The IPO Class of 2021, Five Years of Hindsight Later

The Biggest Window Ever

a IPORoughly 400 traditional IPOs raised more than $140 billion both records and that count excludes the parallel SPAC boom covered in its own entry which more than doubles the total number of companies thatEverything a founder could want was in place: indexes at all-time highs interest rates at zero retail investors flush with stimulus savings and a market paying historic multiples for growth

Five years later the class of 2021 is one of the best natural experiments in market history because we can rate an entire cohort that was listed exactly at the top of a valuation regime

The Yearbook

The class had real star power. Coinbase the cryptocurrency exchange went public in April via a direct listing and was briefly valued near $86 billion the coming-out party for the entire cryptocurrency industry. Rivian an electric truck maker raised about $11.9 billion in November the largest U.S. initial public offering since 2014 and within days its market value surpassed $150 billion.dollars more than Volkswagen General Motors or Ford. At the time Rivian had only delivered a token number of trucks. Robinhood the trading app that fueled the meme stock era listed in July. Dozens of software fintech and consumer names emerged around it priced at revenue multiples that assumed flawless execution forever

A Worked Example: What 150 Billion Dollars Was Actually Assuming

Calling a rating crazy is easy and foolproof. Reverse engineering is better because it turns an opinion into a testable statement about the world. Here's Rivian's number worked backwards

You can't use a price-to-earnings ratio because there were no earnings and you can't use a price-to-sales ratio because there were essentially no sales. So ask the question the other way around. Instead of valuing the company ask what the company would have to become for $150 billion to be a fair price

Step one: Choose the multiple the company ultimately deserves. Auto manufacturing is capital-intensive cyclical and competitive and mature automakers trade on a fraction of their revenue. Ford has typically traded at between 0.3 and 0.4 times sales. Be generous with Rivian and assume it gets a premium multiple of 0.5 times sales better than any legacy manufacturer

Step two withdraw the required income. 150 billion divided by 0.5 gives $300 billion in annual revenue

Step three: convert revenue into trucks. At an average sales price of about $70,000 $300 billion in revenue is 300,000 divided by 70 which is equivalent to about 4.3 million vehicles per year

Now put those numbers next to reality. Ford sells about four million vehicles a year worldwide. General Motors sells about six million. Rivian as of November 2021 had delivered a few hundred

So the $150 billion price tag didn't say Rivian would succeed. It said Rivian would reach roughly Ford's scale with a better valuation multiple than Ford has ever sustained and it said it before the company had proven it could build vehicles at any volume. Even the most flattering row on that table four times higher than any legacy multiple still requires more than a million trucks a year

None of this required knowing anything about Rivian. It required a multiple assumption an average selling price and two splits all of which were available to anyone on the day of listing. That's what makes reverse engineering valuable. It doesn't tell you that the price is wrong. It tells you exactly what you would have to believe for the price to be correct and then you decide whether to believe it

The Grades Come In

The repricing began within a few months. The Renaissance IPO Index which tracks recent listings fell about 57 percent in 2022 about triple the drop in the broader market and many individual names in the class fell 70 to 90 percent from their highs. The mechanism is what this series keeps coming back to. Public growth startups are the longest-lived assets in the market almost all of their value lies in the profits expected many years from nowso the 2022 rate shock affected them more than anything else in the market

But the fees were only half the story. The other half was that 2021 prices had assumed that pandemic-era growth rates were permanent. Companies that go public in a window of euphoria are by selection those whose bankers judged the timing to be optimal meaning that investors were systematically offered shares at the maximum chosen by the sellers. That's not a scandal it's the structure of the transaction and 2021 is the cleanest demonstration.ever recorded. The people with the most information about the company its founders and its first investors chose that moment to sell. The buyers should have asked why

Case Study: Webvan and Being Right Too Expensively

The class of 2021 has an ancestor and the resemblance is close enough to be uncomfortable

Webvan was an online grocery delivery service that went public in November 1999 and raised $375 million. The stock rose sharply on its first day and the company reached a valuation in the billions on revenue measured in a few tens of millions. It was building automated warehouses and a delivery fleet in several cities simultaneously financed entirely by capital markets that over a period of about eighteen months were willing to finance anything that had a growth story.attached

It declared bankruptcy in July 2001 about twenty months after going public. Investors lost virtually everything

Here's the part that matters. Webvan was right. Online grocery delivery with automated fulfillment is now a huge business. The company had identified a genuine change in the way people would eventually buy food and built real infrastructure to serve it. Amazon subsequently hired executives from Webvan and pursued a strategy that rhymed with the original idea

