Can we have a functioning home financing system?

It is beyond any doubt that the retail mortgage lending market is one of the areas that have been most heavily affected by the crisis in Hungary. As a result of exchange rate volatility, the rise in country risk premia and the higher cost of credit, installments paid on CHF-denominated loans have increased by 30-50%. This, on its own, was sufficient to bring many families into a critical situation, and was aggravated by the fact that the unemployment rate also picked up significantly. The combination of the two factors (the loss of employment and the drastic rise in installment payments) did not simply raise questions about risks in the banking sector, but has led to a widespread social drama and a severe macroeconomic risk.

The responsible handling of this situation requires widespread cooperation, social responsibility and sacrifice from all the players involved. We are not simply speaking about the handling of a technical problem, but rather about the maintenance of social safety within Hungary and the functioning of the economic and financial system. The search for a solution requires intellectual courage, while social responsibility principles also need to be heeded. The proposal that we have created aims to contribute to the evolving discussion on this subject and to the (unfortunately and obviously) very painful solution.

Intensive background work has been done in the past months on this subject by other institutions as well. We present our opinion on the main proposal package that is currently being formed in the box at the end of this document.

OUR PROPOSAL

Factors leading to loan portfolio deterioration:

The deterioration in the quality of household FX home loan portfolios and the rise in the share of non-performing loans can be attributed to two distinct factors:

– a rise in households’ debt servicing costs
– and/or the loss of employment due to layoffs, illness and other social reasons.

Below we will primarily examine possibilities to resolve the first factor. This proposal only marginally deals with the handling of those who have become unable to pay mainly due to the second factor. In the case of those who have lost their jobs etc. and thus are unable to pay, a combination of social and budget policies and labor market tools needs to be created. A National Asset Management Company could be such an instrument, which could provide a temporary or permanent solution to those in need (e.g. by taking on the burden of installment payments partially, temporarily or permanently, but not through an immediate purchase of the real estate assets). Based on estimates, 10-20 thousand borrowers belong to this group.

As a general principle, one should start off from the fact that there is no painfree solution in the situation that has come about. Therefore all affected parties need to make sacrifices in order to maintain long-term financial stability. A resolution to the current problems in the Hungarian lending market has to be a result of a close cooperation between the government, the central bank, commercial banks and their clients.

Seeing that we do not have fully reliable figures at hand, our proposal merely present ideas for further consideration. The specific numbers in it (regarding interest rates, volumes and possible subsidies) are not cast in stone. The introduction of the proposal would clearly have to be preceded by a series of professional negotiations and would have to be communicated in detail and in a unified manner to markets and to international financial organizations.

Goal of the proposal:

The main goal of this proposal is to help restore the functioning of the home lending market. The starting point for this is to halt further deterioration in the quality of the existing loan portfolio and, to the fullest possible extent, to restore the solvency of borrowers who are in arrears or are nonpaying. This goal should be considered based on two, tightly connected elements. The first element is the widespread elimination of borrowers’ unhedged FX position and through that, the removal of exchange rate volatility. The second element is the significant easing of loan installment payments, i.e. the reduction of monthly debt servicing costs.

There are two ways in which households’ debt burden could be reduced:

– by lengthening (directly or indirectly) the maturity of their loans
– by lowering the interest rate they have to pay.

If it is true that the main reason behind the nonpayment of FX home loans is the jump in installment payments, then the mitigation of monthly installments to levels seen before the crisis (i.e. a 30-40% reduction in the monthly debt-servicing burden) could bring about a marked stabilization in the home loan/mortgage market.

Parameters of the proposal:

0.) massive evictions are unacceptable and should be avoided at all costs, especially in case of socially disadvantaged groups (e.g. less well-to-do families with children)

1.) however, the moratorium on housing evictions would need to be lifted, as it represents a sizeable moral hazard (an incentive to not pay) and could cause the mortgage lending market to collapse entirely and not recover for several years

2.) in line with point 0.), debtors’ debt servicing costs need to be reduced. The current situation of the government budget does not allow for the takeover of the debts of 100-150 thousand nonpaying borrowers (not even at a discounted value).

