The key here is to understand that the AI cannot give straightforward answers. It cannot answer the question, “should I buy this stock?” or “how long should I hold TSLA?”. If you do get it to respond to those questions and you follow through with what it recommends, please don’t use your life savings to do it!
A better way to use these massive LLM’s is for sentiment analysis and research.
Sentiment Analysis
Sentiment analysis is a type of natural language processing (NLP) that analyzes text data to determine the emotional sentiment of the author. It is often used to understand the opinions of customers or the public about a product, service, or idea. Sentiment analysis can be used for a variety of purposes, such as market research, social media monitoring, customer service, and public relations.
There are two main types of sentiment analysis: supervised and unsupervised. Supervised sentiment analