GaWC Research Bulletin 115

GaWC logo
  
 
  Gateways into GaWC

(A)

The Global Geography of Credit Ratings in Past and Present: Exploratory Steps

L. Nikitin* and A. Romashov**

Abstract

The business of credit ratings is one of the most relevant to the development and functioning of world economy in the age of globalization. At present credit ratings activity plays a vital role in providing a rationale for decisions of transnational investors. The system of credit ratings can be evidently regarded as an informational component and a pillar of the infrastructure of financial flows. Thus, it seems clear that understanding of an inter-systemic correlation between the universe of world cities and activities of credit rating agencies is very important for proper characterizing of a state and trends in the former entity. That is why we intend to compare some types of credit rating information and GaWC data concerning world cities.


1. INTRODUCTION

It would be an exaggeration to say that rating industry and globalization always develop in strict parallels, but their mutual influence is quite obvious. Since its earliest days bonds market was a mechanism of distant operation: private investors could live thousands of miles away from places where a project was physically implemented, while numerous offices in multitude of cities maintained financial flows in various directions. Distant investors urgently needed exact and reliable information, preferably in a simple and standardized form. "The investor's right to know": these famous words of Henry Varnum Poor became a motto of rating business, which turned into essential part of constantly expanding networks. Now global market of debt securities is gigantic transnational entity spreading over dozens of countries and all the continents; its present outstanding amount, according to the most acknowledged valuations (produced by Bank for International Settlements (BIS), Merrill Lynch and some other institutions, is close to 40,000 billions of US dollars (e.g. BIS 2003). In other words, it is approximately four times larger than gross domestic product of the USA in 2002.

Certainly, there are many other aspects of globalization, besides fixed income market and its infrastructures. Since the early 1990ies innovatory works by Anthony D. King (1991) and especially Saskia Sassen (1991, 1994) drew researchers' attention to the global inter-city network where advanced producer services play a vital role. At the following stages Globalization and World Cities (GaWC) Study Group at Loughborough University began to measure inter-city linkages. Initially GaWC created multi-level roster of world cities and started matrix calculations (e.g. Beaverstock, Smith and Taylor 1999; Taylor 2001a; Taylor, Walker and Beaverstock 2002). Then Peter J. Taylor, Gilda Catalano and David R.F.Walker (2002a) produced fundamental 316x100 matrix (i.e. 316 cities and 100 firms in banking, accounting and other sectors; "GaWC 100") that enabled to compute global connectivity values for every city. 123 of them have connectivity from 0.2 to 1.0 that is at least one fifth of the most connected city (London). Variant ranges (from 123 to 316 cities) enabled employment of some statistical methods to general aggregate (e.g. Taylor, Catalano and Walker 2002b; Derudder, Taylor, Witlox and Catalano 2003) and focusing on several regions (e.g. Taylor 2001b; Brown, Catalano and Taylor 2002). So, GaWC studies clarify many important aspects of globalization.

In their turn, there are a lot of notable investigations concerning debt securities and rating business. First of all, it is necessary to mention that importance of the credit rating industry for proper development of the contemporary economy is highlighted by international conferences, which are related to the industry's problems. Among these research forums, the most representative and specialized ones were "International Conference on Credit Reporting Systems" (Miami, June 22 - 24, 2000) and "The Role of Credit Reporting Systems in the International Economy" (Washington DC, March 1-2, 2001) supported by The World Bank Group, University of Maryland Center for International Economics, New York University Stern School of Business and other prominent institutions.

Participants of the Washington conference stressed that "The recent wave of banking crises has led to a greater awareness of the centrality that a sound process of credit risk assessment has for bank stability worldwide"***. It is necessary to note that Credit Reporting Systems was referred here as embracing not only rating agencies, but also a wider range of institutions. That is to say that

"The quality of credit assessment depends . on a set of institutions that we label collectively as the credit reporting system and that belong to the following two categories: the credit reporting industry (Credit rating agencies, Credit registers, Credit bureaus) and the legal and regulatory framework in which credit reporting is taking place (prudential regulation, accounting, disclosure of loans with respect to insolvency, collateral and privacy)"***.

Materials of the conference contributed greatly to our research. For example, we can especially distinguish "A Historical Primer on the Business of Credit Rating" by Richard Sylla (2001), "Rating Agencies: Is There an Agency Issue?" by Roy C. Smith and Ingo Walter (2001), "The Credit Rating Industry: An Industrial Organization Analysis" by Lawrence J. White (2001) and several other papers.

Besides, in recent period many profound studies have been published in several universities and countries. For example, post-Soviet Russia joined fixed income markets not only in practice, but theoretically as well. In this connection works by Dmitry M. Mikhailov (2000) and Yuri A. Danilov (2002) should be mentioned in particular.

So, there are two powerful research flows existing in parallels; the task is to bring them closer. Certainly, we regard this paper as an attempt of just initial steps.

What are research-worthy points of contact between inter-city networks and rating industry? Firstly (and most obviously), this kind of advanced producer services has its own corporative geography. Institutional background of this sector and office networks of the top-raters are main subjects in the first part of this paper.

In the second part we consider mostly rated cities rather than rating firms. The very fact of credit rating assignment is an additional indicator of city's business activity and various linkages. Such rating is closely tied with transnational bonds offering, but not only; it also could be an important signal for more direct investment initiatives. Another basic type of information is level of rating judgment presented in symbolical form like AAA, A+, Baa1, etc. Such level depends on stability (or instability) of municipal budget. Budget, in its turn, depends on local industries, transportation nodes activity, households' consumption, etc; it also highly depends on numerous corporative offices that enable more or less directly city's involvement into the global network of advanced producer services. We can see how this integrated information is presented not in complex budget sheets and tables, but in the form of simple standardized marks, which are quite suitable for further comparisons and generalizations.

