Inside Busan’s Failed Private-Finance Subway Deal

A privately financed screen-door project allowed Busan to build urgently needed safety infrastructure without paying the full cost upfront. Sixteen years later, the operator is bankrupt, the contract has survived it, and the city’s transit corporation faces a reported liability of roughly 32 billion won. The deeper question is not why one clause went wrong, but what happens when an essential public service is financed against decades of uncertain commercial revenue.

In 2010, Busan had a safety problem and a budget problem. Platform screen doors were becoming increasingly difficult to postpone as the city sought to prevent track falls and other accidents, but Busan Transportation Corporation did not have the capital to install them across the network as quickly as it wanted. Its solution was to bring future commercial income into the present. A private project company would finance and install the doors at 10 major stations, while receiving advertising rights for 21 years and nine months. City council records later described the project at 37.2 billion won, with the long advertising concession serving as the commercial foundation of the arrangement.

The logic was easy to understand. Busan could install safety infrastructure before accumulating enough public capital to pay for all of it directly, while the private operator would recover its investment from advertising sold inside some of the subway system’s busiest stations. The project company subsequently raised 27 billion won in project finance from Busan Bank, according to the transit corporation’s own later account, with a three-year grace period followed by seven years of repayment. The doors entered operation in August 2013.

Sixteen years after the agreement was signed, the doors are still operating. The company built around the project is not. The operator entered rehabilitation and later bankruptcy after its commercial model deteriorated, while a dispute over what Busan Transportation Corporation owed under the agreement moved into court. On July 9, 2026, Busan MBC reported that the corporation had lost at both the first-instance and appellate levels and faced a burden of about 32 billion won, including principal and interest connected to the failed arrangement.

The easiest version of the story is also the narrowest: Busan signed an unusual termination clause, auditors warned about it, the clause was not removed, and the public corporation eventually lost in court. That sequence is important, but it does not explain why the provision existed, how a bank came to lend 27 billion won to a thinly capitalised project company dependent on advertising revenue, or why the public institution that had sought to avoid an immediate capital burden remained financially exposed when the business failed. Those questions point beyond one clause and towards the structure of the project itself.

Busan had paired a permanent public obligation with a commercial revenue stream extending almost 22 years into the future. The social value of platform screen doors did not depend on the advertising market; the doors would remain necessary whether advertisers bought more posters or fewer, whether media budgets moved online or stayed in stations, and whether the project company remained profitable or collapsed. Yet the financial capacity of the company responsible for the project depended on precisely those commercial conditions. Once the gap between the stable public need and the unstable private revenue widened, the contract had to answer a question the business model could no longer solve: who would absorb the remaining financial value of the project when the company created to operate it could no longer survive?

The answer was shaped by more than advertising rights. Investor documents show that the financing was surrounded by security over contractual rights, share pledges, bank-account pledges, insurance-related rights, support commitments and a termination-payment mechanism whose projected value was explicitly compared with the outstanding bank debt. The project was not simply a private company taking a commercial bet on subway advertising. It was a project-finance structure built around the idea that several layers of protection would preserve recoverable value if ordinary operating cash flow proved insufficient.

That is what makes the Busan case more than a story about a failed advertising company. It is a case study in what private finance actually does when it is used to solve an immediate public-budget constraint. Private capital can change who pays first, when the expenditure appears and which risks are initially placed on investors, contractors and lenders. It cannot, by itself, eliminate the public obligation that required the infrastructure in the first place. The unresolved issue in Busan is whether the risk was genuinely transferred, or whether the transaction mainly changed the route and timing through which that risk returned.

Safety on Credit

Busan’s turn to private capital did not begin with evidence that a commercial advertising company could provide subway safety more efficiently than the public corporation. It began with the difficulty of financing a large programme quickly. In council discussions, transit officials described a network in which only part of the system had platform screen doors and explained that installing the facilities more broadly would require capital the corporation could not easily secure all at once. Private investment offered a way to accelerate construction at 10 major stations while other installations continued through public financing.

