The Evolution of Venture Capital: Part 3 Era of disruption

Hello friends, and welcome to The Unsophisticated Investor! Brought to you by Mark, Head of UK Expansion & Operations at Shuttle!

If you want to invest alongside the VC funds who've backed breakout companies like Revolut, Monzo, Asana, JustEat, Bolt, Lets Get Checked, Loom, Runna, Charlotte Tilbury, Deel, Aircall, AngelList, Carta, TransferWise and many more, regardless of your knowledge, network or net worth, then sign-up today using the link below.

Now, let’s get into it 👇

In Part 2, we followed venture capital from the late 1970s to the early 2000s, as a change to America's pension rules and a run of tax cuts opened the retirement savings of ordinary workers to the VC funds of Silicon Valley. A sub-$500 million industry grew into one measured in the hundreds of billions, funded the dot-com boom, and came through its collapse with its structure intact. This episode is about what happened next, when the cost of capital itself fell close to zero.

After the financial crisis of 2008, central banks around the world began to cut interest rates to near zero and held them there for years. Capital, the scarce input around which the whole venture model had been built, became abundant and cheap.

The usual account of the 2010s in technology is a story about products: the smartphone, cloud computing, social media. Those were real, but they do not on their own explain what happened to venture capital as a financial activity. The larger factor was monetary. Cheap money increased the supply of capital chasing private companies, and that increase changed three things in particular: how venture funds were governed and how private companies were valued. The clearest expression of the shift was the SoftBank Vision Fund, which raised close to $100 billion in 2016 and 2017 and, within a few years, backed both ByteDance, one of the era's genuine successes, and WeWork, one of its most public failures.

Section 1: When money became cheap

Venture capital had spent its first sixty years operating under conditions of scarce capital, and the disciplines that defined it were adapted to that scarcity. A fund had a fixed life because a limited pool of money had to be returned and put to work again. Investors took board seats partly because capital was valuable enough to command governance in exchange for it. Terms were negotiated hard, and selection mattered, because there was not enough money to back every company, and choosing well was the core skill. None of these features was incidental. Each was a response to the fact that money was difficult to raise.

That condition changed after 2008. In December of that year the US Federal Reserve cut its benchmark rate to a range of zero to 0.25%, the lowest in its history, and kept it there until December 2015. It also bought several trillion dollars of bonds through successive rounds of quantitative easing between 2008 and 2014. The stated aim was to lower the return on safe assets in order to push money into investment and risk-taking elsewhere in the economy.

For institutional investors, this created a specific difficulty. A pension fund with a long-term return target of, say, 7% could not meet it while government bonds yielded 1% or 2%. To close the gap, allocators moved further along the risk curve, into public equities, then private equity, and eventually venture capital. The effect was to increase the amount of institutional money looking for access to private technology companies. And that money no longer came only from the pensions and endowments that had dominated since the 1980s. Sovereign wealth funds, mutual fund managers such as Fidelity, and hedge funds all began investing in private companies, often at late stages they would once have left until after a public listing.

The move was not entirely new. University endowments, following the model associated with Yale, had been raising their allocations to illiquid alternatives, including venture, since the 1990s. What the period after 2008 did was extend that behaviour to a much wider set of institutions and push it further, as the returns available on conventional assets fell and the pressure to find them elsewhere grew. The direction was the same; the volume of capital moving in it was much larger.

The scale of the shift was large. Annual venture investment into US companies, which ran at roughly $45 billion in 2013, was above $130 billion by 2019. The change was most pronounced at the late stage, where the new entrants concentrated their money: rounds of $100 million or more, once unusual, became a routine feature of the market. This was, in part, the policy working as intended. Lowering the return on safe assets was meant to move capital into riskier investment, and venture sat at the far end of that spectrum. What the policy did not distinguish was between capital that funded genuinely new companies and capital that mainly raised the price of the companies already being funded.

The result is simple enough to state. Venture capital, which for most of its history had struggled to raise money, now had more of it available than at any earlier point. Abundance is not in itself a problem. But it removed the condition that several of the model's disciplines had been built around, and the rest of this episode follows three of them, governance, price discovery, and selection, and what happened to each once capital stopped being scarce.

