The Greek financial crisis: growing imbalances and ...projects.chass.utoronto.ca/link/meeting/materials/19.pdf · The Greek crises starting in 2009 has reversed this gain. We consider

Post on 25-Jul-2020

0 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

Transcript

The Greek financial crisis: growing imbalances and sovereign spreads

Heather D. Gibson,  Stephan G. Halland

George S. Tavlas 

The entry

The entry of Greece into the Eurozone in 2001 produced a dividend in the form of a sharp drop in interest rates.

The Greek crises starting in 2009 has reversed this gain.

We consider to what extent these swings are due to fundamentals and how much is due to speculation.

We investigate the determination of the spread between Greek and German 10-year government bond yields

We adopt a single country analysis using both rateing data and fundamental analysis.

Plan of the paper

1 we outline the development of the crises

2 we investigate the extent which risk agencies rating can explain things

3 we investigate the way news about the Greek situation can explain things

4 we draw some conclusions

2. The Greek financial Crises: Origins

Growing Imbalances 2001-2008/9

Benefits of joining in 2001

•Lower inflation and interest rates•No exchange rate fluctuations lower risk premia•Encourages longer planning and investment

���

���

���

���

�����

� �� �� �� � �� �� �� �� �� � ��

������������������� � �

���������

� �������� �

� �������

������

�!�������"

������"����

#���� ������

������

$�%� ���#���%���

"������

Although growth was high (2001-2008 average 3.9%) two main problems emerged

Large fiscal Deficits

Persistently higher inflation than the Eurozone average

&�������'�(%��"��������"�������������)����������������������!����

*+��"�,-.�/

��0 ��0� ��0� ��0� �0� ��0 ��0��0�

�0���0�

��0 ��0 0� �0� �0� 0� 0� 0� �0���0�

10�1��0�

10�1�0�1�0�1�0�

1�0�1�0�1�0�1�0�

1��

1��

��

��

��

��

������

����

���

����

����

����

����

����

���

����

2����

+��"�,-.

&��������"���� (������)��������� (�������!����

&������'�,����������3����4����5�"�����������1����

*����������������������������%����������!������6����������5��)��"�7��������.�����/

1�0�

�0�

�0�

�0�

0�

�0�

�0�

�0�

�0�

����

���

����

���

����

���

����

���

����

��

���

���

���

���

���

���

����

���

����

���

����

��

����

���

����

���

����

���

����

���

����

���

����

��

����

���

����

���

����

���

���

���

���

���

����

��

����

���

, ����� ���������

&�������'� ����3""����!��3)�%����� ����5���)�*����8 �9���/

��������������������������������������

���

���

���

���

���

���

7.5�#���� �$7�#����

&�������'�,�����'�7�������4�������-�"����

�0�

��0

��0�

��0�

�0�

�0�

�0��0�

��0�

��

��

��

��

��

����

����

���

����

����

����

����

����

���

Section 2b The wake up call

The global financial crises began in august 2007

Collapse of Bear Stearns March 2008

Collapse of Lehman Brothers sept 2008

This had little direct effect on Greece, the Greek banking sector had not taken part in the sub-prime markets. Spreads rose a little but not to a significant extent.

But two shocks

October 2009, new Greek Government announced fiscal deficit would be 12.7% of GDP. (more than double previous predictions)

November 2009 Dubai world (owned by the Gulf Emirates) asked for a 6 month debt moratorium. This caused a sharp increase in investor risk aversion.

Financial markets began scrutinising Greece

These two shocks lead to a substantial and unprecedented rise in Greek spreads as shown earlier

Suddenly membership of the Eurozone did not seem to be an absolute guarantee.

Spreads continued to rise despiteMay 2010 support from IMF and EUCommitment to lower the fiscal deficitECB purchasing Greek bonds

Two applicable crises models

1st generationGovernment missbehaves

2nd generationMarkets speculate against a country

Are there ellements of both of these in the Greek case?

3 Spreads and the Risk Premium

The ECB Financial integration in Europe (April 2011) proposed a simple model of the risk premium. We extend this here for Greece.

We may think of the spread as

ttget

grt RR ερ ++=

Now we decompose the risk premium into two parts

Where may be justified on the basis of market information

And represent irrational market speculation

ttgett

grt RR ερρ +++= 12 )1(

1tρ

2tρ

Now in terms of rational information, in this section we focus on the information provided by the rating agencies and the gradual downgrading of the quality of Greek debt

-��:�:�����1

�:��:����; ���<=�����������,���>������"����4�����4�

-����:��:�����1

��:��:����&���%���?��������,���>������"����4@����4

-��:�:����1

�:��:���

������������.���=����?��������,���>�������"����

4@����4

-���:��:����1

���1������

������������.���=����?��������,���>�������"����

4����###@

Hence we may restate the relationship as

In the ECB study this model was estimated in first differences and by recursive OLS

We argue that both of these choices are wrong.

