This paper Focus on the European sovereign debt crisis. Sovereign - - PowerPoint PPT Presentation

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

SYSTEMIC RISK, SOVEREIGN YIELDS AND BANK EXPOSURES IN THE EURO CRISIS

Niccolò Battistini

RUTGERS UNIVERSITY

Marco Pagano

UNIVERSITY OF NAPLES FEDERICO II, CSEF AND EIEF

Saverio Simonelli

UNIVERSITY OF NAPLES FEDERICO II AND CSEF

April 20, 2014

1 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

This paper

◮ Focus on the European sovereign debt crisis. ◮ Sovereign default risk or euro breakup risk? ◮ How do bank portfolios react to these two risks? ◮ Policy implications.

2 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Stylized Fact # 1

0 ¡ 2 ¡ 4 ¡ 6 ¡ 8 ¡ 10 ¡ 12 ¡ 14 ¡ 16 ¡ 18 ¡ Jan-­‑00 ¡ Jul-­‑00 ¡ Jan-­‑01 ¡ Jul-­‑01 ¡ Jan-­‑02 ¡ Jul-­‑02 ¡ Jan-­‑03 ¡ Jul-­‑03 ¡ Jan-­‑04 ¡ Jul-­‑04 ¡ Jan-­‑05 ¡ Jul-­‑05 ¡ Jan-­‑06 ¡ Jul-­‑06 ¡ Jan-­‑07 ¡ Jul-­‑07 ¡ Jan-­‑08 ¡ Jul-­‑08 ¡ Jan-­‑09 ¡ Jul-­‑09 ¡ Jan-­‑10 ¡ Jul-­‑10 ¡ Jan-­‑11 ¡ Jul-­‑11 ¡ Jan-­‑12 ¡ Jul-­‑12 ¡ Jan-­‑13 ¡ Jul-­‑13 ¡ Percent ¡

5-­‑year ¡government ¡bond ¡yields, ¡except ¡Greece ¡

Austria ¡ Belgium ¡ Germany ¡ Spain ¡ France ¡ Ireland ¡ Italy ¡ Netherlands ¡ Portugal ¡

3 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Stylized Fact # 1 (cont.)

0 ¡ 200 ¡ 400 ¡ 600 ¡ 800 ¡ 1000 ¡ 1200 ¡ 1400 ¡ 3-­‑03 ¡ 8-­‑03 ¡ 1-­‑04 ¡ 6-­‑04 ¡ 11-­‑04 ¡ 4-­‑05 ¡ 9-­‑05 ¡ 2-­‑06 ¡ 7-­‑06 ¡ 12-­‑06 ¡ 5-­‑07 ¡ 10-­‑07 ¡ 3-­‑08 ¡ 8-­‑08 ¡ 1-­‑09 ¡ 6-­‑09 ¡ 11-­‑09 ¡ 4-­‑10 ¡ 9-­‑10 ¡ 2-­‑11 ¡ 7-­‑11 ¡ 12-­‑11 ¡ 5-­‑12 ¡ 10-­‑12 ¡ 3-­‑13 ¡ 8-­‑13 ¡ basis ¡points ¡

5-­‑year ¡government ¡CDS ¡premia, ¡except ¡Greece ¡

Austria ¡ Belgium ¡ Germany ¡ Spain ¡ France ¡ Ireland ¡ Italy ¡ Netherlands ¡ Portugal ¡

4 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Risk of euro breakup?

(a) 20 November 2010 (b) 4 December 2010

5 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Risk of euro breakup? (cont.)

(c) 16 July 2011 (d) 17 September 2011

6 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Risk of euro breakup? (cont.)

◮ Nouriel Roubini:

“Greece will leave the eurozone in the next 12 months, and Portugal after. ..There is a 50% chance that the eurozone will break up in the next 3 to 5 years” (January 2012).

◮ Mario Draghi:

“the premia that are being charged on sovereign states borrowings ... have to do more and more with convertibility, with the risk of convertibility” (26 July 2012).

