Part I — Situation overview

The government decree ordering a full review of the state supports and guarantees attached to private-market lending appeared in the Hungarian Official Gazette on the evening of 3 August 2026. According to HVG’s account, the decree explicitly charges Kapitány István — designated as the head of the portfolio responsible for the economy and energy, with the title used by the press — with ordering the review “in order to fulfil the milestones necessary for placing the resources of the Recovery and Resilience Plan in the service of the Hungarian economy”, with the involvement of Finance Minister Kármán András. The aims set out are unambiguous: the level of state risk-sharing should decrease significantly, the banks’ own credit risk assessment should be strengthened, and guarantee coverage ratios should improve.

The schedule is unusually tight. According to press reports, the amortisation schedule of every state counter-guarantee — that is, the secondary suretyship provided by the state behind the guarantee institutions — has to be reviewed by 12 August; by 17 August the draft government decree has to be prepared that undertakes to reduce the ratio of state guarantees to gross domestic product (GDP) to the EU average; at the end of August the reshaping of the Széchenyi Card Current Account Credit MAX+, the Széchenyi Tourism Card MAX+ and the Széchenyi Liquidity Loan MAX+ enters into force, under which the net transaction interest is tied to the three-month interbank rate (BUBOR), while the drawing of the previously compulsory guarantees becomes an optional choice; and likewise by the end of August the official letter goes out to the European Commission in which the Hungarian state undertakes to implement the reform. The numerical core: the government is reducing the upper limit of state counter-guarantees set for this year from HUF 12,800 billion to HUF 11,200 billion, and the saving has to be taken into account already in the planning of the 2027 budget. The state guarantees covering the lending of the Hungarian Development Bank — among them the capital, interest and exchange-rate equalisation schemes — come under separate examination, with a phase-out schedule broken down by year.

According to MIAK’s reading the direction of the reform is defensible, and the financial logic behind it is correct: the state guarantee is one of the least transparently accounted liabilities of the public finances, because an interest subsidy is immediately visible as current expenditure, whereas a guarantee remains invisible for years as a contingent liability — that is, as an item that becomes an actual payment only when non-performance occurs. The policy problem is not the aim but the order. The decree sets a stock figure (the reduction of the guarantee ceiling and of the GDP ratio) within five weeks, while not a single document is available on who draws on today’s guarantee stock — in what breakdown by company size, sector and region — or on how large the expected loss of the system is. Reducing a ceiling figure is not in itself a policy decision: the substantive question is at which group of businesses the HUF 1,600 billion of dropped ceiling will appear as a rejected loan application. There is no public data on this today, and the schedule leaves no time for there to be any.

Part II — Literature foundation

Three conceptual frames give the basis of principle for the proposals. The methodological insight of the work This Time Is Different by Carmen Reinhart and Kenneth Rogoff (Harvard economists, researchers who processed eight centuries of the history of financial crises) is that analysis simply leaves out of the picture those liabilities whose data are hard to access, and thus the actual sovereign debt exposure is systematically underestimated — this is precisely the position of the guarantee stock. The volume Globalization and Its Discontents by Joseph Stiglitz (American economist, laureate of the Nobel memorial prize in economics in 2001, former chief economist of the World Bank) describes in its chapter on moral hazard how the prospect of a bailout reduces the lender’s caution in screening the client — that is, the objective of the present reform, the strengthening of the banks’ own risk assessment, is economically well founded. The volume 23 Things They Don’t Tell You About Capitalism by Ha-Joon Chang (a Korean-born economist at a British university, a researcher of development policy and industrial policy) provides the counterweight: according to his argument governments are indeed capable of taking good economic decisions, and the condition for this is not commitment of principle but the obtaining of better quality information for the decision. The detailed treatment of the literature — author by author, with quotations — can be found in the 6.4 Literature in detail section.