Section 2: Founder-friendly capital and the retreat of governance

The first thing abundant capital changed was the balance between investors and founders, and with it the role of the board.

When capital is scarce, an investor can attach conditions to it, including a board seat and a say in how the company is run. When capital is plentiful, a founder raising money has more investors to choose between and can prefer the ones who ask for less. Investors competing for access to the most sought-after companies therefore had a reason to offer easier terms, including lighter oversight.

An early and influential example came in 2009, when DST Global, a firm run by Yuri Milner, invested $200 million in Facebook for roughly 2% of the company, at a valuation of about $10 billion. That valuation was high for the time, and several established venture firms had passed. What set the deal apart was less the price than the terms: DST did not take a board seat or ask for governance rights. It supplied capital and left management to run the company. For a founder, this was an appealing offer, and the approach, large cheques at late stages without the usual control provisions, became a template for the growth investing that followed. Facebook's 2012 IPO, at a valuation above $100 billion, made the underlying bet look sound.

A more lasting change came through the dual class share structure. Under it, a company issues shares with different voting rights: those sold to outside investors and the public carry one vote each, while those held by the founders carry ten or more. This lets founders raise large amounts of outside capital, and sell most of the company's economic value, while keeping voting control. Google adopted it at its 2004 IPO, and Facebook followed in 2012 with an arrangement that left Mark Zuckerberg holding a majority of the voting power, and Elon Musk would use this multiple times over when building his wealth at Tesla and SpaceX. Over the following years super-voting shares became common in technology listings, and closer to something investors offered founders than something they resisted.

Alongside the mechanics, a set of norms developed that favoured founders more openly. Y Combinator, the accelerator founded in 2005 by Paul Graham and others, took small stakes in early-stage companies and placed itself firmly on the founder's side. The broader view that founders were hard to replace, and that investors should be "founder-friendly," had a basis in experience: the earlier habit of swapping founders out for professional managers had sometimes destroyed value. But as capital grew more abundant, the same view also lowered investors' willingness to exercise oversight, because oversight was now a cost that could lose them access to deals.

It is worth being precise about what the board seat did, because that is what was being given up. A director could see the company's real financial position, question its management, and, in the last resort, vote to replace the chief executive. That final power was the ultimate check on a founder. A dual-class structure put the decision beyond the reach of outside shareholders however much of the company they owned, and the absence of an investor board seat removed the more informal version of the same oversight. Between them, a founder could raise very large sums and remain, in practice, unaccountable to the people who had provided the money.

The board is the main formal mechanism by which investors monitor a company, and weakening it removed a check that served a purpose. The clearest documented case of what could follow was Theranos, the blood-testing company whose founder, Elizabeth Holmes, held roughly 99% of the voting power through super-voting shares. Its board included the former secretaries of state George Shultz and Henry Kissinger, the former defence secretary William Perry, the retired general James Mattis, and the former senator Sam Nunn: figures of considerable public standing and little relevant scientific or medical background. The company raised substantial capital at a valuation of around $9 billion, and the fact that its core technology did not work took years to surface, in part because its governance left no one both willing and able to press the question. Theranos was an extreme case, but the feature it relied on, concentrated founder control alongside a board unable or unwilling to challenge it, had become unremarkable.

None of this is to argue that founder control was always a mistake. Several of the companies that used these structures, Google and Facebook among them, were run well by founders who kept control, and the case that a committed founder can take a longer view than a rotating set of professional managers has evidence behind it. The same arrangements that let a capable founder act decisively also removed the means of correcting a founder who was wrong, and in a market where investors competed to offer those arrangements, the check was given up before anyone could know which kind of founder they had backed.

Section 3: Staying private longer

The second change concerned how, and when, companies were valued.

For most of the industry's history, a successful private company eventually needed to raise more capital than private markets could easily supply, and it went public. The IPO did two things: it raised money, and it produced a price. On the day a company listed, public investors, trading with their own money, set a value for it. That price could be unflattering, but it was determined by a market rather than agreed among insiders.

