Investing in Defense

In 1994, the largest merger in the history of the U.S. defense sector happened out of business necessity. The Lockheed Corporation of Calabasas, CA, merged with the Martin Marietta Corporation of Bethesda, MD, to form a dominant prime contractor with combined sales of $23 billion.

Martin Marietta needed a win. Four months prior, it lost out on the acquisition of the Grumman Corporation to Northrop. With the long-awaited end of the Soviet Union, the defense industry was beginning to consolidate in preparation for a new, more imminent threat: military spending cuts.

On the day of the Lockheed Martin merger, The New York Times praised the merger: “A combination of the two companies would result in the only military contractor able to supply the armed services with virtually every military need, except for tanks and submarines…”

In hindsight, this view of the defense industry was quaint.

The defense budget did not shrink, but rather grew from ~$442bn in FY 1995 to $716bn in FY2019. Relative peace after the Cold War gave way to the emergence of new rivals as well as non-state threats, demanding different types of technological capabilities. But perhaps the most naïve prediction was the belief that Lockheed Martin would be able to supply the U.S. military with every technological need just as the birth of the Internet was beginning to transform every other sector of the global economy.

The belief that legacy defense contractors could meet all the technology needs of the DoD would lead to decades of stagnating defense R&D investment, just as commercial R&D was exploding.

The stagnation of defense R&D is one of the reasons the U.S. is now at “near-peer” technological capabilities with Russia and China, according to DoD leadership. This pronouncement would have been unthinkable a quarter-century ago. How the U.S. managed to lose its technology lead is not only a matter of policy, but also a market failure. In 2017, around 20% of total U.S. federal contracting dollars — not just military spend — went to a handful of defense primes, the five largest of which were founded long before World War II. Known as “The Big Five,” these companies saw their contracting dollars grow by 33 percent between FY 2015 and FY 2017, and now account for 30% of the 2018 defense contracting budget.

(MDAPs (in green) are mostly controlled by the Big Five.)

That an industry of such scale has aggressively consolidated to maintain control of a more than $1 trillion global annual market is not surprising. It’s success in warding off competitors is the market and policy failure: It would be as if the internet today were solely dominated by HP, IBM and Cisco because Congress effectually mandated it.

Twenty-five years after the Lockheed Martin merger, history continues to repeat: United Technologies and Raytheon announced a merger that will make Raytheon Technologies the defense sector’s largest merger yet, valuing the company at $120 billion.

Now is the time to ask hard questions: after decades of consolidation in defense, do we have better technological capabilities than our adversaries? Has it become more expensive to protect the U.S. and our allies? And most importantly, are we more secure now than we were before?

The technology gap in the U.S.’s defense sector is the greatest security threat of our time. It’s time that technologists and Washington work together to ensure that the most talented engineers are working to solve the most difficult problems in defense.

1. Defense Procurement: A Brief, Complicated History

Defense Procurement: A Brief, Complicated History

From the early 19th century until World War II, the government built and developed technology in-house. As World War II began, defense spending soared to previously unfathomable heights, and it became clear that the government could not build the technology — specifically airpower — that it needed for a costly multi-front war. Private companies entered the fray and supplied the country with military equipment. According to Barry Watts and Todd Harrison of the Center for Strategic and Budgetary Assessments, the onset of the Cold War and the Korean War led to a permanent defense industrial base that could support technological needs and win a prolonged Cold War centered upon technological acquisition and stockpiling.

But researchers had their concerns. Watts and Harrison asked two questions that are relevant today: does the defense acquisition structure sufficiently allow for a free market whereby the best technologies and companies can sell to the government? Second, how does the Department of Defense sustain a “technologically vibrant” defense industry if the structure is not a truly free market?

Sources: Obama White House Archives OMB; Wikipedia

The dissolution of the Soviet Union coincided with another technology shift — public accessibility of the Internet — that essentially reset the defense gains the U.S. had made for decades. Whereas the prime contractors of the Cold War were well equipped to build tanks, planes and aircraft carriers, the Internet was about to transform every aspect of the economy and ensure that the global leader in defense would need to adjust capabilities, talent and acquisition practices to build for modern security. Instead of acquiring small startups to weather this paradigm shift, the big players merged.

By the early 90s, the DoD actively encouraged its most valued contractors to merge due to expected cutbacks in federal defense spending. Between 2001 to 2015, 17,000 U.S. companies left the defense industry, according to a report by the Center for Strategic and International Studies. By 1996, The Wall Street Journal reported, “With the Pentagon acting as cheerleader and federal antitrust regulators largely falling in line, the number of major players in the aerospace and defense field has shrunk by more than half since the collapse of the Soviet bloc, to just six major companies.” The FTC didn’t block any mergers, believing it was a cost-saving measure for the federal government. But even at the time of this massive consolidation, critics worried that M&A would stifle the innovation needed to build new technologies.

Center for Strategic and Budgetary Assessments, Watts and Harrison, 2011

2. Why Procurement Must Change

Why Procurement Must Change

Instead of saving the government money, budgets ballooned and ensured that the Big Five were preferred vendors to the DoD. Government procurement has become a structural problem because — beyond its arduous and long sales cycle — the government is incapable of acquiring software because of the nature of software development.