That distinction matters because the financing model solved a timing problem before it solved an efficiency problem. Under ordinary public procurement, the corporation would have had to secure budget authority, raise capital or obtain government funding before commissioning the work. Under the private arrangement, the project company could borrow and build first, while recovering its investment over a concession lasting nearly two decades. The immediate pressure on the public institution was reduced, but the economic burden of the project did not disappear; it was converted into a long sequence of assumptions about advertising revenue, operating costs, financing charges, debt repayment and residual contract value.

The IMF has long warned that privately financed public infrastructure can create a form of fiscal illusion when the attraction of lower initial expenditure obscures longer-term commitments and contingent liabilities. Its guidance recommends that governments identify, quantify and disclose both expected and contingent PPP costs and ensure that budget and accounting frameworks capture the full fiscal consequences rather than focusing on the first-year cash effect.

The Busan project presented a particularly sharp version of that problem because the service and the revenue source behaved differently. Platform screen doors provide a continuous safety function. Their necessity does not fall when advertising expenditure falls, and their maintenance cannot be suspended because a commercial operator has a weak quarter or loses access to financing. Advertising, by contrast, is discretionary and exposed to changes in economic conditions, media technology, consumer attention and competition among platforms. The question was therefore not simply whether advertisements could produce useful supplementary income for a transit system. It was whether a revenue stream subject to long-term commercial uncertainty could reliably carry the debt and operating obligations attached to an essential public facility.

That mismatch does not prove that advertising income was an inherently improper funding source. Transit systems, airports and other public facilities routinely use retail, property and advertising income to supplement revenue. The difficulty increases when commercial income becomes part of the core debt-service structure rather than an additional source of cash. In that situation, a prolonged revenue shortfall affects more than profitability: it can weaken the project company’s ability to maintain the asset, meet construction-related obligations, service bank debt and remain a viable contractual counterparty.

World Bank guidance on risk allocation emphasises that efficient private participation depends on assigning risks to the party best able to manage them, not simply transferring as much risk as possible to private investors. A private operator can control sales effort, pricing, advertising inventory and some operating costs, but it cannot control the long-term evolution of the media market. The public authority, meanwhile, cannot abandon a safety-critical asset merely because commercial conditions deteriorate.

The first years of the Busan project show how quickly the margin for error narrowed. A 2014 investment prospectus reported that, at the end of 2013, Humetro Lix had assets of about 32.3 billion won and liabilities of approximately 45.1 billion won, leaving negative equity of roughly 12.8 billion won. The company reported about 2.2 billion won in revenue, an operating loss of about 1.3 billion won and a net loss of nearly 3 billion won. These figures do not establish that the project was doomed from the beginning; highly leveraged infrastructure companies can show weak balance sheets during construction and early operation. They do show that the company entered the operating phase with limited capacity to absorb years of underperformance.

The same prospectus recorded another sign of strain. Vitzro Sys, a construction investor in the project, had performed substantial work and had 5.5 billion won in construction payments retained under the financing arrangements. A scheduled release of 1.5 billion won due in August 2014 was not paid, and the companies were negotiating a revised schedule. That missed obligation was not the first principal repayment on the Busan Bank PF loan and should not be described as such; it was a construction-related payment held back within the wider financing structure. Even so, it showed that cash commitments were already becoming difficult to meet shortly after the project entered full operation.

A full diagnosis of the business still requires evidence not available in the public record reviewed for this article: the original financial model and the project company’s actual annual advertising performance. Without those data, three distinct explanations remain possible. The advertising forecast may have been unrealistic from the beginning; the model may have been reasonable when signed but later damaged by structural changes in the advertising market; or the company may have generated meaningful commercial revenue but carried too much debt and too little financial buffer to survive ordinary volatility. A serious assessment needs annual forecast revenue, actual revenue, operating expenses, maintenance costs, financing charges and debt-service data. Bankruptcy alone cannot tell which part of the model failed first.

The difference is central to public accountability. If the revenue forecasts were excessively optimistic at financial close, the main failure lies in project appraisal. If the project initially performed and was later overwhelmed by structural change, the question shifts towards resilience and adaptation. If the underlying business generated cash but the capital structure was too aggressive, leverage and lender protections become more important than the advertising market itself. The public debate has so far concentrated on the final termination obligation; the original cash-flow assumptions have received far less scrutiny.

