Should IPOs Be “Exits?”
I have been investing in private companies for almost 13 years and started Four Rivers (www.fourriversgroup.com) in 2007 to invest in high growth, venture backed companies demonstrating market leadership. As a venture investor, I have always been taught to view IPOs as “exits”, or opportunities to sell these private holdings once the customary 6 month lockup period expires. Conventional wisdom indicates that this milestone in a company’s lifecycle represents a natural place for the company to turn over its investor base, trading out its private investors for public ones.
Now that I have gone through a handful of IPOs, I find myself asking a simple question: Should IPOs necessarily represent “exits”? Is opportunity being passed up? At Four Rivers, we decided to pose the question and answer it empirically, capturing both booms and downturns. We looked at every single IPO of IT companies backed by 19 top venture funds since 2002 (we deliberately chose to avoid the 1996–2000 dot com bubble) and generated a data set of 158 unique companies.
The Data Set
For inclusion in the data set, we selected VC firms that had at least 4 IPOs during the period, and we included only IT and consumer facing technology companies. We did not include biotechnology or medical device companies. To ensure meaningful pricing history, we selected companies that have been public at least one year, so our IPO period ranges from January 1, 2002 through December 31, 2013. We therefore captured public company pricing history through January 9, 2015. We populated our data set with all 158 resulting companies and recorded the following information for each company:
- Venture firm(s) that invested
- IPO price
- First day closing price
- Peak price and key milestone prices, including 2x and 3x from the IPO price and lockup expiration price
- Lockup expiration date
- Number of quarters from IPO and from expiration
- And just for fun, whether the CEO or founder attended Harvard Business School or Stanford Graduate School of Business
In addition to aggregating the pool, we also organized the data by vintage year and reorganized the companies into 4 cohorts: 2002–2004; 2005–2007; 2008–2010; and 2011–2013. This allowed us to deal with time series issues that arose, where early companies had more pricing history and time to mature.
Once we assembled the data set, we posed a series of questions. Before sharing our observations, I’d like to offer the following caveats. First, I am not distinguishing between GPs and LPs. Obviously if GPs distribute stock, LPs can continue to hold. GPs may also choose to hold stock for the Partnership on behalf of LPs. I fully understand the importance some place on providing liquidity whenever possible from a GP/LP relationship perspective. Second, I am not writing from the perspective of prudence. No matter what the data says, many people will argue that given the future is unknowable, it may make sense to sell at least a portion as soon as possible to hedge one’s bets. That is a perfectly reasonable approach. Finally, I am not opining on what is or isn’t a VC’s job: some people say that VCs should distribute at the earliest possible date because they aren’t hired to hold public stocks. I am not taking a point a view on that issue. With those three caveats, here is what we learned:
1) It may not make sense to sell at IPO
Bankers sometimes offer venture firms an opportunity to sell a portion of their holdings at the IPO price, with the remainder subject to a 6 month lockup. We wanted to know the percent of companies for which the IPO price was the peak price. For the aggregate pool, only 3.7% of the companies saw their peak price at IPO. We were also interested in looking at how the answers varied when looked at by venture firm, and for 14 of the 19 venture funds on our list, exactly 0% of their portfolio companies peaked in price at the IPO.
2) Don’t get too excited about the first day closing price. The best might be yet to come…
Knowing that bankers price IPOs to build in a first day “pop”, we wanted to know how many companies experienced a first day closing price as a peak price. The answer for the aggregate pool inched up to 5.6%.
3) Lockup expiration doesn’t mean run for the hills…
Many VCs have the ability to sell or distribute their holdings 6 months after IPO. We wanted to know how many companies experienced a peak price at the point of lockup expiration. The answer for the aggregate pool is an interesting 0%. Of course, given the low odds of the peak price falling on a specific day exactly 6 months after IPO, we further sliced the data. We analyzed instead how many companies saw their lockup expiration price as the highest available price (knowing that any higher prices in the lockup period weren’t available for sale). The answer is only 15% of the companies. The majority of companies experienced a price peak after the lockup period expired, and for the minority that experienced a price peak during the lockup period, 85% traded back above the lockup expiration price after the lockup period ended. Finally, we looked at expiration pricing from another perspective and found that for 39% of the companies, the lockup expiration price was lower than the IPO price, and for 56% of the companies, the lockup expiration price was lower than the first day closing price.
Lockup Expiration Price Analysis
4) Good things take time…
We wanted to know how many quarters it took to reach peak price from IPO. The median answer for the entire pool is 5 quarters, and the mean is 9 quarters. Of course, the aggregate pool is subject to an obvious time series problem (older companies have more data), so we looked at our 3 year cohorts. For the first cohort (2002–2004), the median number of quarters to peak price was 13, or approximately 3 years. For the 2nd, 3rd and 4th cohorts, the median number of quarters to peak price were 9, 7 and 1 respectively. The last cohort has the least amount of pricing history, so the small number would make sense. But, the first 2 cohorts are more instructive, given the number of years they contain. These cohorts suggest that waiting 2–3 years allows sufficient time for many newly public companies to mature and appreciate in value.
Time between IPO and Peak Price
5) So what, exactly, is in the land of milk and honey?
