Saturday 13 February 2016

Eurobank Ergasias SA

Never count on making a good sale. Have the purchase price be so attractive that even a mediocre sale gives good results. —Warren Buffett
In his fantastic book - The Dhandho Investor: The Low-Risk Value Method to High Returns – Mohnish Pabrai lists buying distressed businesses in distressed industries as one of his nine frameworks for investing. It would be difficult to find anything more distressed today than a Greek Bank. This is not any ordinary distress – the European banking shares have suffered a severe sell-off with some of the largest names now trading at well below Tangible Book Value (TBV); the Greek economy is still not out of the doldrums, with tough ongoing bailout negotiations, political and social unrest, and a brewing refugee crisis; and, the market seems to have lost all confidence in the Greek banks in spite a recapitalisation exercise just months ago bolstering their balance sheets further – this is distress cubed.


One would think that a 4th recapitalisation exercise in 5 years should have weeded out all balance sheet problems by now. And, apparently, the most recent one was based on worst case stress scenarios over the next 2.5 years. The markets seem to be pricing in an Armageddon with respect to Greek banks. In such situations lie opportunities. 

Eurobank Ergasias
It is one of the four systemic banks in Greece, with over 20% market share in loans

As part of the consolidation of the Greek banking system, it, like the other 3 systemic banks, experienced a material increase in scale with acquisition of and subsequent mergers with New Hellenic Postbank and New Proton Bank, in 2013

The four systemic Greek banks have now been through 4 capital raising exercises in the last 5 years after Asset Quality Reviews carried out by the ECB and IMF. In the most recent Asset Quality Review last year by the ECB, Eurobank came out in relatively better shape compared to the other 3 Greek systemic banks – it needed the least amount of capital, and had the lowest NPE ratio across the 4 banks

In my opinion, Eurobank has the best corporate governance structure and management of all the four banks

Further to the recapitalisation in November 2015, Eurobank has a Tangible Book Value (TBV) of €5.6 billion (or €2.54 a share). Based on the share price at 12 Feb 2015, it trades at just 13% of TBV, 76% of estimated 2015 Pre-Provision Income, and 1% of its total assets. 
Key valuation ratios
€M
per share
Shares in issue (post Nov15 recap)
       2,186

Tangible Book Value (post Nov15 recap)
       5,562
 €2.54
Market Cap (12 Feb 2016)
          708
 €0.32
Total Assets (Q315)
     73,755
 €33.74
Pre-provisions Income (average for last 4Qtr's annualised)
          859
 €0.39



Price / Tangible Book Value
13%
13%
Market Cap / Pre-provision Income
82%
82%
Market Cap / Total Assets
1%
1%

The TBV for a bank is equivalent to liquidation value – it is what an equity holder would expect to get if the bank is to be liquidated today. Therefore, at 12% TBV, the market seems to be pricing in a severe drop in profitability, or massive write-downs in the loan book, or still has concerns about Greece’s ability to stay in the Eurozone. 

Likelihood of Greece exiting the Eurozone
I will examine the last one – Greek exit - first, and briefly. I am not an economist, but my take on the Greek situation – in spite of all the bailout talk uncertainties, social and political unrest owing to the austerity measures, and the brewing refugee crisis – is that Greece is unlikely to leave the Eurozone. The situation between the Greece and its European partners could make for an interesting game theory scenario. As in a game theory scenario, the Greek’s can either cooperate (by implementing the reforms), or defect (by not implementing the reforms); its European partners on the other hand, can either cooperate (by agreeing to debt relief), or defect (by not agreeing to debt relief). 