That omission matters because advertising rights were only the beginning of the financing architecture. The project company had to transform a highly uncertain stream of future commercial revenue into something a bank would finance today. That transformation required contracts and protections whose importance became much clearer after the business weakened.

Making the Deal Bankable

By 2014, investor disclosures described a financial structure more complicated than a company selling advertisements and repaying a loan from the proceeds. The project was described as a BOT arrangement covering 10 stations and an operating term of 21 years and nine months, with estimated project costs of about 36.6 billion won in the prospectus and a total borrowing application of 27 billion won. The same document described a 10-year debt period, a three-year grace period, seven years of principal amortisation and an assumed annual interest rate of 6.7 percent.

The lender protections extended beyond expected advertising receipts. The prospectus listed an assignment by way of security over rights under the implementation agreement, share pledges, security over bank accounts, insurance-related claims, security assignments over project documents and shareholder support commitments under specified conditions. It also described the retention of construction payments for debt-service purposes. These arrangements do not prove that Busan Transportation Corporation guaranteed the bank loan. They do show that the value of the public contract was integrated into the wider security architecture surrounding the project.

That distinction is essential. Project-finance lenders do not rely on a forecast simply because it appears in a business plan. A special-purpose project company generally has limited assets outside the concession itself, so lenders seek protection in the revenue stream, project agreements, accounts, shares, insurance rights and mechanisms that preserve value when an operator defaults. Standard BTO guidance in South Korea similarly treats risk allocation, early termination, termination-payment calculation, transfer and security of rights, project-company changes and refinancing as interconnected parts of a long-term concession framework.

The Busan prospectus is particularly revealing because it explicitly discussed termination value as part of the assessment of recoverability. According to its description of the implementation agreement, a termination attributed to the project company would produce a payment calculated at 85 percent of the present value of agreed future net cash flows, while termination attributed to Busan Transportation Corporation was described at 100 percent. The same document presented a table comparing projected operator-default termination payments with the scheduled balance of the Busan Bank loan. For June 2014, the projected termination payment was 31.163 billion won against a loan balance of 27 billion won; under the prospectus model, projected termination value remained above scheduled remaining debt in later years.

Those figures were projections contained in an investor document, not a judicial determination of what the corporation ultimately owed, and later audit materials appear to describe a different percentage in at least some circumstances. The discrepancy needs to be resolved against the original agreement and all amendments. What the prospectus establishes more securely is that termination value was not an obscure contractual footnote discovered only when the company collapsed. Parties financially exposed to the project were already modelling it alongside debt recovery years before bankruptcy.

That does not make termination payments inherently improper. They are a normal and often necessary feature of long-term private infrastructure contracts. A lender is unlikely to finance an asset with a long concession life if all project value disappears the moment the public authority or operator terminates the agreement. Korean model BTO agreements formally separate termination causes, termination effects, termination-payment calculations, payment methods, transfer of rights and refinancing, reflecting the fact that these questions determine the economic value of the concession under failure as well as under normal operation.

The harder question in Busan is whether the protection remained proportionate to the risk the public side believed it had transferred. A project can transfer genuine commercial and operating risk to private shareholders while still preserving a contractual route through which severe failure creates a large public obligation. The two ideas are not contradictory. The project company can lose money, investors can suffer losses and lenders can spend years pursuing recovery, while the public counterparty remains exposed to a separate termination obligation. The analytical task is to identify which layer of risk sat with which party and whether that distribution was understood before the contract was signed.

The design also raises a question about symmetry. Formal Korean BTO guidance contains provisions not only for early termination but also for refinancing procedures and the sharing of refinancing gains. KDI’s guidance library treats changes in debt terms, financial-agreement reporting, measurement of refinancing benefits and amendments to the concession as continuing contract-management issues. The principle is important: a long-term public-private contract should not be governed only around failure. If financing becomes cheaper or the economics of the project improve, the public interest in that change must also be considered.

