TRENDS IN LENDING. Fourth Quarter Report 2017

Transcription

1 TRENDS IN LENDING Fourth Quarter Report 217 Belgrade, March 218

2

3 Introductory note Trends in Lending is an in-depth analysis of the latest trends in lending, which aims to ensure better understanding of the conditions prevailing in the domestic lending market. It looks into lending developments, cost of borrowing by households and corporates and lending market conditions, by examining factors behind loan supply and demand. Credit aggregates, as a quantified expression of movements in the lending market, are calculated based on banking sector balance sheet statistics as a source of data on the balance of domestic banks loan receivables. Given the relatively high share of foreign currency-indexed loans in loan portfolios, the increment and growth rates are calculated excluding the effect of changes in the dinar exchange rate against other currencies in the loan portfolio. The report also draws on the results of the bank lending survey conducted by the National Bank of Serbia since early 214. Participation in the survey is voluntary. This survey has greatly improved the understanding of developments in the domestic lending market, allowing insight into bankers perceptions of actual and expected changes with regard to loan supply and private sector loan demand. The report also relies on the results of the survey developed by the European Investment Bank in the context of the Vienna Initiative 2 to monitor deleveraging by cross-border banking groups and the resultant constraints on lending activity. This survey, conducted since October 212 on a semi-annual basis, monitors subsidiaries of international banking groups in Central and South-Eastern Europe, focusing on their strategies, market conditions and expectations. The purpose of the survey is to observe the effects of movement in supply and demand on lending activity, and to gauge the impact of domestic and international factors on supply and demand conditions. Ten Serbian banks participate in this survey, their assets making up around 5% of total assets of the Serbian banking sector. iii

4 ABBREVIATIONS bn billion GDP gross domestic product H half-year LHS left-hand scale М month mn million NPL non-performing loan pp percentage point Q quarter RHS right-hand scale y-o-y year-on-year Other generally accepted abbreviations are not cited. iv

5 Сontent Overview... 6 I. Corporate sector Corporate loans Cost of corporate borrowing Assessment of loan supply and demand based on the results of bank lending surveys... 9 II. Household sector Household loans Cost of household borrowing Assessment of loan supply and demand based on the results of bank lending surveys...13 Methodological notes...14 v

6 Overview Past monetary policy easing by the NBS, increased competition among banks, rising economic activity coupled with the recovery in the labour market, a decline in the country risk premium and low interest rates in the euro area contributed to the acceleration of lending in 217. Growth was recorded despite NPL write-offs, which exceeded the ones from the year before. Excluding the exchange rate effect, 1 in 217 domestic loans increased by 7.4%. Household loans rose by 11.6%, аnd corporate by 4.3%, despite the sizeable write-offs of NPLs and their sale to nonbanking sector entities. Excluding the effect of the NPL write-off during the past year, 2 the recovery is even more evident in 217, total lending increased by 1.2%, corporate loans by 7.4%, аnd household loans by 14.%. Corporate loans continued up in, on the back of a step-up in company lending. The amount of new corporate loans in was broadly the same as in 216, but declined relative to the quarter before, which saw the highest quarterly figures since comparable statistics on new loans became available. Current assets loans made up one half of new corporate loans, while owing to higher disbursement, the share of investment loans rose to 26.4%. The results of lending surveys indicate that in banks eased credit standards for SMEs, while corporate loan demand continued to expand, driven mainly by current assets and investment financing and to a lesser extent by debt restructuring. Easing of credit standards and a rise in demand are expected in Q1 218 as well. Household loans also continued up, and same as before, were driven mainly by the approval of cash loans (including refinancing loans). The extended recovery of housing loans was another contributor to this growth. Excluding the exchange rate effect, housing loans rose by RSD 17.3 bn in 217, which is the biggest growth ever since 211, when the subsidised lending programme was in place. The volume of new household loans in increased by 3.7% from the previous quarter, and by 18.7% from the same period a year before. Almost all cash loans (including refinancing loans) were in dinars 1 Calculated at the dinar exchange rate against the euro, Swiss franc and US dollar as at 3 September 214 (the so-called programme exchange rate used for the purpose of monitoring the implementation of the IMF arrangement), according to the currency composition of loan receivables. 2 In 217, banks wrote off RSD 12. bn worth of NPLs, оf which RSD 74.3 bn pertained to corporates and RSD 23.5 bn to households. and made up around 54% of new loans. Housing loans accounted for a little above 18% of new loans. The share of dinar loans in total corporate and household loans (loan dinarisation) edged up by 1.8 pp in 217, to 33.% at end-december. As households continued to borrow predominantly in dinars, the degree of dinarisation of household loans exceeded 5% at mid-year and reached 51.8% in December, up by 4.8 pp from end-216. Dinarisation of corporate loans equalled 17.5% in December, down by 1.9 pp from end-216, largely due to the write-off of dinar receivables in September. 3 Owing to NBS s measures and successful implementation of the NPL Resolution Strategy, the share of NPLs in total loans recorded a sharp fall. In, this share dropped below the pre-crisis level, to 11.1% in November. In 217 alone, it fell by 5.9 pp, while compared to August 215, when the Strategy was adopted, it plunged by 11.2 pp. The share of NPLs was brought down to its multi-year lows, both in the corporate and household sector. Although intensified NPL resolution activities currently drag down the stock of loans and credit growth, the period ahead is likely to see positive effects on this account as this creates room for new lending, which will provide an additional support to economic growth. Lending activity and GDP (y -o-y rates, in %) Real GDP Tot al domestic loans* Tot al domestic loans*/ ** Sources: NBS and SORS. * Excluding the exchange rate effect. ** Excluding the effect of corporate loans write-off since beginning of According to the Decision on the Accounting Write-off of Bank Balance Sheet Assets which entered into force on 3 September 217, banks are required to write off all loans whose allowances for impairment equal 1% of their gross book value. 6

