I recommend a short position in BVA, which I believe has at least 40% downside (quite likely more) from its current valuation. 

I wrote up BVA approximately 12 months ago and despite its share price being down >35% since that time I think it remains significantly overvalued on both an absolute basis and relative to Spanish domestic, Spanish international banks and Western European Banks generally.  I recommend either an outright short position in BVA or a pair trade with Santander for the following reasons:

1. At its current valuation BVA is one of the most expensive banks in Western Europe (1.4x P/TBV, 31x 2011 P/E, 23x 2012 P/E - Bloomberg consensus earnings), a valuation that is neither justified or warranted
2. BVA's Return On Equity is currently well below its Cost of Equity.  This situation is likely to continue to deteriorate in the medium term
3. BVA's loan book is heavily exposed developers and construction and real estate on the east coast of Spain (Valencia and Murcia in particular), markets which are showing signs of significant stress
4. BVA's Pre Provision Profitability (i.e. before impairments which themselves have been significant) has deteriorated significantly.  This will likely continue into the medium term
5. BVA's funding has become increasingly short dated (ECB repos and interbank lending) as it has not been able to access the EMTN market since 2007 and has struggled to grow deposits rapidly enough to plug its funding gap
6. BVA has materially increased its exposure to Spanish sovereign debt (on the asset side of its balance sheet) and thus has increased its exposure to any problems the Spanish sovereign may encounter

I provide more detail on each of these points in the discussion below.  In conclusion, BVA is materially overvalued which limits the risk from a short position and there are a number of reasons, including a need to raise equity if its loan book deterioration continues that could cause the stock to correct sharply toward its fundamental value.

Background

The brief summary is that BVA is a Spanish retail bank with the very large majority of its operations in its home regions of Valencia and Murcia (Spanish East Coast).   BVA grew its loan book aggressively during the property boom both inside and outside its home regions and did this in part through aggressive underwriting standards.  The Spanish property bubble has now burst and BVA's home regions are being particularly hard hit.

More background information for those interested:

• An excellent report by Variant Perception on the state of Spain and the Spanish banking system (including some of the questionable practices being employed by Spanish banks to defer/hide losses - a point which is very applicable to BVA: http://lacomunidad.elpais.com/blogfiles/lentejas/Spain_The_Hole_in_Europes_Balance_Sheet.pdf
• A Fitch report on BVA which includes a good high level summary of the business and loan book: http://www.bancodevalencia.es/cs/Satellite?blobcol=urldata&blobheader=application/pdf&blobkey=id&blobtable=MungoBlobs&blobwhere=1224936803011&ssbinary=true)
• My original write-up from Dec'09 which includes background on Spain, BVA and the original investment thesis (which has not changed materially): http://www.valueinvestorsclub.com/value2/Idea/ViewIdea/20796

1.       BVA is One of the Most expensive Banks in Western Europe, a valuation that is completely unjustified

BVA is one of the most expensive banks in Western Europe, currently trading at 1.4x P/TBV, 31x 2011 P/E and 23x 2012 P/E (these are Bloomberg consensus earnings numbers).   To put this in context BVA is trading at a material premium to Spanish domestic peers as well as Santander and BBVA (Spanish international banks with only approximately 1/3 of their operations in Spain and the remainder in other parts of Europe as well as faster growing Latin American geographies).

This valuation is not only remarkable relative to other Western European Banks (in light of the material challenges the Spanish economy and banking system have ahead) but even more remarkable in light of BVA's specific challenges relative to other Spanish domestic banks (which I outline below).   In my opinion, with the potential exception of Banco Pastor, BVA is the most challenged of the listed Spanish domestic banks across the whole spectrum of applicable metrics (asset quality, funding, profitability, ROE, etc.).  In light of this it's current valuation seems completely unjustified. 

