ABHI IPO: The Gold Behind the Glitter

November 06, 2026 (MLN): Walk into any of ABHI Microfinance Bank’s 114 branches with your gold jewellery, and the bank will lend against it at a loan-to-value ratio that its own prospectus puts at 95%. The indicative yield is 38.3% a year. The typical tenure is about 11 months.

No gold? ABHI will still lend, but the options are narrower and mostly dearer. A branch loan without collateral carries an indicative yield of 45%, the bank’s highest. Small merchants on the DigiKhata and Run Courier apps can borrow at 36%, with half the revenue shared with the platforms.

Salaried staff at partner companies can draw up to half their earned wages for 5–40 days for “fixed fees” the prospectus does not quantify, or take an advance salary at 22%. Taken together, these products are a sideshow: about a tenth of the loan book in June 2026.

The main event is gold. Gold-backed lending now makes up 88.7% of everything ABHI lends and 90.8% of its loan income. It has turned what was FINCA Microfinance Bank, which ended 2024 with negative equity of Rs4.03bn and a capital adequacy ratio of minus 108%, into a bank that earned Rs1.5bn in the first six months of 2026.

Today, ABHI asks investors to buy a piece of that story. Book building for 132mn new shares closes today, 6 November, with a floor of Rs15 and a cap of Rs22.5. Retail subscription follows on 12–13 November. At the floor, the issue raises Rs1.98bn, and every rupee goes into the bank’s capital.

We read the final prospectus line by line, including the audited accounts that appear in it only as scanned images. We found a real turnaround. We also found that almost every part of it leans on the same four supports. The first is gold prices. The second is regulatory forbearance from the State Bank. The third is expensive institutional deposits. The fourth is ABHI (Private) Limited, the bank’s majority owner and parent company, which plays almost every role at once: owner, depositor, product partner, capital provider and member of the bank’s audit committee. If any one of these supports weakens, the others are exposed too.

Credit where it is due

The turnaround is not invented. Net mark-up income more than doubled from the first half of 2025 to Rs4.48bn in the first half of 2026. The bank swung from a Rs1.75bn loss in 2024 to Rs1.02bn of profit in 2025 and Rs1.5bn in six months of 2026. Reported NPLs fell from 10.23% of loans in 2023 to 0.61% in June 2026. Nobody is cashing out in this IPO.

The prospectus is also candid in places. It discloses the auditor’s going-concern warning and explains the change of auditor: a Category A firm becomes mandatory once assets cross Rs100bn. It reveals the 95% loan-to-value and the bank’s write-off history, and it carries risk factors on capital, deposit concentration and non-recurring profit.

The question for investors is not whether the turnaround happened, but what it rests on.

The gold engine, and the interest that hasn't been paid yet

A gold loan at 38% looks like the safest business in banking: if the borrower does not pay, the bank sells the gold. Gold NPLs are just 0.2%, the prospectus says. But one line in the audited accounts complicates the picture.

Mark-up accrued on advances (interest booked as income but not yet received in cash) more than doubled, from Rs2.41bn at end-2024 to Rs5.43bn at end-2025 (Note 16). The prospectus says it rose by roughly another Rs3bn in the first half of 2026 (footnote 85), yet it does not disclose the June balance. The bank attributes the rise to loan growth. It does not say how much of the uncollected interest sits on gold loans, or how often gold borrowers actually pay.

Now connect the dots. If the 95% LTV is measured on principal at the start of the loan, then by maturity principal plus accrued interest can exceed the value of the gold, unless the gold price has risen. Gold has risen strongly through 2024–26, which makes this model look almost riskless. The prospectus acknowledges the danger in its risk factors but never puts a number on what a flat or falling gold price would do. Auction recoveries were Rs111.7mn for all of 2025 and already Rs96.1mn in the first half of 2026.

Is the 95% loan-to-value measured on principal alone, or on principal plus interest? ABHI did not answer.

Physical risk is real too. Note 16.1 of the 2025 accounts (page 136) records a robbery at the bank's Hyderi branch in Karachi that hit 351 gold-loan cases with Rs180.28mn of principal and gold valued at Rs225.35mn. By December 2025, 205 claims had been settled and 146, worth Rs95.69mn of gold, were still pending with the insurer.

The prospectus lists a smaller fake-gold fraud at the Rawalpindi branch (55 loans, Rs62.69mn) among its legal matters, but the Hyderi robbery appears nowhere outside the scanned accounts. Staff turnover, meanwhile, was 55.6% in 2024 and 60.2% in 2025, in a business built on staff holding customers' gold.