This makes one element of Busan’s 2016 audit especially important. The official city audit archive confirms that Busan conducted a special investigation into the PSD private-investment project and published the result in January 2017. Council scrutiny at the time had already raised concerns about the project company’s finances, the cost structure and the scale of possible termination exposure. The significance of the investigation lies less in one isolated finding than in the fact that the city was examining a project in which financial performance, contract design, lender protection and public exposure were increasingly difficult to separate.

Information asymmetry compounds that problem. If a long concession depends on commercial income, the public counterparty needs reliable access to the operating data that determine whether the project is healthy, temporarily weak or structurally insolvent. Without detailed revenue, cost and debt-service information, the public institution is placed in a poor position when deciding whether to accept a request for relief, extend the term, renegotiate the concession or terminate the arrangement. The same information becomes relevant again when the public side must value the contract after failure.

Busan’s later figures show how difficult that valuation became. In December 2020, the transit corporation commissioned an accounting review that produced an “appropriate value” of 7.4 billion won for the termination payment, according to its own administrative-audit response. In January 2021, a city transport official told the council that the corporation’s position was 5.9 billion won while Busan Bank’s was 21.5 billion won; the official expected litigation because the gap was too wide for easy settlement.

Those figures should not be arranged as a simplistic progression from 5.9 billion to 7.4 billion to 21.5 billion and finally 32 billion won. They were produced at different times, by different parties and potentially for different purposes. One may have represented a negotiating position, another an economic valuation and another a lender-side claim involving debt and interest. The methodology of the 2020 valuation is not contained in the public summary, and the judgments reported in 2026 are still needed to explain the legal path to the final reported burden. What the differences demonstrate is that the parties did not merely disagree over the market value of advertising space; they were applying fundamentally different views of what the project was worth when ordinary operations had failed.

The financing structure also changes the way the 2016 call for corrective action should be investigated. It is easy, in retrospect, to ask why the transit corporation did not simply renegotiate the termination provision. A more serious inquiry must first establish whether the corporation had the power to alter the relevant economic rights by agreement with the operator alone. If contractual rights had already been embedded in the lender-security structure, changes affecting termination value may have required lender consent or altered the bank’s recovery position. The original financing documents, any direct agreement and correspondence after the audit are therefore essential.

This does not reduce accountability. It locates it more precisely. If the project was made so rigid at financial close that the public corporation could not reduce its exposure once the operator weakened, the important failure may have occurred when the financing was approved. If renegotiation was feasible after the audit but was not pursued seriously, responsibility moves to the years that followed. The public record currently shows the warning, the operator’s worsening condition and the later insolvency, but not the complete negotiations among the corporation, project company and lender that would distinguish those explanations.

What can be seen is the central tension created by bankability. The protections that made the project financeable in 2010 became increasingly important as commercial performance weakened. Busan was no longer deciding whether private money could help build the doors. It was discovering that the contractual architecture created to bring that money into the project would also shape how expensive it was to leave.

The Exit Trap

The deterioration of the project did not move cleanly from weak revenue to contract termination. As the operator weakened, Busan Transportation Corporation’s options became more complicated. The public corporation was dealing with a counterparty whose balance sheet had deteriorated and whose construction delays had produced a substantial damages dispute, but ending the relationship could activate a far larger termination claim. Continuing the concession therefore carried growing credit and performance risk, while leaving it risked crystallising the financial protections built into the original deal.

The first signs appeared early. By the end of 2013, the project company had negative equity, operating losses and limited room to absorb further weakness. In 2014, a scheduled construction-related payment had already been missed. These facts did not force immediate termination, and they do not show that bankruptcy was inevitable. They meant that a project designed to operate until the mid-2030s had entered its commercial life with a financially fragile counterparty, making continuous monitoring and a credible restructuring strategy increasingly important.

By 2016, the issue had become official. Council records described the project structure, raised concerns about the operator’s finances and questioned the potential termination burden. Busan City then conducted its dedicated PSD investigation in November 2016 and published the result in January 2017. From that point onward, the project was no longer a routine operating contract with scattered warning signs; its financial and contractual risks had become a matter of formal public oversight.