2.       BVA's Return On Equity is currently well below its Cost of Equity.  This situation is likely to continue to deteriorate in the medium term

BVA's LTM ROE is 6.5% (well below its cost of equity), which is down from 11% in Q1'09 (which is still below its cost of equity).  Looking forward and using analyst consensus earnings for 2011 and 2012 (which are in both cases more optimistic than my own forecasts) BVA will generate a ROE of 4% in 2011 and 6% in 2012.   In very simple terms, you can't justify a premium to book value if you are earning well below your cost of equity and are not growing. 

The major reasons for the reduction in return on equity are: (i) a significant reduction in net interest margin without a corresponding reduction in costs leading to a reduction in pre-provision profitability (note: BVA was one of the most efficient banks in the system so its ability to cut costs was/is limited); and (ii) large impairments resulting from the rapid deterioration of the loan book.  Both of these issues will likely continue into the medium term (discussed in more detail below). 

Additionally, the Spanish banking system is likely to contract in the short/medium term.  A shrinking economy, deflationary pressures, high unemployment (currently ~20%) and real wage decreases will likely result in some of the past excess credit creation reversing.  As an example, loans to real estate developers grew 850% in the 8 years to 2008 (see Variant Perception report referenced above) and increased from approximately 6% of total loan stock in 2000 to approximately 19% of total loan stock in 2007.  Importantly, despite the poor economy, appalling conditions for real estate and real estate developers in Spain (some estimates are that there is currently 6 years of housing supply in Spain, see here for example: http://noir.bloomberg.com/apps/news?pid=newsarchive&sid=asP7WozSmx_A) and the enormous increase in non-performing loans to real estate developers (non performing loans to developers have increased from less than 1% of total loans at the end of 2007 to over 11% of total loans currently) loans to developers as a percentage of total loan stock has not decreased from 2007 to current.  This is in part the result of forbearance and other questionable measures being implemented by banks to delay the recognition of loan losses but it acts as evidence of why the total loan stock in Spain will not grow in the foreseeable future and more likely will shrink (i.e. for the loan stock just to remain flat many loans made to developers at the peak of the market need to be replaced with other Corporate/SME/personal loans.  Given Spain is struggling to register meaningful economic growth and currently has 20% unemployment it is difficult to see this happening in the short/medium term).  This problem/dynamic is exacerbated on in coastal areas, including BVA's home regions of Valencia and Murcia, which experienced the worst of the property boom.  

3.       BVA's loan book is heavily exposed developers, construction and real estate on the east coast of Spain (Valencia and Murcia in particular), markets which are showing significant signs of stress.  BVA's asset quality has been deteriorating and this is likely to continue

Approximately 2/3 of BVA's loan book is property/construction related with a geographic concentration in Valencia and Murcia.  Data from www.idealista.com (see table below) shows that these two regions (consistent with Spanish coastal real estate generally) have experienced material declines from peak prices.  Additionally, the stock of unsold housing in these regions is well above the Spanish average and according to Caxia Catalunya is the highest in Spain (see chart 11 on p.21 of 45 here: http://www.caixacatalunya.com/caixacat/es/ccpublic/particulars/publica/pdf/immobmaiges.pdf), meaning additional downward pressure on house prices is likely in future. 

Additionally, BVA's loan book is highly concentrated.  Specifically, BVA's 20 largest exposures represent ~15% of its total loan book or 2.3x Tangible Book Value and:

• Half of the 20 largest exposures are related to regional construction companies and real estate developers;
• One exposure relates to a debt/equity swap for a large RE developer in Murcia, where BVA has ended up with an 18% equity stake in the company (I believe this is Polaris World, which suffered significant financial distress and the banks are now cutting sale prices materially to try and sell stock. See here: http://www.propertyinspain.net/news/2010/10/polaris-world-resorts-in-spain-discounted-property/)
• The 20 largest exposures also included financing to shareholders or board members (~€100m or 7% of TBV). 

