Saturday 19 January 2019

Greek bank update: The importance of regulatory forbearance

In November of last year, I blogged about the Greek banks, and specifically Eurobank. Since that time, Greek bank prices have remained under pressure, falling about 20%, although Eurobank has bucked the trend somewhat, rising 10%. It is nevertheless down about 10% YTD after rallying in late 2018, underperforming what has been a strong rally in global bank share prices this year.

Having given the issues some further thought, I have realised that I may have initially framed the issues somewhat inaccurately in my initial post. While I stand by all the analysis - that the Greek economy is healing, and that the Greek banks are actually robustly profitable at present on an underlying basis, and as a result, are rapidly repairing their balance sheets - in hindsight I believe this analysis sort of missed the point: none of that is relevant in the absence of regulatory forbearance.

The issues can be roughly though of as such: Banks are required by regulators to hold a certain amount of capital against their assessed risk-weighted assets. The Greek banks have written down their balance sheets significantly over the past decade, but only to a point where they still have enough regulatory capital (like Buffett says, provisioning for bad debts is a self-assessed exam where the penalty for failure is death). However, the actual reality is that, while they have made a huge amount of progress in working through legacy bad loans, they have still under-provisioned for legacy exposures (and are also carrying significant quantities of deferred tax assets on their balance sheets as well). What this means is that their 'kitchen sinked' balance sheet - i.e. what they would look like in the absence of regulatory capital requirements, where they were free to take a hatched to the carrying value of their assets and mark them down to conservative levels allowing for 'fresh start' accounting - would feature significantly lower capital levels than they are current claiming. And that level of capital would almost certainly fall short of regulatory minimums.

The situation can be thought of approximately as such: A bank may currently be profitable and internally generating capital of say €1bn a year (through pre-provision profits and negative net bad debt formation, as well as the utilisation of deferred tax assets). However, the book value of their equity may be €7bn (and required capital, say €6bn), but their realistic mark-to-market capital after writting down legacy bad loans may be closer to zero (or even negative - particularly if DTAs are excluded from capital). This is approximately the situation at Piraeus at the moment.

What this means is that the company is both mark to market (M2M) capital deficit, but also mark to market profitable (a flipside often overlooked), and on current trends, the latter is slowly addressing the former. In the example above, one year hence, M2M capital would have risen to €1bn; two years hence to €2bn, etc. Given enough time (and assuming nothing goes wrong operationally or economically), real capital would eventually converge with accounting/regulatory capital, and you'd be left with a healthy bank that is both very profitable and well capitalised (instead of just profitable). Piraeus' diluted market cap is currently about €400m - less than 1x current underlying earnings - but could rise to €5bn+ over 5 years or so if it were to move from the latter bucket to the former.

However, this outcome requires not only that current operational/economic trends continue, but crucially, that the company be given the necessary time to internally rebuild capital by regulators. This introduces the central issue of regulatory forbearance to the eventual outcome. If European regulators turn around and say at any time, yeah, actually you don't have enough real capital, and we need you to raise more, the stock is a zero. This is a perpetual risk so long as the company remains de facto capital deficient.

It is approximately the situation someone earning $1m a year would find themselves in if they had a negative net worth of $5m. If you keep earning $1m a year and using it to pay down debts, you'll end up in a very satisfactory financial position more than five years out. However, if you (1) lose your job (i.e. Greece has another economic downturn/crisis in this analogy); or (2) the banks you owe demand that you repay the full amount immediately at any time in the next five years, you will go bankrupt.

An important implication of this is that the most important variable in the investment equation is the outlook for (and odds of) regulatory forbearance. My initial Eurobank analysis placed primacy on the fact that operationally, things were going great and rapidly improving, and yet the share prices had declined. But that missed an important point - none of that matters if regulatory forbearance is not granted, and the odds investors place on future regulatory forbearance arguably ought to be a much greater share price driver than operational trends. In that respect, significant share price declines even in the face of significant operational improvements could in fact be quite rational - particularly because those operational improvements are not necessarily immediately obvious to casual observers.

A core part of the problem here is that if you haven't got a M2M balance sheet, then you also don't have M2M profits. Continuing our initial example, if you generate €1bn of real capital during the year, increasing your real capital from €0bn to €1bn, but the carrying value of your capital is €7bn, what you're going to do is write off €1bn of fake capital against the new €1bn of real, internally generated capital. You'll end the year with zero reported earnings and a constant level of accounting equity of €7bn, even though the marked-to-market net worth of the business has improved €1bn.

The problem with this is that it obscures from investors, regulators, and journalists real operating improvements in profitability and capital generation, as those real profits need to be written off against fake capital. This contributes to the ongoing sense that the Greek banks are still in financial difficulty and struggling under a mountain of bad debt, when in actual fact they are doing just fine. If they were reporting profits of €1bn, things would look very different, but they are not. And the optics matter here, as they contribute to the regulatory actions subsequently taken. If there is a perception that the Greek banks are still struggling, it increases the likelihood they are forced to raise capital.

In this respect, there were some negative developments in 2018 and into early 2019. Media reports suggest the ECB has been sending memos to European banks demanding that they write off their bad loans more quickly. If they force banks into writing down loans more rapidly, capital shortfalls will be exposed, and the probability of forced recapitalisations/nationalisations accordingly increases.