The delay-damages dispute reveals why the problem could not be reduced to simple enforcement. The transit corporation pursued the operator over project delays, and later records show that even after the amount was reduced through dispute resolution, collection remained difficult. By 2018, council proceedings were questioning how the corporation could recover money from a company whose ability to pay was already in doubt.

Economically, the corporation could be both creditor and potential debtor. It was entitled to pursue claims against the project company for non-performance, yet the failure of that same company could activate the termination-payment structure under which the corporation itself faced a larger obligation. The positions were legally distinct, but they created a practical dilemma. Aggressive enforcement might protect the corporation’s claim without restoring the business model, while allowing the company to continue left the public side dependent on a weakened operator. Termination offered no costless escape.

This is the point at which the Busan story departs from an ordinary procurement failure. A conventional contractor can often be dismissed and replaced. A project-finance concession is held together by the operator, the lender, the public contract, the revenue stream, shareholder obligations and the value of rights upon early termination. Removing one participant can change the economic position of all the others. By the time the company became financially unstable, the public corporation was no longer choosing simply between a good contractor and a bad one; it was deciding among restructuring, enforcement, continuation and exit inside a financial structure whose rights had already been allocated.

The 2016 audit therefore needs to be read against the narrowing of those choices. The key question is not only whether it warned about risk, but whether a credible exit remained available when the warning arrived. A serious investigation must establish what amendment was proposed, whether Busan Bank’s consent was necessary, whether another operator could realistically replace the project company, whether refinancing was possible and what price each alternative would have imposed on the corporation. The absence of those documents is one of the largest remaining gaps in the record.

The next decisive move came from the lender. Busan Transportation Corporation’s own 2021 chronology records that Busan Bank filed for the project company’s bankruptcy on April 5, 2019. Vitzro Adcom responded by seeking rehabilitation in June, and the rehabilitation process formally began in July. When no rehabilitation plan was submitted by the deadline at the end of 2020, the process was terminated in January 2021.

The lender’s decision changed the nature of the project. Before insolvency, the corporation could deal with delayed payments, maintenance, damages and performance through contract management. Once the principal creditor sought bankruptcy, the company’s survival became part of a court-supervised contest over claims and recovery. The termination obligation, once a contingent risk, moved towards becoming an immediate financial dispute.

By then, the corporation knew how far apart the parties were. Its 2020 accounting exercise produced a value of 7.4 billion won. In the January 2021 council exchange, the city described the corporation’s position at 5.9 billion won and Busan Bank’s at 21.5 billion won, while the discussion also referenced a higher bank-side figure. The transport director said the gap was unlikely to be resolved without litigation.

The corporation then prepared for exactly that outcome. Its administrative-audit response says it retained a major law firm for legal review and planned specialist bankruptcy representation, with an anticipated sequence of bankruptcy, exercise of termination rights, negotiation and litigation. These steps are important because they show that the corporation did not stumble into court without recognising the danger. By 2021, it was managing a crisis it already expected.

The important distinction is between crisis management and risk management. Hiring accountants and lawyers after rehabilitation has failed may be necessary and professionally responsible, but it does not answer whether the underlying exposure could have been reduced earlier. The years between the 2016 investigation and the 2019 bankruptcy petition therefore remain the crucial unresolved period. The city had formally identified problems; the project company was financially weak; the corporation was pursuing unpaid claims; and the lender’s financial position remained linked to a large project loan. The unanswered question is whether any workable restructuring path existed before the bank moved.

This is what made the project an exit trap. Continuing with the operator meant depending on a financially unstable company. Enforcing claims meant seeking cash from the same weakened counterparty. Renegotiation could alter lender value and might require concessions from several parties. Termination risked crystallising a substantial payment. Insolvency took the project out of ordinary contract management while leaving the safety infrastructure in public service.

The public nature of the asset made the final layer of risk difficult to escape. Whatever happened to the advertising operator, the subway system still needed functioning screen doors. A private company could leave the market. The public corporation could not leave the service. This does not mean the public sector bore every loss or that private risk transfer was fictitious; the project company failed and investors, related parties and creditors faced consequences. It means that contractual risk transfer and ultimate responsibility for continuity are not the same thing.

