Last year the EU set aside €3 million to conduct a stress test on Turkish Cypriot banks. There are currently three stories doing the rounds as to why it has not happened yet. The Greek Cypriots seem to be blaming the Turkish Cypriot leader Mustafa Akinci, Akinci seems to be blaming the Turkish Cypriot coalition led by Huseyin Ozgrugun and Serdar Denktash (over which he has no control), while Denktash seems to be blaming the Greek Cypriots for dragging their feet over a proposal made in December 2016.
I suppose the truth is somewhere in the middle. While the record on leaks means that it is legitimate that the Turkish Cypriots worry about confidentiality, it is also clear that the Turkish Cypriot coalition has not exactly been supportive of Akinci, nor the negotiations to solve the Cyprus problem.
At the same time, there are certain forces on the Greek Cypriot side who are such hair-splitters on any possible hint of a recognition of northern Cyprus that they end up preventing international organisations from supporting good preparations to put the country back together again.
This absence of information is then lapped up by hardliners, who declare that a solution will send us all to economic damnation.
Fortunately, there is enough information on the Turkish Cypriot banks to make some educated guesses about what to expect from a stress test.
Let’s start with the bad news. At least seven banks in northern Cyprus have assets of less than €100,000. These days, being part of the eurozone means massive investment: not only in capital but also in compliance, technology and know-how.
I doubt very much that the banks that are not branches or subsidiaries of foreign banks will be able to survive without merging or being bought out by a bigger bank.
Impact of reinstatement
The picture is more positive for the larger banks. One of the key questions that often comes up when considering Turkish Cypriot banks is the impact of reinstating Greek Cypriot property.
Some of this Greek Cypriot property has been mortgaged to Turkish Cypriots and others, and some banks have lent money using that property as collateral, albeit at a discount. If that property is returned, then the collateral is suddenly worthless.
Late last year I made some calculations about how this would affect capital in the top five banks, which I published in my monthly Sapienta Country Analysis report. Together, the top five banks account for more than half of all assets.
I made some fairly some strict assumptions. For example, we know that around three-quarters of all private property in northern Cyprus belongs to Greek Cypriots. I therefore assumed that three quarters of all loans (not just housing loans) were collateralised with Greek Cypriot property. This is probably a big over-estimate, but I thought it best to be careful.
The only bank where I know that this is not the case is Turk Bankasi (Turkish Bank) – the bank that just obtained trade financing from the European Bank for Reconstruction and Development (EBRD). Turk Bankasi does not accept Greek Cypriot property as collateral, so it was excluded.
I have been told that Turkish Cypriot banks typically apply a discount of 40% on Greek Cypriot collateral, given the legal impediment. Some apply a discount of only 30%, so I was strict and assumed that all of them applied only 30%.
Then I assumed that one-third of Greek Cypriot property would be exchanged with Turkish Cypriot property in the south, one-third compensated and one-third reinstated (broadly in line with the Annan Plan).
If this is the case, then it means that only one-third of the collateral on Greek Cypriot property would suddenly be valued at zero.
I did not finish there, however. I also assumed that the European Banking Authority (EBA) would demand a Core Tier 1 capital ratio of 14% for all banks, as well as a 50% coverage ratio for any non-performing loans (NPLs). Three of the five banks already have coverage ratios well in excess of 50%; the other two are over 40%.
The results
The result of this exercise was that four of the five top banks had small capital shortfalls but none of these shortfalls exceeded €100m. The total shortfall of all five banks combined did not exceed €230m. If you want to know which banks had which shortfalls, I’m afraid you will have to subscribe to my reports. But suffice it to say, the €230m for the top five banks is not much more than the €175m that the Cooperative Central Bank received during its second state bailout in late 2015. I imagine that the amounts are small enough to be covered fairly easily during a transition period.
There is one caveat to my results. Capital ratios are calculated by taking capital and dividing it by risk weighted assets. I took the risk weighted assets as reported by the banks themselves. In a eurozone environment, the EBA might apply different criteria to calculate risk weighted assets, especially as a lot of housing loans are denominated in sterling. Foreign currency loans are probably the biggest vulnerability for Turkish Cypriot banks right now, as about 45% of loans (based on 2015 data) are denominated in foreign currency.
On the other hand, I was deliberately quite strict with the other assumptions. It is also worth noting that no Turkish Cypriot bank has anywhere near the NPL ratio of the three biggest banks in the south (measured on a more than 90 days past due basis). Finally, the assets of the entire banking system amounted to less than €6bn at the end of 2015, compared with €73bn in the south. This means that, whatever the capital requirements, they are never going to reach the eye-watering requirements of the Greek Cypriot banks, where the bail-in alone amounted to more than €9bn.
If the above is not enough to convince you, others who have done their own calculations have also told me that they are not so worried about the banks.
What we should be focusing on is the fiscal side. After much prodding by Turkey, the Turkish Cypriots are finally beginning to practise a certain amount of fiscal discipline. But they have a long way to go and it will take years to eliminate the impact of all the bad practices of the past.