K-means Classification and Comparable Company Analysis in the Drug Manufactures – General (Healthcare Sector)

In this article, we explore the valuation landscape of leading companies in the Drug Manufactures – General subsector using a combination of K-means clustering and comparable company analysis. By grouping companies with similar valuation metrics and growth characteristics, we can better understand market positioning and derive implied equity values that serve as useful benchmarks. Below, we outline our methodology, interpret our clustering results, and perform a detailed comparable company analysis for two representative stocks: AMGN and JNJ.

1. Introduction

The Drug Manufactures – General subsector is home to some of the largest and most innovative companies in the healthcare space. Our initial data set categorizes companies by market capitalization. For example, large-cap names such as LLY, JNJ, ABBV, MRK, AZN, and others are included in our final selection—with market capitalizations ranging from over US$784  billion for LLY to approximately US$20  billion for BIIB. This diverse group provides a rich universe for applying advanced statistical techniques to identify peer groups and extract meaningful valuation insights.

Figure 1. Company market capitalization.

2. K-means Clustering Methodology

To begin our analysis, we applied a K-means clustering algorithm to key valuation and growth metrics:

• EV/EBITDA

• EV/Revenue

• P/E Ratio

• Revenue Growth

Figure 2. Cluster analysis charts.

Figure 3. 3D bubble chart.

Ticker	Cluster Label	EV/EBITDA	EV/Revenue	P/E Ratio	Revenue Growth
LLY	Cluster 2	46.52	18.86	86.95	0.20
JNJ	Cluster 0	12.38	4.24	26.24	0.05
ABBV	Cluster 0	14.61	6.74	79.48	0.04
MRK	Cluster 0	12.04	4.37	20.86	0.04
AZN	Cluster 0	7.27	2.61	33.64	0.18
NVS	Cluster 0	10.86	4.56	17.97	0.09
AMGN	Cluster 0	16.88	6.35	36.91	0.23
PFE	Cluster 0	11.26	3.48	34.96	0.31
SNY	Cluster 0	6.60	1.70	22.67	0.12
SAN.PA	Cluster 0	11.57	2.99	22.62	0.12
GILD	Cluster 1	9.94	4.90	1088.78	0.07
GSK	Cluster 0	16.68	4.93	23.07	-0.02
BIIB	Cluster 0	8.91	2.67	12.87	-0.03

Table 1. K-means clustering analysis.

Interpretation:

• Cluster 0: Encompassing 11 companies, this group exhibits moderate EV/EBITDA (mean ≈ 11.73) and EV/Revenue multiples (median ≈ 4.24) with modest revenue growth.

• Cluster 1: Consisting solely of GILD, this cluster is an outlier with an extraordinarily high P/E ratio, suggesting unique market conditions or one-off factors driving its valuation.

• Cluster 2: Represented solely by LLY, this group shows a premium valuation environment—with high EV/EBITDA and EV/Revenue multiples alongside strong revenue growth (20%).

By segmenting the universe, investors can focus on comparable peers that share similar fundamental characteristics.

3. Cluster Insights

Our broader cluster analysis provides further quantitative context.

Cluster	Number of Companies	EV/EBITDA Mean (Median)	EV/Revenue Mean (Median)	P/E Ratio Mean (Median)	Revenue Growth Mean (Median)	EBITDA Margins Mean (Median)
0	11	11.73 (11.57)	4.06 (4.24)	30.12 (23.07)	0.10 (0.09)	0.34 (0.34)
1	1	9.94 (9.94)	4.90 (4.90)	1088.78 (1088.78)	0.07 (0.07)	0.49 (0.49)
2	1	46.52 (46.52)	18.86 (18.86)	86.95 (86.95)	0.20 (0.20)	0.41 (0.41)

Table 2. Cluster insights.

Key Takeaways:

• Cluster 0 is the most representative group, with companies trading at relatively lower valuation multiples and showing steady revenue growth.

• Cluster 1 (GILD) stands apart with its atypically high P/E ratio, which may be indicative of special circumstances or market sentiment.

