Average-proxy bid-ask spreads for Nvidia’s stock (and others) over the las six months, using open data.

What we did?

The analysis focused on examining 英伟达 ’s (NVDA) stock performance over the past 6 months. This involved studying the range between the highest bid price (the price at which a buyer is willing to purchase the stock) and the lowest ask price (the price at which a seller is willing to sell the stock), commonly known as the bid-ask spread. The bid-ask spread reflects the liquidity and market dynamics of a stock.

By calculating the daily average bid-ask spread, we obtained a measure of how tight or wide the trading spread was on each trading day. A tight spread generally indicates a liquid market with a small price difference between buyers and sellers, while a wider spread can suggest less liquidity and potentially higher transaction costs.

The visualization, created using ggplot2, displayed the time series of these daily average bid-ask spreads. Adding a linear trendline to the plot helped identify potential trends in the spread behavior over time. This type of analysis is crucial in finance, as it provides insights into how supply and demand dynamics, trading activity, and investor sentiment impact a stock's market behavior.

In essence, this analysis aids investors and traders in understanding how the bid-ask spreads have evolved over the analyzed period, which can inform decisions about market entry, exit, and overall trading strategy. By observing patterns and trends in the bid-ask spreads, financial professionals can glean insights into market sentiment and potential shifts in stock prices.

How we did it?

1. Data Retrieval: We used the tidyquant package to obtain historical data for Nvidia's (NVDA) stock over the past 6 months. This was achieved using the tq_get() function and specifying the desired date range using the from and to arguments.
2. Data Preparation: We added a date column to the dataset to ensure that the values were correctly interpreted as dates. This was done using the mutate() function from dplyr.
3. Calculating Averages: We calculated the daily averages of the bid-ask spreads. We used the group_by() function to group the data by date and then applied summarise() to calculate the average bid-ask spread for each day.
4. Creating the Plot: We used ggplot2 to create the visualization. We employed geom_line() to plot the time series of daily average bid-ask spreads. We then used geom_smooth() with method = "lm" to add a linear trendline to the plot.
5. Labeling and Titles: We used the labs() function to set the plot's title and the labels for the x and y axes.

In summary, the analysis involved obtaining historical data, calculating daily average bid-ask spreads, and creating a plot that displays both the time series and a linear trendline. This type of analysis is common in the financial field to comprehend patterns and trends in market data.

Conclusion?

Over the analyzed time period, the average bid-ask spreads for Nvidia’s stock (NVDA) exhibited noticeable fluctuations, suggesting varying levels of market liquidity and volatility. These fluctuations might reflect changing market sentiment, potentially linked to important events or shifts in supply and demand dynamics.

While bid-ask spreads can impact trading costs, their analysis alone isn’t sufficient to determine whether investors made money during this period. Additional factors such as the stock’s price movements and trading strategies employed would be essential to forming a conclusive assessment of investors’ gains or losses.

What about the rest of the stocks that have performed best in the last 6 months?

Although all the actions had good results, it is evident that Nvidia has a totally different behavior from the rest.

Moore's Law (hardware leveraging software) continues to apply with complete clarity.

BTW: I leave the last 2 papers that I have published in summer 2023:

1 - A Comparative Machine Learning Survival Models Analysis for Predicting Time to Bank Failure in the US (2001-2023). Journal of Economic Analysis here

2 - Buy when? Survival machine learning model comparison for purchase timing. 4th International Conference on Big Data, Machine Learning and IoT (BMLI 2023) Dubai, UAE here