📖 Source: Carmen Reinhart – Kenneth Rogoff: This Time Is Different; Joseph Stiglitz: Globalization and Its Discontents; Ha-Joon Chang: 23 Things They Don’t Tell You About Capitalism

Part III — MIAK’s concrete proposal

MIAK proposes three measurable measures. None of them calls into question the aim of reducing the guarantees; all three concern the order of implementation.

3.1 A quarterly report on the guarantee stock and the expected loss (simultaneously with the draft of 17 August)

Alongside the reduction commitment, the public picture of the exposure should be ready on the very same day. MIAK proposes that the finance portfolio publish a quarterly report containing at least four sets of data: the outstanding stock of state counter-guarantees broken down by legal title, the contract stock managed by the guarantee institutions, the expected loss calculated on the portfolio and the amount of actual calls, and the direct public-finance exposure to the operations of the Hungarian Development Bank. The report should follow the structure of the fiscal risk statements used in international financial practice, and should also appear in machine-readable form. This proposal follows from the G1 data-driven budget and the G23 sovereign debt sustainability framework programme points. The Reinhart–Rogoff argument (see 6.4.1) is directly applicable here: as long as the data on the guarantee stock are not available in a regular, comparable form, the exposure does not even feature in the sustainability analyses — not because anyone would hide it, but because there is nothing to calculate from.

3.2 A public breakdown of today’s user base before the phase-out (within 60 days)

The phase-out of the guarantee becomes a policy decision when we know whom it affects. MIAK asks that together with the phase-out schedule the breakdown of today’s users should also appear: by company size (micro, small, medium), sector, region, the size of the loan amount and the guarantee coverage ratio, together with an estimate of what proportion of the businesses concerned would have received credit even without a guarantee. This latter estimate is not impossible: it can be carried out on the basis of the banks’ risk assessment data and the default rates. The data have to appear in an aggregated form that cannot be traced back to individuals. The proposal realises the ex ante branch of the G20 impact assessment programme point, and answers the question that the present schedule does not pose: out of the ceiling cut from HUF 12,800 billion to HUF 11,200 billion, which segment drops out — the one that would be financeable even without the guarantee, or the one that without it does not obtain credit at all. The two cases are policy opposites: in the first the reform eliminates a deadweight loss, in the second it opens a financing gap.

3.3 A transitional period for the contractual base built on the guarantee, and a compulsory subsequent balance sheet (12–18 months)

MIAK proposes that the reshaping of the Széchenyi constructions and the making optional of the guarantees be given an announced transitional period at those businesses that today build their continuous operational financing on the current account credit and liquidity loan operating with a compulsory guarantee — in the summer months this appears particularly sharply among tourism players. The transition should not be favouritism but a schedule announced in advance and identical for everyone. In addition, a compulsory ex post impact assessment of 12–18 months should apply to the reform as a whole: how many businesses’ loan applications were rejected compared to the earlier comparable period, how much the SME loan stock changed, and whether the standard of the banks’ own risk assessment really did rise, or the circle of lending merely narrowed. This is the ex post branch of the G20 programme point: the comparison of the expected result fixed before the decision with the factual data after 12–18 months, with automatic review if the result falls short. The G21 spending review logic asks the same from the other side: in the case of a discontinued support too, it has to be written down what ceases along with it.

The three proposals can be strung on a single principle: the stock target figure and the policy aim are not the same thing. Reducing the GDP-proportionate level of guarantees to the EU average is measurable, easy to communicate and supported by an external commitment — precisely for this reason there is a danger that the figure will be met while the content behind it will not be examined. MIAK is not asking for the aim to be postponed, but for the public picture of the exposure and of the user base to be prepared simultaneously with the commitment. Five weeks is enough for this if the data exist; if they do not, then that is precisely the most important lesson of this reform.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Economy The hidden fiscal exposure decreases; the banks’ own risk assessment strengthens; tying interest to a market reference gives more transparent pricing The riskiest but still viable SME base is squeezed out of lending, and this shows up in employment
Budget The lower guarantee ceiling makes the 2027 planning more predictable; the risk of calls decreases If the data on the exposure do not appear, the extent of the reduction is not verifiable either — the figure is met, sustainability does not improve
Housing The present fall in market interest rates partly replaces the interest subsidy that drops away; shifting the support to the supply side may moderate the price-raising effect The rapid repricing of housing supports reduces the access of young families without the supply having expanded

