ProCredit Holding — Actively seeking to expand client base

PCB posted a solid 4.4% increase in the gross customer loan book in Q220 vs end-March 2020 (5.3% in H120 ytd), ahead of its closest peers Raiffeisen Bank International (RBI) and Erste Bank which both reported c 2% growth in the gross customer loan book during Q220 (according to our calculations), highlighting slowing retail and corporate loan demand. After a weaker March and April, PCB experienced a pickup in lending activity from May. According to management, PCB’s lending activity was ahead of its competitors as they toned down their lending operations during the initial phase of the coronavirus crisis. Nevertheless, we note that loan book growth across the banking sector was assisted by credit moratoria introduced in response to COVID-19, which also involved suspension of the principal repayment. Given that 30% of PCB’s loan book was under moratoria at end-March 2020, we roughly estimate that these could have had a positive c 1–2pp impact on its Q220 growth.

The fact that PCB’s loan book expansion was mostly driven by longer-term investment loans rather than working capital facilities should be considered a positive sign. Its green loan portfolio grew by a particularly strong 8.4% in Q220 (with renewable energy projects up 28%), bringing its share in PCB’s overall loan book to 17.3% at end-June 2020 (vs 16.6% at end-2019), closer to its mid-term target of 20%. All the above has encouraged management to raise the full-year guidance for its customer loan book growth to 8–10% excluding any fx impact (compared to management’s previous guidance for low-single digit growth).

Region-wise, loan growth was particularly driven by Southeastern Europe (up 4.9% to €3.6bn in Q220), especially Serbia, Bulgaria, Romania and North Macedonia. Meanwhile, PCB’s Eastern Europe business saw an increase of 2.8% to €1.1bn, with relatively limited fx impact from Ukrainian Hryvnia during the quarter (as per our estimates). Here, it is worth highlighting that Moldova posted solid 14% growth in H120 overall net of any fx impact. Ecuador was up by 4.9% to €0.3bn in Q220.

It is also worth highlighting that the group was able to improve its customer deposit base by c 4.2% in Q220 vs Q120 (and 2.6% in H120) on the back of both business and private deposits, which according to the management reflects the attractiveness of PCB’s direct banking offer (the group had completed the onboarding of all clients onto the digital platform at the beginning of 2020). This means that the risk of deposits outflow amid a liquidity crunch triggered by the COVID-19 lockdown has not materialised yet. The share of current accounts and the FlexSave savings deposits increased compared to end-June 2019, which on one hand limits the increase in interest expense, but on the other hand means that a higher proportion of PCB’s deposit base has lower withdrawal restrictions.

PCB incurred expenses for loss allowances of €15.7m in H120 vs €4.1m in H119, translating into an annualised cost of risk of 67bp, in line with management’s earlier guidance of 50–75bp in FY20. Importantly, the €11.6m y-o-y increase was mostly a function of updated macroeconomic assumptions in PCB’s credit risk models triggered by the pandemic (€8.0m, see Exhibit 1). The €15.7m in provisions for credit losses include just €2.7m related to stage 3 loans, with PCB’s share of credit-impaired loans rising only marginally to 2.5% at end-June 2020 vs 2.4% in Q120 (and remaining stable vs end-2019). The default rate in the company’s green loan portfolio continues to be visibly below the group level at 0.3%. Expenses for loss allowances for stage 2 loans stood at €4.7m in H120 and at end-June 2020 these loans represented 5.3% of PCB’s gross loan book (up from 4.5% in Q120). The increase was a function of continuous individual assessment for all exposures and restructurings.

Consequently, PCB’s ratio of allowances to credit-impaired loans rose to 93.6% vs 89.1% at end-2019 (though this was slightly down from 95.5% in Q120). We note that PCB’s total loan collateral at end-June 2020 stood at €3.8bn, covering c 75% of its total gross loan book, and generally consistent with the €3.7bn shown at end-2019. This includes mortgages (c 66% of collateral), financial guarantees (12%) and other (22%, of which 2% is cash). It is worth highlighting that all investment loans disbursed by PCB are backed by collateral, while working capital loans (with no collateral) are only granted to existing clients that PCB knows well.

Management now expects the FY20 cost of risk to be at the upper end of previous guidance ie at 75bp (compared to our forecast published in the initiation note at 80bp). We understand that the primary reason for this is PCB’s revised outlook for stage 2 loans (and to a lesser extent also stage 3 loans). Management expectations remain broadly in line with RBI (guidance maintained at 75bp) and Erste Bank (65–80bp after a slight upward revision from 50–80bp previously).

PCB’s NIM declined to 2.9% in Q220 on an annualised basis from 3.1% reported both last year (Q219) and in Q120, with H120 NIM at 3.0%. This is largely attributable to the recent rate cuts in Eastern Europe, in particular Ukraine, where the central bank’s main interest rate was gradually cut from 17.5% in Q219 to 6% at present (down 750bp ytd, of which 200bp was in June 2020). We assume that some additional pressure came from the recent US Fed rate cuts given that part of PCB’s loan book is denominated in US dollars (11% at group level at end-June 2020). Consequently, annualised NIM in the Eastern Europe segment was down to 4.2% in H120 compared to 4.5% in H119 (and was c 3.9–4.0% in Q220, according to our estimates).

Meanwhile, NIM in the SEE region declined somewhat as well to 2.4% in H120 vs 2.6% in H119 and was also slightly down sequentially as PCB reported 2.5% in Q120. As Bulgarian rates were at 0% even before the pandemic and in several countries a large part of PCB’s portfolio is likely denominated in euro (eg Kosovo, Serbia or Bosnia and Herzegovina), the impact from local central bank rate cuts has probably come from North Macedonia (rate lowered in January, March and May by an aggregate 75bp to 1.50%) and Romania (down 100bp ytd to 1.50%, with the most recent cut by 25bp in August). Given Serbia represented 25% of PCB’s Southeastern Europe loan book at end-June 2020, the local rate cuts (down by 50bp to 1.25%) likely constituted an additional drag on margins.

