Remarks by the Supervisor of Banks at the “Calcalist” Real Estate Conference
Remarks by the Supervisor of Banks at the “Calcalist” Real Estate Conference: Risks in an era of changes.
Good morning, everyone. First, I was pleased to hear the Minister of Finance and the Minister of Housing, and I welcome their efforts to increase the supply of residential homes and to lower their prices. In this context, we are already seeing results—the number of building starts in the past three years has not fallen below 40,000 units, which is the volume that is consistent with the average annual growth of the number of households in the past decade. Therefore, even though we are still not seeing the effect of these measures on home prices, maintaining the pace of building starts that we have seen in the past three years and perhaps even increasing it—particularly in areas of high demand—is expected to stop the increase in housing prices and perhaps even lead to a decline in prices.
I would like to express my thanks for the invitation to participate in the conference and to present you with the other side of the housing market, which is the financial risks—those risks inherent in housing loans—both to households and borrowing contractors, and to the banking system.
I will start with a picture of the current situation as we view it.
Developments in the housing market have led to a change in the composition of the banking credit portfolio and the risks in that portfolio. The change in the composition of the portfolio is mainly reflected in an increase in the size of the housing credit portfolio, and in its share in total credit from about 20 percent six years ago (at the end of 2007) to about 30 percent currently, in parallel with a decline in the share of the business credit portfolio (from about 54 percent to about 46 percent).
This change was not immediate, but took place over a number of years, during which the annual pace of growth in the balance of mortgages ranged from 10 percent to 14.5 percent per year (with the record pace being in 2010). And we are talking about the net balance—new mortgages minus current repayments.
In parallel, during these years (2008–2013), home prices increased by a cumulative 82 percent in nominal terms. One indicator that illustrates the increase in home prices is the ratio between the average price of a home and the average annual salary. We are currently talking about roughly 13 years of average salary to purchase an average home, compared with 8 years in 2007.
The combined significance of these phenomena—an increase in home prices and growth in mortgage volume—is that households have taken out larger loans in order to finance more expensive homes: to purchase a first home, to upgrade their housing, or for investment. Basically, the average level of a mortgage increased both in absolute terms and in terms of the level of household income.
As a direct result of the increase in home prices, households have taken out larger loans, with a higher LTV (leverage), while diverting a larger portion of disposable income to mortgage repayments.
This situation was made possible due to, the increase in new variable rate mortgages, which reached more than 90 percent of current new mortgages during the peak period, and due to loans with terms longer than 25 or 30 years, among other things. In this way, it was possible to artificially lower the burden of current debt by spreading it out over a longer period.
The banks’ past experience in connection with housing credit is better than other credit, particularly if we assess credit losses in retrospect. If so, the question must be asked as to whether it is possible to rely on past experience in regard to mortgages, in a period of economic growth and a consistent marked increase in housing prices. Alternatively, what scenarios are we concerned about?
If a household took out a loan at a low interest rate, and can meet its repayments, why would it not do so in the future as well? After all, there is a chance that it will advance in work and that its income will increase, while at the same time, its burden of debt will decline over the years thanks to regular repayments of the mortgage.
Such a scenario is certainly reasonable and possible, but there could be many other scenarios that are less optimistic, and those must also be taken into account. For example, a household that took out, or that will take out, a long-term mortgage (25–30 years), may over those years have to cope with a situation that is different than their current situation. A situation in which one of the breadwinners may find himself out of work or with a lower salary, a situation in which the interest rate is significantly higher than it is today, or a situation in which the household’s expenses for basic needs are higher, could all leave a smaller portion of income available for the repayment of loans.
One of the tools that we are using to evaluate these risks is stress tests—which are just what their name implies.
During the past year, we have built a macroeconomic stress test at the Bank of Israel, which reflects a downturn in macroeconomic conditions that means quite a serious recession. The scenario, which was carried out for a relatively short horizon of just two years, reflects one possible process out of a whole variety of scenarios. It is also not the most serious scenario, but it is similar in its severity to scenarios carried out by supervisory authorities abroad.
Essentially, the scenario is based for the most part on the recession of 2002, and reflects a decline in GDP, a higher interest rate, and a decline in housing prices. The objective of carrying out the scenario was to understand the main points of risk for each bank and for the banking system as a whole, while emphasizing the implications derived from the housing credit portfolio, as well as from exposures to the construction and real estate industry and from leveraged credit.
I will not get into details of the macroeconomic variables that were included in the “scenario story”, but I will note the general direction of the main variables:
· Contraction of GDP, as stated—negative growth over four consecutive quarters (a decline of GDP by an annual rate of 6 percent).
· A sharp increase in interest rates for both short and long terms. (At its height, the monetary interest rate level was 6.7 percent, and the long-term interest rate was 9.2 percent.)
· A decline in home prices by a cumulative rate of about 20 percent in a year—similar to the International Monetary Fund scenario.
· And the most significant factor—unemployment of twice its current rate (about 12 percent).
