The U.S. market for IPOs (initial public offerings) was red hot in 2020. Excluding special purpose acquisition vehicles, U.S. IPOs last year raised $83 billion in gross proceeds. The prices of these IPOs jumped during the initial day of trading by 36% on average.
Despite these high returns, index funds — including mutual funds and exchange-traded funds — almost never bought IPOs at their initial offering price. Instead, index funds waited to buy IPO stocks until near the date on which they were added to the relevant index — typically at the end of a quarter within six months to a year after the IPO.
Yet as the index inclusion date nears for any IPO, its price typically surges in anticipation of a barrage of purchases — driving up the price that index funds must pay for that stock. For example, the price of Tesla spiked once it became likely that the company would be added to the S&P500
In this article, I outline the data from 2010 to 2018 about the high initial returns for IPOs as well as the concerns holding back index funds from buying IPOs before they are included in the index. Then, I make a path-breaking proposal — allowing any index fund that tracks the Russell 1000 Index to meet these concerns by early buying of IPOs if, and only if, they are large relative to the size of the index.
Like many other studies, the study that I co-authored with two experts on indexing found high returns in IPO stock prices during the initial day of trading. After this initial day of trading, the study evaluated IPO returns by a measure known as the index-adjusted performance (IAP) — the difference between the total return of the security and the total return of the index from the closing price on the first day of trading to the closing price of any following day. For example, a positive IAP would signal that an IPO has outperformed the index from the close of the first day of trading until the date the IPO is included in the relevant index.
The study calculated these two metrics of returns for all 932 U.S. IPOs offered in the nine years between January 2010 and December 2018. Of these 932 IPOs, 115 were included in the Russell 1000 within the first six months of trading.
The study used the Russell 1000 because it includes 92% of the total market capitalization of all listed stocks in the U.S. equity market. The Russell 1000 contains the top 1000 publicly traded U.S. companies according to market capitalization. IPOs are considered for inclusion at the end of each quarter, strictly based on their market capitalization.
The first-day return for these 115 IPOs was highly positive — 22% on average with a median gain of 10%. Similarly, looking at the IAP for these 115 IPOs included in the Russell 1000, the study found a positive trend — with an average IAP of 6.89% and median of 5.24% between the IPO and the index inclusion date.
Both of these trends show that index funds could generate excess return by buying IPOs before they are added to the index. The greatest return could be realized by buying IPOs at the initial offering price and holding them through the index inclusion date. Index funds could also realize significant excess returns by buying IPOs after their first day of trading and holding them through the index inclusion date.
Risks in the returns
Index funds would face several risks associated with such early purchases of IPOs.
First, and most importantly, no one knows which IPOs will be added to the index at the time of the IPO. An index fund might purchase an IPO stock that doesn’t get added to the index. In that event, the fund would have to sell the IPO stock, potentially at a loss. The price of the IPO stock would decline because there would no longer be the expectation that other index funds would be required to buy that stock when it is added to the index.
A second concern is that an index fund would not get a large enough allocation in an IPO to reflect the stock’s future position in the index — for example, when a popular tech company goes public. Even so, an index fund can still generate excess returns by buying more of such a stock on the day following an IPO and holding that stock until it is added to the index.
Third, since the index fund would be holding stocks that are not included in the index — at least for several months — the fund would experience tracking error. Tracking error occurs when the returns on an index fund portfolio differ materially from those of the index it is benchmarked against. But investors would probably not be overly concerned if the fund beat the index it was designed to track.
Of course, the prospectus of such an index fund would have to make clear that it would be buying stocks in the initial offerings of IPOs and after their first day of trading. The prospectus should also delineate the risks involved when the fund buys IPOs before they are included in the index.
To mitigate the most important risk — that the IPO will not be included in the index — I recommend that index funds should purchase an IPO only if its expected weight in the Russell 1000 is relatively large. The expected weight equals the gross proceeds raised by the IPO, divided by the total freely traded float of stocks in the index. Since the Russell 1000 Index is composed of the top 1000 U.S. companies by market capitalization, the larger the IPO is relative to the index, the more likely that the IPO will be added to that index.
This strategy could be adapted to varying risk tolerances of index funds by adjusting the size threshold for early purchases of an IPO. In a conservative strategy, the index fund would purchase only IPOs with the largest expected weight in the index, since they are most likely to be included in the index. In a more aggressive strategy, by contrast, the size threshold for buying IPOs would be lower.
The study examined three thresholds for risk appetite, defined in terms of the expected weight of the IPO in the index: 1.0 basis point for conservative; 0.75 basis point for pragmatic and 0.50 basis point for aggressive. (One basis point equals 1/100 of 1%)
The results, summarized in the table below, show that this strategy would have been successful at generating excess returns without significant risks during the period from 2010 through 2018.
For example, 100% of the largest IPOs that would have been purchased under the conservative strategy during this period were added to the index within the first six months and generated excess returns above 15%. Under the aggressive strategy, 88% of the IPOs that would have been purchased during this period were included in the index within six months and generated excess returns of close to 17%.
My recommendation is that an index fund based on the Russell 1000 buy relatively large IPOs in their initial offerings or, if necessary, immediately after their first day of trading. Although there is a modest risk that such IPOs will not subsequently be included in that index, the excess returns from this strategy outweigh the risks.
I would not recommend that any index fund use this strategy to buy an IPO effected by merging a private company with a special acquisition vehicle. I also would not recommend this strategy for any index fund based on other indices where it is more difficult to predict when and whether IPO stocks will be included in the index, such as the S&P 500, where stocks included are chosen by a committee.
Robert C. Pozen is a senior lecturer at the MIT Sloan School of Management and a former president of Fidelity Investments. Research assistance for this article was provided by MIT undergraduate student Peter Hoffman.