Trends in Common Ownership among Insurers in Medicare Part D

Abstract

Background:

Recent studies document the rising prevalence of common ownership by institutional investors in specific industries. Those investors offer products, such as mutual and index funds, to trade securities on behalf of others, and often own shares of multiple firms in the same industry to diversify portfolios. However, at present, few studies focus on common ownership trends in health care.

Objectives:

This paper examines institutional investors’ common ownership in the major insurers offering plans in the Medicare Part D stand-alone prescription drug plan (PDP) market between 2013-2020.

Research Design:

Using data from the Securities and Exchange Commission (SEC) database and the Center for Research in Securities Prices, we compute the percentages of outstanding shares of each insurer owned by institutional investors. Data visualization and network analysis are employed to assess the trends in common ownership among major insurers.

Results:

We document a high prevalence of and substantial increase in shared institutional investors in the PDP market. From 2013 to 2020, the degree of common ownership increased by 7% on average and the common ownership network becomes more connected. Common ownership also varies across the 34 PDP regions depending on their reliance on listed insurers, that are traded in the stock exchange, offering stand-alone PDPs.

Conclusions:

High and rising common ownership in the Medicare Part D PDP market raises policy questions about potential effects on plan offerings, premiums and quality for consumers.

1. Introduction

Common ownership is defined as a scenario in which large institutional investors hold significant shares in multiple firms in the same industry.¹ Economic theory suggests that when common investors hold portfolios of stocks which include competing firms, these investors may consider the profits of competitors when making decisions, leading to reduced competition in the market and higher prices for consumers.¹ Common ownership has increased dramatically in recent years. Backus et al.² estimate that the “Big Three” institutional investors (BlackRock, Vanguard, and State Street) owned about 6 percent of the average S&P 500 firm in 2000; this statistic was 21 percent in 2017.³

Recent studies document the rising prevalence of common ownership in specific industries^1,2,4-7 and have shown anticompetitive effects in some industries (Schmalz⁸ for a review). At present, however, few studies focus on common ownership trends in health care, except Liu and Yao⁶ for hospitals and Banal-Estanol et al.⁷ for pharmaceuticals. Liu and Yao document that common ownership in the hospital industry has anticompetitive effects on hospital prices. Banal-Estanol et al. report significant overlap of institutional investors in the leading “brand name” pharmaceutical firms, with the degree of overlap increasing between 2004 and 2014. A further example is Fowler et al.⁹, who report that common ownership of acute care hospitals and postacute and hospice facilities increased from 24.6 percent in 2005 to 48.9 percent in 2015. Similarly, Stevenson et al.¹⁰ find that common ownership of hospice agencies and nursing homes rose from 6.3 percent in 2005 to 19.0 percent in 2015.

In this paper, we add to this emerging literature on common ownership in health care by examining common ownership of Medicare Part D insurers. Medicare Part D is voluntary prescription drug coverage available to all Medicare beneficiaries. Part D beneficiaries obtain drug coverage either from stand-alone prescription drug plans (PDPs) that supplement traditional Medicare, or from Medicare Advantage prescription drug plans (MA-PDs), which are operated by HMOs and PPOs and include Medicare Part A-B benefits and prescription drugs.¹¹ The focus of this study is stand-alone PDPs, which cover the majority of Medicare Part D enrollees (ranging from 52-62% of total enrollees during 2013-2020). As of 2023, there were 801 PDPs operating in the US (excluding territories) across 34 PDP regions. The number of PDPs available at the PDP region level varies widely, ranging from about 19 in New York to 28 in Arizona.¹¹

Stand-alone PDPs are offered by private insurance companies, some of which are listed in the stock exchange. Figure 1 reports that as of 2013, the top five insurers in terms of PDP enrollment covered about 81% of all enrollees in stand-alone PDPs in Part D (12 million out of total 14.9 million). The corresponding figure increases to 88% for 2017 (17.6 million out of total 20 million) and remains above 87% until 2020 (17.7 million out of total 20.4 million). Figure 1 also shows that there is little turnover among the top five insurers in this market between 2013-2020.