3.) the budget deficit and the stock of government debt should not be boosted significantly by the proposal

4.) the solvency of the banking system has to be improved notably

5.) mortgage lending growth should be restarted in the economy (which is not possible at the current NPL rates and with the moratorium in place)

6.) the FX market should not be upset by the planned conversions.

PART 1: CONVERSION OF FX-DENOMINATED LOANS INTO HUF-BASED LOANS

How to go about the conversion?

In order to reduce systemic risk in the Hungarian financial sector, the possibility of a conversion of CHF-based loans into HUF-based loans needs to be investigated. A fundamental criterion for the conversion to occur would be that forint rates could not exceed franc rates significantly (which currently equal 7.5% on average). For this to happen, the interest rate on the newly converted loan would be strictly controlled. As it will be detailed later, a long-term (3-5 year) forint loan from the central bank would be provided to commercial banks, who would then provide loans to their customers at the benchmark rate (currently at 6%) and, additionally, a maximum 200-250 bp margin, which would lift the overall customers’ forint loan interest rate to 8-8.5%. Later we will discuss the size of the government interest rate subsidy that would be needed in order to allow both nonpaying and punctual borrowers (those who pay in a timely size and manner) to get an acceptable interest rate.

Given that a forint-franc conversion would clearly result in strong CHF/HUF demand in the market, the execution of the conversion in the open market is undesirable. For this reason, the National Bank of Hungary (NBH), the Swiss National Bank (SNB), the Hungarian government and the Hungarian banking system would carry out the conversions in a tightly controlled and coordinated out-of-market operation, using unconventional tools.

Option 1: Perhaps the most preferable option would be if the NBH, in close coordination with the SNB, could purchase a maximum of EUR 12 bln directly from the SNB, for a portion of its currency reserves. This way, the NBH would acquire CHF 15-17 bln, which would probably be roughly enough to convert CHF loans with a value equivalent to HUF 3000-3500 bln. (The total volume of mortgage-backed retail loans is HUF 4300 bln, but borrowers could probably only convert a part of that sum into forints).

Option 2: The NBH could swap euros for CHF with the SNB.

Option 3: Given its time-consuming nature, probably the least favorable option would be for the NBH to purchase a daily sum of, for example, CHF 50-100 mln for euros in the market over a determined period of time. (It would need to be clarified how exchange rate shifts would appear between the parties involved.)

How would the purchased francs help alleviate financial sector stress?

With the switch in its reserves, the NBH would, assumably, mitigate the macro mismatch (attributable to unhedged CHF positions) by converting euros into Swiss francs. (The proportion of CHF-denominated reserves within the central bank’s total reserves should have already been boosted in the past, independently from the current problems in the lending market.) The NBH would sell CHF in exchange for forints to commercial banks in Hungary outside of the market, but at the normal (daily) CHF/HUF rate. Banks would sell the CHF to their clients, who could then repay their outstanding loan debt. Banks would then use the CHF to pay back their own CHF loans they had received from third parties earlier on and would also close down a large share of their swap positions or CHF liabilities. (Further analysis is needed to predict the potential market and cost effects of a massive closing down of CHF positions.)

Parallel to this, the NBH would provide long-term (3-5 year) HUF loans (either securitized or non-securitized) to commercial banks at a fixed rate, which they could then pass on to their clients in the form of HUF-based loans (a replacement for the previous CHF loans). The term of these loans would equal 3 to 5 years, the interest rate would be fixed and it would be possible to renew the loan. Banks would share the burden by charging a much lower-than-usual margin of 200-250 bps (i.e. they would provide an interest rate subsidy).

The balance at the end of the day:
– banks would reduce their short-term deposits placed at the NBH, as those deposits would be used in exchange for Swiss francs
– the NBH’s FX reserves would drop, as the CHF would flow out to commercial banks, who would then use it to pay off their debts to third parties
– the NBH would provide long-term forint loans to commercial banks (e.g. by discounting securities issued by commercial banks, which would appear under its assets), thereby restoring money supply to its original level

Benefits of the conversion

In terms of the country’s FX balance, currency reserves could drop by a maximum of EUR 12 bln based on this proposal. On the other hand, the country’s gross and net external debt would also drop after the execution of the proposal and the riskiest element (short-term CHF debt) would be reduced. This would also suggest an improvement in Hungary’s risk assessment and an assumable EUR 3-4 bln drop in FX reserve needs. The health of banks’ balance sheets would improve notably, clients’ unhedged FX position would be reduced significantly, and that would also cut banks’ risks in a considerable way. The maturity mismatch and FX risk exposure of the entire banking system (including that of households) would be cut drastically.