When the DoD needs to acquire a new capability, it issues “requirements” or details of what would be needed to meet a capability. But the requirements often outline the “how,” with pages upon pages explaining how the DoD would like to see a company solve a given problem. The opposite is true of Silicon Valley company-building: companies solve problems in the most efficient manner and then sell products to customers. The government, however, seeks to work with companies that 1) build to the requirement spec and 2) are known entities that build slowly and methodically.

Over the years, it’s become clear that software development and innovation processes that have built the modern internet are antithetical to how the government procures battleships. And without significant changes to procurement structure, it is almost impossible for tech companies to compete with primes or sell products to the government.

In addition to the requirements, Silicon Valley investors have historically avoided investing in defense for three practical reasons: mergers, margins and interest rates. Unlike the pharmaceutical industry, which is highly acquisitive of unprofitable venture-backed companies, defense contractors mostly buy small profitable companies — by and large, they aren’t acquiring fast-growing startups with the most talented engineers to augment their R&D efforts. Margins, too, make it difficult for investors to get excited: why invest in a company with cost-plus pricing when you can invest in pretty much any other software company and reap the rewards of software margins? (Cost-plus pricing is when the government looks at the actual cost of goods and affixes a mark-up that is often fixed. Therefore, the incentive is to seek out larger contracts that lead to both time and cost overruns, not iterate solutions as quickly and efficiently as possible.) And finally, historically low-interest rates have made capital cheaper, leading many firms to invest in deep tech; but the amount of capital invested in these companies means outcomes must be an order of magnitude larger to see venture-scale returns: aside from SpaceX, there have not been many significant venture-backed companies that receive the majority of their revenue from the government.

Fortunately, these messages are not falling on deaf ears. The Department of Defense has worried publicly about private sector contracting for nearly a decade. Former Defense Secretary Ash Carter, who served under President Obama, spearheaded the charge to change defense industry mergers and encourage acquisition of startup technology. He championed the Defense Innovation Unit, first as an experiment to help startups interface and receive contracts and now as a contracting vehicle that works with hundreds of startups.

The DoD also implemented new types of contracting vehicles that are now utilized regularly by engineering-led startups, including many in the GC portfolio that do not solely sell to the government. “Other Transaction Authorities,” a new type of contracting vehicle implemented in 2015 that allows the U.S to acquire technology quickly (often in 30 days) have allowed the DoD to work with some of the most cutting-edge startups in artificial intelligence, robotics and aerospace. We are encouraged by these changes and founders are pleasantly surprised with the ease of partnership. But it remains to be seen whether these companies will graduate on to large production contracts that ensure companies will work with the government beyond pilots. To use venture terminology, the DoD is learning how to move from “spray and pray” investing — or giving small checks to thousands of companies — to a more concentrated investment style that ensures the success of the best companies. That shift has to happen if this era of startups is to work with the government for the long-term.

There are many investors in deep tech startups eager to engage with the DoD and we are encouraged by the proposals from the Defense Innovation Board, particularly around software procurement. But more can be done, and we’re committed to doing our part.

3. GC’s Investment in Anduril

GC’s Investment in Anduril

We first invested in Anduril’s seed round in 2017, a time when few companies or people in Silicon Valley were discussing the structural and market needs for new defense contractors. From the beginning, we were impressed with this mission-driven team who had worked in investing, startups and defense across both hardware and software. They had felt the pain of trying to work with the Department of Defense (DoD) before and were not naïve to the structural challenges. Still, they were determined to build products for the government.

Over the last two years, we’ve watched Anduril grow and execute on what seemed like an impossible vision: to build a company that works directly with the DoD on the most critical problems in AI and computer vision. We’ve spent time with their customers and watched as Anduril engineers work side-by-side with talented people at all levels of the armed services. And what’s clear is that their products and speed are not only needed but truly valued. The Pentagon deserves a partner that performs at the highest level, and we believe Anduril is meeting those expectations.

What’s also become clear is that after decades of discussions on procurement reform, changes are now happening: we’ve been continuously impressed with Pentagon leadership across all branches of the armed services and at the DOD’s Joint Artificial Intelligence Center and Defense Innovation Unit, as they work to shore up software and AI capabilities. There’s now room and will for new companies to emerge in the sector. And GC is eager to back defense companies. Beyond Anduril, we recently backed Vannevar Labs, an AI company working with the intelligence community in addition to a number of deep tech companies working with DoD on critical technology. (And yes, all these companies are hiring!)

4. Why Defense Technology Matters

Why Defense Technology Matters

At GC, we look for antiquated markets untouched by technology that have an outsized impact on society. There is no market that touches more Americans than defense. We all fund it as taxpayers. We rely on it for security. And every day, the bravest and most patriotic among us, our men and women in uniform, rely upon defense technology to do their jobs.

A few years ago, we spoke with a fellow investor who had served in Afghanistan. When asked what technology company had the greatest impact on his day-to-day work, he replied, “Amazon. We’d order long antennae from Amazon to probe IEDs in the field. They were often more useful than what was issued to us.”

Silicon Valley was built on defense investment, yet the last twenty years of radical improvements in consumer technology have barely touched the military’s end user: our servicemen and women. It’s time we bring the best technology products to those serving our country every day, at home and abroad.

We’re proud to co-lead the Series B of Anduril and to support this team on their quest to build the future of defense technology.

— Katherine Boyle, Hemant Taneja & the GC team