Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
The Effects of Sovereign Downgrades on the Italian Stock
Market
Angelo Marinangeli*
There are few studies that tak e into consideration the effects of the Italian
rating downgrade in the context of the current crisis. This paper analyzes the
impact of the Italian downgrades on the stock mark et, with the aim to
identify the relation between downgrade announcements and the returns of
Italian stock mark et. More specifically, the study analyses a sample of
shares belonging to the k ey sectors i.e. finance, technology, fashion, food
and beverage and health sectors. In addition, the analysis is conducted in
order to verify if the “border downgrade” have an amplified impact than the
other ones. It was found, through the event study dummy approach
methodology, that the transitions from a grade to another have had a greater
impact than transitions into the same grade. Moreover, the first Italian
downgrade has had an important impact on the stock mark et returns,
because this downgrade represented the alarm bell that the crisis was
intensifying also in Italy. Only in the technology industry the first downgrade
had no effect, because the high tech industry is more robust because it can
be defined the most innovative industries, and so it perceived the crisis
more later.
Keywords: Financial Crisis, Sovereign Rating, Italian Downgrades, Stock Market, Event
Study.
JEL Codes: G01, G12 and G14
1. Introduction
In last years the concept of rating took an important role in the economic world; in fact the
rating agencies assessments on the sovereign credit produced effects on political and
economic sphere.
This paper is interested to the last phase of the crisis occurred in the last two years after the
sovereign rating review by the rating agencies. The first sovereign rating downgrade occurred
in US on August 5, 2011 when Standard & Poor‟s downgraded the United States credit
rating from AAA to AA+ with the motivation that the plan that the Congress and the
Administration agreed could not stabilize the government‟s medium-term debt dynamics. This
announcement caused a financial crisis which spread like a domino on other countries.
Afterwards, other downgrades occurred in European countries that caused an upheaval on
the European financial market. An interesting aspect for this survey is the following: such
downgrades have had different impacts depending on the macroeconomic conditions.
The paper represents an empirical investigation because it applies the inductive reasoning, is
based on the observation of a large number of cases, states the hypothesis and provides a
generalization about the events observed.
_______________________________________________________
*Angelo Marinangeli, University of Rome Tor Vergata, Italy.
1
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
The scientific approach is used because we search for knowledge by starting from the
literature contributions, formulating the question and the hypotheses of research, collecting
data, developing the survey and finally we analyze the findings and make conclusions through
the generalization of our results.
This research is a quantitative analysis because it analyzes quantitative data represented by
the values of the share prices observed in the established time period.
In addition the present research is knowledge oriented because it is focused on generating
fundamental knowledge.
The work wants to fill a gap in the literature because there are few studies that take into
consideration the effects of the Italian rating downgrade announcements, or more generally of
sovereign rating changes, in the context of the current crisis. In addition it wants to extend the
past existing literature that analyzes the impact of the rating announcements or sovereign
rating announcements but not in the current context. Research on sovereign credit ratings can
be divided into two categories: studies which analyze the determinants of sovereign credit
ratings and studies which examine the price impact of rating announcements; the latter group
usually focuses on the sovereign bond market; while this research analyzes the impact of the
Italian downgrades on the stock market and more specifically the key sectors.
The sovereign rating change announcements produce effects not only on the cost of the
sovereign credit but also on the financial market, as also emerged from literature.
This paper come from the desire to answer to the following research questions:
• Which are the effects of Italian downgrades on the stock returns?
So, this survey investigates the relation between downgrade announcements and financial
market trend, and it aims to quantify the strength of this relation (if any).
With more details, the paper analyzes if particular types of downgrades named “border
downgrades” have an amplified impact on financial market compared to others.
Cantor and Packer (1996) show that the impact of rating announcements for belowinvestment-grade sovereign bonds is stronger than the impact for investment-grade sovereign
bonds. In addition, Creighton et al. (2007) analyze the effects of corporate rating changes on
the Australian financial market; they find the greater effects for downgrades from “investment
grade” to “speculative grade” .
It would be interesting to verify if this results are generalizable to sovereign rating changes.
So, the aim of the survey is to verify if downgrades that represent transitions have an
amplified effect on the stock returns of our sample shares.
We expect to replace the results of Creighton et al., that is to find a greater impact in
transitions from a grade to another than downgrades into the same grade.
In addition, by analyzing the rating change announcements of the three rating agencies, so it
is possible identify to which rating agency the Italian financial market is more sensitive.
Norden and Weber (2004) show that financial market is more sensitive to Standard & Poor‟s
announcements than Moody‟s and Fitch Rating ones. We attend to verify that also the Italian
stock market is more sensitive to a rating agency than others and to a specific type of
downgrading than others. It represents the effective and real reason of the study, it could be
the aspect of relevance of the paper if the analysis confirmed the attended results that could
be used to better understand the movement of the Italian stock market to downgrade
announcements. The attended result is that the transitions from upper medium grade to lower
medium grade produce a greater impact on the stock market compared with the transitions in
2
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
the same grade and the transitions from high grade to upper medium grade. We believe that
the transitions from upper medium grade to lower medium grade have a higher impact
because they are relevant for the solvibility and the reputation of a Country and so they are
significant for the financial decisions of investors.
