Month: April 2014

Matt Levine on the latest Bank of America accounting maze – or, why our bank accounting/regulatory rules don’t make sense (Part I)

In this excellent Bloomberg article Matt Levine makes fun of BofA’s latest accounting error, making it necessary to revoke its planned return of shareholder funds. The bank had proudly announced its capital return programme only weeks before with – and now it gets funny – explicit consent of the regulator. The reason for the change was evidence of double counting of profits/losses stemming from decreases/increases in the Bank’s creditworthiness (yes, you read that correctly: under current accounting standards banks are allowed to book a profit whenever they get less creditworthy). Click the link to get the full article:

There are a few points I take away from this article:

1.)    Our regulatory/accounting rules for banks are that complex that (literally) dozen of people at the bank and the regulator got it wrong

2.)    The moral hazard problem with bank management is as prevalent as it has ever been

3.)    The accounting rules, by targeting false precision, conflict with purposes of regulation

In order to keep the post short, I will comment on the first point, with the rest to follow in future posts:

Complexity through over-regulation

Currently Banks are still obliged to follow Basel II rules, but required to perform and disclose additional Information to what extent they currently meet future Basel III requirements. Under Basel III Banks will either have to hold more capital for the same risks (e.g. trading book) and hold capital for risks that so far have not had to be accounted for (e.g. liquidity risk a.s.o.). All in all, under Basel III banks will be required to hold more capital and comply with more rules. So far so good, but here is the problem: both, the implementation date as well as the rules, are subject to constant change. There are many reasons for this, some good, some bad. For example, there is clearly constant lobbying of banks to water down the rules and postpone implementation. You can bet they are very effective as they are in a position to threaten the government (regulator) – always preoccupied with some election and hence fearful of recessions – with a credit crunch or a lack of buyers for the respective sovereign debt. Some of the rules have been hastily implemented and don’t necessarily make sense, punishing one type of bank disproportionately more than others (e.g. treatment of minority shareholders in regulatory bank capital). And last but not least, different regulatory objectives between the US and Europe which culminate in whether the leverage ratio should treat derivatives as they are treated under US-Gaap (Net) or IFRS (Gross). As everyone knows: it is difficult to agree on the right strategy if the objective is unclear!

Now, as a mental exercise, let’s put ourselves into the shoes of a regulator assessing a bank’s budget in the year 2014: apart from the stress test assumptions which are always subject to discussion, the regulator has to decide whether to base his assessment on current requirements (Basel II) or to include, in a forward-looking way, potential future Basel III requirements. The difference is non-trivial: If he wants to include them, which makes sense, he/she is faced with the problem of how to weight the information and with the question on which legal grounds he/she can enforce action on a private enterprise? Enforcing future regulatory requirements today is tricky from a legal perspective, likely requiring further by-laws and the like, effectively resulting in what Ludwig von Mises called a “Interventionsspirale”, i.e. once triggered, market interventions (regulations) make further interventions necessary. As one can imagine, there is plenty room for haggling and backdoor deals between the regulator and bank management and its lawyers. Consequently, there is inconsistent treatment between banks and across countries, only increasing the confusion. Matt Levine probably has good reason to suspect double standards being applied between Citigroup and BofA. There is simply no way to be consistent!!!

Complexity through Internal Models and conflict of interest

Basel III is an extension of Basel II, i.e. its goal is to have risk assessed, not by some objective rule, as under Basel I where you had prescribed risk weights, but according to internal bank models. I remember when back at university our professors, who had a knack for financial models, spoke admiringly of Basel II. Banks, they said, after all knew best what risks they have and how to measure them (their principal agent and moral hazard models they showed us in another, unrelated lecture). Well since the financial crisis, for obvious reasons, nobody would argue that way anymore. So today one hears a different argument: “well, we know models are not perfect, but it’s the best we have at the moment”. Really? As the following graph charts show, banks have used the leeway provided by internal models mainly to reduce risk-weights (i.e. lever up their balance sheets). The chart in the lower right corner is especially informative, as it compares risk weight under the old (Basel I) approach to those applied by British banks under Basel II.


