Gold Loan M&A: 4 Deals, ₹2,058 Crore Bought vs ₹35,000 Crore Built
By Nitin Misra (Co-Founder, indiagold) · 2026-07-16 · 9 min read
Tata Capital announced on 13 July that it is acquiring Yogakshemam Loans, a Thrissur-based NBFC. The filing language: a pre-money equity valuation "not exceeding ₹318 crore," a primary capital infusion of "about ₹93 crore," and a resulting holding of "around 88.6%." The consideration Tata actually pays is not disclosed (AFAIK).
Yogloans has ₹708 crore of AUM, 162 branches across Kerala, Karnataka, Tamil Nadu and Andhra Pradesh, and about 32,000 customers. Tata Capital's net AUM at 31 March 2026 was ₹2.77 lakh crore.
The acquired book is 0.26% of the acquirer's. Below is what has happened in this market, and what the data supports.
The transactions
| Buyer / target | Announced | Value | What transferred | Status |
|---|---|---|---|---|
| L&T Finance ← Paul Merchants Finance (gold business) | 7 Feb 2025 | ₹537 Cr | ₹1,350 Cr book at completion, 130 branches, ~700 employees, 98,000+ customers | Completed 9 Jun 2025 |
| InCred ← TruCap Finance (gold business) | 25 Feb 2025 | ₹330 Cr | ₹650 Cr AUM, 115 branches, 550+ employees, 40,000+ customers | Terminated 26 May 2025 |
| Bain Capital → Manappuram Finance | 20 Mar 2025 | ₹4,385 Cr | 18% fully diluted, joint control, of a 5,357-branch, 50,795-employee franchise | Completed 27 Mar 2026 |
| Tata Capital ← Yogakshemam Loans | 13 Jul 2026 | ≤₹318 Cr pre-money + ₹93 Cr | ₹708 Cr AUM, 162 branches, 32,000 customers | Announced |
Another large NBFC acquired a gold lender but I'll let them break the news 🙂
Four points about that table.
Bain is not a strategic entrant, and its stake is widely misreported. Bain invested ₹4,385 crore at ₹236 a share, a ~30% premium to the six-month average, for 18% on a fully diluted basis. Its press release said the final stake would "vary between 18.0% to 41.7%" depending on open offer subscription. The open offer, priced at ₹248.29 with the stock trading well above it, recorded zero shares tendered as of 20 April 2026. Bain holds ~9.89% of equity, rising to ~18% fully diluted on warrant exercise. The 41.66% figure in the press is Bain's RBI approved ceiling — what it could have reached had the open offer been fully subscribed — not what it holds. (By coincidence, the reconstituted promoter group — the Nandakumar family's 31.77% plus Bain's 9.89% — also totals exactly 41.66%.) Bain's stated rationale references capital for growth, a "professionalization journey," and operational efficiency. It says nothing about speed, branches or time-to-market. It bought an earnings stream.
InCred's deal did not close. It signed a binding term sheet on 25 February 2025 and terminated it on 26 May 2025. No public reason was given.
Books that actually changed hands: ₹1,350 crore (L&T) and ₹708 crore (Tata) — ₹2,058 crore in total.
Both strategic buyers who quantified the purchase quantified it in months. L&T Finance, in a company statement: the acquisition "accelerates gold loan business time-to-scale by 36 months, providing approximately Rs 1,000 crore in assets under management and over 98,000 active customers." InCred, in a company statement: the deal "will provide InCred Finance with a head-start in the gold loans business, accelerating its go-to-market timeline by 24-36 months."
The organic builders are far bigger
Capri Global grew gold AUM 111% year-on-year to ₹16,965 crore across 999 gold loan branches in 15 states and UTs as at 31 March 2026, and crossed 1,000 branches in May. Bajaj Finance's gold book was ~₹12,000 crore at September 2025 and had roughly ₹17,800 crore by March 2026, up ~115% year-on-year, with 900 new branches and 500 conversions planned. Poonawalla Fincorp entered in April 2025 with a plan for 400 branches over four quarters.
That is roughly ₹35,000 crore of gold AUM built in-house against ₹2,058 crore acquired. Capri's book alone is 24 times the one Tata bought. The two organic builders together are about 17 times everything acquired.