The combination of BVA's significant loan book growth during the boom (27% loan book CAGR in the 4 years ending 2007), geographic concentration in problematic regions and large exposure concentration have already started to result in a significant deterioration in asset quality.  BVA's Non Performing Loans have increased from 0.6% of gross loans at 31-Dec-07 to 5.5% of gross loans at 30-Sep-10.  Adjusted for real estate taken on balance sheet (swapping loans for assets to avoid recognition of NPLs) the NPL ratio increases to 7.5% of gross loans and shows no real sign of slowing (e.g. the NPL ratio was 3% at 31-Dec-08 and 4% at 31-Dec-09).  Additionally, BVA's NPL coverage (provisions / NPLs) has decreased from 316% at 31-Dec-07 to 71% at 30-Sep-10, meaning that as its loan book deteriorates further it will be required to take larger provisions through its income statement. 

Decline from Peak (%) (source: Idealista.com)

L'Olivereta (Valencia)                          29.5%

Puente de Vallecas (Madrid)             29.4%

Sant Andreu (Barcelona)                    28.8%

Patraix (Valencia)                                 27.3%

Rascanya (Valencia)                             26.8%

Benicalap (Valencia)                            26.5%

Usera (Madrid)                                     26.5%

Villaverde (Madrid)                             26.3%

Jesús (Valencia)                                    26.2%

Campanar-Benifefri (Valencia)          25.9%

4.       BVA's Pre Provision Profitability (i.e. before impairments, which themselves have been significant) has deteriorated significantly.  This will likely continue into the medium term.  Additionally, as its loan book continues to deteriorate this will be further detrimental to profitability

BVA's Net Interest Margin (Net Interest as a percentage of assets) decreased from 1.80% in Q1'09 to 0.99% in Q3'10.  Because BVA's efficiency (cost/income ratio) was one of the best in Spain it was limited in how much further cost cutting was possible in its operations and as such the reduction in Net interest Margin resulted in a reduction of Pre Provision Profit from 1.96% of assets to 0.72% of assets over the same period.  The reason Net Interest Margin contracted so dramatically was because Spreads on customer loans reduced while the cost of deposits did not.  The reduction in asset spreads is in large part the result of increasing non-performing loans.  More importantly, the increasing cost of deposits is due to a deposit war that is currently going on in Spain as all banks simultaneously try to decrease their loan / deposit ratio (i.e. decrease reliance on wholesale funding), which is the result of strong encouragement from the bank of Spain.  The result of this has been very active competition for deposits (some banks are currently offering up to 4% for 12 month deposits) driving cost of deposits up.  There is no sign that this is abating.   In addition to the cost of deposits increasing the cost of wholesale funding has also increased significantly for those Spanish Banks that can access the capital markets (BVA has not been able to access the EMTN market since 2007).   BVA has approximately EUR1.8bn of term wholesale funding maturing by February 2012 (~72% of its total term wholesale funding outstanding), the average cost of which is Euribor+20bps.  It is unclear if BVA will be able to replace any of this with term funding.  However, if it is able to access the markets it will be at a cost that is significantly above E+20bps (at least E+500bps based on where its debt is currently trading).  Given that BVA is earning well less than E+500bps on the asset side of its balance sheet the asset/liability spread dynamic is hugely problematic for BVA.  At a minimum, replacing this funding will be further detrimental to Net Interest Margin.  Additionally, BVA needs to increase its total deposits from current levels (BVA's Loan / Deposit Ratio is >150%).  Given the cost of new deposits is higher than BVA's current deposit cost this will put further pressure on margins. 

Excluding one-time gains BVA has consumed a large portion of its pre-provision profits through loan loss impairments over the past 2 years.  In Q3'10, income statement impairments were greater than pre-provision profits (excluding one-time gains).  As pre-provision profitability declines for the reasons described above and with non performing loan coverage now down to minimum levels future deterioration of the loan book (i.e. as the non-performing loans become loan losses and new non performing loans are formed) will require income statement impairments.   In an optimistic scenario BVA will struggle to generate any profits in the foreseeable future (i.e. pre-provision profits will be consumed by impairments) in a less optimistic case BVA will deteriorate its capital base through net losses.  On this point it is interesting to note that BVA switched earlier from paying a cash dividend to paying a scrip dividend. 