The results speak for themselves: in every segment the Basel I approach (green) puts a higher weight on risk exposures than the internally modelled one. This result implies that for the average bank portfolio, an 8 percent risk-weight, i.e. more than 12 x leverage, is way, way too conservative. Clearly, managements’ desire to lever up the balance sheet have had an influence on the risk models chosen – the power of incentives at work.

The regulators answer to this state of affairs is typical for our times: instead of replacing a failed methodology as it is based on wrong principles (bank self-assessment) one just pops up further, supportive regulation thereby adding to complexity in disproportional way. Under Basel III there will be a myriad of-add-ons regulators have to apply to the bank’s risk weights on a discretionary basis, if they suspect that they are not conservative enough. Welcome to the world of arbitrary, costly and difficult to enforce regulation!

In my next post I will touch the issue of moral hazard in banking!


Recomended: Ralph Raico on the period between the two world wars

There is much talk this year about WWI and its causes. Contrary to what one might think, at the beginning of 1914 few people could conceive of a serious conflict. The current events in Ukraine have increased the interest even further, since almost daily parallels are drawn between then and now – some justified, some less so. Moreover, with every analysis each author reveals his implicit personal and usually overly simplistic (“the good and the bad guys”) reading of history.

WWI really was a watershed event in history in many ways. The end of the great european monarchies is one of those big game-changers that is often quoted, resulting in multi-decade long chaos in the young democracies on the continent, with the well-known consequences. Are there parallels to be drawn to the turmoil in still comparably young democracies like Ukraine and Russia?

For those interested in the intricacies of banking and finance,  WWI marks the beginning of the end of the classical gold standard: contrary to what many believe, the gold standard did not end during the great depression, when Americans were forbidden to hold gold or when Britain abandoned the gold-peg it had reintroduced in 1925. Nor did it end in 1971, which marks the end of Bretton-Woods system designed by Keynes. Instead, according to the great French economist Jack Rueff, it was quietly abandoned at the 1922 Genoa conference, when the “gold exchange standard” was officially introduced. This is the year when the US-Dollar’s role as world reserve currency was established. This is also the year when the “roaring twenties” took off. Coincidence?

 Just like for most people, my interest in this time of world history is rising by the day. The following three-hour lecture by Ralph Raico is a very, very good start on the topic. A word of caution/motivation: there will be a lot of surprising facts and you might find your established views on the subject challenged more often than not!

Ralph Raico is one of my favorite historians. A student in Ludwig von Mise’s seminar in NY, he is that rare bird, a historian who knows economics and which consequently puts him into a unique position among historians to not just enumerate facts or give simplistic explanations, but thoroughly analyze economic motivations behind events. Besides, I also find him funny and entertaining. Here is the link:

As Ludwig von Mises pointed out in his fantastic book “Theory and History”, historical analysis without a theoretical grasp of economics is useless at best and dangerous at worst, the equivalent of “data mining” in the econometric field. The book, in my view, is a must read for historians and economists as well as the educated layman. It can be purchased here:

I really hope you enjoy the lecture as much as I did.


Marty Fridson at the Grant’s Investment Conference on the next junk-bond implosion

Long-time credit market observer, Marty Fridson, gave an interesting speech at the bi-annual Grant’s investment conference in NY. In the speech he used a few landmark historic datapoints in order to arrive at a spectacular prediction for a cataclysmic default wave. He predicted that as much as 1.6 trillion of leveraged loans and junk bonds could default from 2016 to 2020!

Here is the link to the presentation

Fridson pointed out that junk bond issuance has been growing by 12 percent p.a., at the same time that nominal GDP has been rising by a mere 4 percent annually. The loose monetary policies of the past decades have produced ever rising levels of debt, coinciding with a fall in average quality: whereas in 1999 merely 13 percent of borrowers in the Moody’s universe were rated Caa to -C , the current figure is closer to 22,2 percent! In line with steady rising debt levels, default peaks have increased every cycle: the default rate in 2009 set a record 13.3 percent, up from 10.8 in 2001, again up from the 10.8 percent reached in 1990 during the height of the savings and loan crisis. Makes sense! What was unusual in the last cycle however was its brevity, as can be seen in the chart below.