For calculations in both major sections of this study we used open information that is available on agencies' websites. We try to compare such information with GaWC data concerning the most connected cities.

2. RATING FIRMS

2a. Then and Now

A brief glance at the history of credit ratings will be very helpful in assisting us in this research. We can suppose that there are two main fundamentals for credit ratings development: the first is maturity of diversified bond market and the second is the presence of international multi-lateral capital market flows. That is why until the middle of the nineteenth century the quasi-international capital market lacked the system of credit ratings. The historical data show that the first European business projects were financed mainly through bank loans and so were even the hugest colonial initiatives. Also, the peculiarities of social structure of European states made it difficult to construct a sufficient layer of market-oriented independent private investors. Furthermore, the major share of the borrowed assets came from sovereign needs, and these kinds of issuers had no reasons to develop any system of credit ratings.

The situation entirely changed with the starting up of the first private international business initiative - the U.S. railroads. The outbreak of railroad building in the second quarter of the nineteenth century was accompanied by the substantial need for investments, which were almost impossible to obtain in the form of bank loans because of relatively small capitalization of American banking system and a riskiness of the enterprise. The only solution for this problem was a creation of a huge corporate bond market aiming toward resources of private investors, both domestic and international. By the words of R. Sylla "The corporate bond market, essentially a railroad bond market in its early decades, can be viewed as an American financial innovation that later spread to the rest of the world" (Sylla 2001). The author stresses that by the time John Moody began to rate bonds, the U.S. corporate bond market was several times lager than analogous market in any other country.

The further development of the credit rating industry was marked by emerging of the credit-reporting agencies (like R.G. Dun and Company) and specialized business/financial press on creditworthiness of companies (e.g. Henry Poor's Manual of the Railroad of the United States). The first credit rating agency in proper sense was founded in 1909 by John Moody who began his rating on railroad bonds. In 1916 the Poor Company shifted its business into bond rating and in 1941, after the merger with Standard Statistics, the Standard & Poor's agency (S&P) was formed. It is interesting that since those times and by now these agencies still form the core of the world rating industry. It could be also noted that outside the USA the Japanese ratings agencies are among the oldest and most active ones. The answer to a wide spread concern about such a limited amount of top-raters is easy to convey: while a share market is accompanied by sharp increase or decrease of profitability of owned assets, the bond market has a fixed perspective of the expected profit, and, in opposite to variety of stock advisers on future of exact share, the only purpose of bond ratings is to assure, whether the issuer of the exact securities will default or not. In other words, market actors want just one opinion on their bond assets, and the less the number of bond ratings, the more confidence in their evaluation.

By the time, the set of financial instruments that were subject of attention of credit rating agencies has enhanced greatly. Besides issues, such as different forms of commercial securities, it includes today ratings of sovereign and municipal securities and, most important, rates of companies, states and cities that are taken as an integrated entities. For example, since 1998 Moody's have published its ratings that assess the creditworthiness of a firm, even if the company has no outstanding public debt. These issuer ratings reflect Moody's opinions on an entity's ability to meet its senior financial obligations. In other words, credit ratings have become the universal "language" for international community, that, by the means of short codes, disseminates information on creditworthiness and, in a broader scale, on financial, social and other related characteristics of an exact actor of global economy.

Historically, development of credit ratings faced the most powerful challenge in the 1960ies when economic prosperity and consequent drop in default rate significantly lowered the demand for ratings from business community. Even the major agencies employed up to dozen analysts. The new turn for credit rating industry occurred in the 1970ies when globalization of international capital flow entered the new period of its development. Some experts believe that there is a parallel between the 1970ies and the second quarter of the nineteenth century, the time of the first global movement of speculative capital. Another powerful impact on the demand for credit rating services was an effect of deteriorating of international economic condition, which took place in the 1970-80ies. One of the most curious but nonetheless important factor was a wide dissemination of photo-copier technologies that prompted a shift of a fee burden from investors, who previously were obliged to buy published ratings, to issuers. By now, the most data of credit rating information are public and free of charge, but issuers have to pay a little share of their offerings as agency profit (put aside a case of unsolicited ratings).

In addition, it was in 1975 when the U.S. Securities and Exchange Commission (SEC) began the practice of designation of bond rating firms as "nationally recognized statistical rating organizations" (NRSROs). The first designated agencies were Moody's Investors Service, Standard & Poor's Rating Services and Fitch Inc. The Commission subsequently granted Duff & Phelps (1982) and McCarthy, Crisanti & Maffei (MCM) (1983) as NRSROs (MCM was absorbed by Duff & Phelps in 1991), and designated IBCA (1991) and Thomson BankWatch (1992) as NRSROs for banks and financial institutions. Since then - until the very recent times - no other agencies have received SEC designation, whether they are foreign or domestic ones. So, the IBCA's inability to expand its service beyond bank ratings was a major factor determining IBCA's purchase of Fitch in 1997. In 2000 the SEC's list became even shorter when Duff & Phelps and Thomson BankWatch were taken over by Fitch IBCA (now Fitch). It was only in February 2003 that the SEC granted NRSRO status to Toronto-based Dominion Bond Rating Agency, thus enlarging the list again.

Some other countries have a greater number of credit rating agencies, which assessments are recognized for regulatory purposes. For example, Great Britain officially recognizes data from ten agencies, Netherlands and France - nine, while authorities of Luxembourg and Sweden recognize just three credit rating agencies. It is quite surprisingly that Germany does not use rating data in its financial regulation. In the most of the countries their national systems of official recognition were set up in the 1990ies. Historically, none of the agencies was deleted from the recognition lists by the reasons except from merger or acquisition. It may be taken as another proof of the unique market position of credit rating agencies, symbolizing how closely their earned profit is linked to deserved reputation.

