An earnings season full of surprises

What happened

The full-year earnings season just ended. It was an outlier year regarding how investors reacted to numerous earnings releases.

Overall, this year has been exceptionally positive for The AtonRâ Fund and most themes. Over 60% of companies have announced earnings well ahead of analyst expectations, in line with major indices, e.g., S&P500, Russell 2000 Growth, and higher than a comparable product. On the other side, only 31% of firms have negatively surprised, again besting both indices and peers, with the exception of S&P500.

Nevertheless, markets suffered. Indeed, during the recent months, investors and companies were facing a challenging macro environment, mainly fueled by the FED’s shifting strategy, slowing economic growth, increasing trade tensions, and rising interest rates. All of these triggered even higher scrutiny of companies and their financial position.

Why is it important?

Tracking the earnings announcements and how they evolve for each company are essential to continue having an updated picture of the companies, sectors, and themes. At AtonRâ, we focus on growth stocks, the financial quality of which typically attracts heightened attention during the earnings season, even more so with the more uncertain macro and geopolitical state. This also raises investors’ expectations for each stock. Therefore, even if investors are satisfied by the stock quality, the implicit bar for financial results to be considered “satisfactory” is raised. In practice, what was deemed positive in the past seasons is now regarded as disappointing or mediocre at best. Investors are no longer satisfied by 5% earnings surprise but expect double or triple of that to be a truly “positively surprise” to allow the stock to react accordingly.

Understanding the investors’ sentiment is thus vital to investigate the recent volatility in our portfolios more precisely and to comprehend if there was an overreaction in the market, disconnecting the stock price with the long-term business potential.

Impact on our Investment Case

Seeking surprises through an event study 

The best way to investigate if this earnings season was any different and if investors’ reaction to the event was a deviation from the “norm” is to conduct an event study: an empirical analysis exploring investors’ reaction to earnings announcements of the companies in our portfolios.

An event study measures the effect of some company-specific information release on that company’s stock price. The final result of such analysis, however, does not put any single company in the spotlight (the stock price response of any one company to this specific company’s announcement). Instead, it measures an “average” response to a type of announcement. More details about this procedure may be found in the appendix of this note. 

Empirical evidence points to irrational overreaction

The event study allows us to measure the difference between the realized return and the return expected by the market for all stocks in the portfolios based on the announcement outcome: “good news” (positive EPS surprise), “bad” (negative EPS surprise), and “no news.” The expected return is derived by using a classic Capital Asset Pricing Model with 1 year of data before each EPS release. The release itself happens on day 0. The graph below shows the “Cumulative Average Abnormal Return (CAAR)” on the event day and ten days before/after it.

Using the data from the past ten years, we arrive at a very classical stock price “reaction.” Where negative EPS surprises lead to a negative abnormal return, positive to positive, and “no surprise” stocks end up with zero deviation from what was expected by the market. Historically from the data, we consider the “no news” surprise category to be in the range of ±3.5% for our growth portfolios.

We agree that it may not be intuitive why the abnormal return starts to visibly react a day before the event, making reading the graph much harder. We propose a common explanation that the market reaction one day before the event is most likely due to insider trading, and the market reaction one day after the event is most likely due to the results being published after the stock market closes. This way, we may proceed by normalizing the graph to one day prior to the announcement day to see how the stocks react in unison with their EPS surprise.

We repeat the exact same exercise but consider only the recent earnings season. We immediately observe the changing magnitude of investors’ reactions - a more extreme, more pronounced response to all news types.

Additionally, we may observe that the reaction to “No news” was worse than to “Bad news” and that “good news” did not react as fast as in the past. We conclude that this may be due to an out-of-fear sell-off of “negative and neutral news” stocks, given that the expectation was much higher, and investors panicked. A possible explanation seems to be that this time markets visibly expected more information and wanted “more” visibility, be it good or bad, so the performance of stocks that did not surprise deviated more from what was expected and more so than even “negative news” stocks.

Finally, we notice an interesting behavior of stocks that negatively surprised. In short, investors have exhibited the “sell the news, buy the facts” behavior when, after initially selling the positions with lower-than-expected EPS, they reevaluated for the long-term and recalibrated their sentiment, having understood that the stocks continue to have a strong investment case.