Periods of above average default rates

The last downturn was highly unusual: Although the default rate hit a record 13.3 percent in 2009 the default rate dropped below its long term average (4.6%) already the next year (2010: 2.4%), whereas historically the default rates hovered above the long term average for 4-5 years after the peak. It is not difficult to guess the main culprit: the Fed, through low interest rate policy, saved a lot of overlevered issuers. Fridman also noted that historically default waves occurred every 4-7 years, which is why he expects the next one to start in 2016 the latest. As you might have guessed: the market doesn’t seem to care, the average junk spread trading more than 200 basis points below its long term average.


In my view Fridman’s speech is a bottom-up testimony to the validity of Austrian Business Cycle Theory. Low interest rates not only lead to malinvestment, as evidenced by the three times faster increase in debt compared to GDP and the deteriorating average credit quality, but also hamper the recovery by suspending the forces of creative destruction and keeping Zombie companies alive. This figure sums it up neatly: cumulative defaults between 1989-92 amounted to 31%, between 1999-2003 they raised to 37%, whereas  a paltry 17% defaulted cumulatively during the Great Recession. With that much unfinished business, the slow recovery should come as no surprise to anyone.

Greek Banks: What are David Einhorn and John Paulson thinking? (Part 2)

Assessing the earning power of Pireus bank

After having looked at the balance sheet it is time to see what this bank can earn on a normalized basis. For that purpose I have chosen to reproduce the income statement below:


The net interest income (NIM) has jumped by 62% yoy, suggesting impressive improvement in operating performance. Looking at the components we see that Interest income (i.e. what the bank earns from its customers and its securities portfolio) has increased by 23% – in line with the balance sheet expansion – whereas interest expense has stayed more or less flat, despite an increase in liabilities. How is this possible? The answer: interest rates on deposits have come down substantially in Greece. From this month’s bank presentation to equity investors we get the following chart:


As can be seen time deposit rates have fallen by a third and are currently at 3 percent, meaning the increase in volume was fully offset by a fall in cost of funds leading to an increase in the net interest margin (NIM). Whereas rates could go even lower I do think we are pretty close to the bottom in Greece already, further improvements should be marginal going forward. As a consequence, we should treat the boost in net income as a one-time event, further increases will require balance sheet growth. Applying the net interest income of 1.6 bn. to the average balance sheet size in 2013 (80 bn.) gives us a normalized net interest margin (NIM) of 2%. Doesn’t sound terrific for Greece, but makes intuitive sense: as we have seen before, a large chunk of the balance sheet is represented by “slow” assets, that is by customers who are not paying at all (default) or not paying in time (past due but not impaired).

If we take the 2 percent margin and multiply by the current balance sheet size we get an approximation for the steady state NIM of Pireus bank: EUR 1.8 bn. This assumes neither significant deterioration nor improvement in the average asset quality of the bank.

If management can turn underperforming loans into winners (big if), the interest margin would improve. It is difficult to say by how much but I will be generous and assume that a 2.5 percent margin is achievable in a positive scenario, leading to an optimistic NIM of EUR 2.25 bn.

The biggest item on the income statement is the “negative goodwill” of 3.5 bn. resulting from the acquisition of above mentioned banks. Clearly this is a one-time profit that will not repeat in the future and can be discarded for valuation purposes.

Operating expenses are up by 80% most of which are staff costs. Understandable: lot of redundant staff has been taken over due to the acquisitions. Taken from the above mentioned presentation to the equity investors: management estimates that synergies (read: lay-offs) amount to EUR 300 mln. per year. I will take that at face value and assume operating expenses will come down accordingly.

Provisions are a notoriously difficult to estimate as much of it depends on how real the recovery in Greece is. Without arguing back and forth and being generous I will assume substantially reduced provisions of EUR 1 bn. for the base case and EUR 0.5 mln. for the optimistic scenario.

A further aspect which needs clarifying is tax line in the income statement which was positive both for 2012 and 2013. According to note 16, the positive entry results from an increase in corporate income tax in Greece which increased the tax shield stemming from past losses and was applied retrospectively for 2012 as well. At the end of 2013 the bank had a corporate tax asset of EUR 2.7 bn. which basically means the first EUR 10 bn. in profit will be tax-free!

Finally, after adjusting the income figures we get the following income statement. For items I have not explicitly referred to above I have either extrapolated 2013s figure or ignored it if inconsequential.