Conditions of the conversion:

There would be unavoidable risks when the conversion of borrowers’ loans would take place, as this would not occur at the most favorable point in time for the client (accumulated losses would be realized at a higher CHF/HUF rate). Thus it should be considered by policymakers whether the conversion should be fully voluntary or made mandatory. For all borrowers with nonperforming loans (who fulfill the eligibility requirements listed in part 2 of this paper), a conversion into forints should be made compulsory, on a specific date and, according to the possibilities, at levels between CHF/HUF 200-210. For all other FX home loan borrowers, the option of receiving interest rate subsidies offered by the government and by commercial banks and of getting an indirect and indirect maturity extension would be only be open if they also convert their loans into forint.

Given the current market conditions, the EUR 12 bln drop in currency reserves described above would not be desireable. At the same time, the market could probably be made to accept a 5-6 bln drop from the current EUR 34 bln, given the significant easing in the economy’s external vulnerability. The sale of FX assets in the transferred private pension savings (estimated at around EUR 2.5 bln) and a delay in the prepayment of the IMF loan could mitigate the drop in FX reserves (an additional EUR 1-2 bln hike in reserves could be needed) . This way, reserves could equal a satisfactory EUR 30 bln by the end of the transaction. A EUR 4-6 bln drop in the NBH’s euro reserves and the rise in its forint assets could create, given the forint-euro rate differential, an additional annual income of approximately HUF 65-75 bln (under the current interest rate conditions) at the central bank and, therefore, in the state budget. (The technical issue arising from the timing of the settling of accounts between the central bank and the state budget would need to be dealt with.)

After the transaction, the required risk premium on Hungarian assets could, ceteris paribus, decline. Although the transactions would not be cheap (the realization of exchange rate losses, the drop in banking revenues and potential FX and forint interest rate shifts would all be costly), they would effectively remove a potentially substantial source of risk from the Hungarian financial system.

The first element of our proposal (the CHF-HUF conversion described above) would only eliminate risks from the system that stem from clients’ unhedged FX positions. In order to restore borrowers’ ability to pay and to strengthen the ability of those who pay regularly to continue to do so in the future, debt servicing costs need to be reduced significantly.

PART 2: REDUCTION OF DEBT SERVICING COSTS

Proposal to borrowers:

Although the conversion described above would mean that an exchange rate loss would be locked in by the borrower upon conversion, this could be compensated for by an installment payment that is 35-40% lower through a number of years and by the lack of any exchange rate risk. (Interest rate risk would still be present, however, for the central bank, the government and, after a period of 3-5 years, for clients as well).

Eligibility requirements for the government subsidy:

Strict eligibility requirements would be applicable to those wanting to receive a government subsidy and commercial bank subsidy on their loan converted into HUF. This would make the proposal more aptly targeted to those in greatest need of help and would avoid the favoring of well-to-do groups of borrowers.

1.) Borrowers could only qualify if the total value of the real estate asset in question (based on the initial appraisal) does not exceed HUF 20-30 mln.

2.) Debtors and co-debtors would also need to declare that their combined net monthly wage level does not exceed, for example, HUF 300-350 thousand.

3.) Borrowers applying for a government subsidy would also have to declare that they do not hold any other assets with a value, for example, in excess of HUF 5-8 mln.

4.) Borrowers would need to live permanently in the real estate asset in question (i.e. the asset would need to be owner-occupied).

5.) Debt servicing costs would need to exceed 30% of the combined income of debtors and co-debtors.

6.) The size of the installment payment could not exceed a certain sum, for example a monthly HUF 180-200 thousand.

