ma thesis pres

37
Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies Student: Hami SAKA Advisor: Prof. Mehmet ORHAN

Upload: hami-saka

Post on 12-Feb-2017

97 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: MA Thesis Pres

Analysis of the Rating Discrepancies Assigned byCredit Rating Agencies

Student: Hami SAKA

Advisor: Prof. Mehmet ORHAN

Page 2: MA Thesis Pres

Outline Aim and Main Contributions Concept of Credit Rating

Criticisms to the CRAs

Methodology and Data Used

Results

01.05.2023 2Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies

Page 3: MA Thesis Pres

Main Contributions: Analyse the rating inconsistency between Credit

Rating Agencies (CRAs)

Statistical analysis to reveal the discrepancy of CRAs. CRAs sovereign ratings compared to each other

(Pairwise Comparisons)

First study in the literature to present such comparisons.

01.05.2023 3Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies

Page 4: MA Thesis Pres

The Concept of Credit Rating

Credit rating can be defined as measuring the solvency of an economic entity to pay its present or potential debts.

Credit Rating Agencies (CRAs) are private institutions that undertake the activity of credit rating, and inform both the lenders and borrowers of their financial positions through the credit ratings given.

01.05.2023 4Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies

Page 5: MA Thesis Pres

The Functions of Credit Rating Agencies

The reason why Credit Rating Agencies emerged is to protect market players against risks and ambiguities, and to make economic units behave rationally in the process of making decision because of the fact that interpenetrating financial markets are becoming more and more complicated.

CRAs;

Providing issuers to enter the Capital Market Providing regulators to regulate Providing information for investors

01.05.2023 5Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies

Page 6: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 6

Criticisms to the CRAs

Inaccuracy and Inconsistency of Credit Ratings

Conflict of Interest

Oligopolistic Market and Lack of Competition

CRAs and the Global Financial Crisis

The Negative Impact of Rating Downgrades

Extreme Addiction to Credit Ratings

Lack of Responsibility

Page 7: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 7

Credit ratings by Agency

Almost the whole rating market where there are approximately 150 CRAs is dominated by the Big Three.

Page 8: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 8

Inaccuracy and Inconsistency of Credit Ratings

Credit Rating Agencies have become more questionable due to last Global Financial Crises and the new one which is caused by CRAs might be occurred.

Inaccuracy of ratings and not overlapping with reality are in the center of criticisms of CRAs.

Before many crisis experienced, the fact that the countries in crisis are rated with high ratings by CRAs and that economic crisis rises in the periods in which they get these high ratings bring CRAs under suspicion with a different respect, in terms of sovereign credit ratings.

Page 9: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 9

Sovereign Rating Failures / Asia Crisis

  S&P Fitch Moody's

  Start EndNotche

s Start End Notches Start EndNotche

sIndonesia BBB CCC+ 8 BBB- B- 6 Baa3 B3 6South Korea (a) AA- B+ 10 AA- B- 12      South Korea (b) B+ BB+ -3 B- BB+ -5      Malaysia A+ BBB- 5 BBB- BB 2 A1 Baa3 5Pakistan B+ CC 6       B2 Caa1 2Russia BB- CCC- 6 BB+ CCC 7 Ba2 B3 4Thailand A BBB- 4       A2 Ba1 5Sources: Standard & Poor's, Fitch, and Moody's and Kiff, et al. (2012)*(a,b) South Korea’s rating firstly decreased between Jan 97-Jan 98 and after increased between Feb 98 – Dec 98.

 

Table 2.2. Sovereign Rating Failures during the Jan. 1997 - Dec. 1998

The years 1997-1998 are important as East Asian Crisis broke out in Thailand.

CRAs made downgrade of 8, 6, 6 notches respectively at the end of the crisis whilst Indonesia was rated with rating in the investible level (S&P BBB, Fitch BBB- and Moody’s Baa3).

Page 10: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 10

Sovereign Rating Failures / Global Financial Crisis

Fitch giving Argentina RD rating in the beginning of 2008 made downgrade by 7-notch in the aforesaid period.

The ratings Greece took from S&P, Fitch, Moody’s are BB+, BBB- and Ba1 at the end of 2010 while these rating are A, A, A1 in the beginning of that year; and this indicates 5, 4, 6-notch differences.