As one can see: in the base case – which I personally think plausible – the bank is not able to turn out a profit. A strong recovery indeed would have to take place in Greece for a profit of EUR 890 mln., which translates into a return on tangible equity of 11 percent. Depending with what your cost of equity is or a Greek bank you will come to different conclusions. I, personally would say that, given the risks involved, a cost of equity should at least be 12 percent for a Greek bank, i.e. the bank barely returns its cost of equity in a positive scenario.


Given the significant headwind the bank is facing (liquidity, asset quality, potential haircut Greek debt a.s.o.) I think the optimistic scenario highly unlikely. The potential return if all goes well doesn’t strike me as attractive and I do not know how the market arrives at a valuation of 1.2 x book for this bank. However a word of caution: maybe I am missing something, some hidden assets, some feature of the recapitalisation that I have not accounted for and that I have overlooked.

Further, Basel III regulation which comes into effect in 2015 contains explicit regulation limiting the liquidity risk a bank can take. I could not find any hint with respect to that in the financial statement but just looking at their balance sheet I do not thing they meet the Basel III ratio (Net stable funding and liquidity coverage, but from what I can tell looking at its balance sheet I would assume there is some way to go

Greek Banks: What are David Einhorn and John Paulson thinking? (Part 1)

With great interest I have read that David Einhorn and John Paulson have taken substantial positions in Greek banks Pireus (BPIRF:US) and Alpha (ALBKF:US). In Einhorn’s case it apparently amounts to a substantial position. The link can be found below:

Both Banks have recently been recapitalized and Pireus, which will be the focus of this post, trades at 1.2 x book value based on the 2013 financials. In the first quarter the bank has completed a EUR 1.7 bn. equity offering which would bring the P/B ratio down to around 1. Regardless, the market, it seems, thinks that the worst is over in Greece, also evidenced by its recent return to the bond market. While I do think that certain progress has been made, simply because the government has run out of other people’s money (always a good thing) and as indicated by record tourist bookings, the market feels somewhat overextended.

When it comes to stock investing, I favour a value approach trying to focus on fundamentals rather than macro stories. This served me well in Greece in the past, as it led me to invest in two Greek stocks, engineering company Metka (OSQ: GR) state lottery operator Opap (Opap: GA), during the crisis, providing handsome returns. I still hold both stocks. So regardless of whether I personally think the Greek recovery story is largely overhyped or not, I will try to assess the attractiveness of Einhorn’s Investment in Pireus bank on a company specific “micro” level. The most recent financial report served as a basis for my quick analysis.

I have reproduced the consolidated balance sheet on page 4 of the report in a somewhat stylized format below:


As can be seen Pireus bank expanded its balance sheet significantly in 2013. However, the growth has not been organic but is a result of several mergers/acquisitions. Pireus bank acquired the Greek operations of a Cypriot bank, Millenium Bank as well as parts of Agricultural bank of Greece in 2013. In view of this, it is quite tedious to figure out how the underlying business has developed. I will skip that for now and leave it for a further post.

The increase in total equity is largely the result of the already mentioned recapitalisation which accounted for EUR 8 bn. of the increase. Further, there was a profit of EUR 2.5 bn. in 2013. The high profitability comes as a result of the mergers and the “negative goodwill associated with them (details in my second post on the income statement). After deducting intangibles I arrive at a tangible book value of EUR 8.1 bn. So at first sight, the Bank seems under leveraged compared to European peers (TA/TE around 11.5 x), leaving room for strong growth. As already mentioned above, the equity has increased due to the equity offering at the end of Q1.

A quick look at the Balance Sheet

Book values are tricky, especially for banks and all the more for banks in Greece. In order to assess the quality of book value it is necessary to get a sense for the asset quality of the bank. I naturally start with the biggest item on the balance sheet: loans and advances. The bank provides a useful breakdown in its financial risk management section (page 33 of the financial statements, sorry for the bad resolution):


Out of Gross loans of EUR 76 bn., EUR 23 bn. is deemed impaired leading to a NPL Ratio of 30 percent – reflecting the dire situation in Greece. Against these, provisions of EUR 14.5 bn. have been booked, resulting in the net loan amount of EUR 62 bn. The coverage ratios (Provision/NPL): 65% for the retail book and around 60% for the corporate book, respectively. At first glance this looks conservative, but keep in mind this is Greece: asset values have collapsed since the bulk of these loans have been granted, so it could well be that the provisioning level, albeit reasonable at first glance, is not enough. More detail is required until a final conclusion can be drawn on this one.