We love when numbers provide interesting insights into the market's and investors' sentiment. This is especially helpful during periods of increased volatility to gauge the levels of overreaction to the news. This event study has helped us reconfirm that a part of the recent market movement could be attributed to the investor’s increased, perhaps even, unrealistic expectations and the following overreaction when these expectations were unmet. Additionally, those investors who are building their portfolios for the long term should disregard the market move as a reliable signal to sell, as fundamentals have not changed enough to justify portfolio reallocation. On the contrary – most firms have either reconfirmed or increased their guidance for the next quarter and the next fiscal year. 

A glance at the earnings

How did we fare this year?

Beyond the aggregate picture, we assessed the reaction of our thematic portfolios to the EPS surprises, and provide more color on the earnings season for some. 

What was particularly eye-catching, are the surprises of AI & Robotics, Security & Space, and Sustainable Future, for which we provide a brief comment below.

Artificial Intelligence & Robotics continues to beat expectations

What stands out in the AI theme is the relative absence of extreme surprises (only Snowflake came in as a massive beat but delivered disappointing guidance). None of our Computing Components stocks surprised negatively, consistent with the ongoing strong spending cycle in datacenter infrastructure. Similarly, Chinese stocks have almost all positively surprised after the delicate year 2021, impacted by a massive crackdown. On the other side of the spectrum, advanced AI applications have had a challenging quarter, something which might be explained by demanding expectations from investors convinced by the massive possibilities offered by AI. Overall, the consistency in the earnings points to well-identified and strong underlying dynamics driving the theme

Security & Space surprises to the Moon and back

Except for two stocks that disappointed investors (-10% surprise), over a third of stocks in the Security & Space theme have flabbergasted investors and analysts by record earnings (~15%-300% surprises). This, in turn, is not surprising since cybersecurity stocks and space application stocks, making the bulk of the portfolio, have had a record year in terms of growth, customer acquisition, margin improvement, and, for some, 2021 was the first EPS-profitable year. One of such examples was Maxar, an Earth observation stock that continues making headlines given its importance as a data provider during the current conflict in Eastern Europe. Overall, the earnings surprise shows how analysts usually underestimate the importance of security – for both cyber and orbit dimensions – and their potential future growth, both factors we are strongly confident about.

Sustainable Future has faced a challenging year but faces a bright future

Sustainable Future theme has the highest proportion of stocks that have negatively surprised analysts. Additionally, the theme shows a lot of big surprises as almost half the stocks exceeded expectations by 5% in both directions! The electricity storage surprised positively, especially by its ability to adjust prices to compensate and pass on the cost increases of materials and logistics. On the other end of the spectrum, most of the solar segment showed negative surprises (-5% to -25%), again reflecting logistic and material costs. Despite the record-high backlogs the solar players accumulated, the lack of flexibility that the electricity storage players have shown penalized the sector and disappointed investors. Wind actors also had a challenging quarter due to raw material and supply chain headwinds, which reduced their margins. Overall, we reconfirm that actors at the beginning of the supply chain, i.e., infrastructure and capex suppliers, showed more positive surprises, reinforcing our conviction to include them abundantly in all segments. At the same time, the record level of long-term backlogs for most other actors reconfirm their long-term growth prospects and comfort us with our allocation.

Our Takeaway

We continue to believe that any equity investment should include a growth component, especially in a low-growth environment. At the same time, growth stocks stand up to stricter scrutiny by the market, which usually has heightened expectations and gets quickly disappointed if stocks do not exceed these expectations. And while we cannot control investors’ sentiment and their over/under reaction during the earnings seasons, we can control the quality of every stock we hold and ensure the viability of the investment case in all market environments.

Moreover, we cannot expect companies to amaze investors with revenue and earnings growth every quarter, no matter the stock quality. Still, we are confident in their ability to generate growth above-trend and thus outperform the broader markets. We believe that solid fundamentals and earnings growth are the most important variables investors should consider when investing in equities in the long-term.

Appendix

The procedure we followed is as follows:

1. We aggregated all stocks in the AtonRâ fund, past earnings announcement dates, and classified each announcement into “positive,” “negative,” “no surprise” categories, according to the distribution of past surprises and excluding outliers.
2. We define an EPS announcement as an “event day” (Day 0). The event window consists of 21 days – 10 days before the announcement, 10 days after, and the event day itself.
3. We derive the normal return expected by the market using a classic CAPM model and a year of data before the announcement.
4. Over the event window, we compute the difference between the expected and the real return, which we call the abnormal return. For each stock, we average these abnormal returns by stock by type: “positive,” “negative,” “no news.”
5. Finally, we average the result by type and arrive at three vectors, ready for investigation.