Figure 1:

Major Insurers in the Medicare Part D Market for Stand-alone PDPs, 2013-2020

Source: Authors’ analysis of enrollment data from the Centers for Medicare and Medicaid Services (CMS). https://www.cms.gov/medicare/coverage/prescription-drug-coverage

Notes: Figure 1 shows the major insurers in each year based on stand-alone PDP market share. The market share of each plan in each year is calculated as the number of enrollees in each PDP plan in each year over the total number of stand-alone PDP participants in each year.

In this paper, we focus on the insurance companies enrolling the highest numbers of PDP participants during the period 2013-20 (Figure 1). Our goals are to: (1) identify common institutional investors of these insurance companies; (2) show trends in common ownership; and (3) demonstrate the geographical heterogeneity in common ownership in the Part D market for stand-alone PDPs. The empirical patterns reported in our work will inform policy discussions regarding whether common ownership potentially has anticompetitive effects and hence the societal implications in the Part D market for stand-alone PDPs.

2. Study Data and Methods

To identify the top insurers in the Part D market for stand-alone PDPs, we collect PDP enrollment data for the period 2013 to 2020 from the Centers for Medicare and Medicaid Services (CMS). The market share of each plan is calculated as the number of enrollees in each plan in each year over the number of PDP participants in each year.

The data we use to identify the institutional investors of each insurer are the 13F-HR filings from the Securities and Exchange Commission (SEC) database. This information is publicly available and includes information on the shareholders and their shareholding for each insurer.¹² To compute the percentages of outstanding shares of each insurer owned by institutional investors, we also need the information about the number of outstanding shares for each insurer. We collect this information from the Center for Research in Securities Prices (CRSP) through the Wharton Research Data Services (WRDS).

Since we are interested in the potential anticompetitive effects of common ownership, an implicit assumption made here is that a higher market share owned by institutional investors relates to higher market power. Such a positive relationship between market share and market power is best derived from a market with highly substitutable products¹³. It is not unreasonable to employ this measure as we focus on stand-alone PDPs sold in their regional markets, and this measure is used in the literature^1,2,4-7.

Another limitation of using a market share measure of common ownership is that it does not capture the heterogeneity of institutional investors in utilizing common ownership in exercising market power through coordinating competing firms. For example, Stewart¹⁴ finds that Blackrock published voting rationales for most of its voting decisions in their acquired firms, but Vanguard and State Street published fewer than half of them. The heterogeneity in their disclosure intensity may affect how much coordination they exercise in the invested firms.

3. Study Results

Figure 2 depicts an overall increasing trend of shareholding by the top 5 institutional investors in each of the major insurers. (The major insurers are those shown in Figure 1.) Amid the continued consolidation of the Part D market for stand-alone PDPs, common ownership by a few large diversified institutional investors has emerged as an underlying force behind the concentrating ownership in this health care market. Further, there is heterogeneity in shareholding by the top 5 institutional investors in those major insurers. For example, Humana and United Health have higher percentages of shares held by the top 5 institutional investors than Cigna and CVS.

Figure 2:

Shareholding of Top 5 Institutional Investors in Major Insurers, 2013-2020

Sources: Authors’ analyses of 13F-HR filings from the SEC database (EDGAR) and outstanding shares from the Center for Research in Security Prices (CRSP) through Wharton Research Data Services (WRDS). https://www.sec.gov/edgar/; https://wrds-www.wharton.upenn.edu/

Notes: Figure 2 shows the average shareholding by the top 5 institutional investors across all the major PDP insurers shown in Figure 1. (Centene is not shown because it appears in the year 2020 only in Figure 1.)

