In August 2024, shares in the maker of Pringles surged nearly 20% after Mars announced it was considering taking over the $27bn US food giant Kellanova to create a $60bn total revenue behemoth. Kellanova (NYSE: K), which spun out of Kellogg’s in 2023, produces snacks and packaged foods including the Pop-Tarts, Nutri-Grain and Special K brands. At the time of writing the deal was subject to regulatory approval and would be scrutinised by investors on many grounds, including ESG risks and opportunities. What are these risks, and how can AI tools such as Neural Alpha accelerate identifying and analysing them?
Perhaps most notably Mars is a private company with lesser mandatory reporting and disclosure obligations than the publicly listed Kellanova. However Mars have long been a leader in transparency in the food sector in voluntary disclosures with dozens of disclosures available on their website covering key material sector topics such as human rights, land use change, supply chains and emissions amongst other key topics. Indeed on deforestation, a key material risk for the sector, Mars have scored an impressive #6 ranking in the Forest 500 partly due to strong supply chain policies and disclosures - including publishing regional supplier lists for deforestation risk commodities such as beef, soya and palm oil. So overall a transition from public to private markets in this case may yield positive impacts in bringing Kellanova into Mars’ potentially stronger ESG framework.
How can we assess whether this is a likely outcome however? One approach is taking a top down view and examining ESG scores and other ratings such as CDP scores. Unfortunately given ESG scores predominantly cover public companies this comparison is challenging. LSEG appraises Kellanova with an ESG score of 76 / 100 (as of 9th August 2024) highlighting overall strong performance in most pillars but highlighting Governance weaknesses in litigation risks, auditor tenure, anti-takeover devices and Environmental risks in product recycling / takeback. Unfortunately as a private company no equivalent score or sub-ratings are available for Mars and only SDG alignment is measured by LSEG. So to do any comparison we have to examine raw data in the form of disclosures and third party sources. Whilst this can be challenging to do quickly, GenAI technology and Large Language Models (LLMs) are significantly accelerating the process of running ESG assessments and peer benchmarking. In this case we leverage the Neural Alpha Disclosure Assistant and Bulk Q&A features. Here is what we were able to discover in just a few hours of analysis.
Step 1: key facts
| Kellanova (NYSE: K) | Mars Inc | |
|---|---|---|
| Annual revenue | $12.980B (31 March 2024) | $45B (2022) |
| Legal Entity Identifier (LEI) | LONOZNOJYIBXOHXWDB86 | 54930073HKEO6GZBSS03 |
| Number of employees | 23,000 | 135,000 |
| Key products & segments | Cereal, snacks, frozen foods, beverages, plant-based foods | Petcare, Snacks, Food products |
| Example brands | Pop tarts, Pringles, Cheez-It, Kashi, Morningstar Farms | Pedigree Chum, Maltesers, Dolmio |
| Supply chain extent | Undisclosed, Kellanova operates in 20 countries | 30k suppliers in over 80 countries |
| Primary Standard Industrial Classification (SIC) | Cereal Breakfast Foods (2043) | Chocolate and Cocoa Products (2066) |
| Secondary Standard Industrial Classification (SIC) | Cookies and Crackers, Frozen Specialties, Not Elsewhere Classified, Frozen Bakery Products, Except Bread, Potato Chips, Corn Chips, and Similar Snacks, Macaroni, Spaghetti, Vermicelli, and Noodles (2051) | Candy, Nut, and Confectionery Stores (5441) |
Mars is a more diversified company than Kellanova, which specialises mainly in snacks and cereals. Although the business models of the two companies are distinct, frameworks such as SASB can be useful in identifying common ESG risks for the broader Processed Foods sector in which they both operate.
Step 2: specific risks and controversies
However, the aforementioned risks are very broad and largely theoretical. LLMs can perform a useful function in summarising more specific empirical risks, impacts and controversies companies are exposed to. Using a general knowledge model such as GPT-4o, we can rapidly identify more specific and commonly referenced controversies for both companies in the public domain:
Kellanova (NYSE: K): recent controversies
- Corporate restructuring (Kellogg split and impacts on job security)
- Ultra-processed foods / product health claims
- Worker safety, strikes and union disputes
- Unethical supply chain practices
- Packaging waste and environmental impacts of supply chain
- Labour practices
- Marketing ethics
- Market manipulation and pricing
- Global supply chain disruptions
- Diversity and inclusion
Mars Inc: recent controversies
- Child labour in cocoa production
- Palm oil sourcing and deforestation
- Plastic pollution
- Health and nutrition concerns
- Transparency in corporate social responsibility (CSR)
- Worker conditions
- Marketing ethics
- Supply chain labour rights
- Tax practices
- Animal testing
- Misleading advertising
- Russia involvement
- Human rights in supply chains
Cross-referencing this with real news stories in Neural Alpha News Screen, we can then measure risk exposure using news volume and sentiment for particular topics, as shown in the treemaps below. In news coverage, cocoa supply chain issues stand out for Mars, while Kellanova’s news volumes are more anchored around labour rights issues.
