The Capital Cost Trick: How Inflated Costs Locked In Inflated Tariffs
How Pakistani IPPs were built on books that priced concrete and steel at three times the global rate, and why your bill has been paying for it ever since
By Asad Baig · Lahore · April 2026 · Approx. 8-min read
The number that makes this story real
In 1988, when HUBCO, Pakistan's first major IPP, was being planned, its projected capital cost was approximately $1.2 billion. By 1995, when the plant was financially closed, the stated capital cost had risen to $1.766 billion. A 47 percent increase in seven years.
The reasons offered at the time included the Gulf War, currency volatility, and design changes. The reality was different.
Subsequent investigations would find that capital costs for Pakistani IPPs were systematically inflated. Some 1994-era plants had effective capital costs three to four times higher than comparable plants in Bangladesh and Vietnam.
There is no honest engineering reason for this. Concrete is concrete. Steel is steel. Turbines from Siemens cost the same in Karachi as they do in Dhaka. The difference, the investigations would conclude, was in the books, not in the plants.
This article explains how the trick worked, why it locked Pakistan into thirty years of inflated tariffs, and why fixing it now requires the forensic audits the 2020 Power Sector Inquiry Report recommended and that no government has yet conducted.
How the trick works
The capacity payment in a take-or-pay contract is calculated as a percentage of what the contract calls the "capital cost" of the plant. It is the amount the investor claims to have spent building it.
Notice what this means. The investor has every reason to make the stated capital cost as high as possible. The higher the cost, the higher the payment. Forever.
Suppose the real cost of building a 200-megawatt plant is $200 million. The contract allows the IPP to bill on a 15 percent annual return on equity. The capacity payment is set at $30 million per year for thirty years.
Now suppose the IPP, through over-invoicing on equipment, inflated engineering services contracts, and aggressive depreciation schedules, reports the cost as $400 million. The same 15 percent return on equity now produces $60 million per year for thirty years. The plant is the same plant. The cement is the same cement. The turbines are the same turbines. The annual payment is twice as high.
Multiply this across thirty plants over thirty years. The cumulative excess to Pakistani consumers is hundreds of billions of rupees. Some sources put the cumulative figure for 1994-era plants alone over Rs. 1 trillion.
WHAT THIS MEANS IN PRACTICE
If the real cost of building a power plant was Rs. 100 and the contracts allowed billing on Rs. 200, then for thirty years the consumer paid double what they should have. Perfectly legally, under the terms of the agreement. The fraud, if it could be called that, was committed at the moment of contract signing, and rendered legally untouchable thereafter.
How the inflation actually happens
There are three classic mechanisms.
Equipment over-invoicing. The IPP buys turbines, boilers, transmission equipment from international suppliers. The supplier issues invoices that overstate the price. The difference flows back to the IPP's offshore accounts, the supplier's books, or both. Pakistan's tax authorities have limited ability to verify equipment prices against international benchmarks, especially when the equipment is procured from supplier intermediaries in tax-friendly jurisdictions.
Inflated engineering, procurement, construction (EPC) contracts. The IPP awards the construction contract to an affiliated entity, often a subsidiary or a partner firm. The EPC contract values are higher than they would be in a competitive bid. The difference is captured by the affiliated entity, which is typically owned by the same family or group as the IPP itself.
Aggressive depreciation schedules. The contract permits the IPP to depreciate the plant on an accelerated schedule for tariff purposes. Faster depreciation means more capital cost is recovered in early years, when the IPP's cash flow needs are highest. The consumer pays back the inflated capital cost faster, and the IPP achieves payback in two to four years instead of the contractual fifteen to twenty.
The 2020 Power Sector Inquiry Report found that 16 IPPs from the 1994 policy had collectively invested Rs. 518 billion of equity capital. They had earned profits exceeding Rs. 415 billion and paid dividends of over Rs. 310 billion. Some plants had achieved full payback of the original equity investment within four years and operated at pure profit for the remaining twenty-five years of their contracts. Profit multiples up to 18 times original investment. Dividend multiples up to 22 times.