7 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Stilized Fact # 2

0 ¡ 2 ¡ 4 ¡ 6 ¡ 8 ¡ 10 ¡ 12 ¡ Jan-­‑00 ¡ May-­‑00 ¡ Sep-­‑00 ¡ Jan-­‑01 ¡ May-­‑01 ¡ Sep-­‑01 ¡ Jan-­‑02 ¡ May-­‑02 ¡ Sep-­‑02 ¡ Jan-­‑03 ¡ May-­‑03 ¡ Sep-­‑03 ¡ Jan-­‑04 ¡ May-­‑04 ¡ Sep-­‑04 ¡ Jan-­‑05 ¡ May-­‑05 ¡ Sep-­‑05 ¡ Jan-­‑06 ¡ May-­‑06 ¡ Sep-­‑06 ¡ Jan-­‑07 ¡ May-­‑07 ¡ Sep-­‑07 ¡ Jan-­‑08 ¡ May-­‑08 ¡ Sep-­‑08 ¡ Jan-­‑09 ¡ May-­‑09 ¡ Sep-­‑09 ¡ Jan-­‑10 ¡ May-­‑10 ¡ Sep-­‑10 ¡ Jan-­‑11 ¡ May-­‑11 ¡ Sep-­‑11 ¡ Jan-­‑12 ¡ May-­‑12 ¡ Sep-­‑12 ¡ Jan-­‑13 ¡ May-­‑13 ¡ Sep-­‑13 ¡ percent ¡

Domes,c ¡sovereign ¡debt ¡holdings ¡of ¡periphery ¡vs. ¡core-­‑country ¡banks ¡as ¡propor,on ¡of ¡the ¡ total ¡assets ¡of ¡banks ¡ ¡

core ¡ periphery ¡

8 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Stilized Fact # 2 (cont.)

0 ¡ 100 ¡ 200 ¡ 300 ¡ 400 ¡ 500 ¡ 600 ¡ 700 ¡ 800 ¡ 900 ¡ Jan-­‑00 ¡ May-­‑00 ¡ Sep-­‑00 ¡ Jan-­‑01 ¡ May-­‑01 ¡ Sep-­‑01 ¡ Jan-­‑02 ¡ May-­‑02 ¡ Sep-­‑02 ¡ Jan-­‑03 ¡ May-­‑03 ¡ Sep-­‑03 ¡ Jan-­‑04 ¡ May-­‑04 ¡ Sep-­‑04 ¡ Jan-­‑05 ¡ May-­‑05 ¡ Sep-­‑05 ¡ Jan-­‑06 ¡ May-­‑06 ¡ Sep-­‑06 ¡ Jan-­‑07 ¡ May-­‑07 ¡ Sep-­‑07 ¡ Jan-­‑08 ¡ May-­‑08 ¡ Sep-­‑08 ¡ Jan-­‑09 ¡ May-­‑09 ¡ Sep-­‑09 ¡ Jan-­‑10 ¡ May-­‑10 ¡ Sep-­‑10 ¡ Jan-­‑11 ¡ May-­‑11 ¡ Sep-­‑11 ¡ Jan-­‑12 ¡ May-­‑12 ¡ Sep-­‑12 ¡ Jan-­‑13 ¡ May-­‑13 ¡ Sep-­‑13 ¡ EUR ¡bln ¡

Domes-c ¡sovereign ¡debt ¡holdings ¡of ¡periphery ¡vs. ¡core-­‑country ¡banks ¡

core ¡ periphery ¡

9 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Objective

◮ Final objective: understand how default risk and redenomination

risk affect banks’ sovereign exposures.

◮ Preliminary step: decompose sovereign yields in

country-specific component (default risk) and a common component (euro breakup risk).

10 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Preliminary step: systemic vs. country risk

◮ Our method: Dynamic factor model. ◮ Previous literature: Identification of systemic risk via principal

component (Ang and Longstaff, 2011) or observables (Attinasi et al, 2009; De Santis, 2012; Di Cesare et al., 2013; Caceres et al, 2010; Sgherri and Zoli, 2009; Giordano et al, 2012).

◮ Result 1: The common factor correlates with euro breakup risk. ◮ Result 2: Increase in the euro-area yields (relative to the swap

rate) mainly reflects country specific risk.