In the next section it is presented the literature review, that is divided into four sub-sections. In
the third section they are presented the methodology, model and data divided into four
subsections to explain the methodological approaches used and why they are chosen. In the
fourth section findings are described, in the fifth section conclusions are developed and finally
there are references.
2. Literature Review
Reviewing the literature on the subject, three fields of research can be identified. The first
field analyzes the impact of corporate rating changes on the dynamic of shares or corporate
bonds. The second field studies the effects of sovereign rating changes on the trend of
sovereign bonds or financial market. Finally the third field focuses on the analysis of the
Emerging Countries rating changes and their effects on the development of such Countries.
2.1. The effects of corporate rating changes on the financial market
The first field of literature studies the effects of rating announcements on corporate stocks,
corporate bonds, commercial papers and credit default swaps. For all types of instruments it
highlights the impact of announcements on share prices, the most obvious impact in the event
of a downgrade, although some studies show that a positive rating change produces a
positive abnormal return in the days around the announcements (Barron et. al., Cellier and
Chollet, Creighton et al.), other interesting results are found on CDSs (Micu et al., Norden and
Weber and Crabbe and Post), on the effects of watchlistings (Barron et al., Steiner and
Heinke) and final on the relation between shareholders and bondholders when a rating
change occurs (Abad-Romero and Robles Fernandez).
Unlike the other authors analyzed in this research field, Barron et al. (1997) analyze the
effects of rating changes but also new ratings and credit watches on bonds, stocks and also
the commercial papers; while Cellier and Chollet (2011) analyze the effects of the rating
announcements related to corporate social responsibility. The analysis of Barron et al. is
made on daily data published by Standard and Poor‟s between 1984 and 1992, the results
show that, around the announcement date of a positive credit watch, stock returns are
positive. An interesting aspect of this work is relative to new ratings, which conversely have no
significant effects on stock returns; this result is verified in both the short and long time. We
could explain this aspect by stating that companies submitted to their first rating are not widely
known on the market, so the latter is not very sensitive to new rating announcements. The
analysis of Cellier and Chollet is made on rating announcements published by Vigeo between
2004 and 2009. Vigeo is a rating agency specialized on assessment related to corporate
social responsibility. The sample analyzed by the authors is composed by shares and index
prices; shares‟ dividends are extracted from Datastream while market value and book to
market are obtained from Worldscope. The results, obtained through the event study
methodology with the procedure detailed by Renneboog et al. (2008a), show a significant
positive influence on the two days preceding the announcement and the next two days.
Clearly, as evidenced by studying the content of the announcements, only a few types of
announcements have a significant role on stock returns; in addition, some announcements
3
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
have positive effects and other ones have negative effects on stock returns. The work sheds a
new light on the relationship between corporate social ratings and financial performance. To
tackle possible error measurement, all models include an intercept. The survey conducted by
Creighton et al. (2007) analyze the effects of changes on the pricing of financial assets in
Australia. The interesting feature of their work is the limited role of credit ratings in the
Australian system of financial regulation, then the impact of rating changes is observed
without regulatory effects. The sample of events for which the authors consider both rating
changes and yield spread data includes 33 announcements consisting of 21 downgrades and
12 upgrades. The bonds in question are all issued in Australia and denominated in Australian
dollars. The results, obtained through the standard event study techniques „market model‟,
show that after both positive and negative rating announcements, share and bond prices
highlight small movements towards the expected direction. Rating announcements showing
the most important effects refer to small businesses and rating changes from investment
grade to speculative grade.
Even with respect to CDSs, several works show the impact in the presence of rating
announcements (Micu et al., Norden and Weber and Crabbe and Post), only the work of
Crabbe and Post (1994) shows the lack of effects in the days around the date of
announcement.
The survey is conducted between 1986 and 1991 on the issues of commercial papers and
CDSs (Credit Default Swap) by bank holding companies after a credit rating downgrade, in
order to test the model of Diamond (1991). He argues that the company reputation influences
the choice of raise funds directly from the market or using intermediaries. The Authors show
that in the period 1986-1991 the average cumulative abnormal returns decreased of 6.69% in
the first two weeks after the downgrade and 11.05% in the next twelve weeks. Regarding the
CDSs issued by the member banks, unlike commercial papers issued by bank holding
companies, there are no significant changes in the period around the date of downgrade. The
study shows that the direct borrowing from market through commercial papers is sensitive to
the downgrades of companies, in line with Diamond, the rating being a measure of corporate
reputation. While, taken into account the other paper analyzed on CDSs, Micu et al. (2006)
define what types of rating announcements have an impact on CDSs spreads. The survey
collects daily data on CDSs spreads and rating announcements. Data on CDSs spreads
were obtained from Markit, a London-based distributor of credit pricing data.
