Nvidia, Telefónica, Repsol… How to take advantage of AI to invest accurately during earnings season
Technology is supposed to make life easier. However, when earnings season arrives, the average investor usually resorts to endless reports, graphs and presentations impossible to follow on time. Large managers have entire teams to analyze each figure, but most investors only have a few hours a day. That’s where artificial intelligence comes in. The AI can help maintain the pulse of the market even following dozens of companies at the same time.
In practice, that means that someone who invests in Telefónica should pay attention to the numbers of Ericsson or Vodafone, just as a shareholder of Inditex looks closely at the reports from H&M or Nike. For years, all that tracking required patience, spreadsheets, and time. Today, the process has changed. The AI has turned results analysis in a race assisted by algorithms capable of processing in seconds what previously required full days.
When AI replaces Excel
For years, earnings season was a marathon of reports that analysts dissected sheet by sheet. Today, the process has been transformed. AI does not replace analyst teams, but it does amplify their reach. Can read thousands of pages of corporate presentations and detect changes in tone in a CEO’s voice with a precision that no traditional analyst matches.
An example was shown by Bank of America in its latest earnings report. Your system Sentiment analysis identified an increase of 18% in the use of terms associated with “margin pressure” at conferences of European industrial companies. Days later, the sectoral index fell 3.4%.
These tools also allow you to track keyword mentions in real time. If Inditex or LVMH frequently pronounce concepts such as “inventory”, “online demand” or “transportation cost”the models adjust their profit forecasts in a matter of minutes. The precision is such that some managers have reduced the time by 40% who take time to review a portfolio after the quarterly results.
own Telefónica experiments with similar tools. In its latest presentation, the company used an AI system to create automatic summaries of their responses to analysts and cross them with the reaction in the stock market in the following hours. The objective is to understand how the market every nuance of speech and adjust communication in future presentations.
At the same time, Nvidia and Metatwo of the most followed stocks on the Nasdaq, use algorithms that process thousands of mentions on social networks and financial forums during the days prior to publication of results. This information allows us to gauge market expectations and anticipate the impact of possible surprises.
How an ordinary investor can use it
But the revolution does not stay in the offices of the big companies. technology companies or big banks. It has also reached individual investors. Today, any user can access similar tools from their own computer. Google has gone one step further with NotebookLMa platform that combines the power of Gemini with the accuracy of financial data to transform reports into actionable insights.
Let’s imagine that an investor wants to know how they affect oil prices to Repsol before it publishes results. Instead of reviewing charts and balances manually, open NotebookLM and type a simple order. “Analyze the relationship between the price of the Brent barrel and Repsol’s profits over the last year.” In a matter of seconds, the model crosses the data series, generates a graph and calculates the degree of correlation. If the result exceeds the 80%, the system warns that the margins of the oil company move practically at the pace of crude oil.
The user can go further. If you upload Repsol’s latest results presentations to the tool and ask “Summarize for me how the tone of the messages has changed on margins or production compared to the previous quarter”, the model identifies variations in language and highlights whether communication has become more cautious or more optimistic. In practice, it acts as an analyst who reads and compares reports in seconds.
NotebookLM can also cross-reference reports from different companies. If the investor combines those of Repsol, Telefónica and Inditex, he can ask the system to detect which companies mention terms most frequently such as “inflation”, “energy costs” or “demand growth”. With this you get a quick x-ray of the Spanish business pulse before the earnings season.
The challenge of interpretation
Speed does not always guarantee success. The models are so good like the data they train with. An excess of poorly calibrated information can generate false signals, especially in periods of volatility or in the face of mixed results. Analysts remember that AI does not replace human judgment, but rather complements it.
Last season, high-frequency algorithms reacted erratically to Tesla’s results. The negative headlines on reduced margins caused a massive sell-off automatic, but income exceeded forecasts and the value recovered 9% in the following hours. This discrepancy showed that, no matter how sophisticated the technology is, the context remains fundamental. The change is profound. Investing with precision no longer depends only on instinct, but also on having more and better tools.
