2019 appeared to be a banner year for the technology markets. The year ushered in a boom in public market liquidity, with unicorns such as Uber, Lyft, and Pinterest going public before the end of Q2. Total capital raised in 2019 IPOs was projected to rival the year 2000, the highest in history, when $97 billion was raised.  Uber’s IPO valuation alone was more than all 2018 tech IPOs combined. Spending on both consumer and enterprise technology was robust, reflecting overall market confidence. On the B2B side, the U.S. enterprise software industry expanded by 19% from 2016 to 2018, nearly twice the rate of the overall U.S. economy.  Total U.S. spending on consumer technology was projected to surpass $400 billion in 2019, the largest in history.
However, concerning signals began to arise as the year has worn on. Many of those celebrated IPOs performed poorly in the public markets. Since their respective IPOs, Uber is down 30%, Lyft is down 55%, Slack is down 40%, and Peloton is down 17%.  WeWork’s IPO became a debacle and was withdrawn.
The question everyone is now asking is whether current tech valuations are sustainable, or whether a downturn is imminent. Public SaaS companies have been trading at record highs, at an average of almost 11x EV/Revenue. The appreciation of Bessemer’s Cloud Index, shown below, demonstrates how valuations of SaaS companies have significantly outpaced appreciation of other market indices. The index has appreciated almost 500% in less than six years.
Goldman Sachs sounded an alarm in June, stating that “the valuation premium for growth is elevated today relative to history; software in particular now carries the highest multiples since the Tech Bubble.” Goldman noted that the tech sector carries a valuation premium more than two standard deviations above its 10-year average, across multiple metrics. 
So which one is it? Will the current robust technology market continue, or are we heading toward a tech market correction?
A close examination of both quantitative and qualitative data suggests that the tech market is reaching its peak, and that a correction is imminent.
More importantly, we may be underestimating the extent of a potential correction. This may be because we are using the incorrect metrics to compare it to previous corrections, in particular the crash of 2000, rather than evaluating the factors that are unique to this period.
We need to apply different analyses that take into account the amount of capital that has been invested in the venture capital industry; the size of companies at IPO and their impact on public markets; the quality of their underlying business models; and the new corporate governance reality that has emerged.
This article is an attempt to unpack those factors, explain how they may manifest themselves in the technology market over the next few years, and describe potential implications for technology investors and founders.
1. Weak Business Model Quality of Several Newly-Public Companies, Combined with their Large Size, May Trigger a Public Market Correction
Many industry participants have suggested that the current technology boom is fundamentally different than the Internet bubble of 2000, and they are correct in many ways. As an example, Dana Cimilluca of the Wall Street Journal cited the size and age of newly-public companies to suggest that today’s IPOs are of much higher quality than their dotcom brethren. Compared to 1999, the median newly-public company of 2018 is more than 3x older, and more than 10x larger in revenue, than the median newly-public company of 1999 and 2000.  The chart below from the Wall Street Journal, shows the significant growth in median revenue at IPO between 1999 and today. Others have cited additional factors such as the quality of customers and sizes of addressable markets to argue why today’s crop of IPOs is fundamentally better than that of the 2000 timeframe.
It may be true that that the businesses are of higher quality, but many of the factors cited are misleading. For example, revenue size in itself is not a marker of business quality, in particular because the private markets have subsidized the unprofitable revenue growth of many of today’s unicorns. It makes sense that today’s newly-public companies would be larger than those of 2000, when companies were going public before generating material revenue, but that doesn’t make them better businesses.
The most relevant metric for comparison between the periods is one that reflects business model quality: profitability. On that basis, the two periods have striking similarities.
The 2019 class of IPOs is actually the least profitable since the tech bubble of 1999. Only 24% of the 2019 technology IPO companies have been profitable, compared to 21% of the 2000 IPO class. 
Digging further into business model quality, many newly-public companies have low gross margins and sharply negative net margins, but are nonetheless being valued at over 10x revenues in the private markets. In his blogpost “The Great Public Market Reckoning”, Fred Wilson contrasts 75%+ gross margin companies such as Zoom, Cloudflare, and DataDog, all of which performed strongly in their public market debuts, with sub-40% gross margin companies such as Uber, Lyft, Spotify and WeWork, which have struggled.
Wilson makes the argument that there will be a public market reckoning for companies in the second group. These companies were perceived to be software companies and traded at sky-high multiples in the private market, but were actually not software companies at all. Wilson’s argument is another way of making the point that lack of business model quality, reflected in poor underlying profit margins, may doom many of the Unicorns of the current class.
WeWork is a perfect example. For years, WeWork was perceived as a “technology company” by the private markets. The company has raised more than $14.2 billion of capital as a private company, largely from Softbank’s Vision Fund. However, a review of the company’s S1 filing reveals a business with low gross margins (less than 20%) and questionable unit economics, which may never lead to net profitability (see chart below). Yet somehow the company was valued at greater than 15x its run-rate revenues in the private markets. The public markets sharply rejected it.