Building works. It is slower. Capri needed 999 branches and several years. L&T paid ₹537 crore to skip three of them.
What is being bought: assaying, custody, auction
The failure mode in gold lending is operational, not credit. The regulator's record is the clearest evidence.
RBI barred IIFL Finance from gold loan disbursals on 4 March 2024. Its stated findings: "serious deviations in assaying and certifying purity and net weight of the gold at the time of sanction of loans and at the time of auction upon default; breaches in Loan-to-Value ratio; significant disbursal and collection of loan amount in cash far in excess of the statutory limit; non-adherence to the standard auction process; and lack of transparency in charges being levied to customer accounts, etc." Restrictions were lifted on 19 September 2024, per IIFL's exchange filing, following a special audit.
These findings were not credit decisions related. All were operating controls related.
On 30 September 2024 RBI issued a sector-wide circular after reviewing gold loan practices across banks and NBFCs. It set out six major deficiencies: improper use of third parties for sourcing and appraisal; valuation of gold without the customer being present; inadequate due diligence and lack of end-use monitoring; lack of transparency during auction; weaknesses in LTV monitoring; and incorrect application of risk weights. An annexed illustrative list ran to eleven items, adding, among others, accounts closed within days of sanction with no economic rationale, and loans rolled over at the end of tenor against part payment. Lenders were told to remediate in a timebound manner and report to their Senior Supervisory Manager (SSM) within three months.
On 7 July 2026, Delhi police arrested a gold appraiser at an NBFC. The allegation, per Delhi Police's Economic Offences Wing, is that he created 683 fake loan accounts against counterfeit gold and forged KYC between 2022 and 2025, disbursing ₹3.81 crore. The company's internal audit detected it. The same audit found nine packets of genuine pledged ornaments, worth ₹14.1 lakh, missing. The matter is sub judice.
The appraiser is the control (& failure) point. That is what 162 branches and a Manappuram ex-Deputy CEO are worth to a buyer.
Yields: what the data actually shows
There is a widely circulated claim that Manappuram's net interest margin collapsed from 13.5% to 9.8% in Q4 FY26 while gold AUM grew 99%. I checked it. Manappuram does not publish a NIM line. The 13.5%/9.8% pair traces to secondary outlets carrying AI-generation disclaimers, with no period stated. And Manappuram's own CFO said the opposite on the Q4 earnings call: "If you see the P&L, the NIM improved during the quarter. But the main impact is on account of higher provisions, especially from the non-gold portfolio."
What Manappuram did disclose for Q4 FY26: consolidated gold AUM ₹50,953 crore, up 99.1% year-on-year. Standalone gold AUM ₹48,814 crore, up 98%, at a yield of 17.3%. Cost of funds 8.80%. Gold moved from 59% to 80% of consolidated AUM as the higher-yielding microfinance book was deliberately shrunk. Any margin comparison across that period is measuring a mix shift, not a price war.
Muthoot's Q4 FY26 yield improvement included about ₹85 crore of one-off income — roughly ₹50 crore from auctions and ₹35 crore from ARC sales — per its earnings call.
The competitive pressure is real, but the honest evidence is forward-looking rather than proven in the numbers. Investec, cutting estimates on both Muthoot and Manappuram in June 2026: "competitive intensity is at unprecedented levels, unsecured loan growth is rebounding, and the probability of a +15 per cent gold price rally looks low." Investec also notes the sector's mix is drifting upmarket — loans above ₹5 lakh went from 20% of gold financiers' books in FY24 to 42% in FY26 — and larger tickets sit in the lower-LTV, lower-rate tier by construction. HDFC Securities, reported the same week as the Tata deal: attractive unit economics, but a branch-led business exposed to fraud, theft, auctions and high operating leverage, with rising competition likely to pressure yields. CRISIL, against all of that, still projects gold loan NBFC return on managed assets at 4.25–4.5% through FY27, with credit costs under 1% across the past five years.
A squeeze, then. Not a collapse. And not yet visible in a clean number.
Two variables remain once strategy converges: cost of funds and cost-to-serve. On the first, NBFCs lose to any deposit-funded bank, structurally and permanently. Cost-to-serve is the only variable a specialist controls.