Figure 3 shows the top five institutional investors in the major PDP insurers for 2013 and 2020 in order to provide evidence on common investors in multiple major insurers. BlackRock, Vanguard, State Street, Fidelity, and Capital Group are the five leading institutional investors in these major insurers in those two years. We note that BlackRock, Fidelity, Vanguard, and Price T. Rowe Associates considerably increase their shareholdings across the major insurers. This finding suggests that like the other markets analyzed in previous studies^1,2,4-8, common ownership of institutional investors is also pervasive in the Part D market for stand-alone PDPs.

Figure 3:

Top 5 Institutional Investors of Major PDP Insurers, 2013 and 2020

Sources: Authors’ analyses of 13F-HR filings from the SEC database (EDGAR) and outstanding shares from the Center for Research in Security Prices (CRSP) through Wharton Research Data Services (WRDS). https://www.sec.gov/edgar/; https://wrds-www.wharton.upenn.edu/

Notes: Figure 3 shows the percentage of shares held by each of the five leading institutional investors for the major PDP insurers shown in Figure 1 for 2013 and 2020.

To further illustrate the five leading institutional investors holding greater shares of major insurers over time, we follow Banal-Estanol et al.⁷ and Pelligrino and Ederer¹⁵ to apply network analysis to describe the structure of the common ownership network. Figure 4 depicts the common ownership networks observed in 2013 and 2020. In 2013, there were 4 pairs of 5 major insurers (out of 10 possible pairs) that had two common investors holding more than 5% of their shares. In 2020, all 10 possible pairs of 5 major insurers had multiple common investors holding more than 5% of their shares, among which 3 pairs had 5 common investors holding more than 5% of their shares. These results suggest that the common ownership network became more connected over time.

Figure 4:

Common Ownership Network among Top 5 Insurers

Sources: Authors’ analyses of enrollment data from the Centers for Medicare and Medicaid Services (CMS), 13F-HR filings from the SEC database (EDGAR), and outstanding shares from the Center for Research in Security Prices (CRSP) through Wharton Research Data Services (WRDS). https://www.cms.gov/medicare/coverage/prescription-drug-coverage; https://www.sec.gov/edgar/; https://wrds-www.wharton.upenn.edu/

Notes: Figure 4 depicts the network of common owners among the top 5 PDP insurers in 2013 and 2020. A link between two firms exists if they have at least one common investor holding more than 5% share of both insurers. The weight of the line connecting two insurers depends on the number of such common investors that the two insurers share; that number is shown for each link in the figure.

Turning to the geographical reach of common ownership, we consider the proportion of stand-alone PDPs offered by listed firms across PDP regions. Since institutional investors construct their portfolios by buying and selling stocks from the stock exchange, their influence is mainly through the corporate control of listed firms.⁵ Thus, institutional investors exert a stronger influence in a PDP region if a larger proportion of stand-alone PDPs is offered by insurers that are listed in the stock exchange.

Figure 5 depicts substantial variation in the proportion of stand-alone PDPs offered by publicly traded insurers across PDP regions in 2020. For example, in this year North Carolina had 75% of stand-alone PDPs that were offered by publicly traded insurers. In contrast, Connecticut, Massachusetts, Rhode Island, and Vermont (the Central New England market) had 84% of stand-alone PDPs offered by publicly traded insurers. Moreover, Figure 5 depicts substantial variation in the shareholding of major PDP insurers by the top five institutional investors across PDP regions in 2020. Consistent with our idea, we find that the top 5 institutional investors have higher stockholding in the major PDP insurers in Connecticut, Massachusetts, Rhode Island, and Vermont (28% in 2020) than North Carolina (25% in 2020).

Figure 5:

Average Shareholding of Top 5 Institutional Investors in PDP Insurers v. Proportion of PDPs offered by Publicly Traded Insurers across PDP region markets, 2020

Sources: Authors’ analyses of enrollment data from the Centers for Medicare and Medicaid Services (CMS), 13F-HR filings from the SEC database (EDGAR), and outstanding shares from the Center for Research in Security Prices (CRSP) through Wharton Research Data Services (WRDS). https://www.cms.gov/medicare/coverage/prescription-drug-coverage; https://www.sec.gov/edgar/; https://wrds-www.wharton.upenn.edu/

Notes: Figure 5 shows the percentage of PDPs offered by listed insurance companies over the total number of PDPs offered in that PDP region against the average shareholding of the top 5 institutional investors in the PDPs offered in that PDP region. PDP regions that are U.S. territories are excluded.

