Attica Bank — Refocus and grow

We are at the point in Attica Bank’s transformation where increasing revenue is critical, and this can only be done by growing the balance sheet. The SME lending focus has been refined to the environment, energy and infrastructure segments in addition to targeting the professional personal sector. The bank aims to double net loans in three years (we have from 1.6bn in 2018 to 2.3bn in 2022 in our model) and double the assets in five years. After a decade of under-investment due to austerity and the economic crisis, infrastructure investments are poised to increase. Funding will come from government and EU funds but will also include local lending. The biggest portion is expected to be in energy (PwC forecasts 43%) and green energy will be important as the government wants to decrease its imported fossil fuels reliance. The bank’s aim is not to lend to the big players in these projects, but to lend to the smaller subcontracted companies.

Other key features of the strategic plan include: (1) rebranding and a new image, (2) a new digital platform, (3) further cost cutting including centralisation of back offices and headquarter relocation, and (4) some strengthening of the management team. All of this is planned to be executed by 2020.

Attica is likely to participate in the Hercules scheme, the now ECB-approved Greek government securitisation programme. However, there is some reticence regarding the cost of the Greek government insurance. Management is aiming for zero new net non-performing loans (NPLs) in its legacy book by 2021. It is outsourcing a significant portion of the NPL management and will look into possibly doing another securitisation in the second half of 2020. Our figures do not assume a securitisation; another successful one would be beneficial in reducing balance sheet concerns.

Time needed to finalise and approve the growth strategy has resulted in significant cuts in our forecasts due to lower business volumes. The lower interest rate environment in Greece also prompted lower net interest margin forecasts. We now forecast that Attica will still be loss-making in 2020 (underlying EBT loss of €12.1m, previously €21.6m profit). We cut our 2021 forecast EBT from €33.1m to €9.1m (a return on equity (ROE) of 3.0%). We have raised our fee forecast due to the ambitious bancassurance strategy, but this is overwhelmed by the lower interest income, higher impairments and operating expenses. We also introduce figures for 2022 (ROE of 7.4%). Exhibit 8 details the changes in our forecasts.

Attica now trades at a 2020 P/BV of 0.22x, in line with its Greek peers, which are all trading at depressed ratings due to the coronavirus crises. On our estimates, if the bank raises 400bp equity in 2020, the PBV would rise to 0.41x and it could achieve an ROE of 5.4% in 2022. This implies there is upside if the bank raises equity and matches our forecasts. The bank is at a turning point between loss making and profitability, meaning that estimates are highly sensitive to the pace at which improvements will flow through, with implications for its valuation. Our forecasts are based on management’s strategic plan.

There is clear execution risk as Attica tries to deliver on its game plan in terms of growing revenue, hiring staff, cutting costs, implementing a new digital platform and so on. Lending rates in Greece are falling, so there is risk of further unfavourable interest margin compression. We also note that the fast pace of the loan book expansion carries credit risk that could well bite back in later years. The need for capital is clear and a successful issue is important. Finally, we note that Attica Bank, like most banks, is rather exposed to the macro cycle and unfavourable shocks.

The focus is now on growing revenue after several years of balance sheet clean-up and cost cutting. The traditional focus of the bank has been the SME segment and now it aims to focus selectively in this this segment while eschewing the mass market retail space, where its smaller size and more limited customer reach made it hard to compete against the bigger Greek banks.

Attica believes that with quality service and niche focus it can pick up business. As a small bank (1.8% market share in business lending), there is enough client attrition from the bigger banks to provide scope for Attica’s expansion. The niche approach means that the bank, for example, will not usually aim to make loans to the big corporates in investment projects but to lend to the smaller partners, the smaller subcontractors in the ecosystem.

There have been some management changes, including new chairman Costas Mitropolous and Antonis Vartholomaios being brought in as co-deputy CEO. Besides being chief executive of one of Greece’s most successful management consulting companies, Kantor, and being a senior executive at PwC, Mr Mitropoulos also has banking experience with Eurobank and was the first CEO of Greek privatisation agency, PQH. Mr Vartholomaios was a board member of the Greek privatisation company (Hellenic Republic Asset Development), as served as CEO/chairman at Athens Water Supply and Sewerage Company and his experience also includes management positions in Greek venture capital company EuroHoldings, Egnatia Bank and Pegasus Securities.