Elements of the installment reduction:

Introduction of a bullet payment (indirect maturity extension)

25% of the loan amount would be placed into a bullet payment (an indirect extension of the loan’s maturity), on which there would be a government guarantee in the case of troubled borrowers (whose loans are categorized as nonperforming, i.e. are in arrears by more than 90 days). Seeing that the bullet would be a contingent liability for the state, it would be collateralized by the real estate assets underlying the loans in question. An interest rate of 5.5% would have to be paid on this portion of the loan, composed of the benchmark rate (currently at 6%), 2 additional percentage points charged by commercial banks and a 2.5 percentage point state subsidy (6 + 2 – 2.5 = 5.5%). In exchange for the guarantee, a guarantee fee could be collected from customers by banks and paid to the state. (Hence the total interest rate and guarantee fee to be paid by the customer would be 5.5 + 0.5 = 6%.) With the collection of a guarantee fee, the measure would conform to EU regulations.

The bullet would also be an option for borrowers whose loans are performing (i.e. its introduction would depend on the client’s decision). In their case, however, there would be no government guarantee on the bullet payment. (We do note, however, that a partial or full extension of the guarantee to this group of borrowers should also be considered.)

Interest rate reduction

The interest rate on the newly converted loan, the value of which is reduced by the bullet, would be composed of the central bank’s benchmark rate, plus a 2-2.5 percentage points charged by commercial banks (currently, this would equal 6 + 2 = 8%). For borrowers defaulting on their loans, the government would subsidize 2 percentage points of this interest rate burden, so for them the interest rate of the loan would be equal to 6 +2 -2 = 6%.

It is clear that a ’sweetener’ would also need to be introduced for those borrowers who pay their installments on time. Otherwise, the measures would create an unfavorable deterioration in their payment discipline (ie. an incentive to not pay on time). Borrowers without overdue debts should pay interest rates that are consistently lower than those paid by troubled borrowers. Therefore, for borrowers without payment troubles, a 2.5 percentage point state subsidy would be implemented. Continuously solvent borrowers would thus pay an interest rate equal to 6 + 2 – 2.5 = 5.5%. With the aforementioned changes in the interest rate, the rate to be paid on the new HUF-based loans would be 1.5-2 percentage points below the rates currently paid on CHF loans.

Maturity extension

Thirdly, installment payments would be cut through an extension of the average loan maturity from its current level of 15 years to 25 years in the case of troubled borrowers. Borrowers without any nonperfoming loans would also have the option to have the maturity lengthened to the same extent if they wish.

The bullet would have to be paid upon maturity, after the 25 years of the loan’s term, but the customer would also have the option to refinance.

Effects of the proposal’s elements on installment payments:

1.) charging an interest rate of 8% (6% benchmark rate + 2 percentage points) on the forint loan, a rate that is above the current average CHF-loan interest rate of ~7.5%

2.) provision of a government interest-rate subsidy amounting to 2 percentage points in case of troubled borrowers and 2.5 percentage points in case of borrowers who constantly pay their installments on time (such a drop in the installments could only come about if banks also charge a lower margin)

3.) charging an interest rate of 5.5% on the bullet portion of the loan, with a government interest-rate subsidy amounting to 2.5 percentage points (additionally, a +0.5 percentage point guarantee fee paid by troubled borrowers)

4.) loan maturities would need be lifted to an average 25 years, from the current average of 15 years

In case of an average loan in the value of HUF 6 mln (CHF 40,000) with a maturity of 15 years, the aforementioned measures (interest rate rise by 0.5 percentage points, shift to an 8% HUF loan from a 7.5% CHF loan, 2 percentage point government interest-rate subsidy, 1.5-2 percentage point drop in commercial banks’ margins, loan maturity extended by 10 years, 5.5% rate paid on the bullet portion of the loan with another 2.5 percentage point state subsidy) would cut installment payments by 35-40%. This would practically restore the installment payment size that had typically been seen well before the start of the crisis.

Currently, the total stock of NPL’s equals approximately HUF 250-400 bln . If installment payments are reduced by 25% through the bullet, the amount to be guaranteed by the state would roughly amount to HUF 62,5-100 bln. The main risk for the state would be that if the bank’s client encounters payment difficulties again, the bank could draw on the state guarantee. On the other hand, the state would have senior mortgage rights on the collateral.