Here’s my simple game theory scenario
Greece bail-out scenarios
European partners cooperate by giving debt relief
European partners defect by not giving debt relief
Greece cooperates by implementing reforms
The most likely outcome. There will definitely be hard negotiations on who gives more and there will be pain for both parties. But in the long-term this option brings gain for all.
The Greek’s are unlikely to take a sucker’s payoff and implement reforms without any debt relief
Greece defects by not implementing reforms
The European’s are unlikely to accept a sucker’s payoff and offer debt relief without any reforms
Too much uncertainty, turmoil, and too big a gamble with the future for Greece. The European’s now have the refugee crisis to worry about, in addition to a contagion. Both unlikely to defect.
Neither party obviously wants a suckers pay-off, so one can rule out situations where one party cooperates and the other defects. Obviously, both cooperating will be the best outcome. But Greece has a lot more at stake as a society (read this fantastic blog post by an IMF banker re the situation – particularly around the pensions reforms). In my opinion, Greece is unlikely to cooperate in full by implementing every requested reform. And, Greece’s European partners are unlikely to accept to debt relief without Greece giving in to some reforms. Both parties cooperating will entail some give and take on both sides. 

The question, when examining the probability of a Greek exit, comes down to - will both parties want to defect? I think it’s only this scenario that leads to a Greek exit from the Eurozone. The Greek’s probably have the most to lose by exiting the Eurozone – the uncertainty an exit from bring about Greece’s future, the likely terrible turmoil in the short to medium term its people will need to endure, and, above all, it will be a gamble on an uncertain future. But I don’t think that Greek’s European partners can take defecting as an option either. There is one very important reason for this. Greece is at the epicentre of a brewing refugee crisis - the country was the entry point into Europe for more than 850,000 asylum-seekers last year, and is already becoming a vast holding pen. Would its European partners really want Greece to exit the Eurozone, and plunge its economy and society into a turmoil, right in the middle of a refugee crisis? None of these refugee’s entering Greece want to remain in Greece, their ultimate destination is Germany, France, or the UK. I therefore think that the Eurozone cannot afford a Greek exit, the same way that the Greek’s cannot afford a Eurozone exit. The refugee crisis gives Greece a bargaining chip in the negotiations.

Therefore, Greece is unlikely to leave the Eurozone. 

Let’s now focus on profitability and balance sheet risk.
Profitability
There is a lot to like about Eurobank.

It reported better than expected 2nd and 3rd quarter earnings in 2015. This in spite of Athen’s having imposed capital controls to limit withdrawals in the 3rd quarter.

It has a favourable cost-to-income ratio @ 53%, and this is likely to improve further with its planned cost cutting measures. Also, the online banking and increasing debit card issuance will lead to further reduction in cost of servicing the customers.

Greek banks have experienced over €115 billion of reduction in deposits since the peak in 2009. As and when confidence returns in the Greek economy, bank deposits should improve. This should help improve the cost of funding, and increase net interest income. Eurobank should benefit the most as and when this happens. This is unlikely to happen in the short term.

The bank operates in a consolidated market – if one wants to bet on Greece, Eurobank offers a good route - a bet on it is a bet on Greece.

If one believes that ECB’s Asset Quality Review last year was robust, then the subsequent capital raise should give one confidence in the balance sheet and the current sell-off would appear more sentiment and not fundamentals driven.

At 13% TBV, the share price offers an attractive entry point. 

Upside scenarios
Eurobank’s average Pre Provision Income (PPI) over the last 4 QTR has been c€214.8m, given a TTM PPI of €859m. It had total assets as at September 2015 of €73.75 billion. This gave it a pre provision ROA of 1.14% (€859/€73.75m). The below tables shows the share price projection and returns under a number of different ROA and PE scenarios.
Share price scenarios (€M)



ROA--->
0.10%
0.25%
0.50%
0.75%
1.00%
P/E





2
      0.067
         0.169
      0.337
      0.506
      0.675
4
      0.135
         0.337
0.675
1.012
      1.350
6
      0.202
         0.506
1.012
1.518
      2.024
8
      0.270
         0.675
      1.350
      2.024
      2.699
10
      0.337
         0.843
      1.687
      2.530
      3.374






Return scenarios




ROA--->
0.10%
0.25%
0.50%
0.75%
1.00%
P/E





2
-79%
-48%
4%
56%
108%
4
-58%
4%
108%
212%
317%
6
-38%
56%
212%
369%
525%
8
-17%
108%
317%
525%
733%
10
4%
160%
421%
681%
941%

Based on the current share price, a moderate improvement in ROA (0.5% - 0.75%) and a conservative PE of 4-6 produce a 108% to 369% return on current price. But are these ROA assumptions reasonable?