7 I. Corporate sector 1. Corporate loans In 217, corporate loans (excluding the exchange rate effect) rose by 4.3%. It is noteworthy that this growth was recorded amid a step-up in corporate NPL write-offs (in 217 these write-offs came at RSD 74.3 bn and exceeded the ones from 216 RSD 41.4 bn), which testifies to an accelerated lending recovery. Excluding the effect of NPL write-offs over the past year, corporate loans in 217 went up by 7.4%. 4 It should be borne in mind that NPL resolution should yield positive effects in the period ahead, since the cleaning of banks balance sheets of bad debt creates room for new lending. Favourable trends in corporate lending, in place since June, continued during, so that, excluding the exchange rate effect, corporate loans increased by 1.5%, or RSD 15.7 bn. The growth was recorded for companies, while public enterprise borrowing decreased relative to September. saw an increase despite the write-off of RSD 27.8 bn and sale of RSD 38.6 bn worth of NPLs to entities outside the banking sector. In nominal terms, corporate loans increased by RSD 8.8 bn in, reaching RSD 1,59.9 bn, while their share in estimated annual GDP 5 in December measured 23.7%. More favourable tendencies are confirmed also by the total amount of new corporate loans in (RSD 244. bn). The approved amount is similar to the one from the same period in 216, but lower than in, which saw the highest quarterly figures since comparable data on new loans became available. Same as before, current assets loans held the largest share, accounting for one half of new corporate loans in. The share of investment loans rose to 26.4%, and import loans also continued up, amounting to 5.6% of new corporate loans. The structure of new loans was reflected also on the structure of corporate receivables. Thus, the share of current assets loans increased to 48.8% in December, while investment loans made up 31.5% of corporate receivables. During, as well as on the annual basis, majority of loans were extended to companies in manufacturing, trade and construction, the sectors which gave the greatest contribution to GDP growth in 217. Effect of NPL write-offs on lending growth to corporates (y -o-y growth rates at the programme exchange rate, in %) Corporates Corporates* Total domestic loans Total domestic loans* * Excluding the effect of NPL write-offs over the last year. Increase in corporate lending, by currency (in RSD bn) FX and FX-indexed loans* Dinar loans Total * Excluding the exchange rate effect. 215 Structure of new corporate loans (in RSD bn) Current assets Other Export Investment Import 4 For more details on credit growth in conditions of intensive NPL write-off please see Text box 2 in the Inflation Report November GDP achieved in the last four quarters. 7

