Demand conditions: the micro-level motivations of households 8

The studies investigating the establishment of FX-lending based on aggregated data regularly examine the impact of a few macro-economic variables in the share of FX-loans. Among these macro-economic variables, one can most often find the difference in interest rates on domestic and foreign funds and the level and volatility of inflation and the real exchange rate.⁹

However, the findings produced by varied research are ambiguous. Basso et al. (2011) found that the relative volatility of inflation and the real exchange rate play a role, whereas Neanidis and Savva (2009), and Neanidis (2010) found these variables insignificant. The results in respect to the interest rate differentials are also diverse: Rosenberg and Tirpák (2008), Csajbók et al.

(2010), and Neanidis (2010) identify this impact, while according to Basso et al. (2011), and

8 In the presentation of the international literature we greatly rely on a study by BECKMANN and STIX (2015).

9 BECKMANN – STIX, 2015.

Luca and Petrova (2008) their extent is marginal. In addition to the above, the share of foreign banks¹⁰ and the “dollarization” on the deposit side¹¹ may also induce the spread of FX-loans.

Instead of examining the impact of macro-economic variables on each other, other studies use questionnaire-based surveys. The advantage of this research method is that it is easier to separate the impacts of supply and demand, as macro-economic variables may effect both.

In their research focusing on Central-Europe, Firdmuc et al. (2013) found that it contributes to the spread of FX-lending if households have little trust in the stability of the domestic currency and financial institutions, and progress is further strengthened if they are anticipating the introduction of the euro in the foreseeable future. By examining Austrian households, Beer et al. (2010) found that the households borrowing FX-loans are usually less risk-avoiding, with such people being older, wealthier, and with a greater degree of financial awareness than on average. By examining Hungarian households however, Pellényi and Bilek (2009) observed that those borrowing FX-loans do not have deeper financial knowledge, are not well off, and do not prefer risks more than those borrowing HUF loans.

By taking a closer look at the Hungarian situation, the interest rate disadvantage of the fo-rint was obvious compared to foreign currencies, especially in respect to the Swiss interbank interest rate. In 2005, the average level of interbank HUF interest rates exceeded 7 per cent, while the average of the 3-month CHF LIBOR did not even reach 1 per cent during the same period (Fig. 5). The interest rate differential was about 630 basis points on average that year, which was only moderated to 550 and 520 basis points in the following two years.

Consequently, while at the beginning of the decade the typical 6 per cent HUF interest rates of the period of generous government subsidies were competitive against FX interest rates, FX-loans might have seemed more attractive to households after the termination of these subsidies.

Figure 5: The development of the 3-month interbank yields of HUF, EUR and CHF

Source: European Central Bank, Swiss National Bank, MNB

10 BASSO et al., 2011.

11 LUCA – PETROVA, 2008.

The importance of the interest rate differential can be easily understand if we look at the average parameters of loans denominated in various currencies (Table 1). At the end of 2005, the price of an average 65-square metre dwelling was HUF 11 million, which meant a nearly HUF 7.8 million contract size with a 70 per cent loan-to-value ratio (LTV). If this amount of cre-dit was borrowed in HUF, it was accompanied by a HUF 87 thousand monthly instalment rate with the average interest rate levels of that time (assuming a 15-year maturity), which used up nearly half of the income of a household with two average wages. In contrast, by borrowing a loan denominated in CHF the amount of the monthly instalment rate was HUF 65 thousand, which was only about 36 per cent of the household’s income. From another perspective: those who were indebted in HUF were able to take a substantially smaller credit amount with a safer, 30 per cent payment-to-income ratio than those in a foreign currency. This means that liquidity-constrained households (with less savings) were more likely to choose FX-loans to satisfy their credit demand.

Table 1: Characteristics of a housing loan disbursed in December 2005 in different denominations

Note: Calculations are based on the average price of a 65 square metre dwelling at the end of 2005 with 70 per cent LTV-ratio and a 15-year maturity. We assumed a household with two average wages for the calculation of the payment-to-income ratio, where both wages correspond to the national net average.

Source: Own calculations, based on MNB

Of course, the interest rate differential benefiting the FX-loans does not guarantee by itself that the total financing costs of these loans will be actually lower¹² during the entire maturity than that of HUF loans. Namely, the depreciation of the forint can easily cut back on the benefit that the borrower takes with the lower interest rate.¹³ With this in mind, a well-prepared borrower not only needs to examine the interest rates, but also the exchange rate fluctuations.

However, the HUF/CHF exchange rate was reasonably stable during the years of FX-lending (Fig. 6). The exchange rate was in the range of HUF/CHF 141 to 180 between August 2005 and August 2008. Overall, the fluctuation between the strongest and weakest exchange rate seemed affordable in respect to debt service: during 2005, the difference between the strongest and weakest status was barely 7 per cent, in 2006 it was 15 per cent, in 2007 it was 9 per cent, and in 2008 it was already 20 per cent. So, even if households considered the exchange rate fluctuations of the last few years when deciding on the currency of the credit to take, they might have come to the conclusion that even in the event of a relatively substantial exchange rate depreciation they would pay a lower instalment than if they took the credit in HUF. The expectations regarding

12 Here, total financing costs do not denote the legal term of the annual percentage rate of charge (APR), since this rate inter alia contains neither the “expenses” of the exchange rate fluctuation, nor that of the change in interest rates. See BERLINGER, 2017.

13 KIRÁLY – SIMONOVITS, 2015.

a stable exchange rate were also reinforced by the EU accession, which might have encouraged households to anticipate the introduction of the euro. This would have eliminated the exchange rate fluctuation against the euro within a reasonable time from the borrowing of the credit.

Figure 6: Development of the HUF/CHF exchange rate

Source: MNB

The indebtedness of the households had already commenced before the spread of FX-lending at the beginning of the 2000s, mostly due to generous government subsidies. Two new subsidies were introduced in 2000, the complementary interest rate subsidy and the mortgage interest rate subsidy, which became the most significant form of subsidy within a few years, thanks to their low costs (Table 2). The interest rate subsidies enabled the average interest rate level of mortgages to fall well under the market interest rate level, therefore borrowing became possible for a wide range of society. Following the introduction, several further easing measures were introduced related to the rules of utilization, and in addition the amount of the subsidy was raised, until the Government was forced to reduce its volume in 2002 due to the budget effect. The magnitude of the programme can be well presented by the fact that, at its peak, the subsidy for a HUF 10 million mortgage could reach up to HUF 7.47 million at the present value (Table 3).¹⁴

Table 2: Estimated volume of housing subsidy expenditures (HUF billion)

Source: Hegedűs et al. (2006)

14 HORVÁTH, 2008.

Table 3: Discount from the subsidies and its value compared to the house price

Source: Horváth (2008), page 83