blog Market Intelligence /marketintelligence/en/news-insights/blog/50-years-of-altman-z-score-and-pd-model-fundamentals-case-study-general-motors content

Login to Market Intelligence Platform

New User / Forgot Password

Looking for more?

Contact Us

Request a Demo

You're one step closer to unlocking our suite of comprehensive and robust tools.

Fill out the form so we can connect you to the right person.

  • First Name*
  • Last Name*
  • Business Email *
  • Phone *
  • Company Name *
  • City *

* Required

In this list

50 Years Of Altman Z-score, And PD Model Fundamentals – Case Study General Motors

Ultra-Fast Broadband Services Remain A Niche Offering In Latin America


Power Forecast Briefing: Fleet Transformation, Under-Powered Markets, and Green Energy in 2018

Five Cheapest Triple-Play Bundles In Key Countries In Western Europe

Trading Of US Linear TV Advertising Shifting To Programmatic Trading

Credit Analysis
50 Years Of Altman Z-score, And PD Model Fundamentals – Case Study General Motors

Jun. 11 2018 — The year 2018 marks the 50th anniversary of the Altman Z-score, which was designed to gauge credit strength of publicly traded manufacturing corporates. Until this day, the model has been used by financial practitioners to obtain a condensed picture of the financial strength of a company, and serves as a benchmark for credit risk assessment models.

As a part of providing data and tools for a comprehensive analysis of credit risk, S&P Global Market Intelligence has developed a family of PD Model Fundamentals (PDFN). The PDFN is a statistical model that produces probability of default (PD) values over a one- to more than thirty-year horizon for public and private banks and corporations of any size. The model maps the PD values to credit scores1 (i.e. ‘bbb’), based on historical observed default rates (ODRs) extracted from S&P Global Ratings’ database (available on CreditPro® ) PDFN also offers a global coverage of over 250 countries and more than 20 segments, regions, and industries.

PDFN incorporates both financial risk and business risk to generate the overall PD value. This innovative approach captures, in a statistical PD model, important credit risk drivers as identified by S&P Global Ratings’ extensive experience in corporate credit assessments, and provides users with a well-rounded measure of credit risk, where different sources can be easily identified.

We apply the credit assessment metrics to analyze one of the most publicized bankruptcy events in the last decade, the case of General Motors (General Motors Company, formerly General Motors Corporation). In Figure 1 we present the historical evolution of credit risk for General Motors (GM) from January 2005 to May 2018, accompanied by bankruptcy related Key Developments. We compare assessed credit score by PDFN, Altman Z-score, and corresponding S&P Global Ratings Issuer Credit Rating.

At the beginning of 2005, PDFN indicates a credit risk score of ‘bbb-‘, while the S&P Global Ratings Issuer Credit Rating is ‘BBB-‘. The credit risk score indicates that General Motors had adequate capacity to meet its financial commitments. However, adverse economic conditions or changing circumstances are more likely to lead to a weakened capacity of the obligor to meet its financial commitments. Likewise, the Z-score indicates a rather problematic financial situation, placing General Motors in distressed zone category.

In the following months, the credit quality of General Motors rapidly deteriorated. PDFN signals highly increased probability of financial distress already at the beginning of 2007, more than two years in advance. The implied ‘ccc’ credit score suggests high vulnerability to adverse business, financial, or economic conditions with at least a one-in-two likelihood of default. A few months before default, PDFN indicates a credit score of ‘cc’, thus expecting default to be highly likely. Similarly, the S&P Global Ratings Issuer Credit Ratings shows decaying credit quality, albeit the credit rating changes are more sporadic and have larger increments. The Z-score starts to show a significant deterioration of credit quality one year prior to default, but with a notable lag in comparison with PDFN.

After completion of the post-bankruptcy reorganization, creditworthiness of General Motors improved, and PDFN indicates a fairly stable credit risk profile with an implied score of ‘bbb’. In comparison, S&P Global Ratings Issuer Credit Rating initially shows a greater conservatism in light of the reorganization processes. Since then, the credit rating has improved steadily, converging with PDFN estimate. Z-score shows a somewhat steady estimate of credit risk, with a slight deterioration in the recent years.

Figure 1: Historical evolution of credit risk for General Motors (GM)

The shaded area denotes the period of reorganization between the bankruptcy announcement and reemergence of General Motors (GM) as a public company on the New York Stock Exchange (NYSE). Dashed vertical lines denote bankruptcy related Key Development (see corresponding numbers for details). The Z-score scale has been selected to match the credit score level at the beginning of the period.

Source: S&P Global Market Intelligence (as of May 30th, 2018). For illustrative purposes only.