The court dispute reported in 2026 brought that distinction into the open. Busan MBC said the transit corporation had lost in both the trial and appellate courts and faced approximately 32 billion won in principal and interest. The original judgments remain essential for establishing the precise legal reasoning, the composition of the amount and the procedural route through which the claim was enforced. Until those records are examined, the 32 billion won figure should be described as the reported legal burden, not automatically as the project’s net economic loss to the public.

That distinction should not weaken the investigation. A proper fiscal accounting needs to separate court liability from the residual value of the installed facilities, later advertising income, maintenance costs, litigation expenses, recovered claims and any distributions or offsets arising from insolvency proceedings. The doors were built and remain economically useful. The litigation figure is therefore not, by itself, a complete measure of whether the original project delivered value for money.

But the existence of residual value does not resolve the structural question. A project chosen to reduce immediate fiscal pressure eventually produced a large contractual dispute precisely because the company carrying the commercial risk disappeared while the public service and the financing rights survived. The problem was no longer whether the operator could sell enough advertising. It was who would absorb the value left behind when the original revenue model ceased to function.

Public Risk Outside Public Rules

The deepest weakness in Busan’s screen-door project may lie in the institutional space between ordinary public procurement and formal private-investment governance. The transaction combined a public-safety asset, a commercial concession lasting almost 22 years, a heavily leveraged project company, lender protections and a potentially large termination obligation. Whatever administrative category was attached to the agreement, those characteristics created a level of financial complexity that required sustained scrutiny over the entire life of the project.

Busan eventually subjected the PSD arrangement to a special investigation, but the timing is significant. By the time the city opened its review in late 2016, the agreement had been in force for years, the facilities were already operating and the project company had shown signs of financial weakness. The central issue is therefore not whether the project was ever examined, but whether the most consequential questions—revenue resilience, leverage, lender protections, termination costs and the public corporation’s ability to intervene—were tested rigorously enough before the structure became difficult to change.

South Korea’s formal BTO framework illustrates the breadth of governance required for a long-term private infrastructure concession. KDI guidance separates feasibility and delivery-method analysis from the later management of risk allocation, operator default, early termination, termination payments, transfers of rights, changes in project ownership and refinancing gains. The point of such a framework is not bureaucratic completeness for its own sake. It recognises that a concession is a financial relationship extending far beyond the construction phase, and that public value can be lost not only through excessive building costs but also through weak revenue assumptions, poorly designed exit provisions, information asymmetry and an inability to respond when financing conditions change.

The Busan project raises the question of whether scrutiny followed the economic substance of the deal closely enough. A transaction can appear modest in the public budget because the private side finances construction, yet still create a long-duration public exposure through exclusive commercial rights, lender security, adjustments to the concession and obligations triggered by early termination. The initial absence of a large public payment does not make the arrangement fiscally simple; it may simply distribute the risk across documents, institutions and future years in ways that are harder to see at the moment of approval.

That fragmentation is visible in the Busan record. The transit corporation managed the concession and later pursued claims against the operator. The city exercised audit and ownership oversight. The council repeatedly questioned project cost, the operator’s finances, delay damages and eventual insolvency. Accountants were brought in to value the termination obligation, while lawyers prepared for litigation and the lender pursued recovery through bankruptcy proceedings. These actions demonstrate that the project was not ignored. The problem is that they occurred through separate institutional channels while the financial risk itself was integrated: revenue performance affected debt service, debt protection affected the value of contractual rights, termination provisions affected the cost of exit and insolvency brought all of those elements into the same dispute.

The conflicting valuations that emerged near the end of the project illustrate the consequences of that fragmentation. The corporation obtained a 7.4 billion won valuation, city officials later described the corporation’s own position at 5.9 billion won and the lender’s at 21.5 billion won, and Busan MBC eventually reported a burden of approximately 32 billion won after the corporation lost at two court levels. Those figures were produced at different times and may not represent the same legal or accounting concept, but the gap among them shows how far the parties had diverged in their understanding of the project’s downside exposure. What is still missing from the public record is evidence that one institution had continuously modelled that exposure from financial close through deterioration, restructuring and insolvency.

