Probusinesbank

As the story of Probusinesbank failure back in 2015 made headlines again recently, I decided to complete my favorite exercise as a banking analyst – to look through financial reports of failed bank to see if there were any red flags indicating about potential problems.

While it is always easier to do it in hindsight, my experience suggests that it is still very useful practice. Usually, one would be unable to infer that a particular bank would certainly fail (not least because the eventual outcome is subject to regulator’s decision), but just avoiding investing in equity and debt of such institutions still may save a lot of money.

My interest in this practice was prompted by Warren Buffet’s review of Lehman Brothers 10-k report when the management of the investment bank asked for support back in 2008. The first page of that 10-k with Buffet’s marks is in the title picture.

So let’s go through Probusinesbank 2014 annual IFRS statement.

Red flag 0: Rating withdrawal by Fitch

?In October 2014 Fitch ratings placed Probusinesbank credit rating under watch with negative view. The move was driven by Fitch concerns about bank’s securities portfolio, which was stored in offshore depositories, rarely traded and used for REPO operations, and had relatively long duration. Two months later Fitch withdrew its rating completely. While this move was a red flag by itself, I have put it under zero number as the bank might have not been rated by Big-3 rating agency in the first place. Without that, outside analysts could not have known about offshore depositories as the fact was never disclosed, while two other points (rare trading and REPO and long duration) are hardly could be marked as standalone red flags. So going forward, let’s assume that the bank has never been rated by Fitch.

Red flag 1: Segment reporting (p. 56)

In its 2014 annual report Probusinesbank provided segment results. Particular interest presents the “Financial segment” related to “managing financial instruments portfolio” consisted of securities portfolio, cash and cash equivalents, loans to and funds from financial institutions, and funds in the central bank. That segment accounted for RUB 2.7 bn, or 25% of the total RUB 11 bn loan loss provision for that year, and about half of total corporate cost of risk. Provisions in this segment have been negligible for all years before 2014.

It is logical and clear from the report the bank didn’t make provisions on C&CE, funds in banks and mark-to-market securities. Thus, such relatively big provision could be explained only by loans to financial institutions, which, according to segment loan portfolio breakdown (p. 76), increased sharply in 2014 to RUB 6 bn gross amount from RUB 1.3 bn. RUB 2.7 bn is a big provision against such gross amount. What were those loans? Was created provision sufficient?

Red flag 2: Loan reclassification (p. 53)

In 2014 the bank reclassified about RUB 10 bn gross amount of corporate loans from individually to collectively assessed for impairment, to obscure “Other loans” category. That meant that the bank decided to assess credit risk on such loans by some standard model, rather than consider each borrower on individual basis, despite at origination it had chosen the latter approach. If one compares financial statements before and after reclassification it turns out that 30% of reclassified loans were in delinquency as of the end of 2013. 70% of impaired large corporate loans were reclassified to “Other loans”.

Subsequently during 2014 cost of risk in “Other loans” skyrocketed from 4.2% (old representation) to 16.2% (new representation). By the end of 2014, only about RUB 3 bn of corporate loans were classified as “Other”, most likely they were impaired, but sufficiently provisioned. However, overall movements of loans between categories looked like deferral of loss recognition and misleading presentation of performance of large corporate loans portfolio. What was the risk that the bank tried to defer loss recognition in other parts of the corporate loan book as of the end of 2014?

Red flag 3: Unquoted bonds on the balance sheet (p. 69)

For some reason the bank put RUB 5.8 bn bonds into loans to customers balance sheet item. Those bonds were stated to have BBB- and A- credit ratings but were not actively traded. From geographic assets breakdown we can conclude that those bonds are of some companies in OECD countries. What were those companies? Who rated them? Why they were rated, but not traded? What was the rationale of investing in such instruments? From Q1 2015 quarterly report, we could find out that at least one of those bonds were issued by Atlas Investment Solutions, obscure Luxemburg entity. Later it was mentioned by the Central Bank as one of the problem exposures of the bank.