General Motors (GM) – Key Developments:
(1) Nov 8, 2008: GM heads towards bankruptcy
(2) Dec 31, 2008: GM expects to receive $13.40 billion in funding from U.S. Department of The Treasury.
(3) Feb 14, 2009: GM contemplates bankruptcy
(4) Jun 1, 2009: GM filed for bankruptcy
(5) Nov 17, 2010: GM has completed an IPO and starts trading on NYSE

PDFN incorporates both financial and business risk dimensions to generate an overall PD value as well as an assessment of each individual dimension (financial and business risk). It also comes equipped with a useful analytic tool, the contribution analysis, which allows users to identify drivers of risk, in absolute or relative terms, to define potential paths to creditworthiness improvement or deterioration.

Figure 2 presents the current credit risk profile of General Motors as provided by the PDFN based on last twelve months of data. The contribution analysis indicates that overall business risk is strong, but the company’s financial position is aggressive and is currently the main driver of overall PD estimate. A deep dive analysis shows a weak total equity position which in addition to profitability (EBIT/Total Assets) and efficiency (EBIT/Revenues), resulting in limited financial flexibility (Retained Earnings/Total Assets), represent the risk factors with the largest driver for the assigned credit risk score for General Motors.

Figure 2: Credit risk profile of General Motors (GM)

Source: S&P Global Market Intelligence (as of May 30th, 2018). For illustrative purposes only.

This case study exemplifies the value of PD Model Fundamentals, in providing predictive insights into companies’ creditworthiness and dynamic estimates of PD value and mapped credit score. Our model was trained and calibrated on default flags and is able to signal deterioration of credit quality well in advance of the actual bankruptcy event. The combination of both financial risk and business risk enables a comprehensive overview of a company's creditworthiness, while also providing an in-depth review of a company's credit risk profile to identify and distinguish the main sources of risk. S&P Global Market Intelligence leverages leading experience in developing PD models to achieve a high level of accuracy and a robust out-of-sample model performance. The integration of PDFN into the S&P Capital IQ platform allows users to access a global pre-scored database with more than 45,000 public companies and almost 700,000 private companies, obtain PD values for single or multiple companies, and perform a scenario analysis.

1 S&P Global Ratings does not participate in the creation of credit scores generated by S&P Global Market Intelligence. Lowercase nomenclature is used to differentiate S&P Global Market Intelligence PD credit model scores from the uppercase credit ratings issued by S&P Global Ratings.

Companies And Sectors Most Impacted By U.S.-Chinese Tariffs

Learn More

Credit Market Pulse March 2018 Issue

Learn More

Webinar Replay: Outlook On Credit Markets And The Implications For Systemic Risk

Learn More

Technology, Media & Telecom
Ultra-Fast Broadband Services Remain A Niche Offering In Latin America


The following post comes from Kagan, a research group within S&P Global Market Intelligence.

To learn more about our TMT (Technology, Media & Telecommunications) products and/or research, please request a demo.

Oct. 18 2018 — Increasing availability of triple-digit advertised speeds, elimination of lower speed packages, free broadband speed upgrades and migration from copper to fiber are paving the way to faster broadband speeds in five major Latin American markets. However, 38% of broadband households still subscribe to packages with speeds under 10 Mbps.

Kagan analyzes reported fixed broadband speeds in the top five broadband markets in Latin America in 2017 — Argentina, Brazil, Chile, Colombia and Mexico — in terms of subscriber numbers. Broadband speeds are divided into three tiers, with the lowest tier defined as speeds below 10 Mbps, the middle tier as speeds in the range of 10 Mbps to 100 Mbps and any speed above 100 Mbps considered the highest tier. Our estimates are based on regulator data, which is used to determine the markets' reported speeds.

Broadband households with 10-100 Mbps speeds dominate in the five major Latin American markets, with 59.6%, or 34.1 million, of the 57.3 million subs. This tier is followed by households with speeds below 10 Mbps with a 37.9% share, while speeds above 100 Mbps remain a niche offering with only 2.5% of broadband households among the five chosen markets.

Diving deeper into each of the five markets, the majority of the broadband households in Brazil, Mexico and Chile have speeds in the 10-100 Mbps tier, while most broadband households in Argentina and Colombia have speeds below 10 Mbps. Chile has the highest percentage of highest tier households due to local operators migrating customers to higher speeds.

Out of the estimated 7.6 million residential fixed broadband households in Argentina, we estimate 53.5% or 4.1 million subscribers are in the lowest tier at year-end 2017. The middle tier closely follows this with 3.5 million subscribers, leaving households above 100 Mbps as the least penetrated speed with 11,695 subs as of 2017. In the lowest tier, 74% is made up of households with 6 Mbps speeds. Across the middle tier, most households subscribed to speeds in the 10-50 Mbps range. Almost all of the broadband households in the highest tier have 150 Mbps speeds.