The LTV rule that keeps getting misquoted
Three bands, not two:
| Total consumption loan per borrower | Maximum LTV |
|---|---|
| ≤ ₹2.5 lakh | 85% |
| > ₹2.5 lakh and ≤ ₹5 lakh | 80% |
| > ₹5 lakh | 75% |
More consequential is the first phrase of the clause. RBI's Lending Against Gold and Silver Collateral Directions, 2025, Clause 19: "The maximum LTV ratio in respect of consumption loans against the eligible collateral shall not exceed…" The NBFC Master Direction carries identical wording at para 43 of Chapter IV of the Reserve Bank of India (Non-Banking Financial Companies – Credit Facilities) Directions, 2025.
Consumption loans only. There is no LTV cap anywhere for income-generating loans — defined in Clause 6 as farm credit, loans for business or commercial purposes, or the creation of productive assets. The limit is delegated to the lender's board-approved credit policy: Clause 8 of the Gold Directions, para 32 of the NBFC Master Direction.
The logic is coherent. Collateral justifies LTV up to the tiered caps. Anything above has to be justified by the borrower's cash flow — which is why a detailed repayment-capacity assessment becomes mandatory above ₹2.5 lakh, the same threshold at which the 85% band ends.
The flat 75% cap that NBFCs used to operate under is gone, and the mechanism is worth knowing because most commentary still hasn't caught up. It originated in para 37.1.1(i) of the Scale-Based Regulation Directions, 2023: "maintain a Loan-to-Value (LTV) Ratio not exceeding 75 percent for loans granted against the collateral of gold jewellery" — no distinction by purpose. When RBI consolidated the NBFC rules in November 2025, that cap was carried into a transitional Annex-II, kept alive by a clause that preserved it until each NBFC adopted the new Chapter IV, and no later than 1 April 2026. On 1 April 2026 the clause was deleted, and Annex-II ceased to have effect with it. The phrase "75 per cent" now appears in the Master Direction only as the >₹5 lakh consumption tier. The successor Scale-Based Regulation Directions, 2025 contain no gold-lending or LTV provision at all.
So as of today: an NBFC faces no RBI-prescribed LTV cap on income-generating gold loans.
What has not been relaxed is the operational obligation attached. The prescribed LTV must be maintained on an ongoing basis throughout the tenor, not merely at sanction. Renewals and top-ups are permitted only within the permissible LTV, and only where the loan is standard. Running multiple simultaneous loans to a single borrower is described in the text itself as "prone to misuse and susceptible to fraud" and made subject to "stricter internal audit and supervisory examination." And every lender must now disclose average LTV ratio and gross NPA in its notes to accounts, split between consumption and income-generating loans.
The higher LTV is available. It arrives attached to continuous monitoring and a public disclosure of the number you actually run at.
Separately, on 15 April 2026, RBI deleted the requirement for gold loan NBFCs to obtain prior approval to open branches beyond 1,000. (This was proposed in February and enacted in April; a lot of coverage dates it to February.) Barriers to entry fell.
The price tailwind is flattening, not reversing
Industry gold loan AUM went from ₹6.3 lakh crore in March 2023 to ₹19.4 lakh crore in March 2026, per Experian — and note this is the whole market, banks included, with public sector banks holding about 58% of it. Investec puts the same March 2026 number at ₹18.6 lakh crore. Average ticket size doubled from ₹0.98 lakh to ₹1.96 lakh. Gold loans reached 11.1% of India's retail credit portfolio by December 2025, up from 5.9% in March 2022, making them the second-largest retail credit product in the country after home loans. NBFC gold lending grew 69.9% year-on-year to ₹3.29 lakh crore by May 2026, the fastest of any lending segment in RBI's data. Penetration against household gold holdings is around 10%.
Much of that growth was price. Higher collateral value means a larger loan against the same ornament.
In rupee terms, MCX gold futures went from ₹1,35,891 per 10g on 31 December 2025 to about ₹1,42,100 on 13 July 2026 — up about 4.6% year-to-date. Across FY26, gold rose about 61%. So the tailwind has not reversed. It has flattened, hard.