5.       BVA's funding has become increasingly short dated (ECB repos and interbank lending) as it has not been able to access the EMTN market since 2007 and has struggled to grow deposits rapidly enough to plug its funding gap

As described above BVA has not been able to access the wholesale capital markets since 2007.  As such, the portion of its total funding represented by short term inter-bank lending and ECB repos has increased materially.  Additionally, BVA has increased its deposit base (although this has slowed/stopped recently as the deposit war described above has heated up - i.e. BVA cannot afford to pay too much for deposits because its Net Interest Margin is suffering so much already). 

The net result of these actions is that BVA's liquidity risk has increased because it is living off an increasingly short-dated liability base (i.e. the weighted average life of its liability base is getting shorter without the asset side of its balance sheet shortening correspondingly).  The best case is that BVA will term out its funding (or perhaps shrink the asset side of its balance sheet) over time, although this will likely come at a higher cost than the current short term liability structure it has in place.  Of course, the worst case scenario is a liquidity crisis.  I think an outright liquidity crisis is not high probability because the ECB has been very supportive of all banks in this regard, however, I think it is likely that the ECB / Bank of Spain will require this issue to be addressed (which will come at a cost to BVA#'s shareholders through reduced profitability) not only for the stability of the bank but also for the broader system.

6.       BVA has materially increased its exposure to Spanish sovereign debt (on the asset side of its balance sheet) and thus has increased its exposure to any problems the Spanish sovereign may encounter

Remarkably, instead of shrinking the asset side of its balance sheet since the crisis commenced BVA has instead grown its exposure to credit instruments (on the asset side of its balance sheet) materially.  Specifically, its securities portfolio (predominantly Spanish sovereign debt) has increased from 6% of total assets in Q1'08 to 14% of total assets in Q3'10. The rationale for this behaviour is a simple carry trade that BVA has been running to partially offset the material decline it has experienced in Net Interest Margin (i.e. the Spanish sovereign credit on the asset side of BVA's balance sheet is funded through repo transactions at the ECB or other inter-bank lending and BVA nets the spread 'risklessly' ...). 

Clearly, this has increased BVA's exposure to Spanish sovereign risk and given the size of its exposure (14% of total assets, or >2x book value) this could become a solvency issue.  At a minimum, I believe approximately 1/3 of these instruments are held as Available For Sale Assets and as such are required to be marked-to-market.  Given the significant widening of Spanish Sovereign credit of late, this will have a detrimental impact on BVA's equity and could potentially require additional equity to be raised.  Additionally, BVA could be forced to reduce the size of this portfolio going forward, which will have a corresponding impact on profitability.  

Conclusion

Even in a scenario where BVA is not required to raise additional equity it is difficult to see how its current valuation is justified.  Additionally, I believe that the continuing degradation of its asset quality in conjunction with the other issues I have described above could result in BVA requiring additional equity.

For those looking for a pair trade, I would recommend short BVA vs. long Santander.  Briefly, on Santander:

• It is currently trading at approximately 0.9x P/BV and 6.5x 2011 P/E (Bloomberg consensus)
• 2/3 of earnings come from outside Spain, including over 1/3 from fast growing Latin American geographies (note: the ROE and growth of the LatAm businesses justify a premium to book value, which implies the Spanish business is trading even cheaper than would first appear)
• Santander's Spanish business has performed significantly better than BVA's through the crisis (e.g. SAN's Spanish non performing loan ratio was 3.4% at 30-Sep-10). I expect this to continue going forward
• Santander currently has a Tier 1 Capital Ratio of 9.7%. This combined with Santander's profitability (i.e. capital generation) limit solvency risk as well (importantly) as the risk of further equity raising
• Santander has a better current funding profile than BVA and it has been able to (and continues to be able to) access the capital markets (unlike BVA)

• Equity raising if loan book deterioration continues
• Problems with the Spanish sovereign resulting in large MTM or actual losses for BVA (and potentially requiring more capital)
• Focus on domestic Spanish banks as part of renewed focus on European peripheral economies