Data on rating announcements of Moody‟s, S&P and Fitch are obtained from Bloomberg. The
methodology applied is the event study with the market model method. Four market indices
INDg,t are constructed, corresponding to the whole letter rating categories AA, A, BBB, BB.
The value of the index on a given day is set equal to the median spread for the relevant rating
category.
The analysis reveals that all types of changes, including those in the rating look out, have a
significant impact on spreads. The impact is greater for firms with split ratings, small-cap
companies and those with a score close to investment grade.
The study of Norden and Weber (2004) analyzes the effect of rating change announcements
on stock market and CDS market. This survey is conducted on international level for the
period from 2000 to 2002. The dataset consists of market-wide CDSs spreads,
corresponding stock prices and credit rating data. CDSs data was gratefully provided by a
large European bank which is among the world‟s top 25 credit derivatives counterparties.
Time series of daily common stock closing prices obtained from Thomson Financial
DataStream. Additionally, we add time series for three stock market indices (Stoxx 50, S&P
500, and Topix 100), rating agencies taken in account are: Moody‟s, Standard & Poor‟s, and
4
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
Fitch Ratings. The results, obtained through the event study with the index adjuster and market
model approach, show that no relevant effects correspond to positive ratings (upgrades)
announcements, while significant negative abnormal returns correspond to Standard & Poor‟s
downgrades announcements. Such findings are verified around the announcement day, in
both stock market and CDSs market. For Moody‟s announcements, no abnormal returns are
found in the stock market, while significant changes are identified in the CDSs market in the
days around the announcement. For Fitch rating downgrades, no abnormal returns in both
markets are found. The study shows, therefore, that in the years when the analysis is
conducted there is greater attention by the stock market to announcements by Standard &
Poor's, probably related to the better knowledge of this agency by investors.
Furthermore, taken into consideration the contributions that investigate on the effects of
watchlisting announcements, it emerged that abnormal positive returns occurred for positive
credit watch announcements (see, e.g., Barron et al., 1997). In addition, Steiner, Heinke
(2001) show that significant price reactions can be identified to negative watchlisting
announcements, while no effects on share prices occur for upgrades and positive watchlisting
announcements. This survey, more particularly, identifies the rating influence on Eurobonds
prices, by analysing daily price effects on the announcements of watchlisting and rating
changes made by Standard & Poor's and Moody's. They note that the announcements of
downgrades and negative watchlistings lead to significant reactions on prices, and vice versa
no effects on share prices occur for announcements of upgrades and positive watchlistings. In
addition, they show that type and nationality of the issuer are key factors in determining the
prices reactions after the downgrade. In fact, the prices reactions are particularly intense for
downgrades of speculative securities. Moreover, the announcement impact can be explained
as the effect of the prices pressure caused by regulatory constraints rather than rating
changes.
Finally, another interesting aspect emerged by the analysis of this research field is the relation
between shareholders and bondholders, this linkage is highlighted by Abad-Romero and
Robles Fernandez (2006). They analyze the impact of corporate rating changes on the bonds
and shares of the Spanish stock market. This survey investigates the effects on returns and
systematic risks. This study applies an extension of the event study dummy approach
methodology and is based on the rating changes of Moody‟s, Standard and Poor‟s and Fitch
ratings between 1990 and 2003. The results are characterized by the relation between
shareholders and bondholders. This hypothesis states that there is a conflict of interest
between bondholders and stockholders. Thus, a credit rating downgrade reduces bond value,
which is expropriated from bondholders to stockholders, causing the increase of share prices.
In the case of rating upgrades the wealth redistribution is in the reverse direction.
2.2. The effects of the sovereign rating changes
With regard to the second field of research, sovereign ratings summary the information of
macroeconomic indicators such as Income per person, GDP growth, level of economic
development, inflation, external debt, and they are therefore related to the spread of
securities. Research on sovereign credit ratings can be divided into two categories: studies
which analyze the determinants of sovereign credit ratings and studies examining the price
impact of rating announcements; the latter group usually focuses on the sovereign bond
market.
Unlike corporate ratings, sovereign ratings have no effects on single firms but on the market
as a whole, as evidenced by the survey of Brooks et al. (2004), and on Sovereign bonds. The
5
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
study of Brooks et al. analyzes the impact of sovereign rating changes on the market looking
at both local and foreign rating changes. Consistently with the findings observed for corporate
changes ratings, credit rating downgrades have a negative impact on returns. Downgrades
have a negative impact on both local market and Dollar Area Countries one. Among the four
agencies examined, only Standard & Poor's and Fitch ratings downgrades appear to have a
significant effect on yields. Another interesting aspect are the asymmetric effects to the
sovereign rating announcements, as highlighted by the studies of Klimavičienė (2011),
Ferreira and Gama (2007) and Gande and Parsley (2005). The first contribute investigates
about the relevance of sovereign rating announcements in the Baltic stock market testing the
degree of anticipation and price reaction. The methodology used is the event study with the
market-model-adjusted method (or “market model”) where the return on the market is
approximated by the MSCI EM Small Country, in particular the aim is to analyze the price
impact of sovereign rating announcements by Moody‟s, Standard & Poor‟s and Fitch ratings
on the price. Through this survey it emerged that sovereign rating announcements contain
relevant pricing information; in fact the price impact of negative events is sometimes larger
than the impact of positive events. In addition, the impact of rating announcements is relevant
in the announcements day although some announcements are anticipated. The second
contribute analyzes the effects of Standard & Poor‟s sovereign rating of 29 countries, both
emerging and developed. The results, obtained through the event study methodology, show
that sovereign rating upgrades are associated with a positive effect on stock market prices
relative to the US and downgrades are associated with a negative effect.