The main trade-off is tense between pricing and access. A subsidised loan exerts two effects at once: it gives credit to people who otherwise would not obtain it, and it raises prices where supply cannot expand in the short run. It follows that cutting back the support also exerts two effects: it moderates the price pressure, and it squeezes out participants. Which effect will be the stronger is exclusively an empirical question, with a different answer by sector and by region — and precisely for that reason it cannot be settled either by an argument of principle or by a budgetary one. This uncertainty is eased by the fact that a market-based fall in interest rates is under way right now on the housing loan market: in the wake of the moderation of long-term interbank yields, one larger domestic bank announced an interest cut of between 0.15 and 0.40 percentage points (pp). On a HUF 30 million, twenty-year housing loan with a fixed rate throughout this means a few thousand forints a month, and more than one million forints over the whole term. The market movement therefore partly replaces the support that drops away — but only for those who pass the bank’s creditworthiness filter.

The second trade-off lies between the EU milestone and the domestic impact assessment. The external commitment is a disciplining force, and MIAK regards this as a value: it forces out exactly such a schedule and figure as the internal political logic would have postponed for years. At the same time the milestone relates to the stock, not to the composition — the reform tips to the risk side if fulfilment takes place on the items that are easiest to cut, because those are at the least protected groups of users.

Part V — Measurability and summary

5.1 What is worth tracking? (proposed KPIs)

Four performance indicators (KPIs, Key Performance Indicators) are worth tracking over the next eighteen months:

  1. Guarantee stock and expected loss: the quarterly public disclosure of the outstanding state counter-guarantee stock and of the expected loss calculated on the portfolio — proposed target: the first report simultaneously with the draft government decree of 17 August.
  2. Rejection rate of SME loan applications: the rejection rate of micro and small business loan applications by company size and sector, quarterly — proposed target: a value at most a few percentage points higher than the pre-reform level, otherwise automatic review.
  3. Development of the SME loan stock: the annual change, in real terms, of the below-size-threshold segment of the corporate loan stock — proposed target: a non-declining path at the pace of the guarantee phase-out.
  4. Guarantee coverage ratio and call ratio: the average guarantee coverage ratio and the ratio of actual calls to the contract stock — proposed target: both published annually as comparable time series.

5.2 Summary

MIAK’s key message in a single sentence: reducing the stock of state guarantees is a correct aim, but cutting the ceiling figure does not substitute for examining who is squeezed out of lending. Concretely, it asks that the first quarterly report on the guarantee stock and the expected loss appear simultaneously with the draft government decree of 17 August, that within 60 days a public breakdown of today’s user base by company size, sector and region be prepared, and that a compulsory ex post impact assessment of 12–18 months with an automatic review obligation apply to the reform as a whole. MIAK asked for this same yardstick at the introduction of the earlier housing and family support constructions too: the discontinuation of a support is a decision of the same weight as its introduction, and it deserves the same impact assessment.

Two MIAK foundational values are in play in this matter. Accountability, because a contingent liability by its nature pushes the consequence outside the political time horizon: the decision-maker who undertakes the guarantee is not the actor who pays the call, and therefore the system can be held to account only if the stock and the expected loss are continuously visible — and this is now, at the moment of reduction, just as true as it was at the moment of expansion. And data-drivenness, because the two mutually opposed effects of the reform — the elimination of the deadweight loss and the opening of a financing gap — can be decided exclusively from data; if the decision is taken before the data are ready, then it is not data-driven but deadline-driven, and it is worth saying this openly.