Red flag 4: Funds placed in financial institutions without credit rating (p. 64)

As of the end of 2014, RUB 15 bn, or 65% of total placements with financial institutions were in entities without credit rating. Of total RUB 23 bn placements, 75% were non-interest bearing, and of those 72% were without credit rating. That’s how these funds were described in the annual report: “Not rated placements with banks and other financial institutions are not considered for credit risk by the risk management of the Group as such financial institutions have low credit risk with no past due history and well performing businesses”.

More interesting, is that according to geographic exposure breakdown (p. 121), only RUB 4.4 of 23 bn placements were concentrated in Russia, RUB 15.2 bn were in OECD and the rest in non-OECD countries. The latter increased sharply from virtually zero non-OECD exposure in 2013. However, these figures didn’t not add up with C&CE exposure breakdown (p. 109): according to the latter, RUB 18.4 bn were placed in Russian banks, the rest in OECD banks and no non-OECD exposure reported. Why did the figures not add up? Was this massive OECD exposure according to geographic assets concentration breakdown was actually to Russian banks branches in OECD areas? If so, why did they have no credit rating? What was the rationale of holding so many non-interest-bearing balances in obscure financial institutions? Later, Deposit Insurance Agency would be unable to bring back at least RUB 7 bn of interbank balances from Latvia and Switzerland.

Red flag 5: FX transactions

In 2014, the bank lost RUB 3.1 bn. Its total equity was RUB 15.2 bn, Tier 1 capital was RUB 14.4 bn. However, the loss would be 5.2 bn more without positive FX assets/liabilities revaluation. While it was common to Russian banks report FX gains in 2014 after sharp RUB devaluation, in the case of Probusinessbank the nature of gains is interesting. According to note 17 (p. 65), the bank had exposure of RUB 4.3 bn to “FX contracts”. The details were provided on page 66. Particularly interesting is the contract of RUB 25.4 bn notional value, which is a short USD position at the average rate of 66.5 RUB/USD. The year end USD rate was 56.3, thus, the fair value of this contract is RUB 3.9 bn, or 90% of total FX contracts fair value and 80% of FX gains for that year. Given that USD rose above 60 RUB/USD for only a short period in the end of 2014, we can conclude that the bank entered into this contract right before the end of 2014. This was truly speculative and high-risk deal – the bank shorted USD amidst panic RUB devaluation with a notional value of contract of almost twice of its capital. What if USD continued to rise?

But also interesting is geographic breakdown of fair value financial instruments (which included FX contracts): from very useful p.121 combined with detailed instrument disclosure on p. 65, we can conclude, that these 4.3 bn contracts were concentrated either in OECD or non-OECD countries. (by the way, non-OECD exposure in this category rose dramatically in 2014, to RUB 8.5 bn, despite according to detailed instruments breakdown, it consisted mostly of bonds of US and Russian issuers). From p. 113, we see that those FX contracts were valued using Level 2 fair value hierarchy, which meant that they were OTC contracts. So, who was this foreign counterparty in these contracts? Where was it located? What was its credit risk? Would it be able to settle the contract at expiration? Eventually, CBR would state that these contracts were spurious transactions to decrease reported loss.

?Red flag 6: Holder of subordinated debt

According to p. 90, the key holder of the bank’s subordinated debt was AMBIKA Investments Limited.

Two smaller holders – collection agency “Life” and Bank24.ru were clearly related parties of shareholders. From this, one can suspect that AMBIKA entity may also be somehow associated with shareholders, because why otherwise some obscure entity would be an anchor holder of almost the whole issue of the subordinated debt. Worse, it could be that this entity bought the whole issue using loans from the bank in the popular “ring scheme” of Russian banks in that time. As one might have guessed, in the end AMBIKA Investments turned out to be one of the pocket entities of key shareholders of the bank.

P.S.

Several months later after the publication of the 2014 IFRS Financial Report, the bank failed. Bond holders lost all their money, as did some depositors. The lesson here is that investing in FIGs must go beyond simple DCF or Multiples analysis. Without your own comprehensive DD you might load up much more risk than you assume, even with a public entity with prominent shareholders (in this case they were East Capital and BlueCrest Capital Management).