When a country is downgraded the remaining countries do much worse than the US market.
Negative news in the sovereign debt market, but not positive news, seem to have a significant
impact on the stock markets of non-event countries. Gande and Parsley (2005) show the
effects of sovereign downgrades on sovereign debt markets. They apply the event study
methodology with the market model regression, the expected return on the market index as
predicted from the (OLS) coefficients is estimated in the market model regression. They find
that there is an asymmetric impact on sovereign debt markets: downgrades abroad are
associated with a significant increasing in sovereign bond spreads.
One possible explanation for the asymmetric effects to rating news is that upgrades are
partially anticipated by market participants, unlike downgrades. Sovereign ratings affect not
only the financial system but also political aspects related to macroeconomic variables.
Taking into account, as it emerged, that negative announcements have strong effects and
positive announcements have weak effects, you can understand how much important the
sovereign rating is. At the same time, it is possible to understand why that assessment is
particularly complex, given the often irreversible nature of its effects. A different and more
specific result on the impact of sovereign rating changes emerged by the survey of Cantor
and Packer (1996) that investigates which announcements have a greater effects than
others, in particular they argue that the assessments of rating agencies affect market returns.
The analysis applies the event study methodology and confirms the authors‟ hypothesis that
announcements of sovereign rating changes are followed by significant movements in the
yields bonds (Sovereign Bond and dollar bond spreads as the difference [(yield - Treasury) /
Treasury]) in the expected directions. The regression includes four variables for actual rating
changes, positive events, Moody‟s decisions, or speculative grade sovereigns. Three proxies
are used: the change in relative spreads (in the direction of the anticipated change); the rating
gaps between the agencies (the sign of the gap between the rating of the agency making the
announcement and the other agency‟s rating); the third proxy is an indicator variable that
6
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
equals 1 if another rating announcement of the same sign had occurred during the previous
sixty days. All proxies measure conditions before the announcement.
A final regression adds all three anticipation proxy variables simultaneously to the basic
regression; the results are robust to the addition of the proxy variables.
The sample is composed by dollar-bond spreads and sovereign bond spreads.
The results show that the impact of rating announcements for below-investment-grade bonds
is stronger than investment-grade sovereign bonds. In addition, the rating announcements
already anticipated have a greater impact than less predictable announcements. For this
reason the assessments of rating agencies have a predictable component. Regarding the
effects of sovereign rating announcements on the CDSs (Afonso et al., Arezki et al.), also for
the sovereign rating announcements it emerged that negative events cause significant
reactions.
More specifically, Afonso et al. (2011) analyze the effects of announcements by credit rating
agencies Standard & Poor's, Moody's and Fitch ratings on the returns of Sovereign Bonds
and CDSs. The methodology used is the event study on daily data. The results show
significant abnormal returns on government bonds, especially in the case of negative
announcements, being more tenuous in the case of positive announcements. Moreover,
Countries downgraded from less than 6 months have higher spreads than Countries with the
same rating but not downgraded in the last 6 months.
Another interesting result is that while there is a significant reaction of sovereign yield spreads
and particularly CDS spreads to negative events, the reaction to positive events is much more
muted. Arezki et al. (2011) study the impact of sovereign rating announcements on European
financial markets in the period 2007-2010, the methodology is the event study, the dummy
approach event analysis takes into account the potential linkages between markets through
the Vector Autoregression (VAR) framework, the dummy is equal to 1 at time t and zero
otherwise. The results show that the effects depend from the type of announcements, and
how the country lives the downgrade; in addition new rating downgrades have a stronger
effects than revisions of outlooks which could be explained by banking regulation, ECB
collateral rules, CDS contracts or investments mandates.
Finally, Pukthuanthong-Le et al. (2007) study the effect of sovereign ratings announcements
on stock and bond markets. The methodology is the event study; the market model is applied
by using a world stock index and U.S. Treasury bond returns as benchmark.
The sample contains rating changes and rating reviews of 34 Countries for the period 19902000. The results show that both share and bond prices react to downgrades but not to
upgrades; in addition sovereign bond yields anticipate rating downgrades. With regards to
the Rating reviews, both positive and negative, they do not affect a country‟s stock market, but
they lead to a price reaction in sovereign bond markets. It is consistent with previous studies
in the literature, which generally conclude that only negative credit rating announcements have
significant impacts on yields and CDS spreads.