◮ Result 3: CDS premia also reflect systemic risk, especially in

core countries.

11 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Final objective: home bias in banks’ sovereign portfolios

◮ Our method: Vector error correction model using (observed)

aggregate banks’ exposures and (country/common) yield differentials.

◮ Previous literature: (estimated) banks’ exposures suggest carry

trades especially by periphery banks (Acharya and Steffen, 2012).

◮ Result 1: higher country-specific risk =

⇒ periphery banks increase domestic exposure, core banks reduce it;

◮ Result 2: greater common risk =

⇒ all banks increase domestic exposures.

12 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

This paper

◮ Focus on the European sovereign debt crisis. ◮ Sovereign default risk or euro breakup risk? ◮ How do bank portfolios react to these two risks? ◮ Policy implications.

13 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Systemic vs. country risk

The model

◮ To illustrate the dynamic factor model, consider the case of 2

countries:           ∆y1t ∆p1t z1t ∆y2t ∆p2t z2t xt           =           α1G α1C α2G α2C α3G α3C α4G α4C α5G α5C α6G α6C α1             fGt f1t f2t   + ξt ≡ Λft + ξt,

◮ f G t is a global common factor, f1t and f2t are the country-specific

factors;

◮ Λ is the matrix of factor loadings; ◮ ξt is the vector of idiosyncratic errors

14 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Systemic vs. country risk (cont.)

The model

◮ The latent factors are assumed to have an autoregressive

structure: ft = A(L)ft−1 + ut, where A(L) is diagonal with two lags, so that the factors are

• rthogonal.

◮ The common and country-specific factors are estimated via a

two-step procedure (see Doz et al., 2011).

15 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Systemic vs. country risk (cont.)

• 5

5 10 15 20 25 30 35 40 5 10 15 20 25 30 35 40 4-10 5-10 6-10 7-10 8-10 9-10 10-10 11-10 12-10 1-11 2-11 3-11 4-11 5-11 6-11 7-11 8-11 9-11 10-11 11-11 12-11 1-12 2-12 3-12 4-12 5-12 6-12 7-12 8-12 9-12 10-12 11-12 12-12 1-13 2-13 3-13 4-13 5-13 6-13 7-13 8-13 9-13 Global Factor Google trend

Google Trends index measures how often search-terms related to the collapse of the euro were entered in the Google search engine.

16 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Systemic vs. country risk (cont.)

• 5

5 15 25 35 45 55 5 10 15 20 25 30 35 40 4-10 5-10 6-10 7-10 8-10 9-10 10-10 11-10 12-10 1-11 2-11 3-11 4-11 5-11 6-11 7-11 8-11 9-11 10-11 11-11 12-11 1-12 2-12 3-12 4-12 5-12 6-12 7-12 8-12 9-12 10-12 11-12 12-12 1-13 2-13 3-13 Global Factor Intrade prob(euro breakup)

The probability that any country that used the Euro as of March 12th, 2008 would announce its intention to drop the Euro before the end of 2013, based on Intrade data.

17 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Systemic vs. country risk (cont.)

Variance decomposition: yield differentials relative to the swap rate

Common Country Idiosyncratic Austria 0.05 0.86 0.09 Belgium 0.17 0.45 0.38 Germany 0.36 0.32 0.32 Spain 0.21 0.72 0.07 France 0.01 0.95 0.04 Greece 0.03 0.27 0.69 Ireland 0.00 0.34 0.66 Italy 0.29 0.74 — Netherlands 0.04 0.92 0.04 Portugal 0.01 0.83 0.16 Denmark 0.02 0.97 0.01

◮ Result 2: Increase in the euro-area yields mainly reflects country

specific risk (except for Germany).

18 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Systemic vs. country risk (cont.)