We can mention that the USA has the longest history of operating rating information on behalf of government regulatory purposes. For example, in 1931 banks were obliged to inform market about bonds with low credit ratings, in 1936 they were prohibited from buying any securities rated below exact grade. Expanding of credit rating products led toward a rule, introduced in 1999, enabling every U.S. bank open its financial subsidiary only if it met certain credit rating criteria. Internationally, the recognition of worthiness of rating practice was revealed in proposal by the Bank for International Settlements (BIS) Basel Committee on Banking Supervision (June 1999) to include borrowers' credit ratings in assessments of the adequacy of banks' capital.

At present, there are almost 150 rating agencies around the world, and analysts suggest that this number will grow further, especially in less developed markets, while the global agencies will continue to shrink. Thus, distinction has to be made between national agencies, regionally targeted agencies and globally operating agencies. For example, Capital Intelligence focuses on Central/Eastern Europe and most of the Swedish credit raters are concerned over their domestic companies. By contrast, S&P, Moody's and Fitch are internationally recognized credit rating agencies, while some of the Japanese agencies are recognized globally, but lack domestic recognition. As for the global trade organization for the rating industry, there are no such institutions except the ASEAN Forum of Credit Rating Agencies (AFCRA) that was formed in 1995 by rating agencies from Indonesia, Malaysia, the Philippines and Thailand. It is believed that similar trade organization may be set up soon in Germany.

The influence of credit rating agencies on international market is enormous. In 1996 Thomas Friedman, a journalist, said: "There are two superpowers in the world today in my opinion. There's the United States and there's Moody's Bond Rating Service. The United States can destroy you by dropping bombs, and Moody's can destroy you by downgrading your bonds. And believe me, it's not clear sometimes who's more powerful". While holding something of overvalues, this statement clearly reflects the current role of credit ratings in contemporary economy. But we are sure that, for ill or for good, credit rating agencies can make our period of economic imbalances and shocks a little more "creditworthy" and predictable.

2b. Their Own Networks

It seems obvious that while operating data on global rating agencies we have to clarify their own network system. For the purpose of present exploratory research the "Big Three" of global rating agencies were selected. It includes Standard & Poor's, Moody's and Fitch. This universe of offices of the major raters builds a vital infrastructure for proper functioning of rating service worldwide. No wonder that because of similarity of rating business to other forms of seamless services, rating activity is primarily concentrated in world cities. Nevertheless, some important peculiarities of this business have made the network of global rating agencies rather unlike to the system of corporative networks in other sectors.

First of all, in comparison to insurance or banking service, rating agencies are not required to establish their branches and subsidiaries in every city, even if an exact city plays a significant role in regional or national economy. It is sufficient to the most of the raters to operate several "watching" offices aimed on a wider geographical area. This case is especially actual for offices opened in 1980-2000ies, while traditional locations of agencies, especially in the USA, were denser. The main factor of this difference is the spreading of informational technologies, which made it possible in many cases to collect, analyse and disseminate information distantly. For example, almost 30 percent of the Fitch regional offices are concentrated in cities outside the Top-123 of the GaWC ranking, and 46 percent of them are the U.S. cities, while in Japan and China this agency has just one office per country, which are placed in the national capitals.

That is why, while analyzing structure of rating agencies network we found it useful to probe correlation between GaWC level of global connectivity and presence of offices of global raters. The basis for such comparison is the set of computed data for 123 cities grouped according to their GaWC ranking: 1.0 - 0.5; 0.49 - 0.4; 0.39 - 0.3; 0.29 - 0.2 (besides this section, the suggested gradation will be used in further studies - see Part II). The research process includes summing up numbers of offices presented in these cities that enables to derive the total number for each group. In parallel it is necessary to collect data on exact quantity of cities in these groups where three, two or one office of global raters are presented. It will give us both the quantitative and qualitative measures of rating network and provide a verified material on correlation between values of world cities' global connectivity and the presence of global rating agencies.

The results of the examination are aggregated into a new data set, which enables us to provide a reasonable argumentation for describing the universe of rating agencies. Collected information on distribution of offices through the network of world cities and density of offices of different raters in each major stratum are shown in generalized table on studied formations (Table 1).

Under ideal condition, suppose that each of the three agencies are presented in every city, the total of examined 123 cities would give us 369 offices theoretically possible. The real number collects just 85 offices for 48 cities. Thereafter, 85 reflect 23 percent from 369, while 48 make up 39 percent from 123. Thus, the ratio between cities that have offices of international rating agencies and those that have not, is even smaller as compared to rated / not rated ratio that will be considered further.

The distributed data on separated levels of GaWC connectivity reveal some other outcomes. While comparing the ratio between the real numbers and ideal conditions depicted above in each of the selected groups, we found strong correlation between GaWC ranking and office presence of global raters (Figure 1a). Also, the correlation occurs for data on number of offices of different raters presented in exact cities and GaWC ranks of connectivity (Figure 1b).

Although density of raters' office network is predictably lower than in other advanced service sectors, both global patterns are rather similar. The most connected cities mainly confirm their outstanding status when we consider the world map of rating business; in addition, such situation emphasizes reliability of the previous GaWC data. New York and London are especially important as places where top-raters' headquarters operate. Maximal score of three offices also can be seen in cases of Toronto, San Francisco, Sao Paolo, Frankfurt, Paris, Milan, Tokyo, Hong Kong and Singapore. Each of these cities has connectivity 0.5 or higher. In the lower strata just three cities (Mexico City, Buenos Aires and Seoul) attract all the three considered agencies: it also could be interpreted as a higher role of these cities in rating industry than in banking, advertising and other advanced producer services.

To illustrate these data graphically we use the patterns of clusters map "Global Connectivity of the Major Nodes in the World City Network" (Taylor, Catalano and Walker 2002a: Figure 2). The cities where offices of world raters are located can be highlighted on the "background" images of connectivity. Laying on these maps together will produce initial, but nonetheless exact, scheme of world rating in connection with the universe of world cities (Figure 2) that could serve as a preliminary database for further exploration.