The key pillars of the strategy are as follows:

Of the ongoing actions needed to support and underpin this strategy the following need to be highlighted:

Along with skilling up the workforce, Attica’s cost-cutting efforts will continue. Staff reduction will be done through natural attrition. Attica is targeting another 15% decrease in administrative expenses in 2020, with staff costs stable (new hires offsetting attrition). As part of the business plan, the bank expects to redesign and relocate some of its branches.

Because the bank has been focused in recent years on shrinking its balance sheet, cutting costs and cleaning the balance sheet, it has been difficult to start to redirect attention and effort towards growing the balance sheet again. As a result, growth in recent quarters has not matched earlier expectations. This has resulted in both net interest income (NII) and fees falling short due to lower business volumes. We note that single owner business and SME lending rates fell in 2019 (Exhibit 7) by about 20–30bp and this had an additional adverse factor. Following the clarification, approval and roll-out of the new strategy, the bank now appears well placed to deliver on the necessary planned loan growth and business rebuilding.

Delivery of this growth will depend of the successful implementation of the actions highlighted above.

Underlying revenue in Q319 was down 31% y-o-y due to net interest income being 46% lower than 12 months before. The net interest margin as percentage of financial assets shrank from 2.7% in Q318 to just 1.5% in Q319.

The bank reported underlying EBT losses of €9.8m and €8.2m for Q219 and Q319, respectively. The bank is operating quite below its structural capacity and so the drop in revenue as it repositions results in significant losses.

Reported revenues are affected by gains on securities, which are included in other income. The underlying revenue figures exclude some one-offs such as the significant gains on securitisations (€47m) booked in Q418 from the last loan book securitisations (Metexelixis).

Non-performing exposure (NPE) as a percent of gross loans now stands at 46%. The coverage of the NPE by provisions and collateral now stands at 120%. The annualised average impairment charge fell back to below 1% in Q319.

The statutory CET1 ratio was 11.9 in Q319 and we estimate that the fully loaded CET1 was 8.2%.

The Greek economy is expected to continue on its recovery trajectory for the medium term. The sovereign risk has been reduced (the government 10-year bond yield is currently 1.45%) and economic sentiment has reached a 12-year high. Unemployment, which reached almost 30% in 2013, is expected to be 18.5% by the end of 2019 (according to IMF forecasts) and consumer confidence is the highest since 2000. The IMF forecasts GDP growth to remain above an annualised 2%, with both the current and fiscal deficits well in check.

The economic crisis and austerity measures have led to a decade of infrastructure under-investment in Greece. Greece is spending about 0.7% of GDP annually (equal to €1.4bn) in infrastructure, compared to a 1.4% average in the EU and 3.0% in Greece before the crisis. Going forward, the government has recognised the need to increase spending in this area. According to consultancy.eu there has been a €67bn shortfall in the last decade in Greece. For a bank with a net loan book of only £1.8bn after the balance sheet shrinkage, the growth opportunities are truly significant for Attica Bank.

Consulting company PwC believes that the number of large infrastructure projects will increase from 13 in 2019 to 23 in 2020 and about 52–55 for the 2023–26 period. It sees a pipeline of 88 projects worth €25bn in coming years.

PwC estimates that about 43% of these investments will be energy related. Electricity interconnectors and electricity generation will account for about 80% of this and the remainder will include gas interconnectors and liquid natural gas (LNG terminals).

Much of this investment will be in green energy, one of Attica’s focus targets. PwC estimates that 33% of the energy will be in fuel/ignite, 43% in wind, 12.5% in hydro, 5% in solar and 7% in hybrid (hydro and wind).