Two developments reduced the pressure to reach that point. The first was legal. The JOBS Act of 2012 raised the threshold at which a private company must register with the SEC and disclose its results, from 500 shareholders of record to 2,000. A company could now take on more investors and grow larger while staying private and not reporting publicly. The second was the abundance of late-stage private capital already described, which meant even large companies could raise what they needed without listing. Crossover investors such as mutual funds, hedge funds, and firms such as Tiger Global, took part in private rounds at rising valuations that no public market examined. In 2013 the investor Aileen Lee described private companies worth $1 billion or more as "unicorns," and counted thirty-nine of them; the word was meant to signal rarity, but the number ran into the hundreds within a few years.

The distinction that matters here is between a price and a valuation. A public market sets a price. A private round produces a valuation that is negotiated, and that can be shaped by the terms attached to the money. As covered in the previous episode, a liquidation preference lets an investor agree to a high headline valuation while securing the right to be repaid first in a sale, which lowers the investor's real exposure to that number. A company's stated worth in a private round was therefore not the same thing as a market price, and it could be held up, and raised, for as long as new investors were willing to fund it at higher figures.

Mutual funds are required to value their holdings regularly, including private ones, and in late 2015 and 2016 Fidelity and others wrote down their stakes in a number of well-known private companies, among them Snapchat and Dropbox before in some cases marking them back up. The individual write-downs were not large, but the valuation of a private company was an estimate, and two holders of the same shares could reasonably carry them at different prices at the same time.

These valuations also mattered to the people inside the companies, who were paid partly in stock options priced against them. When a company stayed private for a decade at a steadily rising valuation, its employees accumulated paper wealth they could not easily sell and that rested on the same negotiated figures. If the valuation was later marked down, or failed to hold at a public listing, the loss fell on them as much as on any investor. One consequence of all this was that companies stayed private for much longer. The median age of a US technology company at its IPO, around four to five years in the 1990s, roughly doubled over the following two decades. More of a company's growth, and more of the gains from it, now happened before it reached a public market. When a valuation was finally tested there, the adjustment was sometimes large and quick.

Section 4: SoftBank and the Vision Fund

The largest single instance of abundant capital was assembled by Masayoshi Son, the founder of SoftBank, whom we met in the previous episode as the investor who lost the most in the dot-com crash, a paper decline of around $70 billion, and who had earlier put about $20 million into Alibaba, a stake that eventually returned many times its cost. The lesson Son seems to have drawn from that record was that a single very large success could justify a great deal of aggression.

In 2016 and 2017 he raised the SoftBank Vision Fund, closing at $100 billion. For comparison, the entire US venture industry was raising something in the range of $30 to $40 billion a year at the time. One fund had gathered more capital than the industry raised in two or three years, which gave it the ability to set prices in the late-stage market rather than accept them.

The sources of the money reflected the period. Around $75 billion came from outside investors, most of it from two Gulf sovereign wealth funds: roughly $45 billion from Saudi Arabia's Public Investment Fund and about $15 billion from Abu Dhabi's Mubadala. Apple, Foxconn, Qualcomm, and Sharp each contributed around a billion dollars each, and SoftBank itself supplied the remaining $28 billion or so.

The fund's structure is also worth looking at. About 40% of the outside capital, some $40 billion, came in not as equity but as preferred units carrying a fixed annual coupon of 7%. This was, in effect, a debt-like obligation: the fund had to pay a set return on that portion each year regardless of how its investments were doing. A conventional venture fund can hold positions patiently, because it owes no periodic payment. A fund carrying a large fixed coupon has a reason to keep deploying capital and to support rising valuations, because it needs both cash and paper gains to service that obligation. The structure encouraged fast and continuous investment.

Son's approach followed from the fund's scale. Rather than making early bets on many companies and selecting among them, the Vision Fund tended to identify the leading company in a category and invest a very large sum in it, encouraging the company to spend heavily to establish a dominant position. Founders were reportedly offered more capital than they had asked for, and Son was said to be willing to fund a direct competitor if a company declined. This was a strategy built on the scale of capital rather than on selection: the intention was to make a company win by giving it more to spend than its rivals could match.

