An Analysis of IPO Performance
What’s next for Blue Apron after their abysmal IPO

Blue Apron opened trading at $10/share in one of the most anticipated IPOs of recent weeks. The company has been largely panned by investors due to its high customer-acquisition cost and heavy competition and has since fallen 34.1%, making it one of the worst IPOs of the year. The company has become a favorite punching bag of many commentators since its poor debut, and I wondered if the constant negativity surrounding Blue Apron has served to drive their share price below what is reasonable. Perhaps as the Wall Street echo chamber moves on, Blue Apron’s price will rise. It’s an interesting question worth investigating. In this article, I will analyze trends in the price of companies in the weeks following their initial public offerings in order to forecast the near future of APRN.
Using TradeDefender.com’s python module, I downloaded historical price data for initial public offerings of companies between 3/5/2013 and 4/28/2017. The latter date was chosen because it was twelve weeks from the date of this writing and therefore gave at least twelve weeks of historical performance for analysis. The former date was chosen arbitrarily.
Typical IPO Returns
First, let’s try to put APRN’s performance in context. APRN was down -24.50% in the first two weeks after their June 29th IPO. If we look at the stocks in our dataset, we can see which percentage of them see any given percent change in price at any interval since the stock first began trading. We can use this information to see how Blue Apron’s IPO compares to its peers. Only 25% of stocks have done worse than -8.40% after two weeks of trading, making APRN’s performance unusually bad. In fact, based on our historical analysis, APRN is doing worse than 94% of companies in the dataset. Here’s a table with a percentiles breakdown:

Here’s a more detailed table with percentiles:
Price Change
So we know that APRN’s performance is very poor so far, but might we expect the price to recover? Looking at the IPOs in the data set, in 49.72% of cases, the stock ends its first day of trading up in price. In the other 50.28% of the time, the stock ends its first day with the price having fallen. When the stock is up on the first day of trading, it will be up after one week 78.77% of the time, but up after twelve weeks only 56.70% of the time. Here are tables showing how often stocks are up or down after various time periods:

From this, we can see that a negative initial performance has greater staying power than a positive performance; when a stock goes up on the first day of trading it will still be up twelve weeks later in only 56.70% of cases whereas a stock going down on the first day will still be down twelve weeks later 73.18% of the time.
Of course, the price movement of a stock on the first day of trading is largely dictated by how well the investment bank that took the company public valued the company in the eyes of investors, so I looked at how well a stock’s performance after one week predicts future returns. I found that negative returns as of the first week of trading had even greater staying power than negative returns on the first day of trading. This of course makes sense because there are fewer days of trading between the first week and the twelfth week than between the first day and the twelfth week, thus the price after one week will tend to change less in the aggregate.
Here is a scatter plot showing the relationship between the change in the stock price in the first week and in the first twelve weeks:

For what it’s worth, the volume of trading on the day of a stock’s IPO does not correlate to percent change in stock price on the first day of trading or after one, two, six, or twelve weeks.
Major Losers
Given that early price movement seems to predict gains or losses in later weeks, the future look bleak for APRN. But perhaps this is less the case of major losers such as Blue Apron. Perhaps certain companies are met by unusually strong investor incredulity when they IPO, creating an echo chamber of negativity surrounding the stock which drives the market’s valuation of the company below what it is really worth. In other words, is it possible that stocks off to abysmal starts, such as APRN, may have a tendency to recover?
To test this idea, I looked to see how often a stock’s price twelve weeks after their IPO is greater than its price two weeks after the IPO. IPOs in the top 10% of returns after two weeks of trading see positive price movement in the following ten weeks in 50% of cases in the data set. IPOs in the bottom 10% of returns at two weeks fair slightly better, achieving price improvements 54.16% of the time. Across the entire dataset, regardless of price at the second week, price gains between the second and twelfth week occurred 55.42% of the time.
This isn’t particularly exciting for a few reasons. First, the difference is fairly minor, and second, I summed together the percent change in price from week two to week twelve, and although there is usually a price improvement, the average return from a strategy predicated on buying at two weeks and holding until the twelfth week is negative. For all companies in the dataset, the return from such a strategy would be -1.56%. For IPOs in the top 10%, the return is -6.16%, and for IPOs in the bottom 10%, the return is -4.78%.
Volume
Another interesting question to examine is the effect that volume has on these results. After all, there are plenty of IPOs that fail to attract much attention and would therefore not be subjected to the level of scrutiny as are the Blue Aprons and Snapchats of the world. This could lead to mispricing either in that investors ignore important aspects of the companies receiving less examination or in that investors overreact to prevailing sentiments about IPOs receiving greater than usual attention.
It turns out that low volume IPOs tend to see price improvement from the second week to the twelfth less often than those of higher volume. IPOs in the bottom 10% improve in price 43.06% of the time whereas those in the top 10% rise in price 58.33% of the time. Again, across the entire dataset, there is a price increase 55.42% of the time.
However, even in the highest volume IPOs, there is still usually a reduction in price from the second week to the twelfth week. Summing all of the IPOs in our dataset, there was on average a -1.56% change in price over that period. For the bottom 10% of IPOs in terms of volume, that average is far worse at -15.89%, and for the top 10%, the return is best of all at -0.37%.
The Future of APRN
This all looks bad for Blue Apron’s future, but it’s important to note that the predictive power of this information is dubious. After all, the share price of a company represents what investors believe the value of the company to be, and this valuation is (in theory) a reflection of conditions within the company, their market, and the global economy — it is not dependent upon the change in the share price yesterday or the change in the share price on the first day of trading, though these do seem to be somewhat predictive. The trends in price change that we’ve discussed are likely due to external factors that influence price change. With this caveat in mind, let’s examine what this data can tell us about the future of APRN.
There are two unique things about APRN’s IPO: how high the volume was and how low the share price ended after two weeks of trading. On the first day of trading, volume was 41,055,424, and after two weeks, the stock was down -24.50%. In their respective categories, these values are in the top 3% and bottom 4% respectively. There were no IPOs in our dataset that were in both of these percentiles, but if we expand to look at IPOs in the top 30% in terms of volume and the bottom 30% in terms of two week performance, there are 54 IPOs for us to look at.
For these IPOs, the mean return after two weeks was -16.41%, but the mean price change from two to twelve weeks was 4.23%. In an impressive 70.37% of cases, there was positive price change from the second week to the twelfth week of trading. From this view, there may be money to be made in APRN in the short term, but beyond that is uncertain.
Follow me on Twitter: https://twitter.com/chrisglet