Part VI — Justifications and further sources

6.1 The press framing by spectrum

The economic band carried the technical content, and on this day that was the most substantive framing. Portfolio summed up the change with the opening “the end of the system so far”, and it was this paper that presented the elements of the schedule in the greatest detail: the end-of-August Széchenyi reshaping, the pricing tied to the interbank rate, the making optional of the guarantees, the reduction of the ceiling figure, and the fact that the saving has to be taken into account already in the planning of the 2027 budget. The same paper also brought the fact of the EU consultation on the methodology of the guarantee reduction. In a separate article it reported on the fall in housing loan interest rates and its numerical consequence, and on the renovation concept being prepared for the domestic building stock of nearly one million expensively heatable dwelling houses.

The public affairs band highlighted the governmental logic of the decision. 24.hu summarised the content of the decree published in the Hungarian Official Gazette, including the antecedent of the amendment of the law on the development bank and the chain of ministerial mandates; ATV brought the transport and investment portfolio head’s conception for managing the housing crisis. HVG recorded the EU connection of the decision most precisely: it is in that paper that the decree’s own justification appears, according to which the review takes place in order to fulfil the milestones of the Recovery and Resilience Plan, together with the description of the finance minister’s task on limiting the public finances’ development bank exposure.

The conservative band did not bring the subject into top focus on this day: neither Magyar Nemzet nor Mandiner made the guarantee reform front-page material, while the energy crisis and the water situation featured prominently at both. This silence is in itself information: a decision tied to an EU milestone and affecting a stock of several thousand billion forints practically did not appear in one half of the domestic public sphere. From the comparison of the bands one common gap is visible: not a single paper posed the question of who draws on the guarantees today, and at which group of businesses the dropped ceiling will appear — the reports described the decision throughout from the side of the measure, not from the side of those affected.

6.2 Facts and data

Data Value Source
The upper limit of state counter-guarantees for this year before the decision HUF 12,800 bn Portfolio and 24.hu, 3–4 August 2026
The new upper limit HUF 11,200 bn Portfolio and 24.hu, 3–4 August 2026
The extent of the ceiling reduction HUF 1,600 bn calculation from the two data above
Deadline for the review of the counter-guarantee amortisation schedules 12 August 2026 24.hu, 4 August 2026
Deadline for the draft government decree on the GDP-proportionate guarantee target 17 August 2026 24.hu, 4 August 2026
Date of the reshaping of the Széchenyi constructions end of August 2026 HVG and Portfolio, 3 August 2026
The Széchenyi products concerned Current Account Credit MAX+, Tourism Card MAX+, Liquidity Loan MAX+ HVG, 3 August 2026
The new interest reference 3-month BUBOR HVG and Portfolio, 3 August 2026
The benchmark of the GDP-proportionate guarantee target the average of the EU member states HVG, 3 August 2026
Deadline for the commitment letter to the Commission end of August 2026 Portfolio, 3 August 2026
Housing loan interest cut at one large bank 0.15–0.40 pp Portfolio, 3 August 2026
Example: monthly instalment of a HUF 30 m, 20-year fixed housing loan HUF 236,023 → HUF 231,690 Portfolio, 3 August 2026
The saving projected over the whole term in the example more than HUF 1 million Portfolio, 3 August 2026
Number of expensively heatable domestic dwelling houses close to 1 million Portfolio, 3 August 2026

A single connection deserves highlighting. The HUF 1,600 billion ceiling reduction is not the same as HUF 1,600 billion of foregone lending, and it is not the same as an equal amount of budgetary saving either. The counter-guarantee ceiling is an upper limit: it shows how much liability the state may undertake, not how much it has undertaken, and least of all how much it will pay out. The actual budgetary effect is given by three figures together — the committed stock, the average guarantee coverage ratio and the expected loss of the portfolio — and these three sets of data are not available publicly today in a comparable form. That is why it can happen that the reduction of the ceiling appears as a communicable result both when the real exposure does not change and when it decreases substantively. The difference between the two can be shown only by the report proposed under 3.1.