A key catalyst of the dotcom bust was the collapse of unprofitable public technology companies that went out of business. Certain of these companies are now remembered as representative of the excesses of the era: Webvan, Pets.com, Kozmo, Exodus. There are certainly fundamental differences between tech companies of that era and today. Companies went public much earlier at that time, and were valued based on astronomical assumptions of Internet growth and adoption that didn’t materialize as quickly as expected. These businesses didn’t have low gross margins; they had negative gross margins. Others were highly dependent upon having other dotcom startups as customers, which became a house of cards. 
While unicorns are not as dramatically overvalued on a multiples basis compared to 2000, they are much larger in size. This increased size means that it doesn’t require nearly as many to decline in value or go out of business to impact the market.
As shown in the table below below, in Q2 2019 the proceeds of the average IPO ($669m) were more than 3x the size of the average IPO of 1999 ($197m). And the differences are much more extreme when it comes to the largest unicorns. As an example, Amazon’s valuation at IPO was $300 million, while Uber’s valuation was $82 billion. Uber was valued at almost 300x Amazon’s market cap at IPO.
Because of this increased size, it would require a smaller percentage of public tech companies to decline in value or go bankrupt to have a sizable impact on the public markets today. To put it into perspective, Uber losing 50% of its market cap has a commensurate market cap reduction as 150 companies the size of Amazon going bankrupt back in 2000. What this means is that if the largest tech unicorns continue to experience weakness in the public markets, their poor performance could have a larger ripple effect than expected.
2. Dramatic Growth in the Amount of Venture Capital Raised by Funds and Deployed in Private Companies May Erode Overall Industry Returns, Despite the Power Law Distribution
There is a simple mathematical equation between capital invested in the private markets and the capital realized via public and private exits. What goes in, must come out. Not only that, capital must come out at a multiple of what went in, to account for the risk.
When the amount of capital returned via realizations substantially exceeds the amount invested in funds, the market corrects itself as investors increase allocations to the asset class.
When the amount of capital returned is substantially less than the amount invested, the market corrects via an exodus of capital out of an asset class.
However, there is a long lag time in this equation when it comes to the venture capital markets. This is due to the delay between the time when capital is invested in venture funds, then subsequently invested in private companies, and then finally realized via exit events. The typical life cycle of a venture fund is ten years, and it can actually take twelve to fifteen years for a fund’s total returns to be realized.
This creates a long feedback loop in the mathematical equation. This lag creates cyclicality in the market, as capital flows take time to equilibrate to market realities. This is reflected in venture capital investment fluctuations, as the chart below shows, and makes the venture capital business subject to boom and bust periods. 
The amount of capital raised by VC funds has grown exponentially over the past few years. According to Pitchbook, $135 billion of capital was invested by VC firms in 2018, and 2019 is on pace to exceed that figure. These levels are higher than the year 2000, the previous peak, when $120 billion was invested into venture-backed companies. 
To generate a 3x multiple of money on the amount of capital invested by VC firms in 2018, there would need to be $400 billion of capital realized by venture firms. Assuming VC’s own roughly 50% of total equity per company on average at exit, that would require $800 billion of exit value to yield the required returns.
This seems improbable, given historical patterns of total exit value. To put this in perspective, the first half of 2019 was one of the highest ever for exit value, with $190 billion of exit value.  That is an annualized level of $380 billion of exit value in a peak year for the technology markets.
Even if VCs owned 100% of all exited companies, the class still wouldn’t reach a 3x overall multiple when comparing the amount of venture capital raised in 2018 to cumulative exit values in 2019, a peak year. Future years would have to generate a much greater level of exit value.
This doesn’t mean that all VCs will have poor returns for vintage years between 2017 and 2019. Venture capital follows a power law distribution, where the best funds capture the majority of returns of the asset class. This can be seen in the chart below, which shows the historical Cambridge Associates venture capital benchmarks from 1994 to 2017. As can be seen, top-quartile funds significantly outperform lower-quartile funds across periods.
However, during periods of correction even the returns of the best funds are impacted. The differences in performance between the top funds and the mediocre funds narrows, and a falling tide lowers all boats.
3. Relaxed Financial Controls and Corporate Governance Have Made it More Difficult to Curb Excesses
An additional signal of an overheated private market is corporate governance controls becoming less stringent. Why does this occur? In a market where capital is scarce, investors can insist on governance controls because they have leverage. In a market where capital is abundant, entrepreneurs who don’t want these controls can seek alternative capital from investors who don’t require them. To win deals, investors are willing to set these controls aside.
