Amid the Purgatory of HPE’s Uncertain Juniper Acquisition

I haven’t paid a visit to the networking neighborhood for a while, partly because networking, while essential as IT infrastructure, seems a little beside the point when the overall industry and the larger world is teetering on the edge of feral madness.

I ask myself, and no longer in idle moments: Were people distracted by activity in what passed for technology markets in the years just before World War I? Maybe.

Anyway, let’s pretend, if only for a few minutes, that nothing has changed and we still find ourselves in a universe governed occasionally by equanimity, logic, and reason. Look, you and I know that’s not true — just take a look around, if you dare — but come along with me in a communal reverie, a carefree waltz into an innocuous nostalgia for the simpler times of the, er, post-COVID era.

It was more than a year ago thatHewlett Packard Enterprise (HPE) announced its intent to acquire Juniper Networks in an all-cash transaction for $40.00 per share, which amounted to approximately $14 billion.

HPE boasted that the acquisition would spawn a doubling of its networking business, creating “a comprehensive portfolio that presents customers and partners with a compelling new choice to drive business value.”

That swaggering boilerplate was followed by the usual hyperventilating prose that one only finds in corporate press releases and at nitrous-oxide parties. To wit:

The explosion of AI and hybrid cloud-driven business is accelerating demand for secure, unified technology solutions that connect, protect, and analyze companies’ data from edge to cloud. These trends, and AI specifically, will continue to be the most disruptive workloads for companies, and HPE has been aligning its portfolio to capitalize on these substantial IT trends with networking as a critical connective component.

Combining HPE and Juniper’s complementary portfolios supercharges HPE’s edge-to-cloud strategy with an ability to lead in an AI-native environment based on a foundational cloud-native architecture.

It’s like buzzword bingo on steroids. If your card had the words “AI,” “”edge to cloud,” “disruptive,” “supercharges,” “and “cloud-native,” you may collect your prize in the information-technology afterlife, which is sure to feature enough automated intelligence to render corporeal employment by network engineers entirely superfluous.

Objection from An Unlikely Source

HPE had every reason to think that its pending acquisition of Juniper would come to fruition. The HPE board approved of the acquisition, the Juniper board approved of the acquisition, and the shareholders of both companies gave their full-throated consent. The UK's Competition and Markets authority and the European Union cleared the deal. Even China failed to raise an objection. What could go wrong?

HPE might prove incapable of breathing vivifying life into the vaporous verbiage contained it its press releases, but the Juniper acquisition itself seemed a foregone conclusion.

Well, not quite.

Reuters, among others, reported last week that the U.S. Department of Justice (DoJ), citing competition concerns, has sued to block HPE’s bid to acquire Juniper Networks. Puzzled looks were affixed to faces across Silicon Valley* (*except perhaps on Tasman Drive).

Why is the DoJ taking action to challenge and potentially prevent this combination? It says here that the DoJ is concerned about the stifling of competition.

If you think about it, arguably every acquisition diminishes competition within a given market. As the result of an acquisition, two competitors become one. Where does one draw the line between a slight diminution in the number of competitors in a given market and the more virulent _stifling_of competition, which involves preventing or constraining said competition?

I’m sure a corporate lawyer could draw the legalistic distinction for us, presuming we’d pay said lawyer for his or her valuable time. Well, on principle alone, I’m not going to do that. Corporate lawyers are richly compensated, and they don’t need me to further line their well-stuffed pockets.

The DOJ’s objection, on the ostensible basis of stifled competition, seems arbitrary to me, especially given that many existing IT markets are far more monopolistic or oligopolistic than enterprise or datacenter networking.

The DOJ contends that the acquisition would result in two companies — Cisco Systems and HPE — controlling more than 70% of the U.S. market for networking equipment.

The Networking Market is What the Market Made It

But that statement isn’t true, at least not if you’re looking at the entire U.S. market for networking equipment, which would include Arista Networks and the bare-metal, white-box switches in massive cloud datacenters.

Still, even if the claim were roughly accurate in enterprise networks, and particular in the context of campus networking, I can think of some markets — GPUs, anyone? — where one vendor (whose name starts with an N) has captured far more than 70% of the market and continues to make acquisitions and investments designed to defend its privileged market position. I’m not saying anything should be done even in that market — I think the situation will resolve itself for reasons I have explained previously — but it does stand as an example of a market that is theoretically subject to a far greater degree of monopolistic abuse than is the networking market.

You could pick other markets — web search and related advertising, social media, etc. — and make similar arguments. Networking is far from being the most dysfunctional or non-competitive market in the IT industry.

Besides, the state of the networking market has been cast by the invisible, Ouija-board hand of the market, through which venture capitalists decided to put their money anywhere other than in network infrastructure. Even back in the oughts, Arista’s funding came from its founders, not from venture capitalists, who chose not to attempt to resist Cisco’s dominance or to try to make money from a capital-intensive hardware business such as networking. (Yes, we witnessed VC investments in software-defined networking and SD-WAN, but those involved software overlays, not network hardware.)