As for Brazil, we estimate 26.2 million broadband households are composed of 9.3 million in the lowest speed tier, 16.5 million in the middle tier and around 417,000 subscribers in the highest tier. Aside from being the largest ISP in terms of market share, Claro Brasil also dominates both the middle and highest tiers through its cable service, which operates under the brand name NET. In the middle tier, 15 Mbps is the most penetrated speed with 5.1 million customers, of which 2.7 million are from NET. A huge chunk of the highest tier is in 120 Mbps, with around 317,000 households subscribed to NET.

Despite being the smallest among the five markets, Chile has the most number of broadband households subscribed to speeds above 100 Mbps, as subscribers were recently migrated to higher broadband speeds, particularly from VTR. VTR increased its broadband speeds from 100 Mbps to 120 Mbps in February 2017 with no additional charges.

Unlike the rest of the five markets, dominant speeds per speed tier in Colombia lead by a wide margin, with 5 Mbps, 10 Mbps and 150 Mbps being the most popular contracted speeds in each tier. In the lowest tier, 5 Mbps is dominant with 58.9% or almost 2.7 million homes, around 1.4 million of which are Claro Colombia subscribers. Claro Colombia dominates the middle tier as well, with around 490,000 of its subscriber base in the 10-100 Mbps spectrum, concentrating 70.8% of the middle tier. In the highest tier, 98.8% or 914 households have 150 Mbps, the majority coming from ETB and a very few households from Azteca and Claro Colombia.

Mexico has the highest middle tier percentage among the five markets, with 78.4% or an estimated 11.7 million out of 15 million households subscribed to speeds from 10 Mbps to 100 Mbps. The lowest tier comes next with 19.7% or 2.9 million broadband homes, around 1.1 million of which are from the combined subscriber base of Grupo Televisa SAB's subsidiaries — Cablecom, Cablemás, Cablevisión México, Cablevisión Red SA de CV (Telecable), TVI and Sky México. The highest tier only has 2%, or around 288,000 households with speeds above 100 Mbps.

Average speeds: Ookla and Netflix

Average speeds reported by Netflix Inc. and Ookla should not be taken as the true speed measure of the mentioned markets. Netflix calculates the average prime-time bit rate used when streaming Netflix content across all end-user devices, regardless of the simultaneous internet activity performed on a single connection. Ookla, on the other hand, is limited to reporting broadband connections that actively performed speed tests on its platform.

Demand for higher broadband speeds is increasing as more people prefer to watch content over streaming services. According to Netflix's ISP Index, among the chosen five markets, Chile has both the highest country average and ISP speed.

In Argentina, cable operator TeleCentro had the highest average speed at 3.61 Mbps, while Telefónica Argentina's Speedy was at the lowest with 2.38 Mbps. TeleCentro's broadband plans range from 30 Mbps to 1 Gbps, causing it to reach higher average speeds compared to other providers. On the other hand, Speedy only offered plans from 3 Mbps to 10 Mbps in December 2017.

Netflix ranks TIM Participações SA as the ISP with the highest average speed at 3.12 Mbps in Brazil, while Telefónica Vivo falls at the bottom with 2.16 Mbps. Vivo Internet (DSL) and Vivo Fibra (fiber) were ranked separately but are both under Telefónica, which might have affected the average speeds ranking of Telefónica as a whole. Given the FTTH upgrades planned by Brazil's major operators, broadband speeds are expected to increase, driving penetration in the middle and highest broadband tiers.

GTD Internet's Fiber broadband ranks as the fastest ISP in Chile with 4.02 Mbps, which is also the highest ISP speed among the five major Latin American markets. Again, separating the average speed from its DSL counterpart might have affected the total average of the ISP. Fixed wireless provider Entel Chile, aside from having the lowest number of broadband subscribers in Chile among the operators analyzed, also ranks last in the average speeds with 2.45 Mbps as of 2017.

According to Netflix's ranking, Claro Colombia leads the average speeds only by a few points above ETB, with 2.99 Mbps and 2.9 Mbps, respectively. The lowest is DIRECTV Colombia's fixed wireless broadband service with speed offerings ranging from 2-10 Mbps in 2017.

Grupo Salinas' Total Play Telecomunicaciones SA de CV has the highest Netflix average speed in Mexico at year-end 2017, given its full fiber network. Average speeds from providers vary between 2-4 Mbps, except for Axtel's fixed wireless service Acceso Universal. However, separating the average with Axtel's Xtremo fiber service might have affected the company's overall average speeds ranking.