Variance decomposition: CDS premia

Common Country Idiosyncratic Austria 0.67 0.15 0.19 Belgium 0.56 0.17 0.27 Germany 0.60 0.07 0.33 Spain 0.43 0.47 0.10 France 0.65 0.04 0.31 Greece 0.05 0.61 0.34 Ireland 0.13 0.21 0.66 Italy 0.59 0.35 0.06 Netherlands 0.49 0.06 0.45 Portugal 0.11 0.73 0.16 Denmark 0.56 0.01 0.58

◮ Result 3: CDS premia also reflect systemic risk, especially in

core countries.

19 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Common and country components of the Italian yield differential and CDS premium

20 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Common and country components of the German yield differential and CDS premium

21 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

This paper

◮ Focus on the European sovereign debt crisis. ◮ Sovereign default risk or euro breakup risk? ◮ How do bank portfolios react to these two risks? ◮ Policy implications.

22 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Home bias in banks’ sovereign portfolios

Why this form of home bias may increase in response to higher domestic sovereign debt yields?

• 1. High-risk sovereign issuers may exert “moral suasion” on “their”

banks to buy domestic sovereign debt in time of stress.

• 2. Undercapitalized banks, most in periphery countries, may bet for

resurrection by engaging in “carry trades” (Acharya and Steffen, 2012).

• 3. Domestic banks are better hedged against the redenomination

risk of domestic sovereign debt than foreign banks.

23 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Home bias in banks’ sovereign portfolios: the model

◮ For each country in the sample, the estimated VECM model can

be represented as follows: ∆yt = α

• β′yt−1 + γ dt−1
• + Θ ∆yt−1 + Γ ∆Dt + ut,

◮ where yt =

spreadt sovexpt ′ in the baseline model and yt =

• commont

countryt sovexpt ′ in the factor-based model;

◮ α and β are the matrices of adjustment and cointegrating

parameters, respectively;

◮ dt and Dt are vectors of restricted and unrestricted deterministic

variables, respectively.

24 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Home bias in banks’ sovereign portfolios: results

Unconditional evidence

◮ In almost all countries banks increase their domestic exposures

in response to an increase in domestic yields. Conditional evidence (country/systemic shocks)

◮ In most periphery countries banks respond to increases in the

country risk factor by raising their domestic exposure, while in core countries they do not.

◮ In almost all countries banks increase their domestic exposures

in response to an increase in the common risk factor.

25 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Home bias in banks’ sovereign portfolios: results (cont.)

2 4 6 8 10 12 14 16 18 20 −0.1 −0.05 0.05 0.1 0.15 0.2 0.25 0.3

IT: spread → sovexp

2 4 6 8 10 12 14 16 18 20 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

DE: spread → sovexp

26 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Home bias in banks’ sovereign portfolios: results (cont.)

5 10 15 20 −0.1 −0.05 0.05 0.1 0.15 0.2 0.25 0.3 0.35

IT: common → sov exp

5 10 15 20 0.05 0.1 0.15 0.2 0.25

IT: country → sov exp

5 10 15 20 −0.08 −0.06 −0.04 −0.02 0.02 0.04 0.06 0.08

DE: common → sov exp

5 10 15 20 −0.35 −0.3 −0.25 −0.2 −0.15 −0.1 −0.05

DE: country → sov exp

27 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Home bias in banks’ sovereign portfolios: interpretation

◮ Country shock:

For periphery-country banks only, the evidence supports the “moral suasion” and/or the “carry-trade” hypothesis.

◮ Common shock:

Increased risk of euro collapse and currency redenomination has increased the home bias of banks, especially in core countries.

28 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

This paper

◮ Focus on the European sovereign debt crisis. ◮ Sovereign default risk or euro breakup risk? ◮ How do bank portfolios react to these two risks? ◮ Policy implications.

29 / 30

Motivations Preview of results Factor decomposition Sovereign portfolios Conclusions

Policy Implications

◮ Moral suasion could be tempered by shifting prudential

supervision from national to supranational level (SSM).

◮ Carry trades by banks could be discouraged by change in

prudential regulation: (i) positive risk weights on sovereign debt and/or (ii) limits on the concentration on sovereign debt.

◮ The home bias due to banks’ comparative advantage in dealing

with redenomination risk can only be addressed by reducing the probability of euro breakup (Draghi’s 2012 speech, OMT).

30 / 30