3. RATED CITIES

3a. General Preliminaries

In this part of our research we intend to compare basic rating information and GaWC measuring data concerning 123 cities that have global connectivity from 0.2 to 1.0. All the following comparisons are based on rating judgments produced by three prime agencies: Standard & Poor's, Moody's and Fitch. The moment of observation is January 2003.

Among this range of 123 cities, only 29 are rated directly as issuers by at least one of three agencies. The most obvious addition can be formed by cities of specific administrative status. Indeed, in sphere of Mexican sub-sovereign bonds Distrito Federal can be regarded as synonym for Mexico City; City of Hong Kong could hardly be separated from Hong Kong Special Administrative Region, etc. As a result total number grows to 37.

In the USA (where municipal obligations are highly developed and often diversified for concrete purposes) the closest analogues to issuers' ratings are ratings for general obligation bonds. Such approach is not rare in agencies' own information and analytics. So we can justifiably add 23 more cities.

Some other problems can be solved via sub-sovereign territories that are also rated as issuers. We tried to use this way most accurately. For example, it is rather easy to do in case of Australia, where only states and Capital Territory, but not cities are rated. Besides, Melbourne is obviously predominates over state of Victoria; almost the same could be said about Sydney (New South Wales), Brisbane (Queensland), Perth (South Australia) and Adelaide (Western Australia). In Germany Stuttgart is undoubtedly leading political, economic and financial centre of Bundesland Baden-Württemberg (the biggest regional competitors, i.e. Mannheim, Heidelberg and Karlsruhe, are far behind the global Top-123). So in these and in some other cases extrapolation seems to be reasonable. But it is impossible to use credit rating of Nordrhein-Westfalen separately for Cologne or Düsseldorf (both cities belong to Top-123). Consequently, they are considered as not rated ones.

Finally, in 10 peculiar cases we find it possible to use sovereign ratings: Bermuda Islands (for Hamilton), Bahama Islands (for Nassau), Republic of Panama (for Panama City), Oriental Republic of Uruguay (for Montevideo), Grand Duchy of Luxembourg (for Luxembourg City), Republic of Lebanon (for Beirut), State of Kuwait (for Kuwait City), State of Bahrain (for Manama), Republic of Mauritius (for Port Louis) and Republic of Singapore (for Singapore City). In each of these cases we can see quite small state (no more than 3 - 3.5 million of population) with one indisputable centre of politics and finance. Certainly, these cities control national financial flows and play by far leading role in stability of sovereign budgets and credit ratings. Although these cities are not rated directly, it could be hypothetically connected with such peculiar proportions, but not only with isolation from international rating business. Nevertheless, in some cases we treated statistical contribution of such cities separately.

But even such provisos do not enable to use Israel's sovereign rating for Tel Aviv (which is partly counterbalanced by Jerusalem (its own connectivity is 0.045; No.283) and Haifa), Cyprus' rating for Nicosia (the second most connected city is Limassol: 0.133, No.190), Ecuador's rating for Quito (it is competed by Guayaquil, which is also among the GaWC Top-214) or Arab Emirates' rating for Dubai and Abu Dhabi (each of them has global connectivity over 0.2).

After all these extrapolations and filtrations we can observe 79 cities that are rated directly or indirectly (i.e. 64.23 percent of the general number, 123).

At the following stages of research process two types of information can be taken into account: (i) simple dichotomy of rated / not rated cities and (ii) level of rating judgments produced by agencies.

The latter type needs some other commentaries. Each rating firm uses its own system of marks, symbols and definitions; usually there are specialized and separated scales for issues and issuers, long-term and short-term obligations; also there are tables for bonds in local and foreign currencies, scales with national and regional prefixes, etc. For this study the most important information massifs are long-term issuers' credit ratings both in foreign and domestic currency. Such basic scales of leading agencies could be generalized for statistical and analytical purposes. AAA (Standard & Poor's), Aaa (Moody's) and AAA (Fitch) are conditionally interpreted as 24 points; AA+ (Standard & Poor's), Aa1 (Moody's) and AA+ (Fitch) as 23, and so on (Table 2). If any city is rated in parallel scales by one agency or by more than one agency, we use an arithmetical mean.

Despite some possible provisos, dichotomic information (rated / not rated) arithmetically depicts how much a city is involved into this sphere of transnational business, while the second type reflects stability or instability of urban finance.

Hereafter, two above-mentioned types of information are used and interpreted horizontally (for geographical mega-regions) and vertically (for strata of the most connected cities).

3b. Regional Distribution

In hypothetical demarcation of regional boundaries we followed some of previous GaWC studies (Taylor 2001b; Taylor, Catalano and Gane 2002) with just minimal differences that are due to differences in material itself. So the regions are the following:

  1. North America, including Bermuda Islands (28 cities with global connectivity 0.2 or higher);

  2. Latin America, including Caribbean states (12 cities);

  3. Europe, including Russia, Ukraine, Turkey and Cyprus (42 cities);

  4. North Africa / Middle East that includes African and Asian states with Mediterranean coast plus Arabia and Gulf states (10 cities);

  5. Sub-Saharan Africa that includes all African states without a Mediterranean coast (5 cities);

  6. South Asia that includes all states from Pakistan to Myanmar (6 cities);

  7. Pacific Asia that includes all Asian states with Pacific coast from Japan to Indonesia (13 cities);

  8. Australasia that includes Australia and New Zealand (7 cities).

(i) Rated / Not Rated

While we examine the most simple pairs of information (rated / not rated) the highest possible score for any city is 3 (if such city is rated by all of the three prime agencies); obviously, theoretical regional maximums are: 84 for North America, 36 for Latin America, 126 for Europe, 30 for North Africa / Middle East, 15 for Sub-Saharan Africa, 18 for South Asia, 39 for Pacific Asia and 21 for Australasia. The aggregated global score is 369. What are the real numbers?