Management highlighted that as its competition retreated somewhat during the crisis, it was able to price new loans more favourably in some regions. Having said that, at the consolidated level this was largely offset by PCB’s greater emphasis on the upper medium loan-size segment in some countries (eg in Ecuador or Kosovo), which is characterised by lower margins.

PCB’s net fee and commission income stood at €10.6m in Q220 and fell 18% y-o-y (11% q-o-q) on the back of lower fees from money transfers and card transactions due to COVID-19. However, management highlighted that in June 2020 it experienced an upward trend in the number of transactions by 10% vs May 2020, though these are still below June 2019. Some part of the y-o-y decline most likely came from reduced account maintenance fees due to the departure of non-core customers with low account balances following the full migration of customers to PCB’s digital platform.

PCB’s CIR amounted to 68.5% in Q220 (66.5% in H120) which marks a visible decline vs the Q219 ratio of 71.6% due to a c 3% y-o-y reduction in operating expenses combined with a c 4% increase in net interest income from loan book growth (despite the lower NIM as highlighted above). The lower cost base is partially due to one-off COVID-19-related factors, including lower expenses for transport and travel (down €1.2m in H120), as well as reduced marketing expenses and depreciation of fixed assets. While CIR was up vs Q120 (64.6%), this is mostly due to a seasonal factor as the Bulgarian deposit insurance expense is normally incurred in Q2 (this year it represented €3.5m or 5.8pp in terms of CIR). Management continues to guide to a FY20 CIR of around 70%.

Group net profit reached c €8.0m in Q220 (€21.7m in H120) vs €11.9m in Q219, translating into an annualised ROAE of 4.0% (5.5% in H120). While PCB’s operations in Ecuador again posted a net loss (€0.4m in Q220), it was close to break-even if expenses for loss allowances are excluded with its CIR at c 94%. Management continues to expect ROAE in FY20 to be positive, but below the FY19 figure of 6.9% (we now assume 4.0%, see below).

PCB’s CET-1 ratio in Q220 was broadly stable at 14.1% (vs 14.0% in Q120) and similarly, its total capital ratio (TCR) was 15.7% (vs 15.6% in Q120). These ratios remain well above PCB’s regulatory requirements of 8.2% and 12.6%, respectively. Its risk weighted assets (RWA) also remained largely unchanged vs Q120 despite the loan book expansion, which is mostly due to the partial recognition of the new SME support factors (see our initiation note for details) which reduced the RWA by c €140m as their introduction was pulled forward from 2021 to Q220. Management maintained its CET-1 ratio guidance of above 13% at end-2020. PCB’s liquidity coverage ratio was 142% at end-June 2020 compared to 181% at end-March 2020 and the regulatory requirement of 100%. The decline vs Q120 was mostly due to a slight increase in volume of maturing liabilities over the next 30 days.

Following PCB’s solid loan book growth in Q220 and the upgrade to management guidance for FY20, we have revised our forecast for PCB’s gross loan book growth in FY20 to 9.3% from 2.1% previously (see Exhibit 2). For Q320, we now expect sequential growth of 2.6% (see Exhibit 3), driven especially by the Southeastern Europe segment (3.1%), as growth in the Eastern Europe segment may be dampened by negative base effects from the Ukrainian Hryvnia (assuming the exchange rate remains stable until end-2020) given the strong appreciation in H219 and subsequent depreciation since March 2020. Given that in some countries, moratoria have been introduced for a period of 6 months soon after the pandemic outbreak and recently some of them were extended in selected regions, this should assist loan book growth in Q320 and to some extent also in Q420. PCB’s customer deposit growth in H120 has also encouraged us to raise our forecast for the remainder of the year and we now expect deposits to increase by 5.2% in FY20 (vs a 2.2% decline assumed previously).

Meanwhile, we have reduced our NIM assumptions for FY20 and FY21 on the back of continued central bank rate cuts and PCB’s broadly flat spreads over market rates so far. Consequently, we now expect both FY20 and FY21 group NIM to stand at 2.8% (vs 3.0% previously) and we have also reduced our fee and commission income expenses forecasts given that June 2020 fees has not fully recovered to levels seen last year, according to management. Nevertheless, as PCB’s operating cost base seem to be under control, we still expect a CIR below 70%. We have kept our ROAE forecast unchanged at 4.0% and we continue to pencil in a cost of risk of c 80bp in FY20 and c 60bp in FY21. With respect to risk weighted assets (RWA), we have reflected the remaining positive impact from the SME support factor over the next 12 months, which according to management should be c €110–120m (on top of the €140m recognised in Q220).

We have made only minor changes to our post-tax profit forecasts for PCB and consequently, our PCB fair value estimate changed only slightly from €7.50 to €7.75 due to a lower discount factor associated with bringing the valuation forward.

While PCB has one of the lowest FY20 ROEs in the group (c 4% based on our forecasts), it is also trading at one of the lowest P/BV ratios (0.5x) and as a result is positioned slightly below the regression line in our P/BV-ROE landscape (see Exhibit 4). It is worth keeping in mind that the group is quite scattered across the P/BV-ROE map (eg Georgian banks are traded at quite low P/BV compared to their ROE) making the regression line somewhat less reliable. We believe this may be due to a number of non-financial factors, such as the political risk perceived by investors.