6.3 Policy dimensions

  • Economy (programme points) — the data-driven presentation of contingent liabilities, the ex post impact assessment of the phase-out and the application of the spending review logic to the discontinued support (programme point ID: G1, G20, G21, G23, G19);
  • Construction (programme points) — the data basis of shifting housing support to the supply side and the renovation programme for the expensively heatable building stock (programme point ID: EP2, EP3);
  • Demography (programme points) — the housing access of young families and the impact assessment of family support instruments (programme point ID: DM7, DM2);
  • Digitalisation and AI regulation (programme points) — the machine-readable publication of the quarterly guarantee report (programme point ID: D2).

6.4 Literature in detail

6.4.1 Carmen Reinhart – Kenneth Rogoff: This Time Is Different

The authors worked up the sovereign default, banking and inflation crises of eight centuries on a new database covering several regions. Two of the volume’s findings belong directly to today’s subject. One is methodological, and the authors themselves state it most sharply: because historical data on domestic debt are so hard to obtain, empirical analyses examining debt simply disregarded that debt. From this followed the other: in the mid-2000s many analysts and international institutions concluded that the external default risk of the emerging countries had fallen dramatically, because the share of external debt had moderated — while the domestic debt, which was not measured, was a liability of the same weight. The volume calls this pattern the “this time is different” syndrome: the argument that the actors have learned from their mistakes appears before every crisis.

From the point of view of the Hungarian guarantee reform this frame speaks not against the reform but alongside it. The state counter-guarantee is exactly the kind of liability to which Reinhart and Rogoff’s observation applies: it is not current expenditure, there is no regular public time series about it, and therefore it usually does not feature in the sustainability analyses either. If the government raises the reduction of the exposure into an EU commitment, then the first step is precisely that the exposure should be measurable — otherwise the claim about the extent of the reduction will be just as unverifiable as the earlier expansion was. Proposal 3.1 is therefore not an administrative extra burden: it is the condition of the reform’s credibility.

📖 Source: Carmen Reinhart – Kenneth Rogoff: This Time Is Different

6.4.2 Joseph Stiglitz: Globalization and Its Discontents

In the chapter dealing with the connection between bankruptcy and moral hazard, Stiglitz starts from the fact that in normal market operation whoever grants a bad loan bears the consequence: the debtor may go bankrupt, and legal systems have an orderly procedure for this. According to his argument the international bailout programmes overturn precisely this structure, because in the hope of a bailout lenders pay less attention to whether the debtor is able to repay the loan. The author’s formulation of the mechanism is as concise as possible:

“Insurance reduces the incentive for care and prudence. Bailing out in the event of a crisis is like ‘free’ insurance. If you are a lender, you pay less attention to screening applicants — since you know you will be bailed out.”

This argument is directly applicable to the Hungarian guarantee system, and it explains why one of the stated aims of the reform is correct. If the state takes over a significant part of the credit risk through a counter-guarantee, then the credit institution’s own risk assessment will necessarily be weaker — not out of negligence, but because the distortion is in the set-up. MIAK, however, does not derive from this the complete phase-out of the guarantee: Stiglitz’s argument is about the extent of the incentive, not about the existence of the instrument. The correct answer is the calibration of the coverage ratio and the measurement of risk-sharing, not the settling of whether the state undertakes risk at all. Precisely for this reason proposal 3.2 matters: reducing the coverage ratio improves the quality of lending if we know who those are for whom the guarantee did not spoil the incentive but secured access.

📖 Source: Joseph Stiglitz: Globalization and Its Discontents

6.4.3 Ha-Joon Chang: 23 Things They Don’t Tell You About Capitalism

One chapter of Chang’s volume argues against the widely accepted claim that governments are incapable of “picking winners”, that is, cannot take well-founded business decisions through their industrial policy. According to the author, international experience does not support this: numerous successful government interventions can be found, and nothing justifies the assumption that a government decision affecting companies would necessarily be poorer than what the firms take themselves. The element of his argument most important for today’s subject concerns information: possessing more detailed information does not in itself guarantee a better decision — indeed, the decision may even be harder for someone who is too deep in the thick of things — but the government does have the instruments to raise the standard of its decisions by obtaining better quality information.