The countries listed except from Argentina got rating in investible level before the crisis.

  S&P Fitch Moody's  Start End Notches Start End Notches Start End Notches

Argentina B+ B 1 RD B -7 B3 B3 0Greece A BB+ 5 A BBB- 4 A1 Ba1 6Iceland A+ BBB- 5 A+ BB+ 6 Aaa Baa3 9Ireland AAA A 5 AAA BBB+ 7 Aaa Baa1 7Latvia BBB+ BB+ 3 BBB+ BB+ 3 A2 Baa3 4Lithuania A BBB 3 A BBB 3 A2 Baa1 2Portugal AA- A- 3 AA A+ 2 Aa2 A1 2

Sources: Standard & Poor's, Fitch, Moody's and Kiff J., et al. (2012)

Table 2.3. Sovereign Rating Failures During the Jan. 2008 - Dec. 2010

Page 11: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 11

Analysis and Methodology

Data from the official web pages of CRAs and countryeconomy.com/ratings

Long Term Sovereign Ratings which Standard & Poor’s, Fitch and Moody’s companies gave intended for countries between the years 1994-2014 have been used in the work.

Used Ratings are Long Term Sovereign Credit Ratings and number of countries = 117

Page 12: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 12

Transform to Numeric Scale

Various type of numeric transform used in previous studies.

Becker B., and Milbourn T., (2010), max. grade is AAA= 28 and min. grade is C=4,

Bae, K.H., Kang J.K., and Wang J., (2013), seperated ratings by 7 classes and they determined that max. grade is 28 and minimum grade is 1.

Bennell A.J., and Crabbe D., (2006), used 16 as a maximum grade that is equal to AAA and 1 as a minimum grade that is for B- and below. (for Moody’s B3)

Çalışkan Ö., (2002), grades spread between 0-100.

Bozic V., Magazzino C., (2013), max grade is 21 and min grade is 1.

Valle C.T., and Marin J.L.M. (2005), used decimal numbers to show mid-grades. In their studies max. grade is 8 it decreaes as 7.66, 7.33, 7, …, 1.

Page 13: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 13

Transform to Numeric Scale

S&P FITCH MOODY'S NUMERICSCALE

Characterization

AAA AAA Aaa 25 Highest quality Investment gradeAA+ AA+ Aa1 24 High quality

AA AA Aa2 23AA- AA- Aa3 22A+ A+ A1 21 Strong payment capacityA A A2 20A- A- A3 19

BBB+ BBB+ Baa1 18 Adequate payment capacityBBB BBB Baa2 17BBB- BBB- Baa3 16BB+ BB+ Ba1 15 Likely to fulfil obligations, ongoing uncertainty Speculativ

e gradeBB BB Ba2 14BB- BB- Ba3 13B+ B+ B1 12 High credit riskB B B2 11B- B- B3 10

CCC+ CCC+ Caa1 9 Very high credit riskCCC CCC Caa2 8CCC- CCC- Caa3 7CC CC Ca 6 Near default with possibility of recovery- C - 5R RD - 4 Default

SD DDD C 3D DD - 2- D - 1

Source: In this study, numerical scale based on previous linear transformations [see Afonso A., Gomes P., and Rother P., (2007) and Aizenman J., Binici M., and Hutchison M.M., (2013)].

Page 14: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 14

FITCH Greece Hungary Indonesia Ireland IsraelNovember

97… … … … …

December 97

17 (BBB) 17 (BBB) 16 (BBB-) 24 (AA+) 19 (A-)

December 97

17 (BBB) 17 (BBB) 15 (BB+)  24 (AA+) 19 (A-)

January 98 … … … … …

Data Designing

Data are designed monthly. Assumption: Country has got the same rating in all months until a new rating is

announced. Long Term Ratings given to countries may changed more than once in the same

month. The month in which rating of a country has changed more than once is divided by two or the number how many times the rating has been announced (Table 3.1). Table 3.1. Example of Data Designing

Page 15: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 15

Methodology

Paired Sample T Test is an influential test which is used in comparison of the average of two samples.

The hypothesis of the test,

If is rejected, it means there is a inconsistency between .