I understand why they are doing this. The ECB doesn't care about shareholders or dilution, and nor should it. They care about systemically important European banks having enough capital. But what they perhaps overlook is that their actions are often entirely counterproductive, because what they do is trigger sharp financial market selloffs and erode confidence, thereby impairing underlying economic performance; the ability of banks to generate internal capital; and their ability to raise capital, as share prices get hammered. They are doing more harm than good - especially when organic trends are already favourable, so no action is needed.*

I did think about these issues. One of the reasons I was attracted to (some of) the Greek banks is (1) they have already done several rounds of recapitalisations, and are now in the 6-7th innings of addressing legacy bad debts in my assessment; and (2) they have already agreed with Brussels on NPL/NPE reduction targets over 2018-21, which in my opinion are achievable (and which they are currently on track to meet), provided Greece's economy continues to recover at about 2% a year. My view was that if they met those targets, forbearance on the balance would be granted. However, I may have been wrong to assume that the EU would honour those agreements.

I am in the process of reassessing the probabilities at present. However, one positive development this week was that Italian banks indicated that the EU was actually intending to be more flexible and pragmatic than had been reported in the media (UBI Banca for e.g. said in a press release "... please be reminded - as recently underlined by sources close to the supervisory authority ... - that recommendations are formulated with account taken of the specific situation of each bank concerned", and went on to say that no impact was expected regarding its 2019/20 business plan). However, the situation is fluid, and outcomes are ultimately beholden to the whims of EU bureaucrats.

On the bright side, all of the above vindicates Eurobank's decision to preemptively boost its capital via its merger with Grivalia, and accelerate the workout of its legacy NPEs - albeit at the cost of meaningful dilution to shareholders. No doubt rising back-channel pressure from EU policymakers beared significantly on their decision to do so. This also likely explains why Greece tabled the policy option late last year of setting up an SPV that would transfer up to half of the banking system's bad loans off bank balance sheets, in exchange for releasing the their DTAs - a significant long term source of value now that the Greek banks are legitimately profitable on an underlying basis.** Following this merger, there is a good chance Eurobank will be able to make it through from here without needing to raise further capital, even if the EU is relatively punitive in its policy approach (I estimated in the article Eurobank was about €3bn underprovisioned - the transaction will reduce that by about €1bn, and core profits between 3Q18 and the end of 2019 will likely be another €1-1.5bn).

On Piraeus, they will almost certainly need to raise more capital if EU bad debt resolution pressures increase. That has been reflected in the stock's significant underperformance relative to Eurobank (down 80% over the past 12 months, despite very robust operational progress on par with Eurobank), with the price now trading below 0.5x normalised earnings (closer to 1x fully diluted, if the CoCos convert). I have a small position in Piraeus (30bp) which I'm happy to own at current levels, but only because it is priced like an option, because that's essentially what it is - a long dated call option on the (increasingly unlikely) scenario where they are not forced to raise more capital. I retain a larger position of about 2% in Eurobank, although trimmed some of the stock I bought in the mid 40s in the low-mid 50s in recent weeks as my assessment of the regulatory risks have increased while the price had risen. I am currently still reassessing what position size is appropriate. I do not own National Bank of Greece or Alpha Bank.

These stocks are not for the faint of heart, but I still do believe the risk-reward to be attractive for risk-tolerant investors. If it is indeed the case that the EU is discussing a 7 year NPL provisioning window for the Italian banks, then there is a good chance they do adhere to prior agreements with Greece's banks on the pace of NPE resolution, and seven years should be enough for Greek banks to become fully provisioned (assuming, of course, Greece's economy does in fact continue to recover).


LT3000


*This is another example of some of the problems with the EU, that have created political reactions in both the UK and Italy of late. EU policymaking bodies are democratically unaccountable, and so therefore often care little about the interests of the people in various EU member states - particularly in relatively small countries like Greece.

The Greek (and Italian) populations have already suffered terribly over the past decade as their downturns were severely exacerbated by EU bureaucrats insisting on punishing austerity, to the point where unemployment reached 27% in Greece, and the economy shrank by 30%. And the blows seem to just keep on coming. The EU seems either oblivious to or disinterested in the very real suffering some of their policy actions are inflicting on real people, and there is little countries like Greece or Italy can do about it, as they have relinquished their sovereignty. If it had its own currency and central bank, Greece would not have to be beholden to the whims of EU policymakers, and would not have to suffer the ignominy of being forced unnecessarily into a debilitating depression. It is no wonder Euroskepticism is rising in Europe - it is not without justification.

A preferable option to a breakdown in the Eurozone, however (which also has many positive features), would be an EU policy body that was more sensitive to the needs of its constituent members, more flexible, and more democratically accountable. Italy has said it is attempting to reform the EU from the inside, rather than leave the EU (Greece also attempted this, but was stonewalled; Italy, however, is a larger economy and so carries greater influence), which would be the preferred option. One can only hope that Brexit acts as a wake up call for the EU governing bodies, and reminds them that their policy agenda can only stray so far, for so long, from democratic preferences before a political revolt will occur. 

**The problem with this policy option, however, is that it is hard to see how it would not have the effect of adversely impacting Greek bank capital. The initial draft was for banks to transfer (underprovisioned) bad debts at book value, in exchange for relinquishing some of their DTAs. However, the value of these DTAs is carried on bank balance sheets, and forms a part of capital, so these transactions overall would be capital dilutive.