Insistence on corporate governance controls often follows bust periods, where investors learn from their sins. But the memory of venture capital investors can be short, especially when boom periods return. In the words of Bill Gurley of Benchmark Capital, describing the prevailing mood of investors in Silicon Valley, “I’ve never met a group of people where risk is forgotten so quickly.”
One way that relaxed corporate governance has manifested itself in the current market is via dual-class voting shares. In these structures, founder shares have super-voting rights, typically 10x or 20x the votes of non-founder shares. As an example, in the case of Lyft, the founders owned only 5% of equity at IPO, but owned 49% of voting shares due to super-voting shares.
Many articles have been written about the incentive problems associated with dual-class voting structures in technology companies.  These structures may lead to value destruction, due to founders making economic decisions that benefit them, but do not benefit shareholders as a whole. Furthermore, these structures can be maintained for long periods, even when founders are no longer active in the business or best positioned to make key decisions on behalf of all shareholders.
Kosmas Papadopoulos of ISS Analytics conducted a comprehensive study of dual-class voting share structures across industries.  He found that the percentage of dual-class voting share structures had surged in recent years, in particular in large tech IPOs. As an example, for the period of January to May of 2019, seven of the 10 largest IPOs had dual-class voting structures. 
Not surprisingly, these structures have been correlated with negative results, including lack of independent board leadership, a higher rate of CEO related-party transactions, failure to disclose director evaluation process, and lower gender diversity at the board level, as shown in the table below.  As a result, the S&P and Russell index providers announced policies to partially or fully exclude companies with multiple share classes from their indices , but technology investors continue to allow them in private rounds.
The implications of relaxed corporate governance are that it’s more difficult to curb corporate excesses and prevent corporate fraud. We’ve witnessed this over the past few years with companies such as Theranos and WeWork. There may be more examples to come over the next few years.
4. How These Factors May Result in a Technology Market Correction
Several top tech investors have predicted a technology market correction for some time. Bill Gurley, for example, has been arguing for almost five years now that we are in a bubble, due to excessive capital and high valuations. So why hasn’t it happened yet?
The first reason is that the public markets are a much stronger corrective mechanism than the private markets, and companies are only now reaching the public markets due to the huge amount of capital that has been available in the private markets.
Founders and investors have argued for several years now that the proliferation of late-stage funding has been a good thing, because founders can grow their businesses without the constraints of reporting public earnings and short-term oriented investors. They also argue that the private markets are better at fostering innovation and risk-taking.
This is all true. At the same time, we are finding that the private markets aren’t particularly good at assessing the intrinsic value of business models that aren’t intrinsically technology-driven. Software may be eating the world, but not all companies should be valued like software companies, especially capital-intensive companies with low gross margins. The glare of the public markets appears to be triggering the pricing correction that the private markets haven’t been able to provide.
The second reason for a potential correction is that the long feedback loop described above will soon begin to kick in for venture funds raised between 2010 and 2015. These funds are now approaching years five to ten of their vintages, and LP’s are beginning to look for cash-on-cash returns.
Poor performance by the largest venture funds would have the biggest negative impact on the markets. The Softbank Vision Fund, the largest of them all, is having difficulty raising its second fund , due to the recent WeWork debacle as well as the poor post-IPO performance of Uber and Slack.  The Vision Fund appears particularly exposed to companies with low gross margins and a lack of profitability.
The third reason is due to the how corrections happen. While technology markets overheat very gradually, they correct sharply. Bill Gurley refers to this as the “boiled frogs” phenomenon.  According to Gurley, in a bull market investors gradually take on more risk via higher valuations and more risky deals, but they don’t realize they are slowly losing their risk aversion.
In his words, “Over a 5 year period your VC firm has taken on a tremendous amount of risk. But every day you just moved a little bit so you never felt you were making this massive gap in risk exposure. When markets bust, risk aversion comes on immediately. Overnight. Boom!”
5. Important Implications for Technology Investors and Founders
There are many compelling arguments for why this period is fundamentally different than the dotcom bust, and why a tech market correction will be much more moderate in size in the coming years. At the same time, founders and investors should consider what a new normal might look like after this long period of market expansion.
For founders who have built and scaled their companies in these exuberant times, you may need to make important changes. You will likely have less time to figure out how to make your business model profitable. Once the current private capital excess has been washed out of the system, funding terms from may be more restrictive, and capital may be harder to come by. Investors will be less likely to apply software multiples to companies that are capital-intensive and have low gross margins. Corporate governance controls may tighten.
You may also have to do more to boost employee morale and retain top people. Employees who have gotten used to seeing their stock options increase in value, due to ever loftier private rounds, may see them reduced in value, or even rendered worthless.
For technology investors, norms may likely be different. Mark-ups will be more scarce, and funds may be more difficult to raise. Cram-down and pay-to-play provisions rounds may be more common. Funds will likely be more difficult to raise, and many firms may leave the market, as occurred in 2001.