Page 16: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 16

Methodology

The test statistic is

where = sample mean of difference

= standard deviation of differences = number of pairs (sample size)

That hypothesis is rejected shows that the difference between the

averages of and grades is significant. It then enables us to reach the

conclusion that these two CRAs doesn’t coincide with each other.

Page 17: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 17

Methodology / Examples of Paired T Test Experiments

Data on Table 3.4 are the ratings which were given to Egypt by three big CRAs between June 2012 and August 2014.

Taking these ratings into consideration, we will examine whether or not S&P, Fitch and Moody’s coincide with each other in the years in question.

Country: EGYPT

Time S&P Fitch Moody’s Time S&P Fitch Moody’

sJune12 12 13 11 Sep.13 10 12 11July12 11 13 11 Oct.13 10 12 11Aug.12 11 13 11 Nov.13 10 12 11Sep.12 11 13 11 Dec.13 10 12 10Oct.12 11 13 11 Jan.14 10 11 9Nov.12 11 13 11 Feb.14 10 11 9Dec.12 11 13 11 Mar.14 9 11 9Jan.13 11 13 11 Apr.14 9 11 9Feb.13 11 12 11 May.14 9 10 9Mar.13 11 12 11 June14 9 10 9Apr.13 11 12 11 July14 9 10 9May.13 11 12 11 Aug.14 9 10 9June13 11 12 11 Sep.14 10 10 9July13 11 12 11 Oct.14 10 10 9Aug.13 11 12 11 Nov.14 10 10 9

Table 3.4. Egypt's Sovereign Ratings, Between June 2012 and January 2014

Page 18: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 18

Methodology / Examples of Paired T Test Experiments

Country: EGYPT  Time TimeJune12 -1 1 2 June12 -2 -1 1July12 -2 0 2 July12 -2 -1 1Aug.12 -2 0 2 Aug.12 -2 -1 1Sep.12 -2 0 2 Sep.12 -2 0 2Oct.12 -2 0 2 Oct.12 -1 1 2Nov.12 -2 0 2 Nov.12 -1 1 2Dec.12 -2 0 2 Dec.12 -2 0 2Jan.13 -2 0 2 Jan.13 -2 0 2Feb.13 -1 0 1 Feb.13 -1 0 1Mar.13 -1 0 1 Mar.13 -1 0 1Apr.13 -1 0 1 Apr.13 -1 0 1May.13 -1 0 1 May.13 -1 0 1June13 -1 0 1 June13 0 1 1July13 -1 0 1 July13 0 1 1Aug.13 -1 0 1 Aug.13 0 1 1        -1,33 0,10 1,43    nd=30 sd   0,66 0,55 0,50

Table 3.5. Differences of Egypt’s Credit Ratings and Average of Differences

Page 19: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 19

Methodology / Examples of Paired T Test Experiments

 

We see through t values we obtained that the difference between the averages of the ratings of S&P - Fitch and Fitch-Moody’s significant as hypotheses are rejected and the difference between S&P - Moody’s isn’t significant as hypothesis cannot rejected.

Page 20: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 20

Analysis / S&P and FITCH

Ctr t df p

ARG 10.08 7.60 2.48 11.44 239 0.00AUS 24.45 23.75 0.70 24.34 249 0.00BEL 23.88 23.34 0.55 11.36 273 0.00CHI 19.28 19.91 -0.63 -8.90 231 0.00CYP 18.58 19.86 -1.27 -15.11 167 0.00GEO 11.97 12.50 -0.53 -9.97 95 0.00GUA 13.96 15.00 -1.04 -55.53 116 0.00IND 11.64 12.81 -1.17 -11.41 238 0.00KOR 19.03 19.77 -0.73 -21.88 251 0.00KUW 21.27 21.79 -0.52 -16.03 234 0.00LIB 18.78 17.81 0.96 26.00 26 0.00MAC 14.36 14.94 -0.58 -8.91 119 0.00MON 12.56 11.77 0.78 8.93 123 0.00MOZ 11.60 11.00 0.60 13.21 118 0.00PHI 13.99 14.50 -0.51 -14.64 204 0.00POR 21.01 21.62 -0.60 -12.50 273 0.00RUS 13.91 14.88 -0.97 -8.39 247 0.00SRI 11.77 12.60 -0.82 -23.47 118 0.00SWE 24.52 23.92 0.60 12.28 273 0.00TAI 22.09 21.00 1.09 50.39 171 0.00TUR 12.31 12.86 -0.55 -12.57 273 0.00VEN 11.72 12.24 -0.52 -6.71 235 0.00VIE 13.45 12.78 0.66 14.48 145 0.00