What really stands out however is the EUR 20 bn. of loans that are classified as “past due but not impaired”. This item amounts to 30% of net loans!!! These loans are do not need to be impaired according to IFRS only if the restructuring is done in a “NPV Neutral” way, i.e. if the interest rate foregone in the short-term is added in the future, or if the client has substantial collateral. I have examined enough loan books to know that in most cases assumptions in “NPV neutrality” make Facebook investors look conservative. Nowhere, however, have I seen such a large part of the portfolio classified this way. Further, from then notes we learn that there are EUR 5 bn. in loans forborne (i.e. restructured) that are neither impaired nor are they past due, i.e. they were restructured in time before they were past due. Now it gets interesting: depending on how you want to treat this information, the equity has to be adjusted substantially – clearly from a PV perspective these loans are not worth par. I will be generous: assuming that 25% of the past due loans are really “hidden” NPL and applying a coverage ratio of 50%, I arrive at an adjustment to book value of – EUR 2.5 bn.! Doesn’t look that well capitalised anymore!

Further we can see that public loans with Greek sovereign risk amount to EUR 2.1 bn. To this has to be added the book value of the Greek government bonds in Investment securities (EUR 1.2 bn.) and the EUR 300 Mio. Greek T-bills in the trading book and you get a Greek sovereign exposure of EUR 3.5 bn. If you happen to think – as I do – that Greece needs another haircut, you would want to adjust these values downward as well.

The rest of the investment securities book (EUR 14 bn.) consists of EFSF bonds, guaranteed by EU taxpayers. Although this worries me as an EU tax-cow, at least I do not have to adjust them for now.

Of course Einhorn and Paulson could reason that these loans will benefit disproportionately from the recovery and become money good again and think the book value alright. In this case the bank would be well capitalised and could start lending to Greek customers, presumably at attractive spreads.

Pireus Bank still heavily dependent on interbank (ECB) financing

But things are not that easy, growth needs to be not only supported by capital but has to be funded as well, so let’s turn to the liquidity situation of the bank: what jumps into the eye is the startling amount of interbank financing (due to credit institutions), which, although down from almost 50% of the balance sheet at the end of 2012 to 30% at the end of 2013, is still sizeable for a commercial bank. Of course most of these liabilities are liabilities to the ECB. By the way, if you’ve wondered why there have been no spectacular bank defaults in Greece despite a sovereign debt restructuring, deposit outflows and record unemployment – here is the answer! As a member of the Euro system Greek banks have been able to access Euro liquidity by the ECB, even on an unsecured basis (“ELA” loans). With that “competitive advantage” they could repay scared depositors and bondholders although they were lacking eligible collateral in the amounts required. If the ECB hadn’t loosened its collateral rules, this – and probably most Greek banks – would be long gone.

It is clear from the above stylized balance sheet that the bulk of the cash from the EUR 8 bn. capital increase has gone into repaying the ECB. This is understandable: repaying the ECB must be top priority for bank management and Greek regulators as well as the ECB. First, despite the fact that central bank intervention is considered normal nowadays, there is still negative stigma attached to your institution if you need ECB financing. Second, the sooner the ECB refinancing is gone the sooner the Greek government can get rid of the highly unpopular “Troika” and present themselves as reformers. It is safe to assume that any significant profit the bank is able to earn in the near future as well as liquidity coming in from bond offerings will flow into repaying the ECB rather than into new loans or dividends! Given these difficulties I do not see how Pireus Bank can grow without tapping new sources of funds. Certainly, the liquidty situation has improved after the most recent equity issuance. However, any tapping of external liquidity, be it equity or bonds will be costlier than the sweet, non-market based rates of the ECB facilites they currently have.  External funds would pressure the profitability of the bank.

It could be of course that the bank will be highly profitable after the restructuring and grow its way out of the problems. In the next post I will look at the income statement and the earning power of the bank.