The fund deployed its money quickly and across a wide range of companies. Within roughly two years it had committed most of it, taking large stakes in more than seventy businesses, among them Uber, WeWork, the food-delivery company DoorDash, the UK chip-designer Arm, and the workplace-messaging company Slack. The cheques were large in absolute terms and, often, large relative to what the companies could sensibly absorb; one widely noted investment was a reported $300 million in a dog-walking application. The fund's presence also affected the wider market. A single buyer of that size, willing to pay high prices at the late stage, raised valuations for everyone, and other firms responded by raising larger vehicles of their own in order to compete.

Receiving capital on this scale also changed how a company behaved. A business handed far more money than it needed, and encouraged to spend it to capture a market quickly, tended to prioritise growth over profitability, to subsidise its customers, and to expand into new products and countries faster than it could manage them. For a company in a genuine race for a market that only one firm could win, this could be a reasonable strategy. For a company that was not in such a race, it mostly produced losses. The capital did not simply fund a company's plan; it shaped it. A strategy of this kind — cheap capital deployed at speed and scale in place of selection — will, by its nature, produce a wide range of outcomes.

Section 5: WeWork and the public-market test

WeWork is the case in which the features of the era were most visible, because the company tried to go public and, in order to do so, had to disclose how it was run.

Its business was the leasing of office space. It signed long-term leases on buildings, divided and fitted them out, and rented the space to companies and individuals on shorter terms. This carries an obvious risk, the long-term lease obligations remain when short-term tenants leave, which makes it vulnerable in a downturn. WeWork presented itself as a technology company rather than a property company, with a stated mission "to elevate the world's consciousness." SoftBank invested heavily through the Vision Fund, and a financing round in January 2019 valued the company at $47 billion.

In August 2019, WeWork filed the S-1 required for an IPO, disclosing its finances and governance for the first time. The economics were substantial and unprofitable: the company had reported roughly $1.8 billion in revenue in 2018 against a net loss of around $1.9 billion, and its losses were still growing in the first half of 2019. It also reported a non-standard measure it called "Community Adjusted EBITDA," which excluded not only interest and taxes but also basic operating costs, presenting the business more favourably than standard accounting would. There were related-party dealings: Neumann held interests in buildings that WeWork leased and had been paid around $5.9 million by the company for the trademark to the word "We" a payment later reversed. And control was heavily concentrated: Neumann held shares carrying twenty votes each, giving him a position the board could not override, together with provisions that involved his wife in choosing a successor and made him very difficult to remove.

Public investors responded poorly to the disclosures, and the valuation the company could realistically expect fell over the following weeks, from the $47 billion of the January round toward figures nearer $10 billion. WeWork cut Neumann's super-voting rights from twenty votes per share to ten in an attempt to save the offering; Neumann stepped down as chief executive on the 24^th of September and just days later the IPO was withdrawn. SoftBank then provided rescue financing that valued the company at around $8 billion, well below the $47 billion of a few months earlier. As part of the arrangement that removed Neumann from control, he received an exit package reported to be worth as much as $1.7 billion.

The sequence illustrates the difference between private and public valuation set out earlier. For as long as WeWork raised money privately, its valuation rose. The first time it was exposed to public disclosure and public pricing, that valuation did not hold. The governance that had given its founder near-total control, and which private investors had accepted, was among the things public investors would not.

Section 6: The same method, different outcomes

WeWork is often taken as evidence that the Vision Fund was simply a poor investor. That reading is incomplete, because the same fund, using the same approach, also produced one of the best outcomes of the period.

At around the time it was funding WeWork, SoftBank invested in ByteDance, the Chinese company that owns the short-video app now known as TikTok. ByteDance had been founded in 2012 by Zhang Yiming; its domestic app Douyin and its international version TikTok, grew quickly globally. A 2018 round valued ByteDance at about $75 billion, then the highest figure for a private company, and it went on to become one of the most valuable private companies in the world and a genuinely successful global product. ByteDance and WeWork received capital on similar terms and reasoning and produced very different results.

This is the expected consequence of a strategy based on scale rather than selection. When the method is to find a leading company and give it more money than its competitors can match, the outcome depends mostly on whether the company was actually capable of building a durable business. Some were, some were not, and a large fund deploying capital quickly across many such bets will back both kinds. The Vision Fund's record is not well explained by asking whether Son could pick winners; it is better explained by noting that, at that scale and speed, selection was not really the mechanism at work. Capital was.

