Date: January 17, 2024

Topic: A closer look at daily returns

Recall

Histograms are another possible indicator for a stock’s performance, by tracking daily returns

Notes

Histograms to track occurrences

• Use histograms to see how many times a daily return of a certain percentage occurs

Daily returns over a long time

Daily returns of S&P 500 over a long time

• It is likely that the returns will form a normal distribution

Kurtosis can be used to compare a stock’s daily return histogram with a normal distribution

Statistics with histograms

• We can get common statistics like mean and standard deviations

Kurtosis

• By comparing our graph with a normal distribution at the tails, we know how related the histogram is
• Positive kurtosis: Fat tails and flatter, more occurrences at edges
• Negative kurtosis: Thin tail and stretched at center, less occurrences at edges

<aside> 📌 SUMMARY: Histograms are another indicator for a stock’s performance. We can use kurtosis for evaluation

</aside>

Date: January 17, 2024

Topic: Plotting histograms

Recall

1. Get the daily returns
2. Call .hist() to get histogram

Notes

Plotting the histogram

Histogram with 20 bins

• Obtain the daily returns graph, then run daily_returns.hist()

# First get the daily returns from the stock
daily_returns = (df/df.shift(1)) - 1
daily_returns = daily_returns[1:]

# Plot the histogram
daily_returns.hist() # Default 10 bins

# Increase the number of bines
daily_returns.hist(bins=20)

From the histogram, we can plot the mean and std to get additional information on the graph

<aside> 📌 SUMMARY: Use dataframe operations to discover statistics about histograms or even plot them together

</aside>

Date: January 17, 2024

Topic: Scatterplots