tgett

grt DDDDRR ερ ++++++= 4321)1( 2

Hence we estimate the following model using the Kalman Filter

Where the state equation ist

gett

grt DDDDRR εααααβ +++++= 4321 4321

),0(~1γωϖββ eNttt += −

7��""������ ��0�3���� A1 �������� .���0

α� �0������ ��0����� �0������ �0�α� �0����� �0����� �0������ �0��α �0���� �0����� �0����� �0����α� �0����� �0���� 0���� �0����γ 1�0���� �0����� 1��0���� �0����

&����� ���� ����; 3 A1 �������� .���0

B� 0������ �0���� �0���� �0����

$�����>���%��� 1����0��� 4>��>����"����������� �0������.��������� � �%?��A���������� �0�����-�""���������� � 6�����18 �����������0 �0���

(������'���)�������� ��>�.�������*������������C��������*/�����*�/)

The unexplained risk premia

�0�

�0�

�0�

�0�

�0�

�0�

0�

0�

�0�

�� �� �� � �� �� �� �� �� � ��

B� ��� ; 3

So this would suggest that a large part of the total spread can not be attributed to the downgrading of Greek debt but that instead there is a large degree of speculation going on.

But of course there was a lot of other things going on, so we now turn to a more fundamental explanation

5 A fundamental Approach

Here we use cointegration techniques to investigate the link between economic fundamentals and the equilibrium spread.

Data sample Jan 2000 to Sept 2010.

We focus on the main macro fundamentals which are widely regarded as determining the spread.

In addition we use a new news variable

The News variable

We look at revisions to the European Commissions spring and autumn forecasts for the Greek fiscal debt.

These revisions reflect news in the market

We then cumulate these revisions over time

Fiscal deficit revisions (% of GDP

-8

-6

-4

-2

0

2

4

00 01 02 03 04 05 06 07 08 09 10

Cumulated News

-100

-80

-60

-40

-20

0

20

00 01 02 03 04 05 06 07 08 09 10

Other variables considered

Actual debt to GDP ratioActual deficit to GDP ratioTrade and current account to GDP ratioRelative Greek to German pricesThe rate of growth of the economy (BoG indicator)Oil prices

We initially estimate a cointegrated VAR treating all variables as endogenous except oil prices.

We are interested in identifying the structural relationship so we must consider formal identification

We begin by estimating a standard VAR(3) (results in the paper)

∑−

=−− +++Δ=Δ

1

10 '

p

jt

ssjAA εδβα 1tjtt zzz

∑−

=

−−−

−− ++′+ΔΓ=Δ

1

1

10

10

10

p

jtj AAA εδβα 1tjtt zzz

Tests of cointegration�������������7������������� ��>�(����*(����/

6<���%���A�� (���� �0��D �0��"�73*�/ 3����!���� �������� 7��������B���� .���0EE

D ����E �0������ ��0����� ��0���� �0����4�������� �0��� ��0��� �0���� �0���4�������� �0���� �0����� ��0���� �0����4������� �0����� �0����� 0������ �0��

(�������������������������������������C�*�/�����%���0�����!��E�����������F��������"��%��%<���%���������%���0�����!��EE; ��G�����16���1; ��%�����*�/��1!�����

�������������7������������� ��>�(����*; �)�����3����!����/

6<���%���A�� ; �)13���� �0��D �0��"�73*�/ 3����!���� �������� 7��������B���� .���0EE

D ����E �0������ �0����� ��0���� �0����4�������� �0��� ��0���� ��0���� �0���4�������� �0���� �0����� ��0����� �0���� �

The identified cointegrating vector

B������� 7��""������ ��0�3���� �1 �������� .���0

7 ����0�� ��0��� �0����� �0����

3��������4���!��< 1�0����� �0����� 1�0����� �0����

�����!�������� ���0��� ����0� �0����� �0����

&��������?��*���������/ 1�0���� �0������ 1�0����� �0����

(5; 3 1�0���� �0����� 1�0������ �0����

H ����������*1�/ �0����� �0����� �0���� �0��

The impulse responses of the full cointegrated VAR are generally reasonable

-40

-20

0

20

40

5 10 15 20 25 30

���������"��������������������!��"��������?�

-40

-20

0

20

40

5 10 15 20 25 30

���������"������������������!��������

-40

-20

0

20

40

5 10 15 20 25 30

���������"�������������������������!��<

-40

-20

0

20

40

5 10 15 20 25 30

���������"�������������������

�����������7%����><�H ��� 0-0�5���!������

But our main interest is in the deviations between the equilibrium spread and the actual

1���

���

���

���

1���

���

���

���

���

�����

�� �� �� � �� �� �� �� �� � ��

��������*��"�1%�����)��/

4������*���%�1%�����)��/

&������*���%�1%�����)��/

�������������������� � �

���������

�!�������"

������"����

$�%� ���#���%���

"������

#���� ������

������

� �������� �

� �������

������

There is a period in the middle period where spreads were below the equilibrium, caused by the Euro reputation effect

A very significant overshoot at the end however. Suggesting again that fundamentals can not explain a large part of the spread and again suggesting that

Conclusion

Entry into the Euro provided Greece with a number of benefits

The Crises and news about the Greek fiscal position has removed this benefit

But our results suggest that markets have over reacted to the Greek situation

top related