Quite predictably we can see the highest score for North America: 73, i.e. 86.9 percent in comparison with possible regional maximum. Even excluding sovereign ratings of Bermuda (that could be used for Hamilton) we get 83.33 percent. The United States, which once were a cradle of rating business, have been the centre of this activity so far. Canada also makes its considerable contribution. Many Canadian cities are rated directly, including Toronto, Montreal, Calgary and Vancouver that belong to the global Top-123. It should be additionally noted that each of 10 Canadian provinces is rated by at least one prime agency; besides, aggregated data from Standard & Poor's (for general number of all issuers, public and private) show that Canada takes the third place just after the USA and Great Britain.

The second highest share is in the case of Australasia: 61.9 percent. Fitch is not very active there, but two other agencies (especially S&P) often show their rating judgments for sub-national issuers in Australia and New Zealand. (Moreover, in the above-mentioned aggregated data form S&P, which embrace public and private issuers, Australia takes the 7th place after the USA, Britain, Canada, Japan, Germany and France. New Zealand is the 15th: undoubtedly high position for country with rather small population and GDP.) But it is necessary to note that most of ratings in Australasia are produced not strictly for urban, but for territorial issuers as well.

Australasia is followed by Europe (46.03 percent; or 43.65 percent after exclusion of Luxemburg's sovereign ratings). There are rather surprising omissions, e.g. Amsterdam and Warsaw. The absence of national political and financial centre is contracted by presence of rating judgments for some other Polish cities: Bydgoszcz, Gdansk, Katowice, Krakow, Lodz, Poznan, Szczecin and Wroclaw. Nevertheless, European municipalities and sub-regions are increasingly active not only in western countries, but also in the former Communist block.

The next highest value (41.67 percent) belongs to Latin America, but this share instantly reduces to 22.22 percent after extraction of sovereign ratings (Bahamas / Nassau, Panama / Panama City, Uruguay / Montevideo). Even such global giant as Sao Paolo has no public ratings of its own. On the other hand, a very high activity of Mexican states and cities (although usually rated in national scale) could influence Brazil and other countries.

Region of North Africa / Middle East shows 30 percent value, but it is absolutely due to sovereign ratings (Lebanon / Beirut, Kuwait / Kuwait City, Bahrain / Manama).

For Pacific Asia the value is 23.08 percent; exclusion of sovereign ratings (Singapore) reduces it to 15.38 percent. In fact it could be interpreted as more high activity of Pacific Asia in comparison with the previous region. Nevertheless, some dynamically growing countries including China are little connected with global community of three prime agencies and their rating service.

Owing to sovereign rating of Mauritius (as applied to Port Louis) we can get 6.67 percent for Sub-Saharan Africa that practically has much in common with the absolute zero of South Asian region.

As a result only two regions, North America and Australasia, have figures higher than the global average 48.24 percent; the third one, Europe, is very close to it (Figure 3a). Also it could be illustrated via initial regional contributions (by number of highly connected cities) to the aggregated global amount and derivative contributions (by score of rating judgments for cities) to analogous amount. Such contribution grows in cases of North America and Australasia, remains practically the same in case of Europe and decreases for all other regions (Figure 3b).

(ii) Level of Rating Judgments

Using through scores for parallel scales and taking the highest mark as 24 (Table 2), we can get arithmetical means for cities and regions. New arrangement considerably differs from the previous one. Australasia turns into indisputable leader. Its weighted regional mean (computed directly by rating judgments) is 23.2; its arithmetical mean as calculated for cities is 23.021. Even the lowest urban scores (Auckland and Wellington, 22 each) are quite impressive. After Australasia, the following regions are: Pacific Asia (21.357, or 21.417 as computed via cities), North America (21.19; 21.26) Europe (20.495; 20.635), Sub-Saharan Africa (both 17.5), North Africa / Middle East (both 15.467) and Latin America (11.304; 12.529); certainly, South Asia maintains its zero. The global weighted mean is 19.86 (Figure 4); global average for 79 rated cities is 20.153.

Probably, in cases of Europe (29 rated cities, widely ranging from Istanbul's 9.75 to Paris' and Vienna's 24) and North America (28 cities, ranging from 16.5 to 24) such wide generalizations do not produce very valuable information. On the other statistical pole, value for Sub-Saharan Africa (as calculated for a single city) is not representative. But some other figures could be interpreted in quite reliable way. Sub-national issuers in Australasia and Pacific Asia are few, but strong and stable. In Latin America, especially in Argentina, analogous ratings are under mighty pressure of sovereign debt problems.

Comparing (i) and (ii) arrangements we can see just very weak (if any) connection between region's activity in rating industry and trustworthiness of regional issuers.

3c. Ratings and Connectivity

Values of inter-city connectivity, which were computed, analysed and interpreted in GaWC studies (e.g. Taylor, Catalano and Walker 2002a; Taylor, Catalano and Walker 2002b), enable some different modes of vertical stratification. Trying to combine strict gradation and quite large statistical massifs, we used the following version (as in the first section of this paper):

Stratum 1. Connectivity 5.0 - 1.0 (17 cities);

Stratum 2. Connectivity 0.4 - 0.49 (31 cities);

Stratum 3. Connectivity 0.3 - 0.39 (20 cities);

Stratum 4. Connectivity 0.2 - 0.29 (55 cities).

Like in the case of geographical regions, two types of information could be used and interpreted for analysis of these strata.

(i) Rated / Not Rated

Let's suppose again that any city, which is rated by all the three agencies, gets 3 conventional points; consequently, theoretical maximums for Strata 1 - 4 are 51, 93, 60 and 165.