Strong infrastructure governance is designed to prevent precisely that kind of discontinuity. It separates the public need for an asset from the question of how the asset should be financed and delivered. Busan’s need for better platform safety could have been compelling while the choice of a leveraged advertising concession still required independent examination. The relevant questions were different: whether advertising revenue could withstand long-term market change, whether the project company had sufficient capital to absorb underperformance, whether the lender protections were proportionate to the risks being transferred and whether the public corporation had the information and contractual authority needed to intervene before failure.

The distinction becomes more important when private finance is chosen because public capital is scarce. Fiscal pressure can make an alternative financing structure attractive precisely because it reduces the immediate cash burden, but that is also when independent scrutiny is most necessary. A public institution facing urgent investment needs is vulnerable to evaluating a project by whether it solves today’s funding problem rather than by whether the full contract remains sustainable over two decades. International fiscal-risk guidance therefore emphasises that privately financed infrastructure should be assessed against long-term expected costs and contingent liabilities, not simply the absence of a large upfront appropriation.

Busan’s original constraint was real. Delaying safety improvements also carries costs, and public institutions cannot always wait for ideal fiscal conditions before investing. Private capital can accelerate useful projects and transfer genuine construction, operating and commercial risk. The lesson from the PSD project is not that public enterprises should avoid private finance, but that urgency should raise the standard of diligence rather than lower it. A contract lasting more than 20 years requires the public side to test the original financial model, verify actual revenue, understand changes in debt and ownership, monitor lender protections, capture benefits when financing improves, and compare the costs of continuation, restructuring and termination before insolvency removes most of those options.

That continuing capacity matters because financial close is not the end of risk allocation. Markets change, revenue assumptions weaken, interest rates move and shareholders change. A project that was commercially plausible at signing may require adaptation years later, but renegotiation can itself redistribute value. A weak operator may seek more time or more favourable terms, a lender may resist changes that reduce recovery and a public authority may accept short-term continuity because termination appears more expensive. Unless the public institution has reliable operating data, independent financial expertise and a clear understanding of lender rights, “renegotiation” can become little more than a recommendation without a credible path to implementation.

The 2016 Busan investigation should be judged against that standard. By then, the question was no longer simply whether the original agreement had weaknesses. The more important issue was whether the public side still had the tools needed to correct them. That would have required a clear view of the operator’s actual cash flow, the consent rights attached to the financing structure, the availability of replacement operators or refinancing, and the relative cost of continuing, restructuring or terminating the deal. The public record reviewed so far does not show that those options were analysed together in a way capable of guiding a decisive intervention.

This is the broader significance of the Busan case. The project did not fail because every institution was passive. Oversight occurred, enforcement occurred and professional advisers were eventually hired. The weakness appears to have been one of integration and timing: the public response was divided among institutions and often became stronger only after the financial structure had already narrowed the available choices. By the time the lender moved for bankruptcy, the public corporation was no longer deciding how best to govern a long-term concession; it was trying to limit its losses inside a structure whose principal rights and obligations had already been fixed.

The policy question therefore extends beyond the officials who negotiated one clause. Public enterprises that enter long-term, debt-backed commercial arrangements need governance systems capable of matching the complexity of the risks they accept. That means treating a project’s financial structure, revenue assumptions, lender protections and exit obligations as part of public-risk management from the beginning, even when the initial construction cost sits on a private company’s balance sheet rather than in the current public budget.

Busan found a way to finance the infrastructure earlier. What the later dispute exposed was the absence of an equally durable mechanism for governing the risk that came with it.

The Bill Was Deferred, Not Erased

Busan’s screen-door project did achieve the most immediate objective for which it was created: the doors were built, entered service and continued performing a public-safety function long after the company established to finance and operate them had collapsed. That fact matters because it rules out the easiest version of the story. The project cannot be evaluated as though 37.2 billion won had been spent and nothing of value remained, nor can the roughly 32 billion won burden reported after two court losses automatically be treated as the project’s complete net cost to the public. A serious accounting would have to separate the legal payment obligation from the residual value of the installed facilities, advertising income earned after the original operator’s failure, additional maintenance costs assumed by the corporation, litigation expenses, claims recovered from the insolvent company and any other proceeds, offsets or distributions connected to the bankruptcy process.

