Information & Market data has always been an important aspect for investors-both naïve and sophisticated. Over the years, it has been found that trading activity increases during earnings announcements or corporate presentations. According to William Beaver (1968), he showed that trading volume and return volatility increased at the time of annual earnings announcement. In addition, he documented that earnings reports provided information to investors as measured by abnormal volume and U-statistic. He defined the U-statistic as the announcement week squared residual returns divided by the mean squared residual returns in the non-announcement period. Consistent with the U-statistic, Patell (1976) derived the statistical distribution of the U-statistic under the assumption that the security returns are normally distributed and serially independent. Given these assumptions, he was able to show that the U-statistic has an F-distribution with an expected value of slightly higher than 1. This documentation was done under the null hypothesis that earnings announcement does not contain more information than in the non-announcement period. It is also important to mention that the F-distribution is skewed to the right and therefore the median of the U-statistic is less than one in the null hypothesis. According to Landsman and Maydew (2002), it has been necessary to measure the mean of U-statistic over a three-day period when moving from weekly to daily returns. This is attributed to the uncertainty that comes with the timings of the next earnings announcement. The Three-day U-statistic (TCU), is not expected to be symmetric, Landsman et al focused on the median as well as the mean in order to examine the extent to which relative return variability differs during earnings releases across a cross-section of firms.
Several research studies have been conducted to evaluate whether earnings announcements have significant information content and whether the information content varies with time, profitability, firm size and analyst coverage. The numerator of the TCU captures the information released during earnings announcement. On the other hand, the denominator of the TCU captures the information released during earnings announcement windows. It is crucial to point out that any other factor that increases the information released during earnings announcement windows or decreases the information released during non-report windows will increase the information content of earnings announcement. In order to examine the exploratory factors of the U-statistic it is important to examine the factors that might affect the level of information released within and outside earnings announcement windows.
Prior research has been conducted to examine whether the information content of earnings announcement has increased or decreased over the last three decades. A paper done by Lo and Lys (2001) discovers that information content contained in earnings announcement does not increase over time. However, the value relevance decreases over time. This is as a result of the difference in the findings for the information content and the value relevance to unrecognised disclosures at the same time as earnings. In contrast, Francis, Schipper and Vincent (2002) discovered that the usefulness of earnings announcement have increased over time. However, the increase in the magnitude of the market reaction to their sample data earnings announcements is not directly linked to the absolute amount of unexpected earnings conveyed in the announcements. It is important to mention that, consistent with increasing reaction, Collins et al (2009) finds a trend in the informativeness of earnings announcement. This is attributed to increasing reaction to earnings over the years.
Naïve Trading and the post earnings announcement drift
It has always been thought that individual traders drive the post earnings drift. On the other hand, institutions are big traders and therefore dominate price-setting. One could also argue that it is difficult to argue that the drift could not represent market inefficiency due to the fact that if naïve trading were to induce patterns of mispricing, then smart arbitrageurs would be able to profit from the mispricing. Nonetheless, a literature in behavioural finance and accounting puts it that despite arbitrage by sophisticated investors, the behaviour of imperfectly rational investors can induce mispricing such as the post earnings announcement drift. A model developed by Fischer and Verrecchia (1999) shows that irrational investors can be able to influence the price of an asset in the short run. In addition, a paper by Hirshleifer and Teoh (2003) modelled firms’ choices between alternative means of presenting information, and the effects of different presentations on market prices when investors have limited attention and processing power. Their conclusion was that irrational investors have non-negligible risk-bearing capacity and therefore they affect asset prices. In case naïve investors are exposed to common misperceptions, then in equilibrium, these misperceptions can influence price by an amount that depends on the relative size and risk tolerances of different investor groups. According to Lamont and Thaler (2003), they discuss how limits to short-selling can prevent process from adjusting to reflect the views of sophisticated investors.
Prior research has shown that there is a difference between customers of the discount brokerage and other individual investors. For instance, clients with full-service brokerages are more likely to receive more advice about which stocks to choose and trade and this could eliminate the possibility of them making naïve trades. Again, it is expected that such kind of investors are likely to be relatively sophisticated and also to have the befit of professional advice. According to Hirshleifer et al in their paper “Do individual investors cause post-earnings announcement drift? Direct evidence from personal trades,” they use sample data that has more than 1.25 million clients in a single brokerage firm with a mix of both traditional and online traders. Their observation is that there is no indication that trading by individuals by individuals explains the concentration of drift at subsequent earnings announcement dates.
A group of investors that drive the post earnings announcement drift would trade in a way that opposes a full and rational stock price adjustment in response to earnings surprises. This implies that, typically, after the release of good (bad) earnings reports, these groups of investors would sell (buy) the stock as opposed to placing trades consistent with the naïve traders. Simply, the investors would use a contrarian approach with respect to current earnings news. Hirshleifer et al (2008), therefore, tested whether on average, investors buy after extreme negative news surprises and short the stock after extreme positive news.
Past empirical literature has evidenced that after an extreme earnings surprise; the PEAD manifests itself in the days after each of the next two quarterly earnings announcements. As per Bernard and Thomas (1989), following the drift, there is normally a significant reversal that follows the fourth subsequent announcement. It is important to mention that naïve and sophisticated investors differ in their assessment of the fundamental value of a financial asset. For instance, after receiving good news, a sophisticated investor would believe that the price is too low and that their action to buy the asset would drive prices higher. On the other hand, naïve traders would believe that the prices have moved too much and therefore their action to sell the asset would bring back the asset price back to normal. Consequently, during the following quarter, incoming news data may not resolve suffice to resolve the gap between naïve and sophisticated anticipations. If not, then before the next earnings announcement, sophisticated traders ought to buy and naïve traders ought to sell so as to offset the pressure of rational arbitrage trading and eventually preventing the price from adjusting upwards sufficiently prior to the next earnings announcement. On average, subsequent earnings would be higher than expected by naïve traders, leading to abnormally high average return on the earnings announcement date.
Prior empirical research has shown that the post earnings announcement drift represents market inefficiency. With regard to this, sophisticated investors can exploit this pattern near the subsequent earnings announcement by using a dynamic trading strategy. For instance, after a positive earnings surprise, investors can earn high returns by buying shares a few days prior to the next quarterly announcement and partly unwinding their positions in the days following the announcement. This trading strategy offers a favourable balance between risk and expected return.
Ken is a Quantitative Trader with experience in investments, quantitative finance, financial modelling and algorithmic trading in Global Investable Markets (GIM). He enjoys using Bayesian Statistics, Time Series and Machine Learning in developing Robust consistent Alphas in Equities Market, FX, ETPs and Derivatives instruments. He enjoys deep dives in understanding High Frequency Trading infrastructures and improving how the African financial markets work. He holds a Bachelor's in Actuarial Science from Strathmore Institute of Mathematical Sciences : An Executive Program in Algorithmic Trading (EPAT) certificate in Algo Trading from QuantInsti : A current MSc student in Financial Engineering at World Quant University.