Read those numbers slowly. An equity investment of Rs. 518 billion turned into dividends of Rs. 310 billion in less than the contract life. The original investment was returned, in cash dividends alone, multiple times over, while the underlying asset continued to operate and generate further returns.
This is not what successful private-sector enterprise looks like. This is what a wealth-transfer mechanism dressed in private-sector clothing looks like.
Why no Pakistani regulator caught it
The natural question is why the regulator that approved these tariffs did not catch the inflated capital costs at the time.
The answer involves NEPRA, the National Electric Power Regulatory Authority, established in 1997. NEPRA's job, on paper, was to verify IPP cost claims and approve tariffs accordingly. NEPRA's record, in practice, was to accept IPP cost submissions without independent assessment.
The 2020 inquiry was specific. NEPRA "accepted IPP submissions of capital cost without independent assessment, allowing inflated cost claims to be locked into tariff calculations." There was no engineering audit. There was no benchmarking against comparable international plants. There was no due diligence on supplier invoices.
When the 2020 report was released to the cabinet, NEPRA's response was to claim it had not been "consulted" on the report's findings about NEPRA's own conduct. This was technically accurate and substantively meaningless. The findings concerned NEPRA's historical decisions, which were already in NEPRA's own records. The regulator did not need a copy of the report to know what it had approved.
I have written about NEPRA's conduct in detail at NEPRA: The Regulator That Did Not Regulate.
The short version. The regulator that was supposed to constrain capital cost inflation enabled it. The auditors who could have detected it were never engaged. The political leaders who could have demanded engagement were the same people who later defended the contracts.
Why this is hard to fix now
Once an inflated capital cost is locked into a Power Purchase Agreement, it becomes legally protected. Pakistani regulators cannot easily challenge it without triggering international arbitration.
The fraud, if it could be called that, was committed at the moment of contract signing. After that, every monthly payment that flows from the inflated capital cost base is a contractually owed payment. Refusing to pay triggers arbitration. Arbitration produces awards that include damages plus interest. The damages can be larger than the inflated payments would have been.
This is why the 2020 inquiry recommended forensic audits, not unilateral cancellation. A forensic audit can establish, with documentary evidence, what the real capital cost should have been. That establishes a basis for renegotiation. Renegotiation, supported by a credible forensic finding, is how comparable systems in other countries have been reformed.
The forensic audit was never conducted under the PTI government that commissioned the report. It has not been conducted under the Shehbaz Sharif government that followed. The legal framework permits it. The political will to do it has not existed.
What the audit would find
We do not need to wait for a forensic audit to know what direction the findings would take. The 2020 inquiry already documented some of them. Specific findings included:
- HUBCO's capital cost rose 47 percent between 1988 projection and 1995 financial close, with the increase only partly explained by documented cost drivers
- Sahiwal coal IPP, a CPEC plant, had excess capital costs of approximately Rs. 32.46 billion allowed because of misrepresentation by sponsors regarding the period for which interest charges were calculated. The interest was calculated for forty-eight months, when the plant was actually completed in twenty-seven to twenty-nine months
- Combined excess payments to Sahiwal and Port Qasim alone, from misrepresentation of construction timelines, were estimated at $2.5 to $2.6 billion over the contract life
- 16 IPPs from the 1994 policy had cumulative excess returns far above the 15 percent legally permitted return on equity
A full forensic audit, conducted by independent auditors with subpoena power, would extend these findings across all 1994 and 2002 policy plants. The estimated recoverable amount, per the 2020 report, was approximately Rs. 1 trillion.
Read that. One trillion rupees of recoverable excess profits, identified by a Pakistani government commission, and never pursued by any government since.
What this means for your bill
Capital cost inflation translates directly into your monthly bill. The mechanism is mathematical, not opinion.
A higher reported capital cost produces a higher per-unit tariff. A higher per-unit tariff appears on every bill of every consumer who buys electricity from the grid. Every month. For the contract life of every affected plant.