8 The share of dinar in total loans to corporates measured 17.5% in December, down by 1.9 pp from end-216. This is largely attributable to the write-off of dinar receivables in September when this share contracted by 1.6 pp. 6 Euro-indexed and eurodenominated loans still made up the bulk of corporate loans (8.8%), while the share of US dollar- and Swiss franc-denominated loans continued to decline in December, these loans made up 1.2% and.4% of the loan portfolio, respectively. Implementation of the NPL Resolution Strategy 7, additionally supported with the coming into force of the Decision on the Accounting Write-off of Bank Balance Sheet Assets as of September 217, continued to yield good results in 217, аs confirmed by the assessments of the IMF and rating agencies. NPL amounts dropped down further in, owing to a stepup in banks collection, resolution, write-off and sale of NPLs, underpinned by the NBS s measures. The NPL reduction in absolute amount on the one hand and growth in credit activity on the other led to a continued shrinking of the share of NPLs in total corporate loans. In, this share dropped below the pre-crisis level. The share of NPLs in total corporate loans shrunk by 1.4 pp in October and November (down by 5.4 pp since the start of the year), to 11.8% in November. Looking at companies only, the share of their NPLs in total loans measured 11.9% in November, which is a drop of 1.5 pp from September and of 5.6 pp since end-216. The NPL share decreased in all sectors and is currently at its multiyear lows. The greatest drop is recorded in sectors that were most severely hit by the crisis (construction, manufacturing, real estate). From the standpoint of financial stability, it is important to note that allowances for impairment to total loans in November stood at 69.3% of NPLs, and that regulatory provisions for balance sheet exposure continued to fully cover gross NPLs, at 138.7% in November. Also, following the introduction of Basel III standards, 8 the capital adequacy ratio edged up further, to 22.5% in September, testifying to the high capitalisation of the domestic banking sector. 6 According to the Decision on the Accounting Write-off of Bank Balance Sheet Assets which entered into force on 3 September 217, banks are required to write off all loans whose allowances for impairment equal 1% of their gross book value. 7 Activities envisaged in the NBS s Action Plan ( aimed at boosting banks capacity for NPL resolution and providing a contribution to the development of the NPL market, were implemented in full, some of them even before the deadline. Their implementation was one of the important factors that led to the sharp fall in NPLs in 216 and 217. For more details on the measures undertaken and results achieved in terms of NPL resolution by the end of 216 please see Text box 2 in the Inflation Report from February Basel III regulatory framework is applied as of 3 June 217. Corporate loans by sector, end-of-period (in RSD bn) 1, 4 1, 2 1, Transport and telecommunications Real estate; scient. and serv. act.; arts, enter. and recr. Trade Construction Mining, manufacturing, water management Agricult ure, forest ry, fishing Other Currency composition of bank claims on corporates (in %) RSD EUR USD CHF Other 216 Structure of gross NPLs of corporates (in %) NPL share in total corporate loans Manufacturing, mining Wholesale and retail trade Construction Real estate business Transport, informat ion, communications Agriculture, forestry and fishing * Data for 217, as at November

9 2. Cost of corporate borrowing The downward trend in interest rates on new dinar corporate loans began in September 213 (from 17.7%) after a cycle of NBS key policy rate cuts initiated in May that year. Since then, rates on dinar loans to corporates fell by 12.9 pp to their lowest of 4.7% in December 217. Relative to September 213, rates on euro-indexed corporate loans fell by 4.2 pp, reflecting lower rates in the euro money market, and, to a larger extent, a drop in the country risk premium, owing primarily to the strengthening of domestic macroeconomic fundamentals. Increased interbank competition in the lending market also worked towards a fall in interest rates on both dinar and euroindexed loans. By late 217, rates on all types of dinar corporate loans touched their new lows. In December, the weighted average rate on new dinar corporate loans stood at 4.7%, down by 1.2 pp from September. The rate on current assets loans fell by the same amount, to 4.3% in December. The rates on investment loans and loans for other purposes fell by.2 pp and 1.1 pp to 5.5% and 5.% respectively. In, the weighted average rate on euro and euroindexed corporate loans also declined, by.2 pp to 2.8% in December. The rates were trimmed on current assets loans by.3 pp to 2.6%, euro-indexed loans for other purposes by.6 pp to 2.7%, and import loans by.1 pp to 2.1%. It was only the price of investment loans that edged up somewhat by.1 pp from September, to 3.3% in December. Interest rates on dinar corporate loans and deposits* (av erage weighted, annual, in %) New dinar corporate loans New dinar corporate deposits BELIBOR 3M, period-average Key policy rate, period-average * Excluding revolving loans, current account overdrafts and credit card debt. Interest rates on FX corporate loans and deposits* (weighted av erage, annual, in %) New FX corporate loans** New FX corporate deposits EURIBOR 3M Sources: NBS and European Banking Federation. * Excluding revolving loans, current account overdrafts and credit card debt. ** Euro and euro-indexed. 3. Assessment of loan supply and demand based on the results of bank lending surveys According to the results of the January NBS bank lending survey, in aggregate terms, after easing their credit standards in and, banks kept their corporate credit standards unchanged in. Under the criterion of the company size, standards were eased for SMEs, which corresponds to the assessments stated in the EIB s CESEE Bank Lending Survey. 9 Standards were eased for short-term loans, owing to interbank competition and higher risk propensity, as well as lower costs of funding for some banks. Interbank competition and higher risk propensity Interest rates on corporate loans (av erage weighted, annual, in %) Import * Investment* Current assets* Other* Current assets** Other** * Euro and euro-indexed. ** Dinar. 9 Ten banks from Serbia whose assets account for somewhat less than a half of total assets of the Serbian banking sector participate in this survey: 9