However, there are bright spots to a tech market downturn. For investors, valuations and deal terms become more favorable. Cash burn rates are reduced. Competition will lessen as more firms leave the market.
For exceptional founders, the silver lining is that great technology companies can be built throughout market cycles. As we explained in an article published earlier this year, billion-dollar technology companies are founded throughout both boom and bust periods, and the number of companies hasn’t differed significantly across periods.  In particular, during periods of recession, technology companies that enable enterprises to become more efficient and consumers to reduce costs and will thrive.
- See “IPOs have their best quarter in years in terms of performance and capital raised”, CNBC, at https://www.cnbc.com/2019/06/28/ipos-have-their-best-quarter-in-years-in-terms-of-performance-and-capital-raised.html.
- See “Corporate Tech Spending Helps Lift the U.S. Economy,” Wall Street Journal, at https://www.wsj.com/articles/corporate-tech-spending-helps-lift-u-s-economy-11569367000.
- See “What Really Fueled the 2019 Unicorn IPO Funeral”, Yahoo Finance!, at https://finance.yahoo.com/news/whats-really-behind-the-ipo-unicorn-funeral-142210130.html.
- “Goldman Sachs is sounding the alarm: Tech stocks are overvalued,” USA Today, at https://www.usatoday.com/story/tech/2019/06/17/goldman-sachs-says-technology-stocks-overvalued/1483689001/.
- See “The 2019 IPO Frenzy is Different from 1999. Really.” https://www.wsj.com/articles/the-2019-ipo-frenzy-is-different-from-1999-really-11553918401.
- Kate Rooney, “This Year’s IPO Class is the Least Profitable of Any Year Since the Tech Bubble,” CNBC.com, at https://www.cnbc.com/2019/09/18/this-years-ipo-class-is-the-least-profitable-of-any-year-since-the-tech-bubble.html.
- See “Startups Selling to Other Startups: A House of Cards,” TechCrunch, at https://techcrunch.com/2016/02/11/startups-selling-to-other-startups-a-house-of-cards/”
- “Quarterly U.S. Venture Capital Investments 1995–2017”, Wikipedia, at https://en.wikipedia.org/wiki/Venture_capital#/media/File:US_VC_funding.png.
- After reaching this high-water mark, the venture capital market cratered in 2001, with only $40B invested, and further shrunk to a little more than $20B invested in 2002 as the technology markets collapsed.
- See “Venture-backed Exit Activity Hit a Quarterly Record $138.3 Billion in 2Q 2019”, https://pitchbook.com/media/press-releases/venture-backed-exit-activity-hit-a-quarterly-record-1383-billion-in-2q-2019.
- Two articles are “The Perils of Lyft’s Dual-Class Structure”, by by Lucian Bebchuk and Kobi Kastiel, at https://corpgov.law.harvard.edu/2019/04/03/the-perils-of-lyfts-dual-class-structure; and “The Perils of Pinterest’s Dual-Class Structure”, by the same authors, at https://corpgov.law.harvard.edu/2019/04/10/the-perils-of-pinterests-dual-class-structure/.
- Kosmas Papadopoulos, “Dual Class Shares and Company Performance”, at https://corpgov.law.harvard.edu/2019/06/28/dual-class-shares-governance-risks-and-company-performance/.
- See “S&P and FTSE Russell on Exclusion of Companies with Multi-Class Shares”, Camberview Partners, at https://corpgov.law.harvard.edu/2017/08/05/sp-and-ftse-russell-on-exclusion-of-companies-with-multi-class-shares/.
- See “Softbank’s Plans for Second Mega-Fund Hit by WeWork Debacle”, Reuters, at https://www.reuters.com/article/us-softbank-group-visionfund-insight/softbanks-plans-for-second-mega-fund-hit-by-wework-debacle-idUSKBN1WJ0AA.
- See “Masa’s multi-generational vision is running into a brick wall: the public markets”, at https://www.cnbc.com/2019/09/07/uber-wework-and-slack-public-valuations-show-softbank-vision-flaw.html.
- See “The Twenty Minute VC: Bill Gurley”, transcript at https://davidsnotes.substack.com/p/bill-gurley-on-why-the-abundance-212.
- “The $1.7 trillion dot.com lesson,” CNN Money, at https://money.cnn.com/2000/11/09/technology/overview/.
- See “The 2019 IPO Frenzy is Different from 1999. Really,” Wall Street Journal, at https://www.wsj.com/articles/the-2019-ipo-frenzy-is-different-from-1999-really-11553918401
- See “3 Things Founders & VCs Should Know About Building Billion-Dollar Startups During Market Uncertainty,” 645 Ventures, at https://medium.com/@645ventures/3-things-founders-vcs-should-know-about-building-billion-dollar-startups-during-market-fdabeb72044e.