Aileen Lee, Venture Capitalist and founder of Cowboy Ventures, coined the term “Unicorn” in 2013 - a mythical animal chosen intentionally to depict the scarcity of finding a $1 billion company in the early 2010’s - of which there were only thirty-nine at the time.1 Let’s just say several other terms have been invented since then; Decacorn ($10+ billion), Centaur (100+ million in Annual Recurring Revenue), Centicorn ($100+ billion) and finally Terracorn ($1+ trillion). In just thirteen years, the number of Unicorns has surged to over 1300 as of March 2026.2 In this article I will be unpacking the crucial market dynamics that led to the Venture Capital boom we are currently living in.

Cheap money and the risk curve

The attributes of “extremely rare, and magical” which once defined Unicorns have now been diluted, but I would like to take you back to 2013, when this was not the case, and instead a slow and stagnant recovery from Great Financial Crisis was in process. Desperately trying to boost economic activity in the US, interest rates remained low (at 0.11%), and even though the recession was five years prior, unemployment was still elevated.3 Low-interest rates aimed to encourage US businesses and citizens to take out loans, therefore increasing the volume of trade and transactions across the country rather than sitting on cash piles. This fuelled one sector in particular, and it all came down to one metric: the risk premium.

As five years had passed by since the Financial Crisis of 2008, the “easy money” in public stocks was already long gone, where the S&P 500 could be seen doubling in this time period. Investors now had to think outside of the box for new opportunities they could prosper from in the next 2-5 years, and, knowing that interest rates would soon be on the rise, it was a race against the Federal Reserve. The next obvious choice would be real-estate, and, aside from the fact that this was the literal cause of the crash, expected returns were not forecasted to be as lucrative as other investments.

A natural shift towards private companies came about after the JOBS act of 2012, where previously companies with over 500 shareholders, a count that had included employees as shareholders, would have to release quarterly filings and an annual report, urging companies to go public earlier as it made very little difference administratively. However, through legislation, the threshold had increased to 2000 shareholders, now excluding employees exercising stock options, meaning that companies were able to stay private for longer.4

Investors subsequently reached the conclusion that with more companies being in the private sector, this could open up a new opportunity. They were happy to venture further out on the risk curve as they only had to beat the current 0.11% interest rate to outperform savings accounts. With Facebook’s IPO in 2012 at a monster valuation of $100 billion, and LinkedIn going public a year prior, it was clear that tech was going to be a winner.5 Encouraged by Marc Andreessen’s (Founder of NetScape and A16Z) “software is eating the world” article, investors crowded the software industry in a competitive brawl to find the next Unicorn in the industry.6

The maths behind the boom

With momentum acquired, investing now came down to the numbers, where the DCF was the gold standard of valuation. The benchmark rate is usually considered to be the US 10-year Treasury yield, which was at 1.91% in January 2013, with the bonds treated as the risk-free asset class.7 This risk-free rate is step one in forming the infamous discount rate which is used as a measure of shrinking the future cash flows. It is representative of the return you could get on your money at the time with zero risk, through risk-free asset classes such as the 10-year US Treasury Bond, and by the end of 2013, this yield rose to 2.90%.7

This made start-ups look less attractive for investors, as investments with no risk were returning increasing yields. Where the risk-free rate rose by about 1 percentage point, the risk premium (component number 2 in forming the discount rate) fell. Essentially, this is the extra return an investor requests based on how risky the deal is. Usually, a premium would be added depending on the risk or uncertainty of a company, a modest premium of 2-5% for a proven business and even up to 40% for a start-up. For years, the risk premium had been a mathematical deterrent for many investors. Because the risk-free rate typically moves within a narrow range of a few percentage points, it was the size of the risk premium, not the risk-free rate, that usually decided whether institutional investors thought a start-up's future revenue justified their backing.

With the Fed holding overnight lending rates near zero, cash and other safe short-term assets were paying investors next to nothing, so to generate any meaningful return at all, investors had to venture out on the risk curve. This therefore meant accepting a smaller premium for taking on the same level of risk, pushing investors out of their usual comfort zone and into start-ups. The risk for investing in start-ups did not change, but the risk premium did, since many were craving greater returns

This significantly decreased the overall discount rate, despite the rising risk-free rate. But crucially, this discount rate was applied long-term due to the trajectory of most start-ups, and in the calculation of discount rate, the letter “n” represents the number of years you predict the cash flows for. Knowing this, start-ups’ long-term cashflows were less heavily discounted and a significant portion would be retained – implying a higher valuation. The rise in valuation of start-ups coinciding with the influx of investors entering the market boosted demand for tech start-ups specifically. Tied with the social aspect of Marc Andreessen’s famous article, a mindset shift had been instilled in investors’ minds and the technical aspect, via the DCF model, confirmed it in the numbers.