According to the International Energy Agency (IEA), about 95% of Greece’s energy needs came from fossil fuels in 2017. Nearly two-thirds of this needed to be imported as Greek’s natural energy resources are limited. The government wants to see this decline. It also wants to move to alternatives to fossil fuels to cut down on its carbon footprint. It has approved a scheme (the National Energy and Climate Plan) in Q419 to cut the carbon footprint by 55% by 2030 when compared to 2005. The plan calls for 32% (currently 18% estimated) of renewable energy by 2030 (this matches the EU-wide target for 2030). The National Energy and Climate Plan estimates that the investment amount needed for this is about €44bn over a decade.

Non-energy infrastructure programmes will include rail, tourist infrastructure, waste management and recycling projects, where Greece is still lagging far behind EU goals.

These eye-catching figures underpin Attica’s choice of sector focus. Besides government support, non-government funding will come from the European Investment Bank (EIB), with some from the EU funds, but according to the National Energy and Climate Plan there will also be an emphasis on investment funding at competitive and market terms. Attica’s strategy will not be to take the lead in lending to the major players in the projects, but lending to the smaller sub-contractors where they have relationships and experience.

Greek economic growth has been a very helpful tailwind as the Greek banks have been steadily repairing their balance sheets, shoring up their capital bases and adjusting their operating structures. Of the various challenges facing the sector, managing the large amount of NPLs (€835m gross) on their balance sheets is probably the greatest (stage 3 loans are 46% of gross loans, 38% on a net basis). NPL resolution is therefore an important factor and while steps have been taken to better this in Greece there is still room for improvement. There are still issues with some judges, while out of court resolution success is low (only 3% so far according to government figures). It is not clear if regulation will sufficiently change to significantly improve NPL resolution in coming years. The need to reduce the amount of legacy NPLs is also important. The ECB wants this to drop down to 6% of gross loans, while most banks have non-performing exposures of 30–40% (Exhibit 6). Further massive sector securitisations will be necessary, and the government-sponsored Hercules Asset Protection Scheme plans to reduce these exposures by up to €30bn. This was approved by the EU in October 2019.

Central bank data shows that new business lending activity remains subdued but is improving. However, the stock of lending, both gross (-8% y-o-y, September 2019) and net of provisions (-2% yoy), continues to drop, affected by write-offs, securitisations and provisioning. Attica’s securitisation affected numbers are down 30% and 24% y-o-y for gross and net loans, respectively.

Attica’s new focus on selected smaller business clients is already reflected in the disbursed loans in the first six months of this year. SME and small business lending (SBL) accounted for 83% of new disbursements if we exclude a single loan to a large corporate entity. SME loans accounted for just under a quarter of the loan book at the end of 2018, as shown in Exhibit 3.

Attica’s total coverage ratio (provisions plus collateral) of its NPE was 118% as of H119. The big Greek banks’ coverage ratios were ranging from 100% to 120% at the end of June. We note that Attica continues to have the highest proportion of collateral in its coverage; cash provisions only account for 33% of the NPE as of H119. NPE was reported as 44% of gross loans and this has changed little since the Metexelixis securitisation was completed in Q418 (42%).

Lending rates continue to come down in Greece, but at a gradual pace. Central bank data shows rates on new loans to sole business proprietors at 5.21% in December 2019, compared to 5.5% a year ago, while the SME average rate on new lending was 3.73% in December 2019 compared to 3.93% in December 2018.

We now forecast an underlying loss before tax for Attica of €8.8m (previously €21.6m profit) in 2020. We cut our 2021 forecast PBT from €33.1m to €12.4m (an ROE of 3.0%). We have raised our fee forecast to reflect the benefits from the bancassurance strategy, but this is more than offset by the lower interest income, higher impairments and operating expenses. We also introduce figures for 2022. If Attica Bank can deliver on the growth in its balance sheet, we believe its profitability should start to normalise by 2022, with an ROE of 7.4% (but paying minimal tax due to tax credits). Exhibit 8 below details the changes in our forecasts.

Exhibit 9 shows the earnings sensitivity to movements in the net interest income line. In our base case scenario, we see a 37% CAGR 2019–22e in NII and a 7.4% ROE by 2022. If we increase the NII in 2022 by 10% (a CAGR of 41% 2019–22e), revenue increases by 8%, the underlying PBT by 31% and the ROE estimate increases to 9.7%.