Although Ookla's Speedtest Global Index also shows Chile as the leader in average broadband speeds, it lists Colombia as the market with slowest average speed.

The availability of faster broadband speeds is also attributable to the increasing number of operators deploying their own fiber networks. After the initial trend of fiber rollouts, Kagan forecasts FTTH revenue in the Caribbean and Latin America is set to experience staggered but stable growth until 2022 as ISPs hold off spending until high-end fiber networks become more available, hence incurring lower expenses.

Despite the availability of broadband packages with higher speeds, reported speeds reached remain low. As internet service providers are slowly getting past the hurdles of network upgrades and deployments, the challenge is how to come closer to the maximum advertised speeds in their broadband plans.

Global Multichannel is a service of Kagan, a group within S&P Global Market Intelligence's TMT offering. Clients may access the full article, with detailed breakdown of speed tiers and Netflix ISP average speeds per country, as well as year-end 2017 data available in Excel format, by clicking here.

Learn more about Market Intelligence
Request Demo

Watch: Power Forecast Briefing: Fleet Transformation, Under-Powered Markets, and Green Energy in 2018

Steve Piper shares Power Forecast insights and a recap of recent events in the US power markets in Q4 of 2017. Watch our video for power generation trends and forecasts for utilities in 2018.

Five Cheapest Triple-Play Bundles In Key Countries In Western Europe


Western Europe’s biggest markets represent some of the region’s cheapest triple-play (TV, broadband and fixed telephony) bundles.

Germany, France and the U.K. are the three biggest media markets in Europe by revenue and by total GDP, yet they host some of the cheapest bundles

Scale and the intense nature of competition in those markets are key factors affecting pricing.

Oct. 12 2018 — Western Europe’s biggest markets represent some of the region’s cheapest triple-play (TV, broadband and fixed telephony) bundles. Germany, France and the U.K. are the three biggest media markets in Europe by revenue and by total GDP, yet they host some of the cheapest bundles. Scale and the intense nature of competition in those markets are key factors affecting pricing. Austria is the only nation featured in our cheapest bundles table that is not one of the big five markets in Western Europe. Bundling, incidentally is not popular everywhere: in the Nordics – Europe’s most advanced and competitive sub-region – operators are less inclined to bundle their products with the lack of interest in fixed telephony impacting bundling strategies.

Learn more about Market Intelligence
Request Demo

Technology, Media & Telecom
Trading Of US Linear TV Advertising Shifting To Programmatic Trading

Oct. 08 2018 — Both buyers and sellers of traditional linear TV advertising, not including connected TV or over-the-top video, are moving toward the adoption of programmatic trading. In 2017, Kagan estimates that $690 million or 0.9% of total linear TV spend was traded programmatically. Within the next five years, that figure is expected to climb to $9.76 billion or nearly 12% of total linear TV advertising revenue. MVPDs are forecast to trade the greatest percentage of their ad inventory programmatically in 2022 with 30% of ad revenue from programmatic trading.

Kagan defines programmatic trading as being automated and data-enhanced, not just one or the other. Trading may be through a private or open marketplace and does not have to be through an auction, which is more common in digital video advertising.

There are several issues holding participants back from programmatic trading. Unlike digital programmatic marketplaces, where there is a seemingly unending supply of ad inventory, linear TV has a finite supply. Demand for TV inventory exceeds the supply, so there is still an attitude of "If it isn't broken, don't fix it." TV ads are also bought well in advance, not immediately.

While many agencies have experimented with the programmatic trading of linear TV, not all are on board. Many of the advertisers and agencies are interacting directly with the supplier platform rather than going through a demand-side platform, or DSP, today. In their experiments, the agency needs to use separate platforms to aggregate inventory and tie it together, which is a lot of work.

The lack of inventory is one factor holding back programmatic trading. The only way it takes off is to make linear TV inventory available in some type of buyer platform that can combine the various supply platforms. It is even more complicated when the buyer wants to bring in connected TV (OTT).

Agencies do like the automation capabilities of programmatic, particularly where the process takes a lot of time. An algorithm may do better in areas such as weighting estimation, the first pass at scheduling and the negotiation process as well as postings and billings. The process of buying inventory is not difficult, but computing where a buyer will be able to find its preferred audience is. Therefore, interest in automating the planning and analysis to find an optimal audience is high.

We forecast a gradual uptake for programmatic trading with continued testing in 2018. Broadcast stations and networks, cable programmers, and MVPDs need to add more inventory to programmatic platforms before agencies begin using it in earnest. It will take time for all parties to feel comfortable transacting in a new way.

Learn more about Market Intelligence
Request Demo