SoftBank and the arms race

By 2015, the model was in full swing. It had done exactly what the numbers had suggested. Aileen Lee’s original thirty-nine “rare and magical” unicorns had more than doubled to eighty-four, yet the Venture Capital world was not ready for what was to come. 2013 to 2015 had been the revelation for mutual funds, hedge funds and institutional investors, but in 2017, a man named Masayoshi Son – founder of SoftBank and briefly the richest man in the world in February 2000 – through SoftBank’s dot-com-era stock price, was ready to take another bet in the market.8 A new, record-breaking VC fund, valued at roughly $100 billion.9

The tactic here was simple, this was not a fund that encouraged organic growth in small companies, instead, Son inflated the company with $500m to $1 bn cheques to outsize all other competition in the market. Just through pure capital dominance, any of Son’s bets were destined to come out on top. This was the exact moment which signalled price was catching up with demand.

To further fuel this huge Venture Capital resurgence, Sequoia followed swiftly. Not wanting to be outmuscled by Son again, as his previous $100 million investment into Yahoo! significantly diluted their much smaller stake in 1996, Sequoia began raising its third Global Growth Fund.10 Ultimately raising $8 billion, this was nothing compared to Son’s enormous cash pile.11 The money wasn’t necessarily the most important factor here, the shocking thing was that Son’s fund created a competitive environment where, if no one responded, his fund would solely dominate what looked like a promising future. The interlinked nature of investors raising money and companies raising more frequently and at higher valuations was repeated process carried out for years, gradually shifting power towards the founders rather than the VC firms who could be seen comically squabbling over deals. The weight in the industry was shifted so the investment firms began to outnumber the founders.

This notion was most evident in Adam Neumann’s WeWork, a real estate company leasing office space to SME’s. Although not tech-related in the slightest, SoftBank Group’s Vison Fund poured billions of dollars, fuelling it round after round. This move revolved around the same logic where enough capital deployed as quick as possible would result in market leadership along with a commotion that attracted a sea of customers. By January 2019, this approach had catapulted WeWork’s valuation to above $45 billion, built on SoftBank’s relentless injection of capital instead of reflecting intrinsic value.12 All was fine until WeWork’s attempt to go public in the latter of 2019, where, for the first time, external analysts and forensic accountants reviewed the numbers in financial statements. Their valuation subsequently collapsed by more than 80% in a matter of months.13 This was proof that although Venture Capital money could raise Private market valuations, this was merely superficial, and for valuations to hold, stronger underlying business output was needed.

When the money stopped

After years of smooth sailing and low interest rates, the market had finally caught up. The Federal Reserve was in dangerous territory, facing the highest inflation in four decades and coming out of a world pandemic, so it began raising rates at the fastest pace since the 1980’s.14 Venture Capitalist’s saw what was once their golden window of opportunity rapidly closing. That discount rate which had mathematically convinced them to pursue start-up investments now argued in the opposite direction. The high interest rates favoured secure businesses with stable cash flows and the valuation of riskier public tech companies collapsed, while private companies held on by a thread waiting on a boost of capital from large Venture Capitalists to kickstart a new cycle of Venture Capital. But SoftBank’s Vision fund, the very engine that powered the industry’s decade-long run, posted some of the largest quarterly losses in Japanese history.15 It seemed as if the negative feedback loop between Venture Capitalists and start-up founders had become static. But by 2023 there were signs of hope: an even stronger engine this time.