• Cluster 2 (LLY) is characterized by premium valuation multiples and robust revenue growth, reflecting high investor expectations and possibly superior growth prospects.

These insights form the foundation for a more granular valuation exercise using implied valuation multiples.

4. Implied Valuation and Comparable Company Analysis

With companies now grouped by similar valuation profiles, we next derive implied enterprise and equity values using three valuation multiples: EV/EBITDA, EV/Revenue, and P/E Ratio. The resulting valuations (Data4) provide a framework for comparing market-implied values against current trading prices.

Snapshot of Implied Valuations

For each ticker, we compute:

• Implied EV (and Implied Equity Value) using EV/EBITDA

• Implied EV (and Implied Equity Value) using EV/Revenue

• Implied Equity Value using P/E Ratio

For example:

• LLY (Cluster 2):

  • Implied EV (EV/EBITDA): ~770.7  billion

  • Implied Equity Value (EV/EBITDA): ~743.0  billion

  • Similar valuations are mirrored when applying the EV/Revenue multiple.

• JNJ (Cluster 0):

  • Implied EV (EV/EBITDA): ~347.8  billion

  • Implied Equity Value (EV/EBITDA): ~332.3  billion

Some tickers (e.g., MRK.PA, BMY) show gaps (NaN values) due to incomplete data; however, the bulk of our sample yields robust estimates that enable consistent peer comparisons.

5. Comparable Company Analysis: AMGN vs. JNJ

Focusing on Cluster 0, we perform a detailed comparable company analysis on AMGN and JNJ. Both companies are members of Cluster 0, making them suitable peers for valuation comparison.

Implied Equity Value Per Share Calculations

AMGN

• Shares Outstanding: 537,532,992

• Implied Value/Share:

  • EV/EBITDA: $168.08

  • EV/Revenue: $161.24

  • P/E Ratio: $181.57

• Current Share Price: $289.02

JNJ

• Shares Outstanding: 2,407,620,096

• Implied Value/Share:

  • EV/EBITDA: $138.04

  • EV/Revenue: $148.15

  • P/E Ratio: $141.52

• Current Share Price: $153.49

Interpretation

• AMGN:

  The implied per-share values across different multiples for AMGN range from approximately $161 to $182. In contrast, its current share price is significantly higher ($289.02), suggesting that the market may be pricing in a premium—possibly due to anticipated growth, operational efficiency, or strategic initiatives that are not fully captured by traditional multiples.

• JNJ:

  For JNJ, the implied per-share valuations (roughly $138–$148) are closer to its current share price ($153.49). This smaller premium indicates that JNJ might be more reasonably valued relative to its fundamentals, or that market expectations are more modest compared to those for AMGN.

These comparisons not only highlight differences in market sentiment between peers but also underscore the importance of using multiple valuation metrics to capture a range of perspectives. Investors can use such an analysis to identify potential over- or undervaluation among industry leaders.

6. Conclusion

Our integrated approach—combining K-means clustering with comparable company analysis—provides a structured way to benchmark companies in the Drug Manufactures – General subsector. By grouping companies with similar valuation profiles, we derive meaningful insights:

• Cluster 0 captures the majority of the companies trading at moderate multiples.

• Cluster 1 and Cluster 2 highlight outliers (GILD and LLY) with unique valuation characteristics.

• In a focused analysis of Cluster 0, the comparison of AMGN and JNJ reveals that AMGN's market price significantly exceeds the implied valuations derived from key multiples, whereas JNJ's price is much closer to its comparable estimates.

This multi-dimensional approach equips investors with a robust framework to identify potential investment opportunities and assess market pricing against underlying fundamentals.

Disclaimer:

This research report contains information generated with the assistance of artificial intelligence. While every effort has been made to ensure accuracy and relevance, the analysis presented here may contain errors, omissions, or outdated data. Readers are encouraged to verify the information independently before making any decisions based on the content. This report is intended for informational purposes only and should not be considered professional or financial advice. Neither the author nor the AI service provider assumes responsibility for any inaccuracies or the use of the information provided.