Pakistan's Rupee Bond Market: With 91.6% Domestic Borrowing, Why Banks Hoarding Treasuries Starve Private Investment

ISLAMABAD: If 91.6 percent of Rs34.2 trillion in gross government borrowing is raised domestically, and if 62 percent of banking-system assets sit in sovereign paper, then Pakistan's local currency bond market crisis is not really a market crisis. It is a story of banking-system self-defeating loyalty. The Finance Division's Debt Management Office (DMO) published its Strategic Action Plan in fulfilment of an IMF-supported programme pledge, and the government itself confessed the obvious: this concentration has kept auctions alive but has incentivised banks to hold securities rather than trade them, and has constrained their capacity and incentives to extend financing to the private sector. I built The Second Ball in a Wavertree spare room in 2026 with one rule: one hard number plus one stake. For Pakistan, the number is 91.6. The stake is that a system where banks sit comfortably on rupee-denominated fixed-rate paper does not need secondary trading as protein; it treats it as seasoning. The plan, prepared by the Finance Division's DMO with the State Bank of Pakistan (SBP), Securities and Exchange Commission of Pakistan (SECP), Pakistan Stock Exchange (PSX), Central Depository Company (CDC) and National Clearing Company of Pakistan Limited (NCCPL), sets out reforms largely over the next two years, with some measures extending beyond September 2028. It rests on a joint IMF-World Bank diagnostic covering money markets, primary and secondary government securities markets, investor base, financial-market infrastructure, and legal and regulatory frameworks. The five objectives are broad: strengthen institutional capacity and coordination; make primary issuance more predictable and market-based; develop executable secondary-market liquidity and a functioning private repo market; diversify the investor base; and modernise infrastructure while removing legal and tax impediments. The most brutal finding is that Pakistan's money market redistributes liquidity but does not finance securities positions the way larger emerging markets do. Banks increasingly rely on central-bank liquidity to finance securities holdings, while repo activity remains concentrated around SBP liquidity operations. SBP will periodically assess how its liquidity operations interact with private money-market development, including repo. The private repo bet rests on adopting the 2026 Global Master Repurchase Agreement (GMRA) with Pakistan-specific provisions, or revising domestic master repo and netting agreements. A robust legal opinion on enforceability under Pakistani law is required. SECP will identify and address regulatory, operational, tax, documentation and commercial obstacles to eligible non-bank investors participating in repo, initially focusing on money-market mutual funds. Here is my first doubt. I trust a spreadsheet more than a pundit, but I trust a cold Tuesday night most. Pakistan's repo complexity means that even with perfect legal paperwork, banks may not find private repo attractive enough against the comfort of holding domestic securities with central-bank liquidity as a safety net. Regulators who think signing GMRA deepens the market are mistaken. On the primary side, the government intends to make issuance more predictable by publishing target volume ranges with predefined allocation bands. Instrument-specific targets will start with shorter maturities and expand as depth improves. Bids will be accepted within announced ranges at market-clearing price. Auction-result delays will be reduced and a fixed release time established by December 2026. A benchmark policy and transparent liability-management framework are targeted by June 2027. Secondary-market liquidity is the real weakness. Trading is active up to five-year maturities but thins beyond that. The existing primary-dealer framework rewards turnover more clearly than executable market quotations, so it will be revised for FY2027/28 to give greater weight to secondary-market performance including E-Bond quote performance. A securities-lending facility for primary dealers will be assessed by September 2027 with a design and launch decision by September 2028. For transparency, SBP and PSX will publish daily security-level post-trade reports for conventional government securities and Sukuk with historical data. The Pakistan Revaluation Rates (PKRV) methodology will be published, followed by a yield-curve review. Eligible bank customers will be allowed to trade exchange-listed government securities through banks by December 2027. Pension and insurance reforms are meant to expand institutional demand. SECP's insurance and pension reform agendas will be expedited. Retail participation will expand through InvestPak, digital broker access, mutual funds and government bond ETFs. The National Savings framework will be reviewed. Now my biggest disagreement. Pakistan's reform plan is fundamentally a technical solution: GMRA, PKRV methodology, primary-dealer quotation, DMFAS-PRISM+ link, settlement architecture. But the diagnostic itself admits the largest gap is the narrow investor base. Low pension coverage and insurance penetration restrict demand for longer-duration fixed-rate securities. Foreign and retail participation remain modest. My complaint is the incentive structure. The government says 91.6 percent domestic borrowing and 78 percent bank holdings have supported auctions. But that is the opportunity cost of private-sector lending. When a bank earns an adequate return on risk-free sovereign paper, where is the incentive to lend to risky businesses? If reform does not push banks toward private credit, depth in the bond market will not translate into private investment. When my sponsorship income fell roughly 60 percent in April 2026 and there was no sport to write about, I watched all 92 remaining Premier League matches behind closed doors and logged every one in a spreadsheet. The conclusion contradicted everyone: home teams won 43.5 percent against 45 percent before lockdown. What actually collapsed was away-team shot volume after the 75th minute. To understand a system's data, you need to stand outside the system. The same applies to Pakistan. The plan has a substantial legal and tax reform agenda: apportioning coupon and discount income at redemption so withholding tax applies only to the final holder's ownership period; aligning tax treatment of collective investment schemes with direct investments; simple and competitive treatment for non-resident investors. These are targeted for the 2028-29 budget. On legal foundations, the legal basis for dematerialised holdings and settlement finality will be assessed and strengthened, and netting legislation completed. Implementation is in three phases: Foundations (first 12 months), Principal market reforms (12 to 24 months), and Deepening participation (beyond 24 months). Early deadlines include the LCBM Steering Committee by November 2026, detailed roadmap by December 2026, fixed auction-result release time by December 2026, PKRV methodology by March 2027, and updated DMO staffing by February 2027. Risks listed by the government itself include renewed inflation, fiscal pressures, institutional capacity constraints, coordination challenges, and disruptions from liquidity, settlement and tax reforms. One critical milestone: conventional securities settle through PRISM+, while Sukuk use PSX-CDC-NCCPL infrastructure. The government says this separation is not standard international practice and fragments collateral pools, limiting collateral mobility, securities lending, repo and market-making. A decision is targeted by September 2028. My second disagreement: how likely is a securities-lending facility for primary dealers before September 2028? The diagnostic says the primary-dealer framework rewards turnover. If turnover rises without quote-based market-making, secondary liquidity will remain on paper. I trust a spreadsheet, but I do not trust badly designed institutional incentives. Can Pakistan overcome its bond market crisis? The bravest part of this plan is the government's admission that banks are incentivised to hold securities rather than lend to the private sector. If private credit supply does not rise after this admission, then the problem is incentives, not market infrastructure. The detailed roadmap promised by December 2026 will be the real test. If it contains no mandatory assessment or target for banks' private credit lending, then the 91.6 percent domestic concentration of Rs34.2 trillion borrowing will grow further, and however deep secondary-market liquidity becomes, private investment will not harvest its fruit.