On the other hand, a 10% reduction in the forecast NII in 2022 (CAGR growth 32%), results in a 38% decrease in PBT and an ROE of 4.6%. A 20% reduction in NII results in an ROE of 2.3% and a 69% reduction of PBT versus our base case.

Our estimates are broadly consistent with Attica’s business plans, but we have applied an element of conservatism in our numbers.

Attica will have to eventually raise equity to help fund its balance sheet expansion as it grows into its operating structure. The statutory CET1 was 11.9% on 30 September 2019 and this is well above the currently required 8%. However, the required CET1 will be climbing above 10% based on our growth forecasts, plus there is the phasing in of the transition provisions of ECB’s CRD IV regulation. We estimate the fully loaded CET1 at only 8.6% in Q219. Without any additional capital, we estimate that the fully loaded CET1 could drop to 7.0% by 2021. In our base case scenario, RWA grows 10% in 2019, followed by 25% in 2020 and 35% in 2021.

We believe that Attica will most probably have to live with a CET1 in the 10.5–11.0% range in the future. We therefore believe that 300–500bp of tier 1 capital may have to be raised. This is equivalent to about €117–193m based on the risk-weighted assets (RWA) at the end of 2020 (see Exhibit 12 in the Valuation section).

It is possible that additional securitisations, including the Hercules plan, could reduce the amount of additional capital required. Hidden gains on non-core real estate assets could also reduce this number.

The government’s supported Hercules Asset Protection Scheme was approved by the ECB in October 2019. This could potentially be a game changer for both Attica and Greek bonds. Like the GAPS scheme in Italy, it involves the agency buying up NPLs through an SPV, securitising and then selling them with a state guarantee.

The government believes that as much as €30bn out of the €70bn or so of the Greek bank system NPLs could be removed from bank balance sheets – about 40%. The Greek banks are under pressure to clean up their balance sheets further, including from the ECB. Much has been done, but the banks have NPE ratios in the 30–40% range (Attica has 46%) and we believe the ECB would prefer something closer to 6%. If this is true, it could be that Hercules might have to be followed up with further efforts.

However, we have little clarity on how much capital will be released by these securitisations, especially since some of the loan book could be subject to further write-downs. Attica management is currently assessing its portfolio and aligning itself to prepare to participate in the Hercules plan. The bank needs more visibility on its NPLs before participating. Our estimates assume no gains and no securitisations.

To fulfil the ECB requirements on state aid, the state guarantee is only good for the senior tranche (there will be mezzanine and junior tranches), while the other tranches will have to be sold to private shareholders (at least 50% has to be sold before the state guarantee on the senior is valid). The SPV purchases will be done at market and fair prices. There will be a fee to be paid for the state guarantee.

We highlight the most significant risk factors for Attica Bank:

Execution risk: the bank plans to grow its balance sheet significantly over the next few years. Such a fast pace of growth may bite later if the credit quality turns out to be poorer than anticipated. There is also the risk that the bank is not able to hire the people needed to help drive this growth.

Cost structure requires a lot of growth: with a high cost to income ratio and low profitability, the delta in terms of revenue or cost variance can be high. Recent results have been short of previous guidance.

Capital: the bank will need to raise equity. Although we believe the current share price already reflects capital concerns, the price might be more earnings dilutive than envisaged. In addition, there is no guarantee that enough capital will be raised to deliver on the planned expansion.

Economy cyclical downturn: banks are highly cyclical companies. The Greek economic recovery is expected to continue, and greater slack in the economy offers more space to grow. However, Greece is a relatively open economy and if there is a global slowdown, this would have an impact on Greece, although with a time lag.

Regulation: regulation is always changing and may result in tougher or more expensive compliance standards.

Attica Bank’s share price had tripled over the 12 months since February 2019 (when it was €0.11) until a few days ago. This jump has been driven by country and macro news as well as corporate news. The former includes the continued economic recovery, positive government policy actions, reduction of sovereign risk (the Greek 10-year bond yield has fallen from 4% in February to less than 1.0%) and a new government. Unemployment continues to fall and rebounding house prices help boost bank asset quality. None of the significant banks, including Attica, need emergency liquidity (ELA) funding from the ECB. Attica’s two successful securitisations and its effective cost-cutting efforts have been positive milestones in its recovery.