This argument gives the correct reading of the present reform. The customary argument for phasing out the guarantee is one of principle: the market knows better whom to lend to. In Chang’s frame this argument cannot be settled at the level of principle, only at the level of information — and precisely for that reason the quality of the decision is determined by whether the government carries out its own information gathering or skips it by invoking an external deadline. The sequence of the 12 August, 17 August and end-of-month deadlines carries the risk that the information gathering will be left out, because it does not feature among the milestones. MIAK’s proposal therefore does not speak against the reform, but in favour of the decision being taken in the way Chang describes, with better information — this is the only way for the outcome not to depend on luck.

📖 Source: Ha-Joon Chang: 23 Things They Don’t Tell You About Capitalism

6.5 International comparison

There is no need to invent a new methodology for presenting contingent liabilities. The fiscal transparency practice of the international financial institutions — above all the structure of the fiscal risk statements of the International Monetary Fund (IMF) — treats guarantees, public–private partnerships and state-owned enterprise exposure in a separate chapter, typically with the triad of stock, expected loss and sensitivity analysis. The EU statistical system likewise regularly publishes member state government guarantee stocks as a proportion of GDP — this data series is also the benchmark of the present Hungarian commitment. The Hungarian proposal therefore does not ask for an extra report, but for the domestic filling-in of the already existing international template to be public and quarterly.

In the practice of phasing out SME guarantees, West European experience highlights two elements. One is the transitional period: the reduction of guarantee coverage typically takes place along a path announced in advance over several years, because the switching of the banks’ risk assessment capacity also requires time. The other is targeting: several member states did not reduce the guarantee stock linearly but differentiated the coverage ratio — higher coverage for start-up and innovation financing, lower where market lending works anyway. This latter possibility does not feature in the present Hungarian schedule, even though reducing the ceiling figure and differentiation do not exclude one another.

Economy

  • G1 — Data-driven budget
  • G20 — Economic policy impact assessment system
  • G21 — Systematic review of state expenditure
  • G23 — Sovereign debt sustainability framework
  • G19 — Radical transparency in economic decision-making
  • G10 — State development bank

Construction

  • EP2 — Housing construction data platform
  • EP3 — Energy efficiency renovation programme

Demography

  • DM7 — Housing access programme for young families
  • DM2 — Impact assessment of family support

Digitalisation and AI regulation

  • D2 — Open data programme

Proposed new programme point: A contingent liability register — for the Economy area: a unified, quarterly updated, machine-readable register of state guarantees, counter-guarantees, suretyships and state-owned enterprise exposures, broken down by stock, guarantee coverage ratio, expected loss and actual calls, following the structure of the international fiscal risk statements.

6.7 List of sources

Press sources (MIAK press monitor, 4 August 2026 — topic 4):

Knowledge-base references (books):

  • 📖 Carmen Reinhart – Kenneth Rogoff: This Time Is Different
  • 📖 Joseph Stiglitz: Globalization and Its Discontents
  • 📖 Ha-Joon Chang: 23 Things They Don’t Tell You About Capitalism

Note: the local file path of the books does NOT appear in the visible text of the blog — only the author and the title. The file path is an internal matter of the generation process, not the reader’s.

MIAK internal materials:

  • MIAK policy area: Economy (programme points; programme point ID: G1, G20, G21, G23, G19, G10)
  • MIAK policy area: Construction (programme points; programme point ID: EP2, EP3)
  • MIAK policy area: Demography (programme points; programme point ID: DM7, DM2)
  • MIAK press monitor, 4 August 2026 — topic 4, score: 85/100

Additional public data sources:

  • MNB (National Bank of Hungary) — Lending Survey and Housing Market Report
  • KSH (Hungarian Central Statistical Office) — house price index and housing construction statistics
  • Eurostat — government guarantee stocks as a proportion of GDP, house price index
  • IMF — the structure of fiscal risk and fiscal transparency statements
  • OECD — Affordable Housing Database

Generation metadata