Testing the bullish case for reasonableness
The biggest sensitivity to ROA comes from provisions on the loan book. Although pre provision ROA’s have been positive, provisions have more than wiped out the PPI over the last 4 quarters and before.  

Assuming the TTM PPI of €859m and total assets of €73.75 billion remain constant, a ROA of 0.5% requires that the loan loss provisions be contained at €490m per annum, and a ROA of 0.75% require that the loan loss provisions be contained at €306m per annum.

As majority of the provisions emanate from loans that are 90 days past and due (90 dpd), one way to test the sensitivity of loan loss provisions is by looking at the formation of 90 dpd.

90dpd formation over the last four QTR’s has averaged €228.5m; annualised, this equates to €914m. The bank’s current loss given default rate is 53%. Assuming this holds true, this equates to a provision of €484m on the TTM 90 dpd formation. This would generate a 0.50% ROA. A 0.5% ROA with a PE of 4-6 could generate a 108% to 212% return based on current share price.

However, note that the ROA is highly sensitive to default and loss given default rates, both of which in turn are sensitive to 90 dpd formation. If default and loss given default rates increase from current levels, then the ROA falls. Therefore the bullish case assumes that the 90 dpd formation is contained and reduced, along with an improvement in loan recovery rates. A brief look as 90 dpd formation over the years and the last 4 QTRs is shown below:  
Last 7 years
2009
2010
2011
2012
2013
2014
2015F
90 dpd formation (€m)
      1,510
      1,885
      2,268
          2,866
      2,324
      1,386
         914
The 90 dpd formation was it its highest at the height of the crisis between 2011 and 2013. Since 2013, it has been falling, and for the first time, it is likely to come below €1 billion in 2015. The below table shows the 90 dpd trend over the last 4 QTR’s.
Last 4 QTRs
Q414
Q115
Q215
Q315
Average
90 dpd formation
239
391
118
166
228.5




2015F
914
 The 90 dpd formation has been falling over the last 4 QTR’s from a high of €239m in Q4 2014, to €166m in Q3 2015. This, in spite of the fact that 2015 was a tumultuous year for Greece with the imposition of capital controls and shut down of the banking system. Therefore, based on the trend over the years and the last 4 QTRs, one could argue that the trend in 90 dpd formation augurs well for default rates and ROA.
The table below models ROA under various default rate and PPI scenarios. The table Note: it is not reasonable to assume that PPI will stay constant as default rates increase; net Interest Income, the largest component in the PPI, will likely fall if default on loans increase.

ROA sensitivity





PPI ----> (€m)
859.3
687.44
601.51
515.58
429.65
PPI scenarios-->
current
-20%
-30%
-40%
-50%
Provision (% of 90dpd formation)





53%
0.51%
0.28%
0.16%
0.04%
-0.07%
60%
0.42%
0.19%
0.07%
-0.04%
-0.16%
70%
0.30%
0.06%
-0.05%
-0.17%
-0.28%
80%
0.17%
-0.06%
-0.18%
-0.29%
-0.41%
90%
0.05%
-0.18%
-0.30%
-0.42%
-0.53%
100%
-0.07%
-0.31%
-0.42%
-0.54%
-0.66%
The best case ROA of 0.51% assumes that PPI, 90 dpd formation, and provision for loss given default stay constant at the current levels. Any deterioration quickly trips the net income to negative. It goes without saying that for any ROA over 0.50%, one requires that both loss given default and 90 dpd formation is contained and reduced, or PPI improves significantly.
In summary, the recent improvements in PPI, and the positive trend in 90 dpd formation, augur well for ROA. However, sensitivities remain, and the next few quarters will be a useful indicator.