Table 3.8. The countries whose ratings given by S&P and FITCH are different

Page 21: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 21

S&P and Fitch couldn’t coincide with each other on the ratings they give for 81 countries. (Table shows countries which has d-bar<-0.5 and d-bar>0.5)

The ratings S&P and Fitch gave to the 22 countries which already have well economic conditions such as Germany, Luxembourg and Netherlands.

Analysis / S&P and FITCH

Page 22: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 22

There are more differences in ratings they gave to the countries which have gone through an economic crisis or had various instabilities in fields which will affect their economies. For example, the difference between the averages of ratings which S&P and FITCH gave to Argentina is 2.48.

The discrepancies among ratings averages of some countries such as Turkey, Russia, Cyprus, Taiwan, South Korea and Libya are also higher than others. What these countries have in common is that they were in crisis before or are still in an economic or political instability.

Analysis / S&P and FITCH

Page 23: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 23

Analysis / S&P and MOODY’S

Ctr t df pBAH 18.35 18.90 -0.54 -9.18 143 0.00BAN 18.94 18.45 0.50 4.96 162 0.00BAR 17.76 16.66 1.10 17.93 192 0.00BER 22.91 23.76 -0.85 -38.63 261 0.00BOS 11.57 10.68 0.89 24.54 73 0.00BUL 15.33 14.50 0.83 20.67 215 0.00CAM 11.53 11.00 0.53 10.47 97 0.00CHI 20.28 19.65 0.63 11.71 207 0.00CHI 18.90 20.02 -1.12 -20.77 281 0.00COS 14.00 15.20 -1.20 -46.31 237 0.00ECU 9.16 8.18 0.98 9.94 191 0.00E.S 14.42 15.51 -1.09 -35.99 237 0.00EST 19.79 20.31 -0.52 -6.90 204 0.00FIJ 10.87 12.77 -1.90 -24.43 110 0.00GUA 13.77 14.29 -0.52 -13.77 173 0.00HON 21.87 20.93 0.94 22.68 281 0.00HUN 17.38 18.61 -1.23 -18.16 211 0.00

Ctr t df pICE 19.63 21.11 -1.47 -15.46 281 0.00JAM 10.67 11.41 -0.74 -9.14 200 0.00JAP 23.34 24.13 -0.79 -10.21 281 0.00KUW 21.27 20.73 0.54 6.76 234 0.00LAT 17.03 18.03 -1.01 -14.54 193 0.00MEX 15.92 16.48 -0.56 -16.49 281 0.00MON 12.60 12.00 0.60 9.02 120 0.00MON 14.11 13.20 0.91 12.89 84 0.00PER 14.86 14.24 0.62 15.11 204 0.00POL 17.82 18.73 -0.91 -18.37 263 0.00RUS 13.91 15.18 -1.27 -9.15 246 0.00SAU 21.41 20.15 1.26 10.55 149 0.00SOU 16.63 17.28 -0.65 -18.85 271 0.00SUR 11.61 12.14 -0.53 -6.29 141 0.00TAI 22.80 22.00 0.80 14.26 279 0.00TUR 12.35 12.92 -0.57 -11.55 281 0.00VIE 13.57 12.58 0.99 23.46 113 0.00

Table 3.10. The countries whose ratings given by S&P and MOODY’S are different

Page 24: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 24

The ratings S&P and Moody’s gave to 78 countries don’t coincide with each other.(Table shows countries which has d-bar<-0.5 and d-bar>0.5)

S&P and Moody’s are more distinctly incoherent with each other than others between 1994-2014 examined in the study include Brazil, Russia, Taiwan, Greece, Turkey, Hong Kong, Iceland and Vietnam supports the thesis that CRAs are more incoherent in countries which have gone through a crisis.

The ratings S&P and Fitch gave to the 22 countries.

