Market Anomaly: Holiday Stock Return Effect

Sergey Okun This article “Market anomaly: holiday stock return effect” was written by Sergey Okun – Senior Financial Analyst at I Know First, Ph.D. in Economics.

Highlights

  • A correctly identified market anomaly can generate a profit for an investor that enables them to beat the market.
  • We have tested the Holiday effect anomaly and we have found that exploration of this anomaly cannot provide the desired return for an investor today.
  • The I Know First AI algorithm can identify working market anomalies that could be difficult for an ordinary investor to recognize.
(Source: stockvault.net)

Any investor is interested in finding a strategy that allows him to extract additional returns without taking a corresponding grade of risk. Market Anomaly is a change in the price or return of a security that cannot be directly linked to current relevant information available in the market or to the release of new information into the market. It is a stock behavior that is systemically repeated from one period to another using which an investor can generate a positive return, and where the market efficiency concept does not work. Anomalies can appear, disappear, and reappear with almost no warning. Moreover, before exploiting a market anomaly, a researcher should find supporting evidence that the market anomaly is consistent over reasonably long periods. Otherwise, a detected market anomaly may largely be an artifact of the sample period chosen. Also, wishing to detect a profitable anomaly, a researcher can pitfall in the process of data snooping.

In our previous articles about market anomalies, we tested the weekend effect, and January and April effects. Here we are going to test the holiday effect. The holiday effect says that returns on stocks on the day prior to market holidays tend to be higher than on other days. There are some explanations for this anomaly. One explanation states that short-sellers close their risky positions prior to holidays. Another reason could be investors’ good mood around holidays, indicating greater optimism about future prospects and, therefore, a high probability of positive market moves.

We test the holiday effect by the t-test of unequal variances.

X1-the mean of holidays, X2 – the mean of other days, s12 – the variance of holiday, s22 – the variance of other days, n1 – the number of holidays observations, n2 – the number of other days observations
(Figure 1: The t-test of Unequal Variances)

Let’s analyze the existence of the holiday effect in the S&P500 returns for the recent 5 years period from March 27th, 2018 to March 24th, 2023.

(Figure 2: Testing the Holiday Anomaly of the S&P500 returns for the period of March 27th, 2018, to March 24th, 2023)

According to Figure 2, we can notice that Holiday’s mean is around 0.16% while Other’s mean is around 0.04%. Therefore, we can notice that the average return on holiday is higher than on other days, which, at the first glance, supports the existence of the holiday market anomaly.

The unequal variance t-test, allows us to test the equality of variances, to identify whether returns on holidays are different from returns on other days. Firstly, we calculate the difference between the holidays’ mean and the other days’ mean, which is 0.1199%. Figure 1 enables us to calculate the standard deviation and the t-tests which are 0.2059% and 0.58, respectively. Consequently, the p-value for holidays and other days is equal to 56.06%, which means that we cannot reject the null hypothesis about the absence of the holiday effect. Therefore, we have not found statistical support that the holiday effect has worked for the last 5 years.

Let’s check for an even longer period to validate our findings. Below in Figure 3, we can notice our estimation to identify the holiday effect for the period of January 4th, 2000, to March 24th, 2023.

(Figure 3: Testing the Holiday Anomaly of the S&P500 returns for the period of January 4th, 2000, to March 24th, 2023)

According to Figure 3, we can notice that the Holidays’ mean is around 0.09% while the Other days’ mean is around 0.02%. The p-value for holidays is equal to 44.26%, which means that we cannot reject the null hypothesis about the absence of the holiday effect.

Does this mean, according to our results, that the holiday effect has never existed? The answer is no because the existence of these effects is very well documented in the past periods of the 20th century. Below, we can notice the statistical result of identifying the holiday effect from January 3rd, 1928, to March 24th, 2023.

(Figure 4: Testing the Holiday Anomaly of the S&P500 returns for the period of January 3rd, 1928, to March 24th, 2023)

According to Figure 4, we can notice that the Holidays’ mean is around 0.25% while the Other days’ mean is around 0.02%. The p-value for Holidays is equal to 0.00%, which means that we can reject the null hypothesis about the absence of the holiday effect. Therefore, we can sum up that we have found statistical significance evidence of the existence of the holiday effect in the historical data, which has not appeared since 2000.

Exploit Market Anomalies with I Know First

A correctly identified market anomaly can generate a profit for an investor that enables him to beat the market. However, recognizing a working anomaly is a challenging task where a mistake can cost money. AI can identify a working market anomaly that could be difficult for an ordinary investor to recognize. The I Know First predictive algorithm is a successful attempt to discover the rules of the market that enable us to make accurate stock market forecasts. Taking advantage of artificial intelligence and machine learning and using insights of chaos theory and self-similarity (the fractals), the algorithmic system is able to predict the behavior of over 13,500 markets. The key principle of the algorithm lays in the fact that a stock’s price is a function of many factors interacting non-linearly. Therefore, it is advantageous to use elements of artificial neural networks and genetic algorithms. How does it work? At first, an analysis of inputs is performed, ranking them according to their significance in predicting the target stock price. Then multiple models are created and tested utilizing 15 years of historical data. Only the best-performing models are kept while the rest are rejected. Models are refined every day, as new data becomes available. As the algorithm is purely empirical and self-learning, there is no human bias in the models and the market forecast system adapts to the new reality every day while still following general historical rules.

I Know First has used algorithmic outputs to provide an investment strategy for institutional investors. Below you can see the investment result of our S&P 500 Stocks package which was recommended to our clients for the period from January 1st, 2020 to March 22nd, 2023 (you can access our forecast packages here).

The Investment Result for the period from January 1st, 2020 to March 22nd, 2023

The investment strategy that was recommended by I Know First accumulated a return of 78.55%, which exceeded the S&P 500 return by 54.65%.

Conclusion

Any investor is interested in finding a strategy that allows him to extract additional returns without taking a corresponding grade of risk. Here, we have tested the Holiday effect anomaly and we see that exploration of this anomaly cannot provide the desired return for an investor today. A correctly identified market anomaly can generate a profit for an investor that enables them to beat the market. However, recognizing a working anomaly is a challenging task where a mistake can cost money. The I Know First AI algorithm can identify working market anomalies that could be difficult for an ordinary investor to recognize.

I Know First Premium article

To subscribe today click here.

Please note-for trading decisions use the most recent forecast.