Balance sheet risk
The other major risk which a prospective investor must get comfortable on is the balance sheet risk – whether provisions for loan losses are adequate at current levels or if there is risk of massive write-downs leading to additional capital raising exercises, potentially wiping out ones investment.

As discussed in the introduction, ECB has recently undertaken an Asset Quality Review, which included stress testing under adverse scenarios, after which capital was raised to shore up the balance sheet. If one can place comfort on ECB’s review, then one can place comfort on the balance sheet.

Let’s take a look at the ECB stress test and Eurobank’s balance sheet in more detail. The stress test determined capital adequacy over a 2.5 year time period (H2 2015 – 2017), resulting in the over €2 billion additional capital being raised in November 2015. The TBV increased from c€3.6 billion to €5.6 billion. The key assumptions for the stress test were: cumulative read GPD growth of (6.8%) over the tested period, cumulative residential house price fall of (22.5%), commercial real estate price fall of (8.8%), inflation of (1.1%), and an unemployment rate of well over 27% during the period. Basically, pretty bad scenarios one would have thought. The stress test resulted in additional implied write-offs/provisions on the loan book of €1.9 billion. Eurobank’s existing provisions for loan losses as at 3rd QTR 2015 already covered 97% of the total provisions required under this adverse scenario.

Let’s take a brief look at the key balance sheet numbers.

NPE analysis
Sep-15 (€M)

Gross Loan
       52,693


Provisions
-11,739


Net Loan
       40,954


Total NPE
-22,000






Key metrics



Implied Default rate
42%
[NPE/Gross Loan]
Loss given default
53%
(Provisions/NPE)

The table above shows that as at Q3 2015, the bank had gross loans of €52.7 billion, with a provision for loan loss of €11.7 billion. Total NPE, pre-dominantly consisting of 90 dpd loans was €22 billion. This implies a default rate of 53%, and a loss given default of 42%.

Assuming a constant loan book, the additional write-downs (provisions) that the bank would need to take under various default rates and loss given default scenarios is shown below.
Write-down/provision scenarios (€M)




Default rate--->
45%
50%
55%
60%
65%
70%
Loss given default






55%
-1,303
-2,752
-4,201
-5,650
-7,099
-8,548
60%
-2,488
-4,069
-5,650
-7,230
-8,811
-10,392
65%
-3,674
-5,386
-7,099
-8,811
-10,524
-12,236
70%
-4,859
-6,704
-8,548
-10,392
-12,236
-14,081
75%
-6,045
-8,021
-9,997
-11,973
-13,949
-15,925
80%
-7,230
-9,338
-11,446
-13,554
-15,661
-17,769

Broadly, if default rates rise to 60%, or if loss given default rise to 75%, or both rise by another 10bps-15bps from current levels, the write-downs required would be in the region of €5.6 billion - equal to current TBV. Such a scenario will most likely require additional capital raise, potentially wiping out current shareholders.  It is worth nothing that alternatives to new capital raise exist: e.g. HFSF holds €950 million of preferred stock in Eurobank which can be converted to common equity, and, potentially deposits over €100k could be wiped out to bail-in. But this too, like a capital rise, would have the effect of diluting current shareholders.
In conclusion
An investment at current price of 13% TBV has the potential of offering outsized returns. But requires the following to hold true:
-improving or constant PPI;
-improving trend in 90 dpd formation, default rates, and loss given default; and,
-that Greece reaches a satisfactory agreement with its European partners with respect to its bail-out agreement, and implements structural reforms.

The risk of permanent capital loss is not an improbable event – as discussed above, a 15bps increase in default rates and loss given default could potentially wipe out TBV.
One way to play this situation is to wait and watch the trend in PPI, 90 dpd formation, and default rates over the next one or two QTRs. Waiting may erode some of the potential outsized returns, but should give more comfort on risk.