A larger market share comprised of publicly traded insurers potentially allows a higher level of common ownership in the Part D market for stand-alone PDPs. Vives¹³ develops an oligopoly theory showing that firms operating in markets with a high level of common ownership adjusted market concentration enjoy higher market power because of reduced competitive pressure. According to our calculations, the average markup in Connecticut, Massachusetts, Rhode Island, and Vermont is about 3% higher than that in North Carolina in 2020.¹⁶

4. Discussion

Recent research highlights the growing consolidation and “corporatization” of the U.S. health sector.¹⁷ Some research has focused on private equity acquisitions in health care. Tan et al.¹⁸, for example, document that private equity firms increasingly have been acquiring dermatology practices, with 17 organizations acquiring 184 practices between 2012-2018. Bruch et al.¹⁹ report increased private equity acquisitions of OB/GYN practices in recent years, mostly clustered in Northeastern urban areas. Bruch et al.²⁰ document rising hospital and nursing home acquisitions by real estate investment trusts. Common ownership, however, has become an important antitrust issue⁸ but has yet to receive much attention in the health care literature.

Our results highlight that a small group of institutional investors hold appreciable shares of the insurance companies with the largest shares of the PDP market. This trend has been growing over time, particularly in geographic areas where the major PDP insurers are listed firms. These findings raise concerns that common ownership may adversely affect Part D enrollees.

To our knowledge, there is no research quantifying the association of common ownership with care utilization or costs in the Part D PDP market. Using 2020 data from the CMS landscape files, we find that a higher average shareholding by the 5 leading institutional investors in a PDP region associates with a lower availability of total and new stand-alone PDPs, a higher ratio of premium to income and a higher star rating of stand-alone PDPs. These empirical associations together with the growing trend of common ownership provide implications for antitrust policy and patient care.

Common ownership among insurers may steer their incentive to maximize their joint profits instead of their own profits in the interest of institutional investors. Consequently, those insurers may attempt to exercise market power by reducing existing and new plan offerings, raising premiums, and adjusting plan quality. The reduced plan offerings is consistent with existing evidence on the deterrent effect of common ownership on entry in other industries, such as pharmaceuticals²¹. The pricing effect of common ownership can attribute to unilateral and/or coordinated effects because both can lead to softened price competition²². Finally, the high-quality plans offered and a higher premium together are suggestive of an exercise of market power as well. Firms with market power can offer multiple products to segment their customers and drive high-end consumers to their pricier products, and such strategy is documented in the retail gasoline market²³. However, quality-based price discrimination has mixed implications for consumer surplus²⁴. More research is needed to identify the mechanisms and welfare implications of these potential anticompetitive effects on consumers.

Overall, even in the absence of a merger, insurers could have common investors, which might influence market competition and insurer behavior. Policy makers should ensure that the rising phenomenon of common ownership driven by shared institutional investors does not disrupt long-term priorities in health care delivery for the elderly.

We conclude with a few limitations of our research. First, we employ a single year of data to establish the association between common ownership and market outcomes. Moreover, since the insurers compete in more than one market, coordination among insurers may be implemented across markets of different beneficiaries, so that they can avoid detection of coordination by antitrust authorities. Future research may employ more sophisticated methods to establish the causality between common ownership and market outcomes. Second, our research focuses on the common ownership in the Part D stand-alone PDP market among institutional investors. Future research may extend to examining whether common ownership arises due to private equity and venture capital acquisition. Those investors play an active role in managing their acquired firms and can affect policies of competing firms by appointing agents on the management of those firms²⁵.