In the elitist Stratum 1 there are 7 cities rated by three agencies, 4 cities rated by two agencies and 4 cities rated by a single agency. Also there are two absolute omissions: Amsterdam and Sao Paolo are not rated at all. So, the score is 33, i.e. 64.71 percent to theoretical maximum (51). After extraction of Singapore's sovereign ratings, this share diminishes to 58.82 percent.

The Stratum 2 contains no sovereigns; its share value of rating activity (47.32 percent) dramatically differs from analogous figure for the highest stratum. Among their general number (31), 8 cities are rated by three agencies, but 11 cities (Dublin, Warsaw, Johannesburg, Mumbai, Beijing, Shanghai, Manila, Bangkok, Kuala Lumpur, Jakarta and Taipei) are not rated by any of the prime firms.

In the Stratum 3 there are just 7 not rated cities. Share rises to 53.33 percent that is higher than the above-mentioned global average (48.24 percent) Nevertheless, it reduces to 43.33 percent after extraction of all sovereign issuers (Duchy of Luxemburg / Luxemburg City and Lebanon / Beirut).

The lowest Stratum 4 has also the lowest value of rating activity (41.82 percent) with 24 not rated cities. After exclusion of sovereigns (Bermuda / Hamilton, Bahamas / Nassau, Panama / Panama City, Uruguay / Montevideo, Kuwait / Kuwait City, Bahrain / Manama, Mauritius / Port Louis) it reduces even to 31.52 percent.

As a result, we can see quite a strong, although not absolute correlation between levels of global inter-city connectivity and cities' involvement into transnational rating business (Figure 5). Exclusion of sovereign issuers makes such correlation even stronger.

(ii) Level of Rating Judgments

Generalized scoring scheme from the Table 2, where the highest possible mark is 24, enables to produce the following values (Figure 6):

Stratum 1: 22.152 (weighted mean) and 22.222 (as calculated via cities);

Stratum 2: 18.576 and 18.992;

Stratum 3: 20.273 and 20.987;

Stratum 4: 19.441 and 19.547.

It should be noted again that the global weighted mean is 19.86 (Figure 4) and the global average for 79 rated cities is 20.153.

So, the Stratum 1 maintains its elitist character. Even the lowest individual figure (19.333 for New York City) is not very far from the global average figures.

Comparatively low index of borrowers' reliability in Stratum 2 needs some other commentaries. Such level is due mainly to the single issuer, which is Buenos Aires. Its calculated score (5.4) is very low even for rather unstable finance of Latin America as a whole. Argentinean economic drama is well known and described by many analysts in detail. In the 1990ies currency board system enabled strong peso and low inflation; national and sub-national credit ratings were at moderate 11 - 15 (as transferred to the same conditional scale). Nevertheless, lack of budget discipline and some other factors led to the crisis and default of 2001. Although improvements since then were "next to nothing", it could be assumed that current situation of extremely low ratings is not typical for potentially successful Argentina and its leading territories. Temporally excluding Buenos Aires from these calculations, we can get 19.656 (weighted mean) and 19.715 (average for other 19 cities) in the Stratum 2. As a result downgrading form stratum to stratum becomes a bit smoother, but remains slight, nevertheless.

Besides, the correlation between cities' connectivity and their credit ratings can be measured more precisely by another standard procedure, namely calculation of Pearson's linear correlation coefficient (Pearson r). In such case every city influences the general situation individually and does not dissolve in average meaning for this or that stratum. For all the 79 rated cities r = 0.149, which could be interpreted as a very slight, although positive correlation (Figure 7). Even after the removal of the most obvious and not typically rated city outlier (Buenos Aires) coefficient grows just to r = 0.180.

Such results are important for some conclusive generalizations.

4. CITIES BEYOND THE GLOBAL TOP-123

Although an accurate measurement for these urban and regional multitudes is not among the aims of the present study, a brief glance at such cities could be worthwhile for further comparisons and verifications. Certainly, there are a lot of other sub-national issuers rated by S&P, Moody's or Fitch, but territorial distribution of these issuers remains highly uneven. Just few of them can be seen on positions 124 - 214 in the widened GaWC's ranking (where global connectivity is lower than 0.2). "Black holes", as J.R.Short describes minimally connected or even isolated cities with huge populations (Short 2002), do not cease to be the same "black holes" when we observe the universe of credit ratings. It could be supposed that for many countries and mega-regions shares of rated cities will definitely reduce, if an investigation will also embrace cities beyond the Top-123.

As for levels of rating judgments, their spread is predictably immense. There are some first rank borrowers (e.g. Marseilles, Strasbourg, Bremen, etc.) whose marks are usually maximal and outlooks are stable. However, ratings like A, BBB and BB are more common for majority of other issuers. Again, precise measurements for these cities are outside the tasks of this exploratory study, but it could be validly presumed that average credit ratings for non-global cities are lower than those for the Top-123 as calculated in the previous sections of this paper. Such observations enable to see the 123 cities in focus more clearly.

5. CONCLUSIONS

As it have been noted in the first major section of this paper, there are certain relationships between corporative networks of the prime rating agencies, on the one hand, and the global network of advanced producer firms form the GaWC 100, on the other. Indeed, global cities attract producers of rating service, while leading agencies of such specialization contribute in their turn to the cities' connectivity and competitiveness.

Diverse information about rated cities (the second section) enables some other summarizations. Undoubtedly, national and regional business traditions play a vital role in today's rating practice. National and supra-national regulations, federative or unitary type of a state, governmental policies and grades of financial transparency are of great importance too.

Being initially a purely American enterprise, producing of credit ratings is highly concentrated in the United States until now. Canada, Australia and New Zealand are also well represented in this sphere.

