IPO windows and the IPO berg

– Melinda White, July 2025

With the success of some larger IPOs in the last 12 months, there’s been increasing talk of the return of the IPO window. We were curious to look at the data over a longer period of time to see just how the current period looks relative to our living memories in markets. We discovered several curious aspects to the IPO cycle which we think all investors should be aware of.

Our aim with this exercise was to understand what the shape of the IPO cycle had looked like and whether there was any data to support the idea that IPOs are a good way to enhance returns to investors. Our challenge was to find a dataset that would help us make relatively robust observations around number per year, size of IPO, returns and relative returns from IPOs. We interrogated the dataset for 3 providers and settled on using Factset’s as it was the most complete and had comprehensive data going back to the late 1990’s.

We learnt 3 interesting things from the data. IPO activity appears to be pro-cyclical to market conditions (i.e. when the market is hot, IPO activity is high); very few IPOs are investable for institutional-style investors; and IPOs are not an effortless pathway to riches (in-fact far from it).

Let’s explore that in more detail.

Using our numbers, there have been approximately 2,400 IPOs since 1998 on the ASX. The number of IPOs (rather than value) appears to fluctuate in-tune with positive and negative sentiment in the market. Essentially IPO activity is cyclical and pro-cyclical at that. We can see in the numbers that IPO activity peaked in 2000 at the end of the first Tech Boom, again in 2006 – 2007 during the China-led mining boom, and then again in 2021. During all three periods, the market was relatively expensive, and the economy was in the midst of an economic boom.

Source: Factset, Longwave, Iress, Bloomberg. 

When IPO activity drops away, it tends to happen very sharply and coincides with market conditions being less frothy and less expensive. Like in 2001, 2008, and more recently in 2023-2024. Interestingly, these were good times to buy into the market. Buying in 2003 and early 2009 and were good strategies, and we suspect 2024/25 may be as well.

Given the number of column inches dedicated to talk of IPOs, investors would be forgiven for thinking that they are a relatively risk-free path to market beating returns. This couldn’t be further from the truth. The first issue is that many of our 2,400 IPOs in the last 30 years are just not of an investable size. Of the 2,400 IPOs, only 24% were over A$50m market cap or deal size at the time of IPO. This stands to reason as many of the companies we now see as large and successful (think JB Hi Fi, Northern Star Resources, Carsales) listed when they were firmly small caps.

Although the media (aided by the relentless marketing from under-writing banks) would like us to believe that IPOs are an effortless path to free returns, the data leads us to conclude that investors should be just as selective or more so with IPOs as they are when selecting established listed equities for portfolios. Using the total return data since IPO (a data field provided by Factset), we can make the observation that 58% of our dataset are below their IPO price today or were below their IPO price at delisting. 42% are above their IPO price. Digging a bit further, we calculated the annualised return since IPO to 30^th June 2025 for each stock in the dataset and then compared it to the annualised total return of the Small Ords Index over the same time period. Just 20% of all IPOs in this dataset beat the Small Ords Index.

Source: Factset, Longwave, Iress, Bloomberg. 

What is really interesting is that when we dig into simple average and median returns from the data, we can observe a large skewness in the return profiles. The average total return of the entire dataset (not controlling for investable size) is 121%, but the median total return is -53%. This is due to large outliers in both directions of total return. The maximum downside investors can experience is 100% (a real feature in our dataset). However, the upside is unlimited, and in the case of small caps that survive and become larger businesses, that unlimited upside is significant.

Source: Longwave Capital

If we look at the skewness in total return across our A$50m+ investable sub-set and compare it to the total return of the Small Ords since 1999, we can visually see how that skewness looks. The vast majority of the Investable IPO sub-set haven’t beaten the total return of the Small Ords over that time period. Just a handful of stocks in each IPO vintage year account for the outsized returns. And some returns are stellar. These are often the mid or large caps of today that were once small-caps. They not only survived but thrived through the high grown phase of their lifecycle.

In the process of gathering and cleaning data, we were reminded of one of the great lessons when you study history. The saying “archives replace hands that have vanished and lips that are sealed” which refers to the challenge of keeping historical data that allows for an objective interpretation of history.

In order to look at a few market cycles, we had to go back as far in time as we could. All three portfolio managers at Longwave entered the workforce in the late 1990’s and in our minds, this is not a long history to be asking for when thinking about markets and looking to learn from history. However, we did not slavishly keep our own full database of IPOs and ECM activity since the inception of our careers, so we are reliant on the market data providers to give us the year-by-year snapshots that we know we lived through, remember, but don’t have data for. Despite a large number of records in our dataset, we discovered that there were individual IPOs we remembered that were missing. Among the datasets we looked at, each one had its own quirks – so one dataset might have captured a particular IPO but another one hadn’t.

Much of this data issue has to do with the arcane subject of ticker and name mapping. Companies list then change names, are taken over by another listed or private company, or delist. Tracking them via a unique identifier through history should in theory be simple and error-free, but in practice appears to not be. It certainly isn’t made easier by the ASX practice of ticker recycling (where a ticker that has been given up by a since delisted company is then assigned to another company coming onto the boards).

As incumbents in financial markets, the large global market data providers are effectively the librarians and archivists of market history. However, we need to be mindful of their shortcomings. Indeed, it provides an opportunity for us to add value as we wrangle their raw data into useful information.