Examining the DOJ’s Challenge to Hewlett Packard Enterprise’s Acquisition of Juniper

Hewlett Packard Enterprise (HPE) proposed a $14 billion deal in early 2024 to acquire Juniper Networks. HPE, it should be noted, is not the Hewlett Packard (HP) of printer and ink fame, as it was spun off from the larger HP corporation in 2015. Instead, HPE is, like Juniper, involved in connectivity—specifically, wireless technology, including software and hardware to deliver internet access to large enterprises and campuses.

After a nearly yearlong investigation under the Biden administration, the U.S. Justice Department (DOJ)—under President Donald Trump’s interim head of the DOJ’s Antitrust Division—issued a complaint Jan. 30 seeking to block the acquisition. This decision is particularly notable because the acquisition has already been approved by 14 other international regulators, including the European Commission (which found that “HPE and Juniper are not each other’s closest competitors”) and the United Kingdom (which explained that the acquisition “does not give rise to a realistic prospect of a substantial lessening of competition”).

The DOJ’s complaint lays out several assertions about the deal. First, it declares that Cisco, HPE, and Juniper together represent a combined 70% of sales in the “enterprise wireless networking solutions” market. Second, the DOJ quotes internal documents wherein HPE and Juniper were head-to-head competitors for certain customers. Finally, the DOJ dismisses the parties’ efficiency claims.

Before considering each of these claims and assessing what the district court will likely examine during the trial, it is worth discussing the key technologies involved in this acquisition.

Most households are familiar with Wi-Fi and the use of routers to set up a wireless network to deliver internet access throughout the home. Imagine, however, delivering wireless internet over a much larger area, such as a university campus, hospital complex, or corporate headquarters. Such a setup involves establishing a local area network (LAN) and, more specifically, a wireless connection over that area (WLAN). The types of products and services involved to deliver this connectivity include routers, hubs, switches, and software to coordinate the entire network. Undoubtedly, much of the value that suppliers offer include factors such as internet speeds, reliability, minimal maintenance, and straightforward troubleshooting.

As with nearly all antitrust cases, the complaint must establish the area of commerce that this acquisition allegedly impacts. In this case, the “relevant market” is declared to be “enterprise commercial wireless solutions.” Delineating the relevant market in antitrust is more of an art than a science and can generally follow one of two paths (which are not necessarily mutually exclusive).

The first uses more qualitative evidence, called “practical indicia” or Brown Shoe factors—named after the Supreme Court case that first laid out these factors. One notable factor is how customers and competitors perceive the industry and the market players. This is largely the approach detailed in the complaint.

The second path to define a market is the use of the more technical “hypothetical monopolist test” (HMT), which normally involves some quantitative evidence on the sensitivity of consumers to price increases for a proposed market. Regardless of the path, defining the boundaries of the relevant market is often hotly contested. The parties dispute the DOJ’s market definition, which omits “a broad set of players.” At trial, this is where much of the battle will be fought and, as demonstrated by the FTC’s loss in its challenge to the Tempur Sealy’s acquisition of Mattress Firm, there is no guarantee a judge will agree with the government when the markets are perceived to be gerrymandered for litigation purposes.

Further, even if adjacent markets, such as consumer-grade WLAN, are “outside” the relevant market defined by the DOJ, the suppliers in those markets could represent potential entrants if market conditions change. Additionally, even if the DOJ successfully defines the geographic market as the United States, some global players may either expand their existing presence or enter into the U.S. market.

Even if we accept the DOJ’s proposed market definition, a simple examination of the complaint’s market shares raises some questions. Puzzlingly, nowhere in the complaint does the DOJ actually provide the market shares of Cisco and HPE. Rather, the complaint repeatedly asserts that Cisco, HPE, and Juniper combined will represent 70% of the market. Given the complaint’s concentration calculations, one can, however, back out the market shares. It appears that Cisco, HPE, and Juniper have shares equal to 36.5%, 27%, and 6.5%, respectively.

The elephant in the room is that, even if the top three firms represent 70% of the market, there is still the remaining 30%. Who are these competitors? The complaint is silent on this key question. While Juniper may be the “number three” player, that may create a false impression regarding Juniper’s influence on the market—especially considering the remaining 30%. Another way to frame the market shares is that nearly 67% of sales occur outside of a combined HPE-Juniper.

The complaint asserts that Juniper has strong brand recognition with its Mist product; yet, as with most differentiated product markets, there is always some level of loyalty and differences based on proprietary technology. According to HPE, they face competition from eight competitors in the United States. Further, even the DOJ’s complaint admits that Juniper’s growth has been “swift” since purchasing Mist Systems in 2019. If Juniper, as a smaller player, can grow swiftly, then what prevents the suppliers comprising the remaining 30% of the market—or even a supplier currently outside the narrowly defined DOJ market—from likewise expanding rapidly?

What about the head-to-head competition documents that the DOJ highlights? Undoubtedly, there is some overlap between the firms, or we would have never reached this point in the case. Even if two firms have a 10% and 2% share in a market, there will be episodic cases of direct competition. The point is to consider the larger competitive environment and whether the competitive process has been materially impacted due to the acquisition.

On this point, while the combination would certainly result in the loss of whatever competition existed between HPE and Juniper, the question is whether other suppliers will be able to fill this void and whether the potential efficiencies are sufficiently strong. The complaint is short on details quantifying the extent of the lost competition relative to the overall quantity of commerce in the market. Picking out a handful of documents that mention direct competition between two firms is an insufficient basis to condemn a merger, or else most horizontal mergers would be effectively illegal.

The questions that a judge will ask include:

1. What is the percentage of times that the parties compete head-to-head?;
2. What is the trajectory of the remaining competitors that comprise 30% of the market; and
3. How have shares changed, including due to entry, over the past 5-10 years?

Additionally, what the complaint omits is the critical question of whether a combined HPE-Juniper will create more frequent competitive scenarios with Cisco, the market leader. This potential for a greater intensity of competition between Cisco and HPE-Juniper could drive significant consumer gains. For instance, according to one assessment:

…integrating Juniper’s AI-powered networking capabilities with its own, HPE aims to create a next-generation networking portfolio that enhances automation, security, and performance for enterprise customers, which is expected to expand customer choice rather than limit it and offer them a stronger alternative to Cisco’s long-standing grip on the sector.

Further, if “the DOJ blocks the HPE-Juniper deal, Cisco stands to gain the most.”

Finally, considering the efficiencies will play a major role in the case. The parties have advanced that the acquisition will result in numerous complementarities. Within antitrust law, however, if the benefits occur outside the relevant market, they do not “count” in terms of offsetting any competitive harm that is found. The efficiencies must be “within” the market. This means the inquiry will be tailored to the enterprise commercial wireless market—or whatever market the court determines.

Given the importance of reliable, secure internet access for institutions such as hospitals, universities, and corporate headquarters, improvements in quality and innovation will play a central role in the inquiry. The potential procompetitive effects include incorporating Juniper’s integration of AI in its networking products. The conceptual gains from spreading that technology to HPE products are enormous.

Ultimately, the DOJ claims that “such efficiencies are unlikely to be timely or substantial enough to mitigate the risk to competition posed by the transaction.” Yet it is not entirely clear that the risk to competition is at the level asserted in the complaint.

The case will boil down to whether the enterprise commercial wireless market matches market realities, and whether a combined HPE-Juniper can materially bring greater competition to the market leader, Cisco.