Activity of European issuers is higher than average, but their geographical distribution is rather uneven. It is not only in Eastern Europe, but also in the EU countries that we can see some peculiarities. For instance, British corporative issuers traditionally are quite active, but in the field of municipal obligations United Kingdom is less visible than Italy and France. In Germany (unlike another federative state, Canada) only federal regions (Bundesländer), but not exactly cities are usually rated. Besides, countries of Eastern Europe (including Russia and some other CIS members) are behind the EU as a whole, but their involvement into rating business grows rapidly.

Participation of Latin America and Pacific Asia is even more irregular and fragmentary, but it is not those regions that stand last on the listing. Many African, Arabic and South Asian cities and countries are absolutely or almost absolutely outside the rating industry. Some recent studies and measurements (e.g. Taylor 2001b; Brown, Catalano and Taylor 2002; Short 2002) emphasize how much uneven the waves of globalization are; now we can see a confirmation based on other data and sources.

Examinations of cities' ranking by values of global connectivity can bring some other influences and correlations to clear light. Usually, the higher average connectivity in a certain group accompanies the higher share of rated cities, and their reductions are approximately parallel. It could be partly attributed to regional peculiarities (e.g. African cities do not have very high connectivity values; at the same time they are rather isolated from the rating industry), but it doesn't abolish this hypothetical trend.

While we can expect the highest rating activity at the upper rows of connectivity table, levels of rating judgments (i.e. financial stability grade) grow not so obviously: as it was mentioned above, Pearson's coefficient (r) is 0.149.

All these data could be interpreted in the following way: almost every city with considerable status in the global network (i.e. with connectivity 0.2 or higher), if it is not isolated form rating industry because of local business traditions or other factors, is really a global city possessing good credit ratings and stable finance of its own. Indeed, our "average city" gets 20.153 conditional points that is slightly higher than A+ according to Fitch or S&P and A1 according to Moody's scale. Usually the agencies describe such issuer as an obligor, which has "strong", although not "very strong" or "extremely strong" "capacity to meet its financial commitments". If connectivity grows from 0.2 to any other value, such change plays some strengthening role, but individual factors for this or that city are more important. If these preliminary assumptions are true, we can see some types of complex correlation between rating information and global connectivity.

Certainly, figures, correlations and conclusions could be rectified in a possible widened study on basis of information from many other agencies. However, as Standard and Poor's, Moody's and Fitch are indisputably elitist leaders in this sector of transnational business, the present calculations concerning these three top-raters reflect (as we hope) integrated entity of its own importance.

REFERENCES

Bank for International Settlements, BIS (2003), Quarterly Review: March 2003, Basel: BIS

Beaverstock J.V., Smith R.G., Taylor P.J. (1999) 'A Roster of World Cities', Cities, 16 (6), 445-458

Beaverstock J. V., Smith R. G., Taylor P. J., Walker D.R.F., Lorimer H. (2000) 'Globalization and world cities: some measurement methodologies', Applied Geography, 20, 43-63

Brown E., Catalano G., Taylor P.J. (2002) 'Beyond World Cities: Central America in a Global Space of Flows', Area, 34 (2), 139 - 148

Danilov Yu. A. (2002) Markets of Public Debt: Global Tendencies and Russian Practice, Moscow: VShE (in Russian)

Derudder B., Taylor P.J., Witlox F., Catalano G. (2003) 'Hierarchical Tendencies and Regional Patterns in the World City Network: A Global Urban Analysis of 234 Cities', GaWC Research Bulletin 88, www.lboro.ac.uk/gawc

Friedman J. (1986) 'The World City Hypothesis', Development and Change, 17, 69-83

King A.D. (1991) Global Cities: Post-Imperialism and the Internationalization of London, London: Routledge

Mikhailov D.M. (2000) The Global Financial Market: Trends and Instruments, Moscow: Examen (in Russian)

Sassen S. (1991) The Global City: New York, London, Tokyo, Princeton, NJ: Princeton University

Sassen S. (1994) Cities in a World Economy, Thousand Oaks, CA: Pine Forge Press

Short J.R. (2002) 'Black Holes and Loose Connections in the Global Urban Network', GaWC Research Bulletin 76, www.lboro.ac.uk/gawc

Smith R.C., Walter I. (2001) 'Rating Agencies: Is There an Agency Issue?', The Role of Credit Reporting Systems in the International Economy, Washington DC: World Bank Group, www1.worldbank.org

Sylla R. (2001) 'A Historical Premier on the Business of Credit Ratings', The Role of Credit Reporting Systems in the International Economy, Washington DC: World Bank Group, www1.worldbank.org

Taylor P.J. (2001a). 'Specification of the World City Network', Geographical Analysis, 33 (2), 181-194

Taylor P.J. (2001b) 'West Asian/North African Cities in the World City Network: A Global Analysis of Dependence, Integration and Autonomy', The Arab World Geographer, 4 (3), 146 - 159.

Taylor P.J., Walker D.R.F., Beaverstock J.V. (2002) 'Firms and their Global Service Networks' in Sassen S. (ed) Global Networks, Linked Cities, New York - London: Routledge, 93-115.

Taylor P.J., Catalano G., Walker D.R.F. (2002a) 'Measurement of the World City Network', Urban Studies, 39, 2367 - 76.

Taylor P.J., Catalano G., Walker D.R.F. (2002b) 'Exploratory Analysis of the World City Network', Urban Studies, 39, 2377 - 94.

Taylor P.J., Catalano G., Gane N. (2002) 'A Geography of Global Change: Cities and Services, 2000 - 01', GaWC Research Bulletin 77, www.lboro.ac.uk/gawc

White L.J. (2001) 'The Credit Rating Industry: An Industrial Organization Analysis', The Role of Credit Reporting Systems in the International Economy, Washington DC: World Bank Group, www1.worldbank.org

NOTES

* Leonid Nikitin, Chelyabinsk State Pedagogical University, Russia.

