One of the ghosts haunting Europe these hours is that Brexit may provoke a wave of uncertainty in financial markets and a flight from the sovereign debt of the most “fragile states”. In recent days, Mario Draghi has tried to reassure markets declaring to be ready to take the necessary measures to stabilize markets, in addition to the massive current monthly purchases of sovereign debt. The question is whether such assurances will be effective, and to what extent a Brexit victory will raise the perception of sovereign risk for several European countries. What are the effects of the Brexit vote? What would have happened if Remain had won? To answer these questions, we calculate the implied probability of sovereign default implicit in the 5-year CDS contracts for insuring against sovereign default, traded daily on the financial markets between April 1 and June 21, 2016 for all traded EU countries (Source: DataStream). We study how these default probabilities have been influenced by the results of the daily polls preceding the Referendum result, and we recover the default probabilities conditional on a “Leave” or “Remain” outcome. The data are taken from the Financial Time (See the Technical Appendix at the end of post). These estimates enable us to calculate the change of default probabilities after a Brexit vote, relative to what would have happened had Remain prevailed (see Appendix for details).
The figure below shows on the horizontal axis the probability of default if Remain wins, while on the vertical axis the probability of default conditional on a Brexit vote, for the different countries. The countries where the outcome of the British referendum should not affect the probability of default are located on the right at 45 degrees. The countries that are above the 45 degree line are those where the risk of default in case of Brexit are larger than those in case of Remain. Conversely, countries under the 45 degree line are those where the default risk should fall after a Brexit win.
The figure highlights some interesting results.
1. The effects of Brexit on CDS spreads are limited for all the countries , and the distance between the spreads in the two scenarios do not exceed 14 basis points ( bp ).
2. In the United Kingdom risk perceptions are not affected by the outcome of the consultation.
3. The countries where the sovereign risk increase more in the case of Brexit, relative to Leave are: Italy ( + 7 bp ) , Spain ( + 5) , Portugal ( +14 ) .
4. The default probability is only marginally higher in Germany (2.4 bp)
(this post is joint work with Giulio Trigilia)
Appendix: The theoretical model assumes that the probability of default implicit in sovereign CDS, p, which are observed before the Referendum results are known, reflects the default expectations in case of Brexit victory, and the default expectations in case of Leave:
(1) p = E ( p | Brexit ) x Prob ( Brexit ) + E ( p | Remain) x [ 1 – Prob ( Brexit ) ]
We estimate the empirical model, country by country :
(2) p = a + b Prob(Brexit) + u
where Prob(Brexit) is obtained from polls as follows: Prob(Brexit)= % for Leave/( % for Leave + % for Remain), so we exclude the undecided from the calculations. Combining the two (1) – (2) expressions, we can express the estimated parameters a and b as follows:
(3) a = E ( p | Remain), b = E ( p | Brexit ) – E ( p | Remain) , and then a + b = E ( p | Brexit )
The chart above shows, for each country, on the x-axis the estimated value for the parameter a , and ,on the y-axis, the estimated value of a + b . If the parameter b is not significantly different at least 5 % , it’s value is set equal to zero.