For a fund, the distribution of outcomes matters as much as any single result. A traditional venture fund expects most of its investments to fail and relies on a few large successes to carry the whole, selection is the attempt to improve those odds. A fund that substitutes scale for selection is making a different bet: that writing large enough cheques across enough leading companies will capture the successes anyway. ByteDance suggests that this can work. WeWork, and a number of the fund's other investments, indicate that the cost of being wrong at that cheque size is correspondingly large. Whether the approach paid off is a question about the whole portfolio over time, not about either company on its own.

Uber sits between the two examples. SoftBank became its largest shareholder through an investment of around $7.7 billion at the end of 2017, in a company that was genuinely reshaping urban transport while losing very large sums to do it, and whose founder, Travis Kalanick, had been removed earlier that year in one of the period's rare instances of governance working as intended. Uber was neither a clear success nor a failure of the kind WeWork became. It was a large, real business financed to grow faster than its economics justified, which was characteristic of the era.

Section 7: VC and the new frontier

The two decades from 2000 to 2019 did not change the design of the venture fund. They changed the conditions under which it operated, and in doing so they showed how much of its discipline had rested on capital being scarce.

Each of the model's main features had been a response to scarcity. Fixed fund lives forced the return of a limited resource. Board seats reflected the value of capital that could command oversight in return. Hard terms and careful selection followed from there not being enough money to back everyone. When capital became cheap and abundant, these features became optional. Governance was often given up, because in a competitive market for deals, oversight cost investors access. Price discovery moved later, because companies could raise what they needed privately and postpone the point at which a public market set a value. Selection mattered less at the largest funds, because a fund of sufficient size could try to buy a market rather than choose within it.

Fund economics moved in step. A 2% management fee, applied to a $100 billion fund, produces $2 billion a year in fee income regardless of how the investments perform. The observation made in the previous episode, that it had become possible to earn a great deal from managing a venture fund without necessarily being good at venture investing, held at a much larger scale. The industry built to finance scarce, high-risk innovation had also become an effective means of gathering, and charging fees on, abundant capital, whatever the eventual returns.

The imbalance fell, as before, on the investors rather than the managers. LPs absorbed the losses when valuations were written down, while fees were charged on committed capital regardless of performance. This was the same asymmetry described in the previous episode, now operating at a larger scale and across a wider set of participants: not only pensions and endowments, but sovereign wealth funds and, through their mutual funds and retirement accounts, ordinary savers who had been drawn into the asset class in the search for returns that cheap money had made scarce everywhere else.

By 2019 the industry was larger and better capitalised than at any earlier point, and it had absorbed the failure of its most prominent single bet without structural change. WeWork's founder left with a large settlement, SoftBank raised a second Vision Fund, and pension money kept flowing into venture because its return targets still could not be met elsewhere. The open question at the end of the period was what would follow once cheap capital had made money itself abundant. If capital was no longer scarce and was increasingly similar from one source to the next, then the basis on which funds competed would have to move from the money to everything attached to it. That is the subject of the final episode The New Frontier. 

As capital comes to behave more like a commodity, the industry divides. At one end, large firms such as a16z build media operations, publications, podcasts, and content to generate the founder relationships that money alone no longer secures. At the other, a wave of emerging managers and solo GPs strips the model back to speed and founder alignment. We look at what remains and what the future of VC might look like.

What we’ve been working on at Shuttle

• Finalising on bringing US dealflow and investors to the platform 🌎

• Getting ready to onboard our new payment partner 🤝

• Updating our compliance flows for our UK launch 🧑‍💻

The Unsophisticated Investor is brought to you by Mark, Head of UK Expansion & Operations at Shuttle. Shuttle's mission is to break down the barriers to private markets and make wealth-building opportunities accessible beyond the ultra-wealthy elite. To do it, we're building Europe's private market infrastructure platform for running investment operations. By stripping away the legal friction and admin for deal managers, we make it easier than ever for high-performing opportunities to open up to a wider, modern network of investors.

Mark
Shuttle’s Head of UK Expansion & Operations