If you aren’t selling at IPO, or at lockup expiration, is the wait worth it? What is the median percent gain between the IPO price and the peak price? The answer for the pool is 120%, or approximately 2.2x. For those too quick to conclude that 2.2x doesn’t sound like much, note that the return is 2.2x from the IPO price, not from your cost basis. 2.2x from the IPO price might represent another 5x or 10x on your investment! And what about the peak price compared to the lockup expiration price? The median percent is still approximately 100%, or 2x. We then re-looked at the data from a cohort perspective to address the time series problem. For the first cohort (2002–2004), the median percent gain from IPO to peak price was 207%, or slightly more than 3x. For the remaining 3 cohorts, the median gains were 129%, 156% and 107% for a range of approximately 2x to 3x for all 4 cohorts. Considering this data includes both the highs and lows of the current and prior economic cycles, it takes into account a range of market environments. Finally, we looked at the same cohort analysis from the time of lockup expiration to peak price. For the 4 cohorts, the median percent gains were 142%, 128%, 80% and 94%, or approximately 2x to 2.5x from expiration.
6) But, the road to paradise contains thorns…
We don’t want to leave anyone with the impression that IPOs simply progress up and to the right. We wanted to know what percentage of companies trade below their IPO price in the first year. As it turns out, a whopping 66% do! The range, when looked at by venture firm, was 20% to 91%, and for 16 of the 19 venture funds, at least half their companies traded below the IPO price at least once in the first year.
7) And requires patience and an iron stomach…
Is there relief in the 2nd year? The answer is no. 59% of the companies have traded below their IPO price at least once in the 2nd year. In fact, 59% of the companies have also traded below their IPO price prior to a peak, assuming the IPO price is not the peak price.
Price Path of Median Scenario per Cohort
8) OK, but are outliers driving the returns?
We know that venture is a hits driven business, so we wanted to understand if the above referenced returns were driven by a handful of companies. We found that the returns were fairly broad-based. 65% of companies traded to at least 2x from the IPO price and 36% traded to at least 3x from the IPO price. In addition, 56% traded to at least 2x the lockup expiration price and 25% traded to at least 3x the lockup expiration price. When analyzed by cohort, we found that the results were reasonably consistent across market environments. For the 4 cohorts, the percent of companies that traded to at least 2x the IPO price were 70%, 67%, 67% and 62% respectively. For a 3x, the percents were 55%, 42%, 42%, and 30% respectively (noting that the last cohort has allowed its companies that least amount of time to mature). Looking at the same question from lockup expiration, the percent of companies that traded to at least 2x the lockup expiration price, by cohort, were 65%, 58%, 46%, and 51%. For a 3x, the percents were 45%, 26%, 29%, and 21%, again noting that the last cohort has given its companies the least amount of time to grow.
Percent of Companies Appreciating by 2x and 3x
9) Yes, but as Keynes said, in the long run we’re all dead.
Clearly time matters. We wanted to know how many quarters were required to achieve these 2x and 3x returns from the IPO price and from the lockup expiration price. Here we worked with averages because we didn’t feel the median captured the variation we saw in the numbers, and the median made it seem like the returns were achieved more quickly that was reasonable. From IPO to 2x, by cohort, we observed that the number of quarters required were 4, 6, 4 and 3. For a 3x, the averages increased to 9, 12, 6 and 5, or approximately 1.5 to 3 years. Looking at the same question from the lockup expiration price, for a 2x by cohort, we observed that the number of quarters required (from the time of expiration) were 2, 8, 4 and 2. For a 3x, the averages increased to 7, 16, 4 and 6, or approximately 1 to 4 years. It’s worthwhile nothing that most of the IPOs in cohort 2 took place in late 2006 and 2007, on the eve of the financial crisis.
Time to Reach Returns per Cohort
10) And just for fun… Does attendance at HBS or Stanford’s Graduate School of Business help?
Nope. Only 15.8% of CEOs or founders attended. The vast majority did not, which is no surprise.
One of the trends taking place in the last few years is that that private companies have been staying private longer, as they have been able to raise larger sums of private capital at sky-high valuations. Consequently, one of the objections to this data might be that currently, an increasing percentage of a company’s value is accruing to private investors. Thus, while at one point it made sense to hold public company stock of venture backed companies, these days, there isn’t as much upside left.
There is some merit to this objection. However, a few comments are worth making. First, the music will stop at some point. Just like it stopped in 2000 and 2008. Private company valuations are currently out of step with public company comparables for high growth companies, and as long as private investors pay an illiquidity premium, private companies can raise capital at high prices and stay private. But, like all prior cycles, this will end, and private valuations should come back down. This should make public capital relatively cheaper again, meaning that rather than stay private, companies will opt to go public, presumably at lower valuations.
Secondly, and somewhat contradictory, technology companies are simply able to get bigger today than they were able to 5–10 years ago (and presumably, this observation is what is sustaining the high private valuations). TAMs are increasing as technology permeates a wider and larger part of the economy. Thus, a billion dollar valuation at IPO doesn’t necessarily mean that the company is done growing. More companies will achieve mid cap and large cap status than in the past.
Our take-away from this work is that selling stock at the 6 month lockup expiration period is among the least optimal times to trade out of a position. Many newly minted public companies continue to grow well for a few years post IPO, and substantial additional returns can be made with even one or two years of additional patience.
By: Farouk Ladha, Managing Partner at Four Rivers Group;
Salmaan Javed, University of Southern California