By the analysis of this research field it emerged that emerging markets are particularly
sensitive to rating changes. This result provides the opportunity to understand how the third
field of research, explained in the next section, was born.
2.3. The effects of sovereign rating changes of Emerging Countries
The third field of research analyses the impact of emerging countries sovereign credit ratings
and studies the probability to reproduce these effects, at macroeconomic level, to other
countries. In addition, it examines the role of rating agencies in international finance, in the
7
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
equilibrium of international production processes and in the development opportunities of
emerging countries after positive or negative evaluations. Analyzing the third field of
research, it emerged that investors with a high risk profile are careful to the assessments of
rating agencies in order to invest their assets in the assessed countries. As evidenced by
Biglaiser et al. (2008), the effects of rating changes are more marked during the crisis and
ratings contribute to instability of financial markets, like Kaminsky and Schmukler highlight,
and finally rating agencies can reduce capital flows to emerging countries when they are
excessive (Larrain et al.). Biglaiser et al., (2008) study panel data for 50 developing
countries between 1987 and 2003 to test the effects of rating changes on the flows of foreign
portfolios. The investment in Emerging Countries combines a high risk to a proportionally high
yield, the study therefore predicts that investors with a high risk profile are careful to the
assessments of rating agencies in order to invest their assets in the assessed countries. The
work, which studies rating changes together with the “democracy” factor, using a two-stage
Heckman model, determines:
- The importance of ratings, which have positive effects on the countries that receive most of
the flows of assets;
- Countries where new political institutions arise, and countries in a period of economic
challenge, have a greater chance of being selected by investors because they offer higher
risk premiums.
Ratings and democracy have, therefore, a stronger link in the poor Countries.
Kaminsky and Schmukler (2002) analyze the influence of changes in soverei gn ratings of
Emerging Countries, discovering effects which have a direct impact not only on sovereign
bonds but also on the stock market. An interesting feature emerged from the work is that the
effects of rating changes are more marked during the crisis, and downgrades occur during
crisis. This finding could suggest that ratings contribute to the instability of financial markets. In
addition, downgrades occur in recession periods, because the assessments contain macroeconomic aspects. This is what leads to believe that credit rating announcements cause
turmoils and financial instability.
Larrain et al. (1997), through the methodology panel data analysis and event studies, find
conclusions in line with the previous studies. Among the full data set on government dollar
bond yields, obtained from Datastream, Bloomberg, JP Morgan, Merrill Lynch and the
Federal Reserve Bank of New York, with a sample of 26 countries against a total of 60
countries whose sovereign debt has been rated during part of the observation period. The
event study with the market model approach is applied with the aim to investigate the shortrun impact of press releases observation window spanning from 40 trading days before the
press release (day 0) to 40 trading days after. They used a Granger causality tests based
on an unbalanced panel data set with yearly averages for ratings and yield spreads.
They demonstrate that changes in credit ratings have a significant impact on international
financial markets. The innovative aspect of their findings is the demonstration that sovereign
rating can reduce capital flows to emerging countries, when they are excessive, through
negative rating announcements.
Sovereign ratings have, thus, a more relevant role as indicators for the investors attracted by
high risk. They act as a moderator, through negative rating announcements, when there are
excessive investment flows to emerging markets. In addition to ratings, also democracy has
an important role in financial markets, as well as the establishment of new governmental
institutions due to higher investment returns. Finally, another interesting aspect is the role of
the rating agencies on size and volatility of Emerging Country market as Kräussl (2005) find
analyzing the impact of sovereign credit ratings on financial stability in emerging economies.
8
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
The results, obtained through the event study, show that rating agencies have an important
influence on size and volatility of Emerging Countries credit markets. Effects are more
pronounced in the case of downgrade.
2.4. Literature review: the focus on the methods used
The studies described in the literature review subsection use the event study methodology
because it represents the typical statistic method to identify the impact of unexpected events
on financial series. With more details, some differences are found in the approaches of the
event studies used.
The most common models are: the constant-mean-return model and the market model. The
former assumes that the average return on a given security is constant over time; the latter
assumes a stable linear relationship between market returns and security returns. In addition,
there are alternative statistical models, developed in recent decades. They are: the market
adjusted return model, applied when the data set is restricted; the factor and multifactor
models, applied when the sample is composed by companies similar for capitalization or
activity industry, because they reduce the variance of abnormal returns by explaining a greater
proportion of the
variation of normal returns; the dummy approach, that extends the
«estimation window» up to contain the «event window» [Binder (1998) ] associating to
dummy variables the value 1 for the event window days and the value 0 for otherwise; the
generalized least squares, that consider standardized abnormal returns, applied when the
specific event date cannot be identified; finally the ARCH and GARCH models, applied in the
case of heteroshedasticity. The economics models are statistical models in which some
constrains are imposed in order to obtain a more specific estimation of returns. So, these
models apply the economic hypothesis to statistical model parameters. The capital asset
pricing model (CAPM) establishes that, in competitive markets, expected returns on
securities are directly proportional, in an equilibrium position, to expected returns for market
risk indicated by the β coefficient. The arbitrage pricing theory (APT) determines the
risk/return relationship through risk sources different from market. By the analysis of literature
focused on the methodology used, it emerged that the method mainly used by the authors is
the market model because it generally is the most used event study approach. Other
methodological approaches are used to the presence of particular conditions or to take
advantage by the opportunities that have particular methodological approaches. For instance,
Cellier and Chollet (2011) apply the procedure detailed by Renneboog et al. (2008a) to tackle
possible error measurements; Cantor and Parker (1996) make an event study through a
multiple regression that includes four indicator variables; while Abad-Romero and RoblesFernandez (2006) and Arezki et al. (2011) use the event study dummy approach because this
method allows to consider multiple events. Due to its advantages this approach, better
explained in the next section, it is chosen for this analysis.