The coronavirus-related market crash in the last few days has seen the share price drop as low as €0.20p and as of 11 March 2019, it is €0.23p.

Attica is now trading on a 2020e P/BV of 0.22x, which is in line with the Greek banks average of 0.22x). However, the valuations of the other Greek banks are also relatively depressed due to the higher risk premium on the back of the coronavirus crisis – Greek banks’ share prices dropped 15–28% over the last week. We forecast that Attica’s profitability will lag that of its peers based on consensus estimates. Attica’s forecast ROE is 3.0% for 2021, which compares to a peer average of 5.5%; peers range from 3.6% to 8.0%. Attica’s 2021e P/E multiple of 7.2x is higher than its peers because profitability will only start to normalise in 2022. Attica’s 2022e P/E is only 2.8x, which looks attractive. However, this does not include the rights issue, which would be highly dilutive, as discussed below. It could more than double this P/E ratio (see Exhibit 12).

We therefore believe that the market is pricing in a good degree of success in the new strategy and in the rebuilding of Attica’s revenue base.

Attica’s multiples in Exhibit 11 do not take any right issue into account. However, we think it is unlikely that the bank will be allowed to expand its balance sheet in line with our forecasts (net loans from €1.6bn in 2019 to €2.9bn by 2022; management’s plan is to double the size) without raising capital.

Therefore, we think it is useful to look at the bank’s multiples in the context of three different scenarios of 300bp, 400bp and 500bp equity rights issues at the price of €0.23.The amount to be raised as equity in these three scenarios is €116m, €155m and €194m. These are significant levels to be raised compared to the current market capitalisation (€106m). The shares are trading at a relatively low price and the EPS dilution is therefore considerable. We estimate 52–65% 2021 earnings dilution, respectively.

We are assuming that the equity is raised in 2020 in our analyses and we calculate the dilution on 2021, which is when we see the full impact of the fresh capital and when we expect the bank to make significant earnings. Exhibit 12 shows the impact on the earnings, book value and net tangible assets. Exhibit 13 shows Attica’s multiples with the rights issue. The 2021 ROE drops from 3.0% to 2.2% with a 400bp capital hike, for example, in our estimates. The ROE for 2022 falls from 7.4% to 5.4%. We think that the ROE of 5.4% is more indicative of the sort of ROE Attica could be earning in 2022 with a more supportive capital structure. With a 400bp rights issue (we estimate that this would take the fully loaded CET1 to 11.3% in 2021), we estimate it would be trading at a P/BV of 0.40x in 2021e and 0.38x in 2022e, which is not demanding for a capitalised bank well on the path to recovery and offers good upside potential.

The P/E ratio for 2022e would increase from 2.8x to 6.9x with a 400bp capital hike (€155m) in our sensitivity analyses. We are being conservative with the yield forecast of 0% on the fresh money that is assumed to be parked in cash. It is likely that the Greek government would give a little bit of a yield, but this would be low and not that needle-moving in our sensitivity analyses.

We think it is conceivable that besides securitisations, Attica Bank may seek to raise some hybrid capital structured to be accepted as common equity tier 1 by the regulator and this may also reduce what is raised as straight equity.

The bank is currently losing money and needs capital and there is a need to take the longer view. Using a simple (ROE-g)/(coe-g) with about 2% (close to inflation figure) and a COE of 12%, then the required ROE is c 5–6% by about 2021 to justify a PBV of 0.36–0.43x (after rights issue dilution) as seen in Exhibit 13. We forecast 3% before a capital hike, or 2.2–2.4% with capital. However, by 2022 we are forecasting 5.1–5.7% (Exhibit 12) with a rights issue and we think that 7–8% for Attica could be sustainable in the longer run if the company delivers on its expansion plan while retaining good control on credit quality and if the Greek economy remains buoyant. Small business lending can be a profitable niche, particularly with good relationships, and smaller banks can make difference in service quality that sometimes bigger banks do not always deliver to the smaller business clients. Again, this suggests there is upside potential at the current price of €0.23 if the company raises equity and meets the forecasts.