Essentially, my criticism here involves the DoJ’s lack of clarity, consistency, and transparent process in its handling of mergers and acquisitions. At this point, the entire exercise is opaque, leaving companies without a clear and actionable understanding of how, when, or whether to bother with time-consuming and enormously costly acquisitions that complicate the lives of customers, partners, and employees.

Understandably, as the Reuters article notes, HPE and Juniper would like a word with the DoJ:

"The DOJ's claim that the WLAN (Wireless Local Area Network) market is composed of three primary players is substantially disconnected from market realities," the companies said in a joint statement.
The firms added that they would defend the deal, arguing that the acquisition will bring together two complementary networking offerings to better compete with global incumbent players.

From an Associated Press article, we get a little more of the Justice Department’s perspective:

In its objection, the DoJ alleges that HPE’s salespeople were deeply concerned about Juniper as a competitive threat, even noting that one former HPE sales executive, rising perhaps a bit too emphatically to heights of competitive hyperbole, said “there are no rules in a streetlight.” This gentlemen — he is identified as male — even encouraged his staff to “kill” Juniper in competitive engagements.

Angst of the Salespeople

Frankly, I’m not sure where the DoJ is going with this line of argumentation. What the DoJ describes is standard operating procedure for most tech-industry salespeople, who, contrary to conventional belief, have extremely vivid imaginations. Salespeople often envision themselves as characters in genre movies based on poorly written screenplays. Consequently, salespeople often allude melodramatically to life-and-death trench warfare in enterprise accounts, even though nobody dies — to the best my knowledge — and the only injuries accrue to commissions and egos.

Perhaps the personnel in the DoJ are not that worldly, so allow me to reveal that salespeople are inveterate drama queens. For salespeople, every account is an agonizing contest between good and evil or, more strikingly, a “bloodbath.” It’s no such thing, of course, but you have to let salespeople indulge in what amounts to motivational and essentially harmless psychic cosplaying.

Meanwhile, on the stock market, investors are confused because they assumed that the Trump Administration, unlike the Biden Administration, would let anything short of a watertight monopoly — and perhaps even that, too — sail forth on uncontested seas. Investors expected Trump to put the “laissez” back in _laissez faire, and now they’re struggling to ascertain how much different the new administration, at least in this domain, will be from the ancien régime.

The Wall Street Journal reported that most of the work on the HPE-Juniper file had been done by the Biden Administration, though it’s now been inherited by the Trump administration. Interestingly, according to that report, the companies declined in recent days to extend an agreement that would have given the Justice Department more time to consider whether to intervene. Well, perhaps that was a mistake, because more time is being added to the clock now.

We can argue all day and all of the night about whether the DoJ’s objection to HPE’s proposed acquisition of Juniper is justified. That might be a stimulating debate; we’d might enjoy wallowing in it, but let’s simply cut to the chase and acknowledge that there’s a real-world possibility, if not a probability, that this deal will be scuppered.

Twisting in a Sustained Wind

Even if the transaction proceeds as planned, all of Juniper and HPE’s networking business will have been twisting in the winds of uncertainty for nearly two years. If you’ve been through an acquisition, you know that the longer the process takes, the more difficult life becomes for those operating within the company subject to the acquisition. People don’t like uncertainty. Some employees leave, as do some executives. Sometimes key people leave. In fact, they often do leave.

This deal was announced a year ago, and I would suggest that most of Juniper’s staff back then expected the deal to move forward at a stately pace and be consummated by about now. Instead, they’re still twisting in a cold wind, unsure whether they’ll become one with HPE or will remain a separate company. Similarly, there are employees in HPE’s networking unit who don’t know whether they’ll be displaced by Juniper staff. Keeping a company together under such strain is not a simple proposition. Juniper’s executive team — and to a lesser extent, HPE’s — must walk a fine line and they must keep walking it over a marathon distance.

Fortunately for Juniper, it’s still putting up decent numbers amid the uncertainty. Yesterday Juniper announced that its fourth-quarter financial results surpassed revenue and profit estimates. Enterprises and cloud customers came through for Juniper, but business from service providers remains chronically weak. (Juniper’s not alone there.) These numbers suggest that Juniper isn’t yet coming apart at the seams. Still, the company might have many more months to go before it knows whether it will become part of HPE or continue to exist as a separate company.

In the Reuters article, Evercore ISI analyst Amit Daryanani observes that “HPE and Juniper will undergo its pretrial and trial processes over the next . . . eight months prior to absolute walkaway date in October — which we think gives suffice time for a trial to commence.”

As for the market’s verdict, Junipers share price — as I write this missive — is $35.69 per share, which gives the company a market capitalization of $11.81 billion. That means the market, at this point, is leaning moderately against HPE’s consummation of the acquisition. That can change, of course.

I know there are nominal laws against insider trading, but we’re in a post-ethical period of history, and I would wager that the market will be your best guide to assessing whether the HPE-Juniper combination will come to fruition.