3. The Methodology and Model
The methodology applied is the event study. It represents, as viewed in the literature review,
the typical statistic method to identify the impact of unexpected events on financial series.
More specifically it is applied a particularly model based on the event study dummy approach
studied in order to investigate the impact of the Italian Downgrades on the stock market.
9
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
3.1. The SHES methods
The first step of an event study analysis is to define the time horizon amplitude, generally
named “estimation window”, on which we have to estimate the normal return; while the event
window is the period that contains the event date.
Binder (1998) finds the common use of an estimation windows of about 250 days, for the
daily data analysis, and about 5-7 years for the analysis based on monthly data.
In this analysis the short time horizon is used, because the SHES (Short Horizon Event Study)
method is correctly specified and sufficiently powerful when the specific dates of events are
known; the sovereign rating announcements generally are spread by media (news,
newspaper, etc.) so everyone come to know the announcements even people who do not
follows the financial news. In addition, the effective identification of the abnormal returns
decreases when the time horizon amplitude increases. Brown and Warner (1980, 1985) find
that the long time horizon decreases the powerful of the statistic test.
3.2. The event study dummy approach
The event study dummy approach is chosen because this methodology allows to consider a
multiple event and to solve the «volatility clustering» problem.
In this approach, differently to the event study, the estimation window is extended up to contain
the event window [Binder (1998)].
Fig.1 The timeline of an event study dummy approach
Source: own processing
The dummy variable is equal:
• to 0 for the observation of the estimation window, period of the time line (Figure 2) that
precedes the event window;
• to 1 for the observation of the event window
3.3. Impact of Italian downgrades on the stock market: The model
10
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
In order to answer to the second research question on the Impact of the Italian downgrades on
the stock market we try to verify one of the following hypotheses:
H0: there is not a relation between the Italian downgrade announcements and the Italian stock
market;
H1: there is a relation between the Italian downgrade announcements and the Italian stock
market;
We used the following model:
- the dummy variable regression, developed through the methodology described above, can
be represented as:
Rit = α + βRmt + γt Dt + ξit
(1)
Where:
α is the regression constant;
Rit is the return on the stock i at the time t;
Rmt is the return on the market at time t, approximated through the FTSE Italia All-Share
index;
D t is the dummy variable,
γt is the coefficient of dummy variable and represents the abnormal return on stock i at day t;
ξit is the stochastic variable.
The estimation window is 487 days and contains eight event windows, we chose to use event
windows of three days following Afonso et.al. (2011) and Ferreira and Gama (2007) because
this time horizon is long enough for a correct analysis.
We apply this model to analyze multiple events, so the regression is:
Rit = α + βRmt + γ1t D 1t+ γ2t D2t + …+ ξit
(2)
Where: γ1t , γ2t , etc… are the coefficients of the dummy variables D 1t, D 2t and so on.
As better specified in the next section, we have six coefficients γt relating to six dummy
variables, because we take into account six announcement events. So, we have: γ1, γ2, γ3,…
showing the effects of the first, the second, the third announcement and so on… of sovereign
rating downgrade on stock returns.
With the aim to verify the significance of the regression coefficients we use the t-test, which in
the event stud y analysis can be used also to test the abnormal returns, and the Durbin
Watson test to verify the absence of autocorrelation among variables; in addition the VIF
coefficients are analysed to test the absence of multicollinearity.
3.4. Data
Table 1 represents all the Italian downgrade announcements occurred in the period between
th
4 October 2011 and 9 July 2013. Among all downgrades we select the first one, the third
one, the fifth one, the sixth one, and finally the seventh and the eighth one. We discard the
second downgrade because it is closed to the first one so it is difficult to separate the effects
11
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
of the first and those of the second downgrade; therefore for the same reason the fourth
downgrade is discarded.