Our loan growth forecasts reflect management’s strategic aim of doubling the loan book over three years and the balance sheet over five. We are taking the company guidance that net NPLs from the legacy book will be neutral and we see the non-performing exposure falling from 46% in Q319 to 27% in 2022. We are not factoring in any securitisations, although these are likely to happen, nor are we assuming any significant sell-off.

The considerable balance sheet expansion will result in RWA also increasing significantly. We assume no securitisations, then we are estimating RWA expands from €3.5bn in Q319 to €4.4bn by 2021. With the bank generating little profit, this results in Attica’s statutory CET1 dropping from 11.9% (Q319) down to a forecast 9.3% in 2021. This decline also reflects the amortisation of some of the deferred tax assets (DTAs) and transition capital items. The fully loaded CET1 declines less since the transition items are already removed, but the starting point is lower (we estimate 8.2% in Q319). By 2021 we estimate it will be at 7.0%.

The December 2019 requirement is 8% and as mentioned we think the bank will probably need to be in the range of 10–10.5% fully loaded by 2021 and we therefore expect that between 300bp and 500bp of capital might conceivably have to be raised, as mentioned in the Capital for growth and Valuation sections. However, in our base forecasts we have not included any rights issue, even if for valuation purposes we have looked at likely dilutions.

We are forecasting a CAGR growth of NII of 37% for 2019–22. We are estimating that Attica’s net interest margin (NIM) will widen from the forecast 1.7% in 2019 to 3.0% by 2022. This will mostly be driven by loan growth (including a higher rate lending mix) but we are also expecting improvements on the liability side. The change in loan mix is important to emphasise since SME loans and small business loans have higher interest rates (and of course risk) than corporate loans and mortgage loans that were part of lending in recent years.

Once again, we stress that our numbers reflect Attica’s business plan and that we have applied some conservatism. There is significant implantation risk, especially given the ambitious growth targets. Other risks include competitive dynamics, falling interest rates, macro risks and unfavourable surprises in the legacy NPL book.

The average cost of financial liabilities is estimated to drop from 174bp in 2019 to 93bp in 2022. This is expected to still be higher than its peers. Reducing the cost of funding is one of the target areas of action in Attica’s business plan. As Exhibit 14 shows, Attica pays a significant premium for its client deposits and it is possible that the company might do better on the liability side, especially if it can securitise and sell down its legacy NPL book and return to profitability. Conversely, the risk on the interest income side is higher on the downside as rates come down, the economy improves and banks become aggressive on the lending front. We are forecasting that Attica’s average yield on financial assets goes from 3.5% in 2019 to 3.9% in 2022.

We expect fees to grow strongly in 2020 and 2021 due to greater business volume (more loans and letters of credit) as well increases in account fees and the development of its bancassurance business. Net fee growth is boosted by the fact that a good portion of the fee expenses are relatively fixed (mainly the fees Attica must pay the Greek government for the guarantee on the securitisations).

We expect operating expenses to fall by 7% in 2019 and then by 5% in 2020 and 1% in 2021. Staff costs are expected to be relatively stable with no VEP, as previously mentioned. The bank hopes to cut other admin expenses by 15% in 2020 and we have allowed for some efficiency gains in this respect. Savings in other areas such as central office consolidation will offset the cost of paying for a new digital platform (Attica Bank is already in advanced negotiations with vendors).

A key objective for Attica is to have net zero losses from its legacy book from 2021. The legacy stage 3 loans are 46% of the loan book and this is a clear risk factor if asset quality does not develop as planned. We are assuming the impairment charges ratio as a percentage of net loans to be 170bp in 2020, followed by 150bp in 2021 and 2022. As a percentage of performing loans (stage 1 and stage 2), the ratios are 130bp, 125bp and 135bp, respectively. However, there is relatively little visibility and therefore it could easily turn out differently and would have an impact on earnings. A 10% increase in the forecast impairment charges in 2021 would result in a reduction of about 10% to pre-tax earnings. We assuming the bank will be using the ample carry-forward tax losses in its DTAs to avoid paying any tax in the coming years.