Our objective was to focus on the effects of the housing and real estate market, by looking at past events, in order to create a set of macroeconomic variables that are consistent with the economic models and with the “actual story” which, in our case, focuses on the domestic market.
The general picture that we get from the results of the stress scenario is quite clear, and I again emphasize that we are talking about a stress scenario, the results of which are sensitive to the assumptions upon which it is based.
· An increase in the volume of problematic credit, and high credit losses on loans issued to the construction and real estate industry, as well as to other sectors that are characterized by high leverage and borrower concentration (current provisions that reach 3 percent compared to something on the order of 0.3–0.4 percent in recent years).
· Repayment difficulties and high credit losses in view of households’ difficulties in repaying loans previously taken out, when the interest rate was low, employment was stable and wages were high. The population that is most harmed by the scenario we carried out are households who took out loans in recent years, meaning those who purchased homes at high prices and high leverage rates that mean high repayments in relation to their monthly income. The most significant parameter that affected them is the unemployment rate.
· The bottom line is that a recession, with interest and unemployment rates similar to those that were prevalent in 2002, will lead to difficulties in repaying housing loans among about 23,000 borrowers—which is a very high number. This means 23,000 families, households; 23,000 homes that were provided as collateral to banks; 23,000 debt restructuring arrangements that will be faced by the banks, but mainly, that will be faced by Israeli society as a whole due to the societal ramifications—that is what concerns us. By way of comparison, the banking system is currently reaching debt restructuring without needing to vacate properties on a large scale, where the volume of properties being vacated does not exceed a few hundred per year. But here, we are talking about volumes that are completely different.
· Of course, in this stress scenario, the banks will absorb losses, and a decline in capital ratios, but they are far more resilient than any household.
If so, what are the measures we have taken, and what is their effect on risk levels?
In view of these concerns, we acted to limit the risks to both borrowers and the banks and to strengthen the ability of the banks to deal with this type of scenario by way of higher capital levels, strict underwriting process requirements, and a higher level of disclosure to customers. In addition, we imposed quantitative restrictions that reduced the risk to households and, as a result, reduced the risk to banks, although the risk still exists and remains significant.
· In May 2011, we limited the variable rate portion of the loan to one-third;
In this context, I note that when the mortgage is a variable rate mortgage, the risk of a change in the interest rates applies to the borrower and not to the bank, precisely when interest rates are at low points.
· In November 2012, we imposed restrictions on the loan-to-value (LTV) ratio according to the type of buyer: investors, those upgrading their housing, and those purchasing a first home.
We must remember that investors are competing with those purchasing a first home for the same market segment—the less expensive homes (up to NIS 1.2 million), and sometimes the investors are the parents of those same young couples.
· In March 2013, we increased the capital requirement and provision in order to increase the cushion for absorbing losses at banks.
· In September 2013, we restricted the payment-to-income (PTI) ratio to 50 percent of disposable income, limited the part of the mortgage at variable rate to two-thirds, and limited the period to final repayment to 30 years.
Basically, the borrowers and banks are focusing on the ability to make the monthly payments and, as we know, the monthly payment can easily be reduced by lengthening the loan period, thereby bearing a larger debt burden. We saw young couples committing to a mortgage until almost retirement age.
More than once I have been asked about the effect on housing prices of the measures taken by the Bank of Israel. I again clarify that the measures we have taken were not intended to stop the granting of mortgages altogether, but to reduce the risk implicit in each and every mortgage and in the mortgage portfolio as a whole. They are intended to protect both borrowers and the stability of the banking system.
Even though these steps reduce the risk, they are still not managing to restrain the pace of new mortgages being taken out, which in March reached a record of NIS 4.5 billion.
If the difficulties in repayment are due to the downturn scenarios in the macroeconomic situation, it will increase the negative impact to banks in view of the ramifications on all borrowing companies. We must remember that a negative impact on the robustness of the banks has implications on the level of the entire economy, and we cannot minimize this. A strong banking system can support the business sector such that it enables it to survive crises and lead the economy back to a path of growth. However, the objective is not only to identify the risks and estimate their possible ramifications, but also to act—to be proactive—in order to reduce the negative impacts resulting from the realization of those risks. That is what we have done, and that is what we will do.
In conclusion, despite the restrictions we have imposed, and which have limited the risk that borrowers can take upon themselves, the individual household may still be exposed to high risk levels. A household is much less able to protect itself from the ramifications of changes in economic conditions than is a bank with capital adequacy, liquidity, and complex tools to manage risks. Therefore, to anyone who wants to take out a mortgage in order to purchase a home, I suggest that they make a proper assessment of their ability to repay the loan over the years, and that they not be tempted to take out credit that cannot be repaid. A higher mortgage over a longer repayment period may enable one to purchase a home, but the future price may be too high.
In parallel with the need for proper management of the households and of the financing banks, the government must continue to act in order to ensure attainable housing by way of supporting an increase in supply.