Table 1 The events taken into consideration by the survey
Event dates
Rating agencies Downgrades Transitions grade
10/04/11
Moody‟s
Aa2 to A2
High to Upper medium
10/07/11*
Fitch
AA- to A+
High to Upper medium
01/13/12
S&P
A to BBB+
Upper medium to Lower medium
01/27/12*
Fitch
A+ to AUpper medium (no transition)
02/13/12
Moody‟s
A2 to A3
Upper medium (no transition)
07/13/12
Moody‟s
A3 to Baa2
Upper medium to Lower medium
03/08/13
Fitch
A- to BBB+
Upper medium to Lower medium
07/09/13
S&P
BBB+ to BBB Lower medium (no transition)
* discarded downgrade; source: own processing
The sample is composed by shares belonging to various sector with the aim to capture the
rating downgrade impact of different industries.
With more details, it includes 106 shares of companies belonging to: finance (bank,
insurance, financial services), technology, fashion, food and health sectors. They are chosen
because they represent the Italian market key sectors. More specifically, the sample is
composed by the company included in the indices FTSE Italia fashion, FTSE Italia finance,
FTSE Italia food and beverage, FTSE Italia health and care and FTSE Italia technology.
Total return prices (which include stock dividends) are extrapolated by Thomson Reuters
Datastream, stock returns are calculated through the logarithm of the ratio between the price
at time t and t-1.
(3)
Where:
pit : is the total return price at time t;
pi(t-1) : is the total return price at time t-1.
4. The findings
The findings show that the Italian downgrade announcements caused different effects on
different sectors analysed, especially regarding the border downgrade and more specifically
the transition from upper-medium grade to lower medium grade.
More particularly by the analysis emerge that there is a negative relation between the Italian
downgrades announcements and the stock market returns but this relation is especially
related to specific downgrades.
It‟s an interesting result because we started this analysis with the aim to understand which
downgrades could demonstrate more pronounced effects in the crisis context and, then, to
study the develop of the current crisis.
12
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
Table 2 shows that downgrades that represent a transition from upper medium grade to lower
medium grade have had a significant impact on the Italian stock market. In addition, it also
has had an important impact on the stock market returns the first Italian downgrade of October
4, 2011.
The possible explanation of it is that this downgrade occurred in a context of general review of
the sovereign credit rating so it represented the alarm bells that the crisis was intensifying
also in Italy; another important aspect is that this downgrade from Aa2 to A2 is a transition
from high grade to upper medium grade.
The first Italian downgrade has expressed its impact, more specifically, on the sectors:
fashion (γ1: -0,004), finance (γ1: -0,011), food and beverage (γ1: -0,020) and health and care
(γ1: -0,010); only in the mobile and communication industry and in the technology industry the
first downgrade had no effects maybe because the high tech industries are more robust
because they can be defined the most innovative industries, and so they perceived the crisis
more later.
In all industries analysed the transition from upper medium grade to lower medium grade had
a negative impact, specifically, in addition to the first downgrade analysed previously:
- fashion industry: the 2nd (γ2: -0,03) and the 5th (γ5: -0,03) downgrade;
- finance industry: the 4th (γ4: -0,02) and the 5th (γ5: -0,03) downgrade;
- food and beverage industry: the 2 nd (γ2: -0,002) and the 5th (γ5: -0,004) downgrade;
- health and care industry: the 4th (γ4 : -0,01) and 5th (γ5: -0,01) downgrade;
- technology: the 4th (γ4: -0,01) and the 5th (γ5: -0,015) downgrade.
The VIF test (always close to 1) and the Durbin Watson test (always close to 2) show that
there is no correlation and multicollinearity among the variables.
SECTOR
Table 2. Summarize of the regressions results
γ1 D2
γ2 D 2
γ3 D3
γ4 D 4
γ5 D5
γ6 D6
INDUSTRIE
10/04/11
01/13/12
02/13/12
07/13/12
03/08/13
07/09/13
Fashion
-0,004**
-0,003***
0*
0
-0,003*
0
Finance
-0,011***
0
0,001**
-0,02****
-0,003*
0,001
Food and
beverage
-0,020****
-0,002***
0,002*
0,003
-0,04*
0,001
Health and
-0,010***
0
0
-0,01**
-0,012***
0,001
0,004**
0,010
0,003
-0,010**
-0,015***
0*
S
care
Technology
Durbin Watson test is always close to 2 and VIF is always close to 1
Significance level: *90%; **95%; ***99%; ****99,9%. Source own processing
13
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
5. Summary and Conclusions
This paper is interested to the last phase of the crisis occurred in the last two years after the
sovereign rating review of the rating agency. The first sovereign rating downgrade occurred in
US when Standard & Poor‟s downgraded the United States credit rating from AAA to AA+.
This announcement caused a financial crisis which spread like a domino on other countries.
Afterwards, other downgrades occurred in the European countries that caused an upheaval
on the European financial market.
The aim of the survey is to verify if those downgrades that represent a transition have an
amplified effect on the stock returns of our sample shares. We believe that the transitions from
Upper medium grade to lower medium grade have a higher impact because they are relevant
for the solvibility and reputation of a Country and so they are significant the financial decisions
of investors.
The findings highlight that there is a negative relation between the Italian downgrade
announcements and the stock market returns and this relation is especially referred to
specific downgrades.