10 should contribute to moderate easing of standards in Q In the terms of borrowing improved for corporates due to lower margins and associated costs, and an increase in the maximum loan amount. In case of FXindexed loans, payment terms were also extended, which is probably why collateral requirements were slightly tightened. As it was expected, loan demand of corporates continued up in. According to the EIB survey, corporate loan demand in Serbia was stronger than the region s average. Demand was led mainly by the financing of current assets and investment and, to a lesser extent, debt restructuring. These factors are likely to remain the main drivers of loan demand in Q Impact of individual factors on changes in standards for the extension of loans and credit lines to enterprises (in net percentage) Total * Costs of funding** Competition from other banks** Risk on the collateral demanded** Expectations regar di ng general economic activity** Risk propensity** Uncollectibility of receivables** Q1 217 Q1 218 Q1 217 Q1 218 Q1 217 Q1 218 Q1 217 Q1 218 Q1 217 Q1 218 Q1 217 Q1 218 Q1 217 Q1 218 Achieved Expected * Positive values indicate tightening of conditions and negative easing, relative to the previous quarter. ** Positive values indicate the contribution of individual factors to tightening, and negative values indicate the contribution to easing of lending standards. Changes in conditions for the extension of loans and credit lines to enterprises (in net percentage) Dinar FX 1% % -1% -2% -3% -4% -5% -6% -7% Interest margin Fees and commissions Maximum loan amount Collateral requirements Maturity Interest margin Fees and commissions Maximum loan amount Collateral requirement s Maturity Achieved Expected * Positive value indicates tightening of conditions and negative easing. ** The intensity of change is not shown. Changes in enterprises' credit demand (in net percentage) 9% 8% 7% 6% 5% 4% 3% 2% 1% % -1% To tal Dinar shortterm Dinar long-term FX shortterm FX longterm SMEs Large enterprises Fa rmers Achieved Expected * Positive value indicates an increase in demand, and negative value indicates a decrease. ** The intensity of change is not shown. 1

