A while ago, I read an article of WSJ written by Katie Martin concerning the market tranquillity with respect to what happens in Cyprus; the probable bank deposits haircut and the non-attainment of any agreement among EMU member countries. Among the explanations of this calmness reported by apparently successful analysts I found one, that of Mr. Beat Siegenthaler, a UBS analyst, who more or less states that the risk of diffusion of a crunch in Cyprus is perceived to be low. I really do not know if that is indeed a general sense- although several other rather interesting interpretations are listed- but that is not the case at all and claiming the opposite is naive especially in the aftermath of Lehman Brothers collapse. The risk of contagion following a crash in the Cypriot financial system is high and most certainly non negligible. I shall elaborate.

After doing some trivial digging I found an article that matches my needs; Billio et al. (2012)[1], concerning the estimation of the extent of interconnections developed among financial institutions based on four periods from January 1994 to December 2008. As a matter of fact, they found that any given bank can "Granger-cause"- 0.05 significance level- the performance, i.e. the yielded returns, of 114, minimum, and 328, maximum, other financial institutions (hedge fund, banks, brokers and insurance companies). Following, these estimations let us use some mathematics.

Let the linear map (discrete) x(t)=φ*x(t-1) be the number of financial institutions that default at each time step t=1, 2,... . Coefficient φ can be regarded as the product of two variables: θ*λ; θ is the number of institutions that will be affected and is derived from the Billio et al. (2012) estimations; λ is the mean exposure to Cyprus' financial institutions, i.e. the arithmetic mean of ratios of liabilities issued by Cypriot financial entities to the total value of assets at the time of the crash. Even with the most conservative estimation of the one Granger-causing 114 others and with a ratio of exposure at 1% (φ=1.14>1) the mapping tends to infinity; at each forward time step t the number of organizations in default gets larger and larger at a pace depending on φ.

In addition, let as consider a non-linear case, as well. Let, x(t)=φ*[x(t-1)^ε] be a non-linear map (a difference equation, once more), where ε is the elasticity of further/new defaults to the sum of the defaults that have already occurred; the rest of the variables as previously. The non-linear case although more complex shows that the propagation of the adverse scenario depends only to the elasticity ε. In other words, if ε>1 defaults move towards infinity and if ε<1 any contagion remains bounded and eventually disappears.

Although I might was quite tiring in the last couple of paragraphs, the point that I am trying to prove is that only under specific circumstances we have a chance not to be led to a severe crunch. If you argue for the non-systemic importance of Cyprus financial sector, keep under consideration that the interconnections among Greek and Cypriot banks is rather dense and hence the systematic involvement of both countries may not add up to be insignificant. Therefore, playing down the situation in Cyprus and the potential outcomes could only serve to prevent a self-fulfilling prophecy from becoming true.

In case the deposits are taxed that largely in the end, and if measures for capital movements are successfully implemented and bank run in Cyprus is avoided, there are still a few more euro-countries with no remarkable budgetary soundness who host some billions in bank deposits. These depositors may begin to concern and later on may panic because there is nobody to guarantee that their deposits will not be levied a similar tax. Now, consider this: provided this extended concern a fall might also begin from any financial entity within the EMU, as well, disseminating as described previously. On the other hand, there is always the monetary authority that will definitely take actions, if not to prevent, to mitigate the distresses and isolate the diffusion of any mishap.

Please note that I am not advocating for the probability of a financial breakdown but only for the possible results of such an event provided its occurrence. In conclusion, if anything goes wrong, Cyprus will most probably be the "ground zero" and not solely a source of fluctuations.

[1] Billio, Monica; Getmansky, Mila; Lo, Adrew W.; Pelizzon, Loriana (2012): "Econometric measures of connectedness and systemic risk in the finance and insurance sectors". Journal of Financial Economics, vol. 104: pp. 535-559.

After doing some trivial digging I found an article that matches my needs; Billio et al. (2012)[1], concerning the estimation of the extent of interconnections developed among financial institutions based on four periods from January 1994 to December 2008. As a matter of fact, they found that any given bank can "Granger-cause"- 0.05 significance level- the performance, i.e. the yielded returns, of 114, minimum, and 328, maximum, other financial institutions (hedge fund, banks, brokers and insurance companies). Following, these estimations let us use some mathematics.

Let the linear map (discrete) x(t)=φ*x(t-1) be the number of financial institutions that default at each time step t=1, 2,... . Coefficient φ can be regarded as the product of two variables: θ*λ; θ is the number of institutions that will be affected and is derived from the Billio et al. (2012) estimations; λ is the mean exposure to Cyprus' financial institutions, i.e. the arithmetic mean of ratios of liabilities issued by Cypriot financial entities to the total value of assets at the time of the crash. Even with the most conservative estimation of the one Granger-causing 114 others and with a ratio of exposure at 1% (φ=1.14>1) the mapping tends to infinity; at each forward time step t the number of organizations in default gets larger and larger at a pace depending on φ.

In addition, let as consider a non-linear case, as well. Let, x(t)=φ*[x(t-1)^ε] be a non-linear map (a difference equation, once more), where ε is the elasticity of further/new defaults to the sum of the defaults that have already occurred; the rest of the variables as previously. The non-linear case although more complex shows that the propagation of the adverse scenario depends only to the elasticity ε. In other words, if ε>1 defaults move towards infinity and if ε<1 any contagion remains bounded and eventually disappears.

Although I might was quite tiring in the last couple of paragraphs, the point that I am trying to prove is that only under specific circumstances we have a chance not to be led to a severe crunch. If you argue for the non-systemic importance of Cyprus financial sector, keep under consideration that the interconnections among Greek and Cypriot banks is rather dense and hence the systematic involvement of both countries may not add up to be insignificant. Therefore, playing down the situation in Cyprus and the potential outcomes could only serve to prevent a self-fulfilling prophecy from becoming true.

In case the deposits are taxed that largely in the end, and if measures for capital movements are successfully implemented and bank run in Cyprus is avoided, there are still a few more euro-countries with no remarkable budgetary soundness who host some billions in bank deposits. These depositors may begin to concern and later on may panic because there is nobody to guarantee that their deposits will not be levied a similar tax. Now, consider this: provided this extended concern a fall might also begin from any financial entity within the EMU, as well, disseminating as described previously. On the other hand, there is always the monetary authority that will definitely take actions, if not to prevent, to mitigate the distresses and isolate the diffusion of any mishap.

Please note that I am not advocating for the probability of a financial breakdown but only for the possible results of such an event provided its occurrence. In conclusion, if anything goes wrong, Cyprus will most probably be the "ground zero" and not solely a source of fluctuations.

[1] Billio, Monica; Getmansky, Mila; Lo, Adrew W.; Pelizzon, Loriana (2012): "Econometric measures of connectedness and systemic risk in the finance and insurance sectors". Journal of Financial Economics, vol. 104: pp. 535-559.