The downgrades representing a transition from upper medium grade to lower medium grade
have had a significant impact on the Italian stock market. In addition, the first Italian
downgrade has had an important impact on the stock market returns, maybe because this
downgrade occurred in a context of general review of the sovereign credit rating so it
represented the alarm bells that the crisis was intensifying also in Italy. The first Italian
downgrade has expressed its impact, more specifically, on the sectors: fashion, finance, food
and beverage and health care. Only in the technology industry the first downgrade had no
effects, maybe because the high tech industry is more robust because it can be defined the
most innovative industry, and so it perceived the crisis more later.
References
Abad-Romero P. , Robles-Fernandez M.D. (2006) Risk and Returns Around Bond Rating
Change: New Evidence from the Spanish Stock Market; in Journal of Business Finance
& Accounting, June/July.
Afonso A., Furceri D., Gomes P. (2011) Sovereign Credit Ratings and Financial Market
Linkage. Application to European Data, European Central Bank Working Paper Series,
June.
Barron M.J., Clare A.D., Thomas S.H. (1997) The Effect of Bond Rating Changes and New
Ratings on UK Stock Returns, in Journal of Business Finance & Accounting, April.
Biglaiser G., Hicks B., Huggins C. (2008) Sovereign Bond Rating and the Democratic
Advantage Portfolio Investment in the Developing World, in Comparative Political
Studies, August.
Binder J.J. (1998) The Event Study Methodology Since 1969, in Review of Quantitative
Finance and Accounting, Vol. 11.
Brooks R., Faff R.W., Hillier D., Hillier J. (2004) The National Market Impact of Sovereign
Rating Changes, in Journal of Banking and Finance, Vol. 28.
Brown S. and Warner K. (1980) Measuring Security Prices Performance, in Journal of
Financial Economics, Vol. 8.
Brown S. and Warner K. (1985) Using Daily Stock Returns: the Case of Event Studies, in
Journal of Financial Economics, Vol.14.
14
Proceedings of 3rd European Business Research Conference
4 - 5 September 2014, Sheraton Roma, Rome, Italy, ISBN: 978-1-922069-59-7
Cantor R., Packer F. (1996) Determinants and Impacts of Sovereign Credit Ratings, in
FRBNY Economic Policy Review, October.
Cellier A., Chollet P. (2011) The Impact of Corporate Social Responsibility on Stock Prices:
An Event Study of Vigeo Rating Announcement, International Conference of the French
Finance Association (AFFI), May 11-13.
Crabbe L., Post M.A. (1994) The Effect of a Rating Downgrade on Outstanding Commercial
Paper, in The Journal of Finance, March.
Creighton A., Gower L., Richards A.J. (2007) The Impact of Rating Changes in Australian
Financial Markets, in Pacific-Basin Financial Journal, January.
Diamond D.W. (1991) Monitoring and Reputation: The Choice between Bank Loans and
Directly Placed Debt, in Journal of Political Economy, August.
Fama E.F., Fisher L., Jensen M., Roll. R. (1969) The Adjustments of Stock Prices to New
Information, in International Economic Review, 10.
Ferreira M.A., Gama P.M. (2007) Does sovereign debt ratings news spill over to international
stock markets?, Journal of Banking & Finance 31 (2007) 3162–3182
Gande, A. & Parsley, D. C.(2005), "News Spillovers in the Sovereign Debt Market," Journal of
Financial Economics, Elsevier, vol. 75(3), March.
Hill P., Faff R. (2010) The Market Impact of Relative Agency Activity in the Sovereign Rating
Market, in Journal of Business Finance & Accounting, November/December.
Kaminsky G., Schmukler S.L., (2002), Emerging Market Instability: Do Sovereign Ratings
Affect Country Risk and Stock Returns?, in Oxford Journal World Bank Economic
Review, Volume 16.
Klimavičienė A. (2011) Sovereign Credit Rating Announcements and Baltic Stock Markets, in
Organizations and Markets in Emerging Economies, Vol. 2, No 1 (3)
Kräussl R. (2005) Do Credit Rating Agencies Add to the Dynamics of Emerging Market
Crises?, in Journal of Financial Stability, April.
Larraìn G., Reisen H., von Maltzan J. (1997) Emerging Market Risk and Sovereign Credit
Rating, OECD Development Centre Working Papers No124
Micu M., Remolona E., Wooldridge P. (2006) The Price Impact of Rating Announcements:
Which Announcements Matter?, Bank for International Settlements (BIS) Working
Papers, No 207.
Norden L., Weber M (2004) Informational Efficiency of Credit Default Swap and Stock
Markets: The Impact of Credit Rating Announcements, in Journal of Banking & Finance,
November.
Pukthuanthong-Le, K., Elayan, F.A., & Rose, L.C. (2007). Equity and debt market responses
to sovereign credit ratings announcement. Global finance journal, 18, 47-83.
Steiner M., Heinke V.G. (2001) Event Study Concerning International Bond Price Effects of
Rating Actions, in International Journal of Finance & Economics, April.
15