If the real cost of Pakistan's IPP fleet was, say, Rs. 1 trillion, but the contracts allowed billing on Rs. 1.5 trillion, then for thirty years Pakistani consumers have been paying tariffs based on the inflated figure. The cumulative excess, across the fleet and across decades, is in the hundreds of billions of rupees.
Recovery of that excess, through forensic audit and legal process, would not return the money already paid. It might however reduce future tariffs by the inflated component, lower the per-unit cost going forward, and ensure that the same trick is not repeated in future contracts.
This is one of the ten steps in the reform roadmap I describe in my pillar on the IPP system. It is achievable. It requires courts willing to act, regulators willing to investigate, and political leadership willing to back the process. The legal framework exists. The choice is whether to use it.
In closing
The capital cost trick is a quiet kind of corruption. It does not look like a suitcase of cash. It looks like an engineering services contract issued to an affiliated firm at a premium. It looks like an equipment invoice that nobody outside the IPP can fully verify. It looks like a depreciation schedule approved by a regulator that did not interrogate it.
Once the inflated cost is locked into the contract, every subsequent payment is legal. The system cleans itself, on paper. The fraud, if it was fraud, becomes contract performance.
This is why I keep using the word architecture in my writing about the IPP system. The capital cost trick was not an isolated event. It was a designed-in feature of the way Pakistani power plants were procured, financed, and tariffed. It was made possible by a regulator that did not regulate, by a political class that did not interrogate, and by a technical complexity that kept ordinary citizens at a distance from the decisions being made in their name.
You no longer have to be at a distance. The mechanism is now explained, in plain English, in this article. Now you know.
Thank you for reading.
, Asad Baig, Lahore, April 2026
Frequently asked questions
What is capital cost pass-through in Pakistani IPP contracts? Capital cost pass-through is the contract provision that calculates the IPP's capacity payment as a percentage of its claimed capital cost. The higher the claimed cost, the higher the lifetime payment. There is no mechanism to recover excess if the claimed cost was inflated.
By how much were Pakistani IPP capital costs inflated? The 2020 Power Sector Inquiry Report found that some 1994-era IPPs had effective capital costs three to four times higher than comparable plants in Bangladesh and Vietnam. Specific findings include Sahiwal coal IPP having approximately Rs. 32.46 billion in excess capital costs.
How much was HUBCO's capital cost inflation? HUBCO's projected capital cost in 1988 was approximately $1.2 billion. By 1995 financial close it was $1.766 billion. A 47 percent increase, only partly explained by documented cost drivers.
Why was NEPRA unable to prevent capital cost inflation? The 2020 inquiry found that NEPRA accepted IPP submissions of capital cost without independent assessment. There was no engineering audit, no international benchmarking, and no due diligence on supplier invoices. NEPRA's response to the report was to claim it had not been consulted, which was technically accurate and substantively meaningless.
Can the inflated capital costs be recovered now? The 2020 inquiry recommended forensic audits of all 1994 and 2002 policy IPPs to identify recoverable amounts, estimated at approximately Rs. 1 trillion in total. The audits have not been conducted under the PTI or Shehbaz Sharif governments. The legal framework permits them. The political will to conduct them has not yet existed.
Sources and notes
- Power Sector Inquiry Report 2020, Government of Pakistan, headed by Muhammad Ali, former SECP Chairman (ARY News mirror)
- IEEFA Reports on Pakistan Power Sector by Haneea Isaad (2024-2025)
- World Bank PPP Knowledge Hub, Lessons from the Independent Private Power Experience in Pakistan
- The Diplomat, China in Pakistan's Power Sector: The Hidden Costs (January 2025)
- NEPRA State of Industry Reports 2015-2024 (nepra.org.pk)
Related reading from Asad Baig
The pillar this explainer supports
Sibling explainers in this cluster
- What Is a Take-or-Pay Contract? A Plain-English Guide for Pakistanis
- The Working Capital Scam: NEPRA's Quietest Subsidy to IPPs
- NEPRA: The Regulator That Did Not Regulate
Asad Baig