Analysis / S&P and MOODY’S

Page 25: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 25

Analysis / FITCH and MOODY’S

Table 3.10. The countries whose ratings given by FITCH and MOODY’S are different

Ctr t df pARG 7.60 10.22 -2.62 -11.55 239 0.00AUS 23.73 24.24 -0.51 -15.38 256 0.00BUL 14.93 14.35 0.58 10.17 225 0.00COL 15.47 14.90 0.57 14.64 273 0.00COS 14.18 15.21 -1.03 -95.38 224 0.00CYP 19.86 18.83 1.03 9.98 167 0.00DOM 10.59 11.11 -0.52 -5.77 148 0.00ECU 8.83 8.22 0.61 6.08 157 0.00E.S 14.64 15.51 -0.87 -22.68 237 0.00EST 19.41 20.31 -0.90 -12.39 204 0.00FRA 24.97 24.42 0.55 18.31 273 0.00GUA 15.00 14.44 0.56 12.25 116 0.00HON 22.21 20.99 1.23 17.83 273 0.00HUN 17.58 18.61 -1.04 -14.36 211 0.00ICE 19.45 21.10 -1.65 -14.33 197 0.00IND 12.81 12.28 0.53 13.11 238 0.00

Ctr t df pISR 19.33 20.07 -0.74 -26.80 257 0.00JAP 23.41 24.11 -0.70 -12.63 273 0.00KOR 19.77 18.92 0.85 18.87 225 0.00KUW 21.64 20.50 1.14 15.70 256 0.00LAT 17.13 18.03 -0.90 -13.75 193 0.00MAL 20.36 19.58 0.79 20.68 250 0.00MEX 16.00 16.67 -0.67 -18.82 261 0.00NEW 23.69 24.90 -1.22 -30.43 166 0.00PER 14.90 14.26 0.64 13.96 201 0.00PHI 14.50 13.64 0.86 11.98 204 0.00POL 18.17 18.77 -0.60 -8.65 259 0.00SAU 21.45 20.71 0.74 7.42 131 0.00SOU 16.40 17.28 -0.88 -17.65 271 0.00SRI 12.77 12.00 0.77 12.27 46 0.00SUR 11.41 12.14 -0.73 -18.66 137 0.00VEN 12.24 10.98 1.26 24.30 235 0.00

Page 26: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 26

There is a significant difference between these two CRAs is found in the 77 countries which Fitch and Moody’s rate. (Table shows countries which has d-bar<-0.5 and d-bar>0.5)

The values representing this difference in countries such as Argentina, Costa Rica, Cyprus and New Zealand are more than 1 and these differences are critical ones for a large time period like 1994-2014.

The difference among the ratings of 22 countries isn’t found significant. The ratings of 13 countries out of 22 countries to which Fitch and Moody’s gave ratings are same.

Analysis / FITCH and MOODY’S

Page 27: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 27

There is an incoherency in ratings given to group of countries by these three CRAs.

The lowest rating average belongs to Turkey and the highest belongs to Taiwan subsequent to Czech Republic, Chile and China.

It can be seen that those countries have an average between 16-17.This rating means countries taking part in Emerging Markets have enough payment capacity together but they take part in the last group of the investable level

Analysis / Emerging Markets Ratings

  t df p-val.

S&P, MSCI Emerging Markets - Fitch, MSCI Emerging Markets

16.41 16.59   -

0.18-

13.99508

7 0.00

S&P, MSCI Emerging Markets - Moody's, MSCI Emerging Markets

16.57   16.77 -

0.20-

12.29534

9 0.00

Fitch, MSCI Emerging Markets - Moody's, MSCI Emerging Markets

  16.55 16.63 -0.08 -4.99 488

4 0.00

COUNTRY S&PFitc

hMoody'

s

Brazil 13.76

13.72 13.26  

Chile 19.88

19.59 19.65  

China 18.90

19.91 20.02  

Colombia 15.13

15.47 14.90  

Czech R. 19.62

19.48 19.59  

Egypt 14.69

14.84 14.04  

Greece 16.85

17.26 17.43  

Hungary 16.93

17.42 18.61  

India 15.05

15.45 15.45  

South Korea

19.29

19.77 18.92  

Malaysia 18.79

17.94 18.69  

Mexico 15.92

16.00 16.48  

Peru 14.76

14.90 14.24  

Philippines 14.06

14.50 13.80  

Poland 17.82

18.17 18.73  

Russia 13.91

14.88 15.18  

South Africa

16.63

16.39 17.28  

Taiwan 22.81

21.00 22.00  

Thailand 17.70

16.88 17.54  

Turkey 12.35

12.86 12.92  

Table 3.16. Average of Emerging Markets Countries' Ratings (Between 1994-2014)

Table 3.15. Emerging Markets Ratings

Page 28: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 28

At Table 3.20, before and after crisis, the one year period ratings were analyzed and according to those ratings, ANOVA test results were given.