11 II. Household sector 1. Household loans Excluding the exchange rate effect, household loans picked up by 11.6% in 217, even though banks wrote off NPLs worth RSD 23.5 bn in the course of the year. Excluding the NPL write-off effect, household lending rose by 14.% in 217. The stock of household loans equalled RSD bn in December, accounting for around 45% of bank receivables from the nonmonetary sector. Their share in estimated annual GDP equalled 2.1% in December. In the course of, excluding the exchange rate effect, household loans rose by 1.9% or RSD 16.9 bn. Of this, cash loans (including refinancing loans) increased by RSD 14.2 bn and housing loans by RSD 4.9 bn. In 217, housing loans incraesed by RSD 17.3 bn, up by 2.3 times from the year before. The volume of new household loans in (RSD bn) was 3.7% higher than in and 18.7% higher than in the same period last year. Citizens still opted mostly for dinar cash and refinancing loans. In, these loans accounted for 53.8% of new loans to households, and 71% of those had repayment terms longer than five years. Housing loans continued to recover, as confirmed by a high amount of new housing loans (RSD 2.7 bn), up by 19.4% y-o-y. In a low interest rate environment, in addition to the approval of new loans, we also saw increased refinancing of current housing loans. Credit card borrowing stayed unchanged, while current account overdrafts declined slightly relative to. In terms of purpose, housing loans remained the dominant category of household loans at 39.6% in December, followed by cash loans with 38.5%. Consumer loans accounted for 2.4% of household loan receivables. These were, for the most part, dinar loans for purchasing mobile phones and household appliances, as well as FX-indexed car purchase loans. Households continued to borrow predominantly in dinars (on average, in 217, 71% of newly-approved household loans were in dinars). As a result, the dinarisation of household lending trended further up, exceeding 5% in mid-217 and arriving at 51.8% in December. Relative to end-216, dinarisation increased by 4.8 pp (of which by 1. pp in ). The share of euro-indexed and euro-denominated loans fell to 4.6%, and that of Swiss franc loans to 7.7%. Effect of NPL write-offs on lending growth to households (y-o-y growth rates at the programme exchange rate, in %) Households Households* Total domestic loans Total domestic loans* * Excluding the effect of NPL write-offs over the last year. Increase in household lending, by currency (in RSD bn) FX and FX-indexed loans* Dinar loans Total * Excluding the exchange rate effect. Structure of new household loans (in RSD bn) Housing Cash loans Consumer Other* Source. NBS. * Until 214, the 'other loans' category implied cash and other loans together. Currency structure of bank claims on households (in %) RSD EUR CHF

12 As in the case of the corporate sector, the fall in the NPL share in total household loans accelerated further in 217. This was due not only to increased lending, but also stepped-up NPL write-offs, which significantly exceeded the levels recorded in 216 RSD 23.5 bn in 217 vs. RSD 4.3 bn in 216. A significant improvement in indicators concerning NPLs is also highlighted in the latest EIB s CESEE Bank Lending Survey. As stated in that Survey, the improvement in the household sector in the past period was faster than the region s average, and the trend is expected to continue in the coming period. The share of NPLs in household loans equalled 6.% in November, down by.5 pp from September and 3.3 pp from end-216. Inclusive of entrepreneurs and private households, the share was 6.4% in November, which is by.7 pp lower than in September and by 3.7 pp than at end-216. All loan categories registered a drop in the share of NPLs. In October and November, the share of NPLs in cash loans dipped by.1 pp to 4.6%, in housing loans by.4 pp to 6.4%, and in consumer loans by 4.1 pp to 6.2%. 2. Cost of household borrowing Owing to significant monetary policy easing by the NBS, a fall in the country risk premium, low rates in the international money market and increased interbank competition in the lending market, the cost of household borrowing halved in comparison to 213. Relative to May 213 when the cycle of monetary policy easing by the NBS began, rates on new dinar loans contracted by 1. pp and on euroindexed loans by 3.8 pp. Household borrowing at lower interest rates reduces the costs of the repayment of existing loans, which positively reflects on the households disposable income. Continued monetary policy easing in 217 contributed to a further decline in rates on dinar household loans in. The weighted average rate on dinar household loans edged down.2 pp to 1.6% in December. The sharpest fall was recorded for rates on consumer loans (by.7 pp to 8.3%), while the price of other loans fell by.3 pp to 9.%. The average rate on cash loans remained broadly flat from August, measuring 11.1% in December. The average rate on euro and dinar euro-indexed household loans increased in by.2 pp to 4.2% in December. The rate on consumer loans went up by.1 pp to 5.%. The rate on housing loans (3.%) and other loans (6.5%) were at a similar level as in September, while the price of cash loans fell by.4 pp to 3.%. Structure of gross NPLs of natural persons (in %) I 211 III NPL share in total household loans Cash Credit cards Housing Current account overdraft Consumer * Data for 217, as at November Interest rates on dinar household loans and deposits* (av erage weighted, annual, in %) Key policy rate, period average BELIBOR 3m, period average New dinar household loans New dinar household deposits * Excluding revolving loans, current account overdrafts and credit card debt. Interest rates on FX household loans and deposits* (weighted average, annual, in %) New FX household loans** New FX household deposits EURIBOR 3m Sources: NBS and European Banking Federation. * Excluding revolving loans, current account overdrafts and credit card debt. ** Euro and euro-indexed. Interest rates on household loans (av erage weighted, annual, in %) Consumer** Consumer* Housing** Housing* Cash** Cash* Other** Other* * Euro and euro-indexed. ** Dinar. 12