In dollars the picture looks worse: gold set twelve all-time highs, peaked intraday at $5,595/oz on 29 January, and was down about 7% year-to-date as of late June. Forecasts diverge. The World Gold Council expects gold to trade within ±5% of $4,100 in the second half. Other analysts see a further fall of up to 16%, to $3,400–3,500 — though several of those same analysts remain bullish on a twelve-month view.
The point is not that gold is about to crash. It is that a business whose AUM tripled on a ~61% annual move in the collateral price is about to find out how much of that was lending and how much was revaluation.
Summary
- Four transactions in seventeen months. Two closed, one terminated, one announced. Only ₹2,058 crore of loan book actually changed hands.
- Organic builders have added roughly ₹35,000 crore of gold AUM — about 17 times what was acquired. Buying is not beating building. It is beating waiting.
- Both strategic buyers who quantified the purchase quantified it in months: 36, and 24–36.
- One buyer signed a binding term sheet, ran diligence, and terminated it. No reason disclosed.
- Bain bought 18% fully diluted and joint control, not 41.7%. The open offer drew zero tenders.
- The regulator's findings against gold lenders are operational, not credit: assaying, LTV monitoring, auctions, cash handling, custody, end-use.
- The "Manappuram NIM collapse" datapoint that everyone is citing is not in Manappuram's disclosures, and its CFO said margin improved during the quarter. Competitive pressure is real but is not yet proven in a clean number.
- The tiered LTV caps (85/80/75) apply to consumption loans only. Income-generating gold loans carry no RBI-prescribed LTV cap. The old flat 75% NBFC cap died with a transitional annex on 1 April 2026.
- The higher LTV comes attached to continuous LTV maintenance and mandatory public disclosure of your average LTV.
- Gold is up about 4.6% in rupees year-to-date after a ~61% FY26. Capital, licences and branch approvals all became easier. The operating layer did not.
Sources
- Tata Capital BSE filing (Regulation 30), 13 Jul 2026 — Yogakshemam terms ("not exceeding ₹318 crore" verbatim). Business Today & Business Standard, 13 Jul 2026 — deal coverage; Tata Capital net AUM. Market size, ticket size, penetration data (Experian; HDFC Securities) via Business Today, 13 Jul 2026.
- RBI/DOR/2025-26/347, 28 Nov 2025 (updated as on 1 Apr 2026) — NBFC (Credit Facilities) Directions, 2025: Chapter IV paras 32, 33, 34, 43, 44, 48; transitional para 31 deleted w.e.f. 1 Apr 2026, whereby Annex-II ceased to have effect. Original flat 75% cap: SBR Directions 2023, para 37.1.1(i).
- RBI/2026-2027/11, DOR.RAUG.REC.No.3/23-27-013/2026-27, 15 Apr 2026 — NBFC (Branch Authorisation) Amendment Directions, 2026, clause 5(v).
- Investec via Business Standard, 19 Jun 2026; HDFC Securities via Business Today, 13 Jul 2026; CRISIL Ratings, 2 Mar 2026.
- Capri Global Q4 FY26 results and May 2026 branch-milestone release; Bajaj Finance Q2 FY26 and Q4 FY26 disclosures; Business Today, 15 Apr 2025 — Poonawalla Fincorp.
- TransUnion CIBIL, 14 Apr 2026 — gold loan share of retail credit (as at Dec 2025). RBI data via Business Standard, 7 Jul 2026 — NBFC gold loan growth.
- The Tribune and PTI, 10 Jul 2026 — Delhi gold appraiser arrest, 7 Jul 2026 (allegations; matter sub judice).
- World Gold Council, Gold Mid-Year Outlook 2026, 1 Jul 2026; Business Standard, 13 Jul 2026 — gold price forecasts.
- Gold prices: Goodreturns, 31 Dec 2025 (MCX close ₹1,35,891); India TV, 13 Jul 2026 (MCX August contract); daily 22K benchmark endpoints (keralagold.com) for the FY26 change.
Rupee gold year-to-date figures are MCX gold futures, per 10g. The FY26 change is computed from daily benchmark endpoints (April 2025 → March 2026). The USD year-to-date figure is WGC's, measured to late June.