Factor Spotlight
Factor University

IPO Season is Heating Up, What's in Your Risk Arsenal?

It’s IPO season! Despite the pandemic-induced IPO lull early in 2020, we’re now seeing momentum from the IPO train, particularly in the tech space. Snowflake’s IPO (SNOW) this past week made headlines when the stock shot out of the gate on Wednesday with its shares up 130% within a few minutes. Two other tech names, Unity (U) and JFrog (FROG) also debuted this past week with major price jumps on their first day of trading. With popular companies like Airbnb, Palantir, Robinhood and plenty of others expected to IPO later in 2020, there is no shortage of exciting IPO activity despite the challenging economic environment.

Given this expected swell of IPO activity on the horizon, it’s an opportune time to better understand how to analyze IPOs from risk modeling perspective and investigate tools available to limit the potential downside. This week, we’ll lay out a general framework of how to approach IPOs from a risk modeling perspective and continue the series next week by exploring smarter hedging techniques that can help you maximize the alpha of your IPO bets.

Risk “Guess-timation” for IPOs

Since there is insufficient returns history and public information to accurately model every single potential factor exposure for IPOs, risk model providers generally use statistical methods to create a synthetic returns history. They then approximate the IPO’s factor exposures using weighted averages within a comparable universe defined by market cap, sector, and other characteristics. Over time, as the asset builds a true daily return history and fundamentals become publicly available, the asset’s actual characteristics become the driver in the factor exposure and specific risk calculations. Understandably, it can require some time before the exposures and specific risk can be reliably estimated using only the asset’s characteristics alone, so extra caution should be taken when evaluating risk on a security in the initial time period after trading.

The Risk Profile of an IPO

With this caveat in mind, we’ll take a look at some of the top IPOs from 2010-2019 based on IPO proceeds, focusing on the tech/tech-related and health care names: Alibaba (BABA), Facebook (FB), Uber (UBER), HCA Health Care (HCA), and Snap (SNAP).


We examined risk estimates from a medium-horizon risk model (global for BABA and US for the rest) to understand how the annualized risk estimates for IPOs stack up to the actual 1-year realized risk, as well as how these estimates change 3 months and 6 months after the IPO date. Presumably, the estimates should get better the further away from the IPO date as the model is able to account for more of the company’s actual fundamentals and return history.

Generally, our presumption proves to be true for our IPO examples, though there are some exceptions.

Screen Shot 2020-09-19 at 7.40.25 PM

The risk estimates for BABA and HCA as of their IPO dates are very much in line with the actual 1-year realized risk for these securities. However, for FB, UBER, and SNAP, there’s an apparent spread between the actual realized risk and the estimated risk. It is important to note that the COVID-19 market downturn from March 2020 is included in the 1-year period after UBER’s IPO, so that could contribute to UBER’s estimated vs. realized risk spread.



If we re-run the analysis 3 months after the IPO date, we see that the estimated vs. realized risk spread tightens for certain names and deviates further for others. Once we move 6 months following the IPO date, the estimated risk is much more in line with the realized risk across the board (note: UBER is excluded in the 6 month variant of the analysis due to a lack of return history).

There did not appear to be concrete trends in the effectiveness of the risk estimate based on IPO size, sector or IPO date, but more work can certainly be done to dig in further with a larger sample size of IPOs.

Looking Forward for IPOs

While there is no clear conclusion on when risk estimates for IPOs become the most effective, the analysis gives us the sense that the risk estimates unquestionably improve the further you get from the IPO date.

Next week, we’ll look at how we can use the risk model, along with other hedging techniques, to make your IPO bets less risky.

US & Global Market Summary

US Market: 9/14/20 - 9/18/20

Screen Shot 2020-09-19 at 9.40.08 AM.png
US Stock Market Cumulative Return: 9/14/2020 - 9/18/2020
  • The major US indices finished the week on a lower note, marking a third straight down week as investors wrestled with uncertainty around the pace of economic recovery, the availability of a COVID vaccine and passage of fiscal stimulus, and simmering Sino-American tensions.
  • On Wednesday, the Fed indicated that the path to economic recovery will be a long one, and that they intend to keep interest rates near-zero for the next 3-4 years.
  • US consumer spending decelerated in August as extended unemployment benefits ended for millions of Americans. The Commerce Dept. reported that core retail sales were down -0.1% last month vs. consensus estimates of +0.5%.
  • Meanwhile, the U of Michigan reported that preliminary readings of US consumer sentiment suggested that it was 78.9 in September, up from 74.1 in August, and better than consensus expectations of 75.9.

Normalized Factor Returns: Axioma US Equity Risk Model (AXUS4-MH)

Screen Shot 2020-09-19 at 10.03.11 AM.png
Methodology for normalized factor returns
  • Earnings Yield was again the biggest winner, up another 0.38 standard deviations and nearing Overbought (+1 SD) space.
  • Volatility continued to recover from its recent trough of -2.18 SD below the mean (8/4), up +0.32 SD this week.
  • Size continued its slow upward climb on a normalized basis, as it creeped higher above historical trend.
  • Growth descended further towards Extremely Oversold territory, albeit at a slightly slower pace than the preceding two weeks.
  • Momentum fell again by -0.34 SD towards Oversold space, now sitting at -0.86 SD below the mean.
  • Profitability tumbled further into negative space from a recent peak of +0.58 SD above the mean on 8/24.
  • US Total Risk (using the Russell 3000 as proxy) reversed its recent trend and declined by 15bps.

Normalized Factor Returns: Axioma Worldwide Equity Risk Model (AXWW4-MH)

Screen Shot 2020-09-19 at 1.33.44 PM.png
Methodology for normalized factor returns
  • Volatility saw some incrementally positive movement (+0.18) as it continued to climb away from its former Oversold label.
  • Market Sensitivity crossed into positive territory as it enjoyed a +0.11 SD move.
  • The rally in Size stalled out, as it ended exactly flat on the week.
  • Value appears to have hit a bottom of -1.12 SD below the mean on 9/16 and ticked back up to -1.07 SD below the mean.
  • Profitability continued to decline towards the mean, down -0.35 SD on the week.
  • Momentum was again the week’s biggest loser as it fell by -0.38 SD standard deviations, earning an Oversold designation. This factor was at +2.58 SD on 8/5.
  • Unlike the US, Global Risk (using the ACWI as proxy) declined by 21bps.


Related Insights
See All Insights

What Forces Are Impacting Your Performance? Find Out Now...

Schedule a Call