Capital built on permission

ABHI’s capital story is a chain of regulatory permissions, and every link has to hold.

Capital is a bank’s own money, the cushion that absorbs losses before depositors are hurt. The regulator measures it through the capital adequacy ratio (CAR), and requires about 15%. ABHI’s is below zero.

The sponsor’s deposit. When ABHI (Private) Limited bought the bank, it promised the SBP Rs4bn of fresh capital. Half came as a rights issue in 2025. The other Rs2bn was placed as a deposit in ABHI (Private)’s own account at the bank. A deposit is money a bank owes, not capital it owns. On 11 March 2026, the SBP nevertheless allowed the bank to count it as capital for its December 2025 ratios (CY25 accounts, Note 50). Even with that help, CAR stood at −20.28% in December 2025 and −6.26% in June 2026.

The IPO. Investors might assume their money fixes the hole. It does not. On the bank’s own arithmetic, the Rs1.98bn lifts CAR only to about −4.3% (section 4.4). Reaching 15% depends on profits of Rs2.93bn in 2026 and Rs4.13bn in 2027. It also needs Rs3.1bn of “other capital movements”, mainly Tier 2 recognition and deferred tax adjustments, which the prospectus does not break down.

The deadline. The SBP’s CAR relaxation expires on 11 October 2027. The bank’s plan shows CAR at 17.77% only in its December 2027 figures. It does not show where the ratio stands on the day the relaxation ends.

The gold cap. SBP rules limit gold lending to 35% of a microfinance bank’s loan book. ABHI’s exemption runs to November 2029. It was granted on 11 November 2024, the same day the SBP approved the acquisition and the CAR relaxation (section 3.12). Yet the bank’s own forecast keeps gold at 70.2% of the book in 2029 and 60.8% in 2030, beyond the exemption. Footnote 127 explains why:

“…given prevailing industry dynamics, there is a reasonable possibility that the regulator may allow some relief in the 35% limit.”

In plain words, the forecast assumes a favour that has not yet been granted.

This is not a side detail. Gold loans carry a 20% risk weight, against 100% for unsecured loans. As the prospectus itself puts it, the same capital “can support close to five times more gold-backed lending”. That is why IPO money is to be “deployed predominantly into the gold-backed portfolio” (section 2.5), even as another section promises to “gradually reduce” gold (section 2.6). The 45% return on equity behind the valuation is averaged over 2026–30, a period built on the gold exemption and, in its final years, on relief the SBP has not yet granted.

The foundation. The bank’s balance sheet carries a Rs3.75bn deferred tax asset, recognised “largely against carried-forward tax losses” (risk factor 5.1.15). It is not cash. It has value only if the bank earns enough future profit to use those losses, a point the auditors flag in an Emphasis of Matter. Yet it is larger than the bank’s entire Rs3.11bn equity. Strip it and the intangibles out, and tangible equity in June 2026 was still negative. Risk factor 5.1.12 is frank: the bank’s ability to continue as a going concern “remains dependent on completion of this Issue and achievement of forecast profitability.”

For an investor, the bet is simple to state and hard to price the regulator must keep saying yes, and every forecast must come true.

The Rs10 shares that arrive after you pay Rs15

On 13 October, the same day it approved the prospectus, the SECP approved the issue of 200mn shares to ABHI (Private) at Rs10 each against the Rs2bn advance. They will be issued "on or before 31st December 2026". That is after IPO investors have paid at least Rs15 and the stock has listed.

The prospectus's ownership tables still show a 13.65% free float. Only a second row in the peer table shows what happens after the sponsor shares are issued. The float shrinks to 11.31%, and earnings per share fall from Rs2.40 to Rs1.99.

The lock-in is looser than it looks. The prospectus cites Regulation 3(1)(ii), under which sponsors must keep 51% of their shares only until the bank posts two consecutive years of profit, and 25% for three years after that. ABHI was profitable in 2025. If 2026 is profitable too, the 51% floor could fall away once those accounts are published. We asked the bank which lock-in exactly applies. It did not respond.

Who really funds a microfinance bank

ABHI calls itself digital-first and inclusion-led. Its deposit book tells a different story. Deposits rose from Rs36.2bn at end-2024 to Rs86.2bn in June 2026, but the number of individual depositors fell from 2.09mn to 1.50mn (3.8.6). The growth came from about 4,000 institutional accounts holding Rs43.3bn, half of all deposits.