A new engine

This vehicle was not powered by mathematical reasoning or low interest rates, but by the potential for historic productivity increases in the form of large language models (LLM’s). In November 2022, an out-of-the blue release from OpenAI of ChatGPT caught users by surprise.16 Although large language models had been built for decades, the difference was that Sam Altman’s team utilised architecture (Transformer architecture, introduced by Google) which allowed AI to read pages of text simultaneously instead of processing one word at a time.17 Secondly, while older models were processing hoards of information from the internet that was unfiltered, including incomplete thoughts and sarcasm, causing inadequate responses to your prompt, reinforcement learning from human feedback was used in ChatGPT to teach the models how to interact naturally with humans. This also helped to commercialise the LLM, as the revolutionary human factor did not exist before, acting like a digital personal assistant. Finally, ChatGPT was a huge success as it was built at a time period when Nvidia GPU’s (graphical processing units) were exponentially growing in terms of processing power, allowing gargantuan data sets to be consumed to train the model.

Bubble or breakthrough?

From this major market move, new start-ups began to emerge at an increasing rate, this time however, valuations sky-rocketed. Even pre-revenue, with a mere idea to incorporate AI into their business, founders were raising at multi million dollar valuations. The early signs of a bubble were clear, and the funds were buying into this technology, but whether the provision of value matched these insane valuations was a question yet to be answered. VC’s appetite for risk had completely shifted since 2013. Another notable difference included the larger institutions backing this market move. The mutual funds were now happy to buy into the more steady public market which had rallied after COVID, but now the infrastructure providers were the ones financing these projects, recycling cash they received from LLM providers in the form of chip purchase. The behemoths providing computational power to fuel the growth of the industry, such as Microsoft and Nvidia, were simultaneously pumping cash into it at unprecedented rates.

Not only would this money trickle down to smaller start-ups in the same niche, but OpenAI and Anthropic would go on to develop frontier tools which allowed non-technical entrepreneurs to code via an agent, further lowering the barrier to entry for founders.

The result is a cycle of circular funding that feeds itself: the infrastructure giants fund the model builders, the model builders buy those very infrastructure giant’s chips, and the tools they create spawn a new wave of start-ups competing for the same capital. And that is both where the opportunity and problems of the future lie.

Is all of the capital in one insulated bubble? And will that bubble deflate or pop?

Whether this ends like 2021, with valuations subtly marked down, or like the early internet, where only a few survivors justified the hype many times over, will depend on one thing the market has so far been willing to overlook: whether these companies can turn attention into revenue. The unicorn, once a rare creature, is no longer rare. The question now is how many of them are real.

Footnotes

  1. TechCrunch, Welcome To The Unicorn Club: Learning From Billion-Dollar Startups (opens in a new tab), 2nd November 2013.

  2. CB Insights, The Complete List of Unicorn Companies (opens in a new tab).

  3. Federal Reserve Bank of St. Louis, Federal Funds Effective Rate (opens in a new tab).

  4. U.S. Congress, H.R.3606 – Jumpstart Our Business Startups Act (opens in a new tab), 5th April 2012.

  5. The Christian Science Monitor, Facebook to debut at $104 billion (opens in a new tab), 17th May 2012.

  6. Andreessen Horowitz, Why Software Is Eating the World (opens in a new tab), 20th August 2011.

  7. Federal Reserve Bank of St. Louis, Market Yield on U.S. Treasury Securities at 10-Year Constant Maturity (opens in a new tab). 2

  8. The Spectator, The rollercoaster ride of the world's most reckless investor (opens in a new tab), 19th October 2024.

  9. SoftBank Group, SoftBank Vision Fund's First Major Closing (opens in a new tab), 22nd May 2017.

  10. Global Venturing, The 10 best corporate VC investments of the digital age: Part One (opens in a new tab), 1st June 2023.

  11. Bloomberg, Sequoia Capital Raises Bulk of $8 Billion Venture Fund (opens in a new tab), 26th April 2018.

  12. Bloomberg, WeWork Rebrands Itself as SoftBank Commits Smaller Investment (opens in a new tab), 8th January 2019.

  13. Bloomberg, SoftBank Unveils $9.5 Billion WeWork Rescue, Gets 80% Stake (opens in a new tab), 23rd October 2019.

  14. Federal Reserve, Federal Reserve issues FOMC statement (opens in a new tab), 16th March 2022.

  15. CNBC, SoftBank posts a $21.6 billion quarterly loss on its Vision Fund, one of the highest in its history (opens in a new tab), 8th August 2022.

  16. OpenAI, Introducing ChatGPT (opens in a new tab), 30th November 2022.

  17. arXiv, Attention Is All You Need (opens in a new tab), 2nd August 2023.