Results show that not only during crisis but also after crisis, there are significant differences among three big CRAs' ratings.

Before crisis between June 2007 and June 2008, S&P's, Fitch's and Moody's' averages were 17.50, 17.84 and 18.34 respectively. However, after crisis, those averages decreased and became 17.25, 17.67 and 18.25.

Analysis / Global Financial Crisis

    Descriptives ANOVA    N Mean F Sig.   

BEFORE GLOBAL FINANCIAL CRISIS (BETWEEN JUNE 2007 AND JUNE 2008)

S&P 1811 17.50 11.282 0.000***FITCH 1634 17.84MOODY'S 1566 18.34Total 5011 17.87

DURING GLOBAL FINANCIAL CRISIS (BETWEEN JULY 2008 AND JULY 2009)

S&P 1773 17.25 14.947 0.000***FITCH 1555 17.67MOODY'S 1488 18.25Total 4816 17.69

Table 3.24. OECD Countries, ANOVA results, Between August 2008 and August 2009

Page 29: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 29

Ratings given by three major CRAs to Turkey for periods before and after November 2000 are comprised in Table 3.28.

When averages of the ratings before and after the crisis are analyzed, it is observed that there is about one-grade difference in between ratings of S&P and Moody’s and Fitch gave higher ratings in comparison with S&P. This discrepancy is significant and it demonstrates that these two CRAs are incoherent before the crisis.

Analysis / 2001 Turkey Crisis

  t df pTurkey S&P, Between November 1997 and November 2000 – Turkey Fitch, Between November 1997 and November 2000

11.17 12.15   -0.98

-45.00

45 0.00

Turkey S&P, Between November 1997 and November 2000 – Turkey Moody's, Between November 1997 and November 2000

11.17   12.00 -0.83

-14.62

45 0.00

Turkey Fitch, Between November 1997 and November 2000 – Turkey Moody's, Between November 1997 and November 2000

  12.15 12.00 0.15 2.84 45 0.01

Turkey S&P, Between November 2000 and November 2002 – Turkey Fitch, Between November 2000 and November 2002

10.35 11.61   -1.26

-15.75

30 0.00

Turkey S&P, Between November 2000 and November 2002 – Turkey Moody's, Between November 2000 and November 2002

10.35   12.00 -1.65

-12.91

30 0.00

Turkey Fitch, Between November 2000 and November 2002 – Turkey Moody's, Between November 2000 and November 2002

  11.61 12.00 -0.39

-2.56 30 0.02

Table 3.28 Turkey’s Credit Ratings Before and After 2001 Crisis

Page 30: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 30

Analysis / Turkey (1994-2014)

    Descriptives ANOVA

    N Mean F Sig.

   

TURKEY RATINGS (According to Announcement

Dates)

S&P 68 12.40

3.867

0.023

FITCH 64 12.95MOODY'S 68 13.01Total 200 12.79

19941995

19961997

19981999

20002001

20022003

20042005

20062007

20082009

20102011

20122013

9

10

11

12

13

14

15

16

17

S&P FITCH MOODY'S

Table 3.29 ANOVA Results Regarding Ratings Given to Turkey according Announcement Dates

Page 31: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 31

Analysis / Paired T Test for All Data

The results of the test demonstrate that the differences are significant even if the differences among the rating averages of Big Three are very small.

S&P and Fitch catches 61% coherency rate (the ratio of the months in which the same ratings are given to all the months) by giving different ratings 7852 times in the 20307 matched data.

S&P and Moody’s give different ratings in 11067 of 21860 data whilst the coherency rate of this dual is 49%.

Fitch and Moody’s give different ratings in 8607 of 18459 duals and they get 59% coherency rate.