Step 3: disclosure analysis
We can also use Neural Alpha Disclosure Assistant to assess exposure and mitigation steps for particular issues of concern. For example, Mars has significant exposure to the Ukraine-Russia conflict. Will the Kellanova acquisition add to this exposure or ease it? We can use Disclosure Assistant to analyse Kellanova’s 96 financial and sustainability disclosures to answer this question. As the screenshot above shows, Kellanova completely exited Russia in July 2023, so acquiring Kellanova would reduce the share of Russia in Mars’ revenue mix.
Another capability within Neural Alpha is examining to what extent companies acknowledge risks identified by external parties, and what steps, if any, they are taking to address them. Interestingly, the two companies approach two widely recognised risks for their sector, ultra-processed foods and deforestation, very differently.
In written evidence to the UK Parliament, Kellanova made it clear that it does not believe the term ultra-processed food is helpful for consumer guidance, and it does not acknowledge this as a specific risk in its disclosures.
Nonetheless, Kellanova does recognise broader marketing risks and elevated compliance costs linked to unhealthy food sales.
Step 4: benchmarking
Uncovering the deforestation blind spot
LLMs are also very strong at peer benchmarking, another feature of the Neural Alpha platform available via our Assessment tools. Assessments can be run on an entire universe of disclosures or on individual disclosure types for more specific comparisons, which is very useful when comparing the ESG profiles of a target and an acquirer and analysing lengthy framework disclosures. Here we look at CDP Forests, for which Mars discloses a lengthy 235-page report against a 75-page disclosure for Kellanova. Analysing these line by line would normally take days, which is exactly the task for an ESG AI co-pilot.
Mars achieves a stronger BABB rating and Kellanova a B-CC rating in CDP Forests, but the reasons are not immediately obvious without delving into the detail of the disclosures. Using Neural Alpha’s AI Co-Pilot, we distilled the headline differences between the two companies:
- More of a commodity specific focus by Mars
- Mars’ traceability and risk management is more subnational in focus for commodities than Kellanova
- Considerably higher traceability and certification levels at Mars, across many more tiers of suppliers (up to tier 4 for palm oil)
- Considerably more ambitious target setting by Mars with 18 targets measured vs just 3 for Kellanova
- Mars is making significant use of satellite monitoring and geospatial analysis, whereas Kellanova’s disclosure suggests this is more limited
- Mars also uses a classification system to determine risk of deforestation and engages with a much wider group of governmental and non-governmental stakeholders and in landscape / jurisdictional approaches
- Mars is aiming for 100% responsibly produced cocoa by 2025, with significant progress in traceability and supplier engagement
- Mars has more detailed and higher-performing disclosures on deforestation- and conversion-free (DCF) volumes
Many of these findings are shown more clearly in the tables below, with key statistics extracted from each company’s CDP disclosures. Kellanova’s poor performance on cocoa, for which no traceability or certification processes are in place, is perhaps the most significant blind spot in the deal. Given the historical press coverage and controversies around Mars’ cocoa supply chain, this will undoubtedly be an area of focus for investors and others.
Mars: CDP Forests 2023
| Timber products | Palm oil | Cattle products | Soy | Cocoa | |
|---|---|---|---|---|---|
| Traceability system in place | Yes | Yes | Yes | Yes | Yes |
| % revenue associated with input | 71-80% | 21-30% | 41-50% | 21-30% | 31-40% |
| Supply chain mapping level | Partial | Full | Partial | Partial | Partial |
| Tiers of supply chain covered | 1,2 | 1,2,3,4 | 1,2 | 1,2 | – |
| Recycled content | 65% | N/A | N/A | N/A | N/A |
| % production traceable | 48.51% | 97% | 35% | 55% | 68% |
| % production certified | 92% | 100% | None | 12.40% | 68% |
Kellanova (NYSE: K): CDP Forests 2023
| Timber products | Palm oil | Cattle products | Soy | Cocoa | |
|---|---|---|---|---|---|
| Traceability system in place | Yes | Yes | N/A | None | None |
| % revenue associated with input | 81-90% | 6-10% | N/A | 31-40% | 6-10% |
| Supply chain mapping level | Partial | Partial | N/A | Partial | None |
| Tiers of supply chain covered | 1,2 | 1 | N/A | 1 | None |
| Recycled content | 100% | N/A | N/A | N/A | N/A |
| % production traceable | 99% | 32% | N/A | 99% | None |
| % production certified | 27% | 50% | N/A | None | None |
Figures extracted from 2023 CDP Forests reports. Where multiple traceability statistics under different regimes are available for the same commodity, we have taken the highest percentage per commodity.
Conclusion
Historically, analysis of this complexity would take an analyst several days to perform. However, using Neural Alpha, we were able to uncover salient differences between these two companies, and identify potential sustainability risks and opportunities. Both investors focused on these issues, or companies interested in assessing such risks and opportunities in the context of a strategic transaction, stand to benefit from leveraging Neural Alpha AI-enabled analytics.
Register for a free trial now at app.neuralalpha.com