The need for that distinction does not make the project’s central problem smaller. It clarifies what went wrong. Busan turned to private finance because the cost of expanding a safety programme arrived before the transit corporation believed it could finance the work through conventional public spending. The transaction resolved that immediate timing problem by placing construction on a project company, borrowing against almost 22 years of advertising rights and surrounding the future revenue stream with contractual protections intended to preserve value for investors and lenders. The physical infrastructure could therefore be delivered earlier, while payment and risk were stretched across a much longer period.

For much of the project’s life, the commercial risk appeared to sit where private finance was supposed to place it. The project company had to sell advertising, control costs and service debt; as the business weakened, shareholders, related companies and creditors were exposed to real losses and recovery disputes. The private sector did not escape the failure unharmed. Yet the collapse of the operating company did not eliminate the public need for the infrastructure, and it did not extinguish the legal rights created when the financing was arranged. The transit corporation eventually confronted a different layer of exposure, one that emerged only after ordinary commercial performance had failed: the contractual cost of ending a concession whose safety function had to continue even after its revenue model no longer worked.

That is why the Busan case is more consequential than the discovery of an unfavourable termination clause. The clause mattered because it formed part of a larger financing decision. A stable public obligation was paired with an uncertain commercial revenue stream, and that stream was then embedded in a debt-backed structure that required protection against failure. Once advertising income could no longer sustain the company, the dispute ceased to be about whether the business could recover and became a contest over how the remaining value of the project should be allocated among the public corporation, the lender and the failed operator’s insolvency estate.

Several important parts of that calculation remain unresolved. The original advertising forecasts still need to be reconstructed against actual annual performance. The relationship between the implementation agreement and the lender-security structure cannot be fully understood without the underlying financing documents. The record of any substantial renegotiation after the 2016 audit remains incomplete, while the corporation’s own valuations, the lender’s claim and the amount later reported after the court defeats need to be compared against the judgments themselves. Those documents may substantially change the allocation of responsibility among the officials who approved the original arrangement, the managers who inherited it, the city that audited it and the lender that financed it.

They are unlikely, however, to alter the broader economic lesson. Private finance can transfer construction risk, operating risk and part of the commercial downside to investors and lenders. It can bring forward useful investment that a public authority might otherwise delay, and the losses suffered by private participants in Busan show that those transfers were not fictitious. What private finance cannot do is remove the public character of an essential service. When the revenue supporting the financing structure weakens, the public authority still has to decide whether the service continues, who maintains the asset, whether the concession can be restructured, how lender rights are treated and what it will cost to terminate the arrangement.

That residual responsibility is particularly important in infrastructure whose operation cannot simply stop when a business model fails. A private advertising company can enter rehabilitation and disappear through bankruptcy. A subway operator cannot suspend platform safety while creditors, shareholders and lawyers determine how the remaining claims should be divided. The public institution responsible for the system therefore remains the actor of last resort, even where significant risks have been contractually transferred to private parties.

The deeper weakness emerging from the Busan record is therefore not simply that the corporation borrowed against future advertising income. It is that the capacity to deliver the project appears to have developed more quickly than the institutional capacity needed to govern it over its full life. A contract lasting nearly 22 years required a public system capable of testing the original revenue assumptions, monitoring actual cash flows, understanding changes in the financing structure, assessing whether lender protections were still proportionate, intervening while restructuring options remained credible and calculating the cost of exit before insolvency turned a commercial problem into a legal dispute.

The doors survived because the public need for them survived. The company failed because the commercial model did not. Between those two outcomes lay a financing structure that had postponed the corporation’s immediate capital burden but had not eliminated the consequences of failure. By the time those consequences returned to the public side, the project was no longer a debate about how quickly Busan could install safety infrastructure. It had become a test of whether shifting the timing of payment had been mistaken for shifting the ultimate responsibility.

Private finance allowed Busan to pay later. The project now shows why paying later is not the same as making the cost disappear.