What killed Webvan was not being wrong about the future. It needed an uninterrupted supply of cheap capital to reach that future and trade at a valuation that assumed the supply would never be interrupted. When the window closed in 2000 a company burning cash in warehouses in six cities had no path through the stock market and the stock market had gone home

Compare that to a pre-revenue automaker valued at $150 billion in November 2021 which needed to finance factories supply chains and a network of services outside of capital markets that were about to face the fastest rate hikes in four decades. The thesis about electric vehicles could be entirely correct and the investment could still fail for reasons that have nothing to do with the thesis being correct

Capital intensity plus a closing window is the specific combination that destroys companies that were right about everything else

The Survivors and the Lesson in the Spread

Five years later the class has not uniformly failed and diffusion is education. Companies with real unit economics network effects or category leadership survived the downsizing rebuilt and in several cases finally justified meaningful valuations. Those priced purely based on history pre-revenue automakers growth-at-any-cost software consumer apps with no path to margin mostly stayed low were acquired cheaply orThey disappeared. The 2021 price said almost nothing about which group a company belonged to. The business model told you almost everything

That's the practical skill hidden in this retrospective. Reading an IPO prospectus for burn rate unit economics and insider selling behavior would have separated the two groups in advance much better than any price chart and still does

Where the Insiders Sold at the Top Story Breaks

I wrote previously that more informed people chose that time to sell and that buyers should have asked why. That framing is popular it's partly true and it's vaguer than it seems

In a traditional IPO most insiders did not sell. The vast majority of shares sold in a conventional offering are primary shares issued by the company with the proceeds going to the corporate treasury rather than the founder's pocket. Insiders are often subject to lock-up agreements for six months afterwards meaning they could not have sold at the top even if they had wanted to and many saw their own holdings fall by seventy percent without being able to act contractually. The clearest counterexample is Coinbase which used a direct listing precisely so that existing holders couldsell and it's worth noting that structural difference rather than lumping all 2021 quotes together

The IPO buyer and the loser are usually different people. Stocks quoted in the offer often rose substantially on the first day. Investors who bet on the offer price did well in many cases. Losses were concentrated among people who bought in the secondary market at the close of the first day or later which is a different transaction with a different counterparty and blaming the issuer for it confuses both

Windows also opens for real reasons. A portfolio of companies had been private for a long time and in 2021 a real backlog was cleared. Some of those companies were mature and were going public because it was time not because a banker discovered a top. Treating every listing in a hot year as evidence of cynicism is a heuristic that would have also deterred you from some of the best companies of the decade

The index is not the cohort. A 57 percent decline in a capitalization-weighted index of recent listings is dominated by the few names that were larger which in 2021 were among the most extreme. The midsize company in the class did poorly. It didn't do too badly

My view is that the signal of the moment is real and weaker than the count and that the lasting lesson is about price and capital intensity rather than who was selling

How I Actually Read an IPO Prospectus

The prospectuses are long and almost all of their length is legal text designed to be skipped. I read them in a fixed order that arrives quickly at the decision

I start with using earnings because it tells me what the company thinks it needs. Money going to selling shareholders is a different signal than money going to the business and money going to pay down debt is also a different signal

Then I go to the cash flow statement instead of the income statement and calculate the months of operating loss against the pro forma cash after the raise. Webvan is the reason. A company with eighteen months of history and a capital-intensive build is making a bet on the capital markets not on customers and I want to know that before I read a word of the growth narrative

Next I look for a cohort table if there is one. Companies with really good retention publish cohorts because it flatters them. Companies without them have a reason

Then I do the reverse valuation of the previous worked example on the back of an envelope. One multiple one unit price two divisions and I have a sentence of the form: At this price I assume that this company is approximately the size of that company. If I can't say that sentence out loud without laughing I'm done

The last thing I check is the lockup expiration date and the share count including anything that hasn't been exercised because dilution is when a good business quietly turns into a mediocre investment

That's how I would approach it and it's a description of the method rather than advice on a particular list

The Bottom Line

The class of 2021 entered the most generous market in history and spent the next few years being rated by the least generous. Record-breaking IPO volume was itself the signal and buyers who treated an active listing window as validation rather than a warning paid the tuition

Rivian is the arithmetic to remember. A $150 billion valuation calculated back to a generous half of sales implied about 4.3 million vehicles a year roughly equivalent to Ford. That calculation required two divisions and was available to anyone on listing day. Webvan is the other half of the lesson: the idea was right online grocery delivery is now huge and the company went bankrupt in twenty months because it needed a capital window that closed

Judge companies by their economics and judge windows by who is selling in them while remembering that in a traditional IPO insiders were mostly locked in and people who lost money usually bought on the second day instead of the first. The Class of 2021 teaches all that better than any textbook