Pakistan counts 7.14 million business establishments in its Economic Census, and only 9.8 per cent of them sit in manufacturing. Wholesale and retail trade accounts for 45.1pc, nearly five times as many. Twenty-seven years of measuring small and medium enterprises (SME) policy success by headcount and GDP share, never by whether small firms are actually financed for production, has led the government to finance consumption.
Trade and consumption-facing services recycle domestic demand; they do not build export earnings, productivity growth, or foreign exchange. An SME policy that channels financing toward that segment by default, because it is sector-blind, not because it is deliberate, is financing consumption growth and calling it SME development.
This is in a scenario where consumption as a percentage of GDP has consistently exceeded 97 per cent, while investment as a percentage of GDP has lagged at an average of 12pc over the last five decades. There does not exist a single example of a middle-income economy that graduated to middle-income status solely on the basis of consumption, without producing or exporting much.
The value chain that actually sustains growth runs the other way, from export-oriented large-scale manufacturers, down through the SME suppliers, component makers, and processors that feed them. That value chain is where Pakistan’s SME financing is not going, and the numbers say so on their own.
Nothing in the current design distinguishes a rupee of credit reaching an exporter’s supplier from a rupee reaching a neighbourhood retailer
A decade of lending data demonstrates that only 5.5pc of all bank credit extended to the manufacturing sector reached SME borrowers. In trade, the same figure rose from 33.4pc to 40.6pc. A bank lending into trade today is more than eight times as likely, proportionally, to be lending to a small firm as a bank lending into manufacturing, and that gap has only widened in the last decade.
Manufacturing’s position within the SME loan book has deteriorated further, more recently. Manufacturing’s outstanding SME credit stock increased to a peak of 46pc in 2021, and it has fallen every year since then, closing May 2026 at 30.4pc. A 15.6-point collapse in five years, even as the national SME loan book grew at a compound 12.5pc a year. The money is flowing somewhere, and that somewhere is trade and services.
Pakistan’s manufacturing base is large in firm count and thin in scale. Roughly, 95pc of manufacturing establishments employ fewer than ten people, as per the Economic Census 2023; a formally registered factory is outnumbered 28-to-1 by an informal production shop.
The magical number of ten can be linked to policy or legal arbitrage, where, as the number of employees increases beyond ten, multiple regulatory and legal requirements begin to shape up, increasing the cost of doing business. Instead of scaling up, firms increasingly prefer to split into multiple entities after hitting the sweet spot of ten employees.
The lending data explains why formalisation does not happen. A firm that cannot borrow against its own production has no route to invest in the certification, equipment, or working capital an anchor exporter’s order size would require. It stays small, stays informal, and drifts toward trading, where credit is actually available.
Manufacturing SME credit does not even track large-manufacturer credit in any stable way, as the year-on-year correlation between SME and non-SME manufacturing lending is a weak 0.19, and its 24-month rolling version has run negative over the past two years.
In trade, the same correlation is 0.59, three times as coherent. The credit market does not connect Pakistan’s SME sector to large-scale manufacturing. A robust SME sector remains a function of strong linkages with large-scale manufacturing.
Germany’s SMEs, called Mittelstand, generate 68pc of national exports from a base that is over 99pc small and medium firms. These firms are embedded as specialised suppliers within large industrial value chains, financed through a network of regional guarantee banks explicitly built to substitute for the collateral that a growing SME does not yet have. Independent studies of those guarantee banks find that roughly 60pc of supported loans would not have been made at all without the guarantee.
Taiwan built the same instrument alongside its subcontracting system from the 1970s onward, guaranteeing SME loans with up to 95pc coverage and no collateral required, so a small supplier could take on a subcontracting relationship with a large exporter before it had the balance sheet a bank would normally demand. That subcontracting layer is where Foxconn and TSMC came from.
Bangladesh increased the value addition in its ready-made garment sector from 19pc to roughly 75pc in knitwear by making backwards-linkage textile mills as bankable as direct exporters through bonded warehouse and deemed export status.
Textiles is the sub-sector in Pakistan most directly comparable to Bangladesh’s story, and it remains Pakistan’s largest export earner. Its SME credit book has grown at a compound rate of 8pc a year since 2015, slower than the national SME average of 12.5pc and trade’s 14pc. The sub-sector most obviously suited to backwards-linkage financing is not receiving it at pace.
SME policy in Pakistan is sector-blind by design, but that needs to change if we actually want any kind of sustainable growth, rather than doubling down on consumption-oriented growth, which hasn’t really worked well. Can’t expect different outcomes while making the same mistakes time and again.
A trading firm and an export-grade component manufacturer qualify for the same refinance schemes, the same subsidised credit lines, the same policy attention, as long as their headcount or revenue fits the arbitrary count that defines a SME.
Nothing in the current design distinguishes a rupee of credit reaching an exporter’s supplier from a rupee reaching a neighbourhood retailer. Both count identically toward the aggregate SME credit growth figure, but the economic impact of credit extended to an exporter remains considerably more than a retailer.
We need to rethink SME policy here. Any policy or intervention for SMEs needs to align with an industrial policy, which may or may not exist. National priorities need to focus on increasing the investment-to-GDP ratio to 30pc, which requires reallocating all policy and financial capital to export- and production-oriented interventions, rather than more trade.
None of this requires new legislation or new institutions, but it does require SME policy to stop treating a rupee of credit as fungible across sectors that share nothing but headcount. A thriving SME ecosystem requires a robust large-scale manufacturing space, strong linkages between the two, and policies that incentivise graduation and institutional scaling, rather than incentivising lemming-like firms.
The writer is an assistant professor of practice at IBA and CEO of National Credit Guarantee Company Limited
Published in Dawn, The Business and Finance Weekly, July 13th, 2026