Benefits of the proposal:

The main benefits of the aforementioned setup would be the following:
– 100-150 thousand families could avoid eviction
– the package would ease the burdens on those whose payment difficulties arise purely from the fact that their installment payments have risen and would distinguish them clearly from those whose payment troubles are related to a loss of employment
– seeing that the state provides support in the form of a guarantee, the measures do not boost state debt based on EU accounting methods and on a cash basis
– the state could earmark the guarantee fees paid by banks into a designated fund, which could be tapped when banks draw on the guarantee
– Although banks’ margins would shrink from a current level of 3-4% to 2-2.5% as a result of the process described above (and this would entail a loss of HUF 50-70 bln for the sector), the package would also lower banks’ cost of credit and ease capital requirements. The HUF 50-70 bln loss would roughly match the annual losses which occur due to the lack of interest rate income on nonperforming loans.

The main disadvantages of the proposal would be the following:
– Borrowers converting their loans would realize their accumulated exchange rate loss for good.
– Borrowers still could not avoid the forint interest rate risk and would continue to be exposed to it, similarly to the NBH and the government (this risk cannot be eliminated based on the current situation of the capital market).
– Drawdowns of the state guarantee, although probably limited in size, would still represent an unforeseeable budget risk in case drawdowns exceed the sum of the accumulated guarantee fee payments within the state guarantee fund (the state’s potential guarantee on the bullet of borrowers who pay regularly would further increase the uncertainty of this element). If borrowers of the restructured loans become unable to pay again, banks would have to initiate and carry out the foreclosure procedure.

Proposal comparison:

The more favorable conditions would be available to borrowers only as long as they pay their reduced installments fully and in a timely manner. As soon as payment discipline deteriorates and installments become overdue, the easier conditions on the forint loan (also listed above) would cease to exist (no more maturity extension would be available, the government’s interest rate subsidy would be eliminated and banks would stop providing their own 1.5-2% interest rate subsidy as well; however, the loan would remain forint-based). This would incentivize responsible behavior on the borrower’s side.

Effect of the proposal on average installment payments:

The costs involved:

Central government budget

Although the government would be expected to subsidize interest rate payments, its costs of about HUF 70-80 bln** would be more than made up for by the additional profits made by the NBH on the assets it holds (see part 1 about the conversion of loans).

The total fiscal burden involved in the proposal would be composed of a HUF 125-175 bln state guarantee (which is a non-cash burden) and the aforementioned annual cash cost stemming from the interest rate subsidy (HUF 70-80 bln). The state would, on the other hand, receive guarantee fee payments, in the amount of about HUF 0.5-1 bln*** and would receive additional revenues through the NBH’s profit (HUF 65-70 bln).

Central bank

The conversion described in part 1 of this proposal would result in a replacement of the NBH’s EUR assets (with a yield of 1%) with forint assets (that have a yield of 6%). Additional central bank income roughly equaling the drop in reserves would be generated if EUR 12 bln of reserves are used and if all borrowers decided to convert their loans. However, such a sizeable reduction of reserves would not be desirable. If only EUR 5-6 bln worth of reserves are used for the conversion, the additional central bank income would equal 75 bln, roughly enough to compensate of the costs of the government’s interest rate subsidy.

Commercial banks

Although commercial banks’ net interest income would suffer as a result of the proposal described above, their cost of credit would drop. Also, the government guarantee on the bullet portion of nonperforming loans would allow for lower provisioning needs among banks. At the same time, banks’ margins would shrink from a current level of 3-4% to 2-2.5%, entailing a loss of HUF 50-60 bln for the sector, but the proposal would also prevent direct losses in the amount of approximately HUF 150 bln and further indirect losses in the amount of several hundreds of billions. At certain banks, the closing down of swap positions would be costly, but this would also be fully compensated for by freed-up provisions.