The top 25 depositors hold more than Rs37bn. Fixed deposits pay up to 23% (Note 19.2), against a policy rate the bank's own projections put at 10.5% for 2026. Branchless, or digital, deposits are just Rs0.96bn, about 1% of the total, and smaller than six months earlier.

Follow the money the other way. Of the Rs11.2bn ABHI keeps with other banks (Note 8, p.184), Rs7.1bn, or 64%, sits with three microfinance banks: LOLC (Rs4.51bn, rated A-), Khushhali (Rs1.51bn, A-) and Halan (Rs1.12bn, BBB+) (Note 45.2.2, p.206). ABHI’s own peer table shows LOLC and Halan were loss-making in 2025, with returns on assets of −2.2% and −9.4% (p.78). Its risk factors warn that the microfinance sector’s “aggregate capital position has become negative” (risk factor 5.2.7, p.84). Meanwhile, banking companies hold Rs8.0bn of deposits at ABHI (Note 19.3, p.192), even as ABHI places money with other banks. Whether any of these flows are reciprocal is not disclosed.

Related parties are in the mix as well. Deposits from TPL Insurance, a group company of the bank's second sponsor, rose from Rs895mn to Rs2.24bn in the first half of 2026 (Note 43, page 204), almost 90% of all related-party deposits.

Profits, with the shine examined

ABHI's first profitable year was not all operating profit. Of 2025's Rs1.67bn pre-tax profit, Rs627mn was a one-off gain from restructuring a subordinated loan from PMIC, which waived accrued mark-up and granted a mark-up holiday. Rs193mn came from securities gains, and Rs242mn from credit-loss write-backs. Risk factor 5.1.18 flags the first two but not the third.

The clean-up of bad loans was mostly write-offs. NPLs fell by Rs1.61bn between 2023 and June 2026, while the bank wrote off Rs2.97bn (3.6.7). Section 2.6 credits "improved recoveries".

The first half of 2026 again ended with a net write-back of Rs125mn, driven by Rs191mn recovered on loans written off in earlier years (Note 34, page 198). That is a finite pool. Meanwhile the loss allowance fell 11% to Rs540mn while gross loans grew 31%, cutting cover for bad loans from 235% to 176%.

The sponsor in every room

ABHI (Private) Limited, a fintech founded in 2021 whose ultimate parent is Abhi Limited of the Cayman Islands, plays at least six roles in this bank:

• Owner. It holds 73.7% today, rising to 78.8% of pre-IPO capital once it receives its 200mn Rs10 shares.
• Depositor. It moved Rs184.4bn in and Rs180.9bn out of its accounts at the bank during 2025.
• Product partner. The bank "integrated ABHI Fintech's digital offerings", but the revenue-sharing terms are not disclosed.
• Capital provider in kind. The SBP approved a Rs1.41bn property transfer from the parent as capital on 3 July 2026. It appears only in a note to the accounts (Note 49, page 212), with no description, no valuer and no place in the capital plan.
• Overseer. Its CEO Omair Ansari and co-founder Ali Muhammad Raza sit on the bank's audit committee, which reviews related-party dealings.
• Buyer. With TPL Corp, it bought Kashf's 3.97% stake in June 2026, two months before the floor price was set. The price was not disclosed, and neither was what the consortium paid FINCA.

The second sponsor, TPL Corp, took up none of the 2025 rights issue or the Rs2bn advance, and its stake has been diluted from 33.17% at acquisition to 26.27% today, and will fall to 21.19% of pre-IPO capital once the sponsor shares are issued.

At the top

Note 41 of the half-year accounts (page 202) shows two people held the President/CEO position in the first half of 2026. The prospectus does not explain the change. The current CEO received SBP clearance on 26 October, four days before the prospectus was dated, and the CFO is acting.

While the bank’s capital ratio was below zero, money kept flowing to the top. Key management pay rose from Rs78.9mn for 12 people in the first half of 2025 to Rs144.0mn for seven in the first half of 2026: roughly Rs20.6mn a head against Rs6.6mn, including Rs36.7mn in cash bonuses (Note 41, p.202).

Loans to key managers jumped more than tenfold during 2025, from Rs10.7mn to Rs115.9mn. The then CEO borrowed Rs49mn, of which Rs25mn was still outstanding at year-end (p.25). The bank also spent about Rs0.9bn on fixed assets and software in six months (p.171), signed Rs0.7bn of new leases (Note 20, p.192), and plans Rs2.0bn of capex in 2026 (p.222).

 Pay is set by an HR & Remuneration Committee of just two members (p.238). Its chair, Jehangir Shah, is a nominee of ABHI (Private) Limited and chairs the sponsor’s own board (p.229). Investors are being asked to rebuild the bank’s capital; the insiders’ pay has already been settled.