  N Mean Rank Z p

Fitch-S&P

Neg. Ranks 3443 3952 -9.92 0.00Pos. Ranks 4409 3907    

Ties 12455      

Total 20307      

Moody's-S&P

Neg. Ranks 5165 5343 -9.56 0.00Pos. Ranks 5902 5701    Ties 1079

3      

Total 21860      

Moody's-Fitch

Neg. Ranks 3960 4357 -5.86 0.00Pos. Ranks 4647 4259    Ties 9852      

Total 18459      

Page 32: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 32

CDS Spreads and Credit Ratings

Credit Default Swap (CDS) demonstrates the economic risk level of the countries and the companies, and it has become popular lately.

It is a specific agreement between the two groups one of which is the buyer who pays premium in certain periods and the other one is the seller who provides protection for the buyer.

Hull, Predescu and White (2004) examined the relationship between the CDSs and rating announcements and they concluded that in the analysis that was practiced, with the data of 98-02, the estimations of changes of negative credit rating CDSs give high-possible information for the CDS that are sold for companies and the credit ratings that announce.

Norden and Weber (2004) shows that CDSs estimate the downgrade of credit ratings 90-60 day in advance.

Finnerty, Miller and Chen (2011), they state that upgrade rating changes have significant influences on CDS.

Page 33: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 33

Data and Methodology

CDS data are taken as a reference from Thomson Reuters DataStream on a daily basis.

Rating events encompass the rating changes of the same period carried out by Fitch Ratings.

In this study, it was used the standard event study methodology (Hull et al., 2004; Norden and Weber, 2004; Finnerty et al., 2013) to test CDS spreads changes whether have any impact on Rating Downgrades or Upgrades.

Page 34: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 34

Data and Methodology

In the study, 6 downgrade events are dealt for Croatia, Cyprus, Egypt, Greece, Ireland and Portugal.

As a result, it can be seen that while in some countries, CDS spread changes can estimate Rating Changes in advance, in other countries, a significant relation cannot be found between rating events and CDS spread changes.

Country [-90,-61] [-60,-31] [-30,-2] [-1,-1]Croatia (p-value) 0.213 0.111* 0.236 0.109*Ireland (p-value) 0.262 0.262 0.362 0.109*Egypt (p-value) 0.349 0.544 0.016** 0.285Cyprus (p-value) 0.047** 0.280 0.031** 0.593Greece (p-value) 0.060* 0.086* 0.076* 0.109*Portugal (p-value) 0.629 0.382 0.538 0.285

Page 35: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 35

Data and Methodology

CDS spreads are generally on the rise before the days of rating changes and that the changing periods start to become clear mostly 45-60 days ago

Table 4.3 includes the test results in which 6 Sovereign Downgrade Events, which were examined separately beforehand, are evaluated together this time.

The daily-period changes [-60,-31] and [-30,-2] in CDS spreads are found significant.

[-90,-61] [-60,-31] [-30,-2] [-1,-1]

# of

EventsWilcoxon

Ranks Test

0.84 0.06* 0.00*** 0.98 6

n 180 180 174 18

Sign Test 0.50 0.33 0.01*** 6

n 180 180 174 18  

Page 36: MA Thesis Pres

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 36

Conclusions

The coherency of the credit ratings that CRAs give to the countries in the long term has been analyzed and it is concluded that the three big CRAs contrast with each other.

CRAs are generally contrast with each other whether country in economic crisis or not. But in crisis periods, rating downgrades more than other periods and CRAs couldn’t predict many economic crisis.

The discrepancies among ratings averages of some countries such as Argentina, Chile, Brazil, Turkey, Russia, Cyprus, Taiwan, South Korea, Estonia, Venezuela, and Libya are also higher than others. This countries were in crisis before or still in an economic or political instability.

Developed countries such as USA, UK, Australia, Netherlands, Luxemburg etc. ratings’ are consistent with each other.

CDS changes can predict downgrade rating events before 30 days but upgrades are not.

CDS are able to follow the market with more sufficient schedule than Ratings.

Page 37: MA Thesis Pres

Thanks

01.05.2023Analysis of the Rating Discrepancies Assigned by Credit Rating Agencies 37

Comments

Suggestions

Contributions

Welcome