** Alexander Romashov, Chelyabinsk State Pedagogical University, Russia.

*** The World Bank Group, http://www1.worldbank.org/finance/html/credit_report_about.html>  


Table 1: World Cities Strata and Distribution of Top-Raters' Offices (Standard & Poor's, Moody's and Fitch)

Connectivity/Number of offices

1,0 - 0,5

0,49 - 0,4

0,39 - 0,3

0,29 - 0,2

Total offices

3

11

3

0

0

42

2

3

3

3

0

18

1

1

12

5

7

25

Number of cities

15

18

8

7

48/85

Sources: Standard &Poor's, Moody's and Fitch's websites; authors' calculations


Table 2: Rating Parallels

Score

Standard & Poor's

Moody's

Fitch

24

AAA

Aaa

AAA

23

AA+

Aa1

AA+

22

AA

Aa2

AA

21

AA-

Aa3

AA-

20

A+

A1

A+

19

A

A2

A

18

A-

A3

A-

17

BBB+

Baa1

BBB+

16

BBB

Baa2

BBB

15

BBB-

Baa3

BBB-

14

BB+

Ba1

BB+

13

BB

Ba2

BB

12

BB-

Ba3

BB-

11

B+

B1

B+

10

B

B2

B

9

B-

B3

B-

8

CCC+

Caa1

CCC+

7

CCC

Caa2

CCC

6

CCC-

Caa3

CCC-

5

CC

-

CC

4

C

Ca

C

3

-

-

DDD

2

SD

-

DD

1

D

C

D

Sources: Standard &Poor's, Moody's and Fitch's websites; authors' calculations


Figures 1a, 1b: Office Geography of Global Rating Agencies

Figure 1a: GaWC Ranking and Office Presence of Global Rating Agencies (in %) 

Figure 1a

Sources: Standard &Poor's, Moody's and Fitch's websites; authors' calculations

Figure 1b: GaWC Ranking and Number of Offices of Global Rating Agencies (per city) 

Figure 1b

Sources: Standard &Poor's, Moody's and Fitch's websites; authors' calculations


Figure 2: Global Connectivity of the Major Nodes in the World City Network and Office Presence of Global Rating Agencies 

Figure 2

The codes for cities are:
AB Abu Dubai; AD Adelaide; AK Auckland; AM Amsterdam; AS Athens; AT Atlanta; AN Antwerp; BA Buenos Aires; BB Brisbane; BC Barcelona; BD Budapest; BG Bogota; BJ Beijing; BK Bangkok; BL Berlin; BM Birmingham; BN Bangalore; BR Brussels; BS Boston; BT Beirut; BU Bucharest; BV Bratislava; CA Cairo; CC Calcutta; CG Calgary; CH Chicago; CL Charlotte; CN Chennai; CO Cologne; CP Copenhagen; CR Caracas; CS Casablanca; CT Cape Town; CV Cleveland; DA Dallas; DB Dublin; DS Dusseldorf; DT Detroit; DU Dubai; DV Denver; FR Frankfurt; GN Geneva; GZ Guangzhou; HB Hamburg; HC Ho Chi Minh City; HK Hong Kong; HL Helsinki; HM Hamilton(Bermuda); HS Houston; IN Indianapolis; IS Istanbul; JB Johannesburg; JD Jeddah; JK Jakarta; KC Kansas City; KL Kuala Lumpur; KR Karachi; KU Kuwait; KV Kiev; LA Los Angeles; LB Lisbon; LG Lagos; LM Lima; LN London; LX Luxembourg; LY Lyons; MB Mumbai; MC Manchester; MD Madrid; ME Melbourne; MI Miami; ML Milan; MM Manama; MN Manila; MP Minneapolis; MS Moscow; MT Montreal; MU Munich; MV Montevideo; MX Mexico City; NC Nicosia; ND New Delhi; NR Nairobi; NS Nassau; NY New York; OS Oslo; PA Paris; PB Pittsburg; PD Portland; PE Perth; PH Philadelphia; PN Panama City; PR Prague; QU Quito; RJ Rio de Janeiro; RM Rome; RT Rotterdam; RY Riyadh; SA Santiago; SD San Diego; SE Seattle; SF San Francisco; SG Singapore; SH Shanghai; SK Stockholm; SL St Louis; SO Sofia; SP Sao Paulo; ST Stuttgart; SU Seoul; SY Sydney; TA Tel Aviv; TP Taipei; TR Toronto; VI Vienna; VN Vancouver; WC Washington DC; WL Wellington; WS Warsaw; ZG Zagreb; ZU Zurich

Sources: GaWC; Standard &Poor's, Moody's and Fitch's websites; authors' calculations


Figure 3a, 3b: Rated and Not Rated Cities in the Global Regions

Figure 3a: Shares of Rating Judgement Scores (in comparison with theoretical regional maximum when every city is rated by three agencies) 

Figure 3a

Sources: Standard &Poor's, Moody's and Fitch's websites; authors' calculations

Figure 3b:  

Regional Distribution (by number of the most connected cities)

Figure 3b1

Regional Distribution (by number of rating judgements)

Figure 3b2

Sources: Standard &Poor's, Moody's and Fitch's websites; authors' calculations


Figure 4: Regional Weighted Means (by Level of Judgements) 

Figure 4

Sources: GaWC; Standard &Poor's, Moody's and Fitch's websites; authors' calculations


Figure 5: Shares of Rated Cities in Connectivity Strata 

Figure 5

Sources: GaWC; Standard &Poor's, Moody's and Fitch's websites; authors' calculations


Figure 6: Levels of Rating Judgements in Connectivity Strata 

Figure 6

Sources: GaWC; Standard &Poor's, Moody's and Fitch's websites; authors' calculations


Figure 7: Connectivity and Level of Rating Judgments 

Figure 7

Sources: GaWC; Standard &Poor's, Moody's and Fitch's websites; authors' calculations


Edited and posted on the web on 4th June 2003