The price tag

The joint lead managers value ABHI at Rs27.90 a share on price-to-book against ROE (p.74) and Rs30.80 on discounted cash flow (p.75). Both rest on the same profit forecasts, which produce a 45% average return on equity. Footnote 43 concedes that figure “is on the higher side due to the Bank’s low equity base in initial years” (p.74). The cost of equity assumes a beta of 1.00, the market average, for a bank with negative capital (p.74).

The headline claim is a P/E of 5.40x, “a discount of 28.40%” to the banking sector’s 7.54x (p.76). That figure counts only the 835mn existing shares. The prospectus’s own table shows P/E rising to 6.25x once the IPO shares are added, and to 7.54x, exactly the sector’s weighted average, once the 200mn sponsor shares are issued by 31 December (p.77). On that fully diluted basis, price-to-book is 2.69x, against a sector weighted average of 1.88x and a median of 0.99x. The JLMs argue that the premium is justified by ABHI’s 2.3% return on assets, nearly double the sector’s (p.78). But the 2.69x itself is calculated on forecast 2026 equity, not equity the bank has yet earned (footnote 72, p.77). Once all the shares are counted, the discount disappears.

The bottom line

ABHI is a real turnaround, but every part of it still depends on something going right. The profits rest on one product, gold, lent at 95% of the metal’s value in a market that has only gone up. The capital rests on regulatory permissions: a sponsor’s deposit counted as equity, a CAR relaxation that ends before the bank’s own plan reaches compliance, and a gold-cap exemption that the forecasts assume will be extended. The equity rests on a Rs3.75bn tax asset that is worth something only if the forecasts come true. The deposits rest on a few dozen institutions paid up to 23%.

The IPO does not change that. At the floor, Rs1.98bn lifts the capital ratio only to about −4.3%. The 200mn shares the sponsor receives at Rs10 by 31 December will shrink your share of the bank. Once those shares are counted, the advertised discount to the sector is gone.

To buy at Rs15 or above, an investor has to believe four things at once: gold prices will hold, the State Bank will keep saying yes, the institutional deposits will stay, and every profit forecast will be met. If all four hold, the upside is real. If one fails, the rest of the structure is exposed too, and investors would be the first to absorb the loss.

What ABHI has not answered

Our questions were grouped into eight themes, each covering the points in this article where the prospectus leaves the most open.

We began with the regulator. We asked whether the SBP has signalled any relief on the 35% gold cap after November 2029, or any extension of the CAR relaxation beyond October 2027. We also asked what would happen to earnings, capital and the bank itself if neither is granted.

On gold lending, we asked how the 95% loan-to-value is measured, how much interest remains uncollected, and how much of it borrowers actually pay in cash before maturity. We asked at what price fall auctions are triggered and what losses they have produced. We also asked how the bank protects customers’ gold when staff turnover is close to 60%, and why the Hyderi robbery does not appear among the prospectus’s legal matters.

On earnings, we asked what the bank’s profit would be without one-off gains and write-backs. We asked why the loss allowance fell while the loan book grew, and how much of the clean-up came from genuine recoveries rather than write-offs.

On capital, we asked for a breakdown of the “other capital movements” the plan relies on, and where CAR would stand if profits fell short. We asked about the deferred tax asset’s recovery timeline, the cost of the PMIC loan once mark-up resumes, and how the bank is performing in the current quarter.

On funding, we asked who the largest depositors are, and whether the money placed with LOLC, Halan and Khushhali comes back as deposits. We also asked on what terms TPL Insurance holds Rs2.24bn at the bank.

On ownership, we asked why the sponsor’s 200mn shares are being issued at par after the IPO, which lock-in applies to each sponsor, and what was paid for the stakes bought from FINCA and Kashf.

On governance, we asked about the agreements between the bank and its parent, the Rs1.41bn property the parent transferred, and how the audit committee handles related-party conflicts. We also asked about the two CEOs in the first half of 2026, the acting CFO, and how bonuses and spending were approved while capital was negative.

Finally, on the business model, we asked whether digital lending makes money, what earned-wage-access borrowers really pay, and whether the gold-tokenisation and crypto-corridor plans described in July are part of the bank’s future.

We sent ABHI our questions and CEO Muhammad Farooq Kamran promptly acknowledged them, promising a reply “at the earliest”. We held this article for an extra day to allow time for a response, but none had arrived by publication. We will gladly publish the bank’s answers in full as soon as we receive them.

Copyright Mettis Link News