13 3. Assessment of loan supply and demand based on the results of bank lending surveys According to banks estimates, growth in household lending reflects the continued positive effect of factors on both the demand and supply side. Credit standards were eased further in, while demand continued up. According to the results of the January NBS lending survey, as it was expected, banks further eased their household credit standards in. The easing concerned dinar cash loans, refinancing loans and FXindexed housing loans, and was under the strongest impact of interbank competition and cheaper costs of funding. Banks are expected to continue easing credit standards in Q1 218 primarily in response to interbank competition and brighter prospects in the real estate market. As suggested by survey results, in banks continued to offer more favourable terms for both dinar and FXindexed loans. Payment terms were extended further and banks continued to lower interest rates margins and associated costs. Compared to, there was an increase in demand for dinar cash and refinancing, followed by housing loans. The main drivers of demand were the refinancing of current obligations, rising employment and wages and real estate purchases. Banks expect that demand will continue up in Q1 218 as well. Impact of individual factors on changes in standards for the extension of loans and credit lines to households (in net percentage) Total * Costs of funding** Competition from other banks** Expectations regarding general economic activity** Uncollectibility of receivables** Q1 217 Q1 218 Q1 217 Q1 218 Q1 217 Q1 218 Q1 217 Q1 218 Q1 217 Q1 218 Achieved Expected * Positive values indicate tightening of conditions and negative easing, relative to the previous quarter. ** Positi ve values indicate the contribution of indi vidual factors to tightening, and negative values indicate the contribution to easing of lending standards. Change in conditions for the extension of household loans (in net percentage) 1% 5% % -5% -1% -15% -2% -25% -3% Interest margin Dinar Fees and commissions Collateral requirements Interest margin FX Fees and commissions Collateral requirements Achieved Expected * Positive value indicates tightening of conditions and negative easing. ** The intensity of change is not shown. Change in household credit demand (in net percentage) Dinar Total 7% 6% 5% 4% 3% 2% 1% % -1% -2% -3% Housing Consumer Cash Refinancing Housing Consumer Cash Refinancing Q1 `17 Achieved Expected * Positive value indicates an increase in demand, and negative value indicates a decrease. ** The intensity of change is not shown. FX 13

14 Methodological notes Loans imply bank receivables under the loan principal. Placements imply receivables under loans, interests and charges, paid deposits, securities and shares of companies. All types of receivables are expressed according to the gross principle, i.e. not reduced by allowances for impairment. Placements in dinars are placements approved in dinars without an FX clause. The FX clause implies a currency clause that defines hedging against changes in the dinar exchange rate. When excluding the exchange rate effect, the calculation is based on the original currency composition and the exchange rate of the dinar to the euro, the US dollar and the Swiss franc as at 3 September 214. New business includes all financial arrangements (credits and deposits) the terms of which are agreed for the first time during the reporting month, as well as all existing contracts the terms of which were reagreed (through annexes), with the active participation of the client. The sectoral classification of monetary statistics is used. The corporate sector includes public enterprises, companies and the non-financial sector in bankruptcy, while the household sector includes citizens, entrepreneurs, private households with employed persons and registered farmers. By way of exception: with newly-approved loans, the household sector includes non-profit institutions serving households (in accordance with the ECB methodology); with non-performing loans, the sectors are presented separately, but are aggregated for the sake of comparison with the monetary statistics data. The term non-performing loans implies the stock of the total remaining debt under individual loans (including the amount of arrears): Non-performing loan represents the total outstanding debt under an individual loan (including the amount of arrears): where the payment of principal or interest is past due (within the meaning of the decision on classification of balance sheet assets and off-balance sheet items) over 9 days, where at least 9 days of interest payments have been added to the loan balance, capitalized, refinanced or delayed by agreement, where payments are less than 9 days overdue, but the bank has assessed that the borrower s repayment ability has deteriorated and doubts that the payments will be made in full. 14