemes can collapse, as seen in Amravati.
Moreover, pooling is not suitable for every type of project. Infrastructure such as highways or transmission lines often requires long, narrow strips of land. For underground pipelines, the requirement may be limited to a right of way to lay and maintain the pipe, rather than full ownership.

A structural challenge

To some extent, land acquisition challenges are unavoidable; most countries grapple with similar issues.
But there are better and worse ways to handle them. In India, the state has repeatedly relied on the most forceful tool compulsory acquisition often because of failures elsewhere: poor data, weak institutions, complex regulations and misaligned incentives.
In other words, the core difficulty in assembling enough land in India is not the physical scarcity of land itself, but the institutional, legal and administrative hurdles that stand between land and its productive use.