Borrowers

The timing of the conversion of CHF loans into HUF loans would surely be imperfect, i.e. there will always be exchange rate levels compared to which the actual exchange rate at which they convert (and lock in their losses) would seem unfavorable. On the other hand, borrowers would be able to pay an installment rate over the years that is about 40% lower. The start of liquidations and the potential loss of favorable loan conditions would strongly incentivize borrowers to maintain their payment discipline.

Other issues for discussion:

i.)
Simultaneously to the execution of the aforementioned two-step proposal, the elimination of the house auctioning and eviction moratorium would be unavoidable. At the same time, it is in the common interest of the parties involved to prevent massive evictions from taking place at the same time. It should, of course, be noted that after the introduction of the proposal, the number of potentially endangered families would drop significantly from the current 10,000-100,000 range. Despite of all this, the unavoidable and always very painful social and family effects would need to be mitigated (the mitigation would also be in banks’ self-interest). Therefore a gradual elimination of the moratorium should be examined, either based on banks’ self-regulation or through intervention by the government/supervisory authorities. One possible method for this is if evictions in the first 3-4 months only affect real estate assets with a value in excess of HUF 35 mln. The value limit would then decrease by HUF 4-5 mln every 3-4 months. Also, evictions could be put on hold in the November-April period, for example. Additional protective measures would need to be worked out in the case of families with children.

ii.)
Policymakers would need to consider the option of making a portion of mortgage interest rate payments by borrowers deductible from the personal income tax. This option could replace the 2 percentage-point government interest-rate subsidy (it would have similar costs). Although it might be technologically problematic to introduce such a measure, it could lead to a whitening of the underground economy.

iii.)
Our proposal highlighted the need for a state guarantee only in the case of the bullet portion of troubled borrowers’ loans. It should be considered whether the guarantee could be extended to the bullet payment of borrowers without any payment difficulties (the size and degree of the necessary guarantee is debatable, i.e. whether a guarantee for the entire sum would be necessary, etc.). Although an extension of the guarantee to this larger group of borrowers would boost the government’s contingent liability, we would expect even fewer actual guarantee drawdowns within this group of loans (these clients have, so far, paid their installments regularly).

iv.)
In the case of borrowers whose payment ability deteriorates due to a loss of employment, the government should avoid purchasing the real estate assets underlying the troubled loans right away (within the framework of the National Asset Management Company), as that would entail a significant and unnecessary upfront expenditure for the state. We have severe doubts whether the state would be able to appropriately manage the enormous portfolio of real estate assets and to bear the costs associated with it, possibly resulting in chaotic real-estate market conditions. Experience with similar government-owned real-estate management companies from the Socialist era shows that financial resources were insufficient for the maintenance and renovation of run-down assets, which, in the end, had to be sold at price levels not reflecting their original value (thereby causing additional financial damage to the state). Mixed-ownership houses (which include both unused state-owned apartments and private apartments) are also more prone to degradation and potentially irreversible slumification.

A preferable option would be for the state to contribute to the troubled borrowers’ installment payments (partially, or fully) on a temporary basis. In return, the banks’ collateral rights would be transferred gradually into state hands. Assuming that about 10-20 thousand borrowers would be in this category (the category of borrowers who lost their job and cannot pay their installments), an annual expenditure of about HUF 2.5-5 bln could be the result in case the state provides for, for example, 50% of installment payments on the troubled loans. A full takeover of the real estate assets would, on the other hand, boost the deficit immediately by hundreds of billions of forints and, therefore, would also increase state debt.

v.)
Besides the restoration of the functioning of the current system, the entire framework of mortgage loan supply should be rethought. If we wish to avoid the severe situation that has come about in the past ten years (and especially during the crisis), it would be recommendable to avoid FX-based lending. A mere prohibition of FX-based lending would not be sufficient, however: to maintain the system’s functioning, the creation of long-term forint-denominated and realistically priced credit supply would be necessary . These aims cannot be reached without a cooperation between the NBH, the government, commercial banks and the Hungarian Development Bank. It is necessary that the banking sector, the NBH and the Hungarian Development Bank carry out a maturity transformation using unconventional tools, that could result in sufficient long-term credit supply. It is also important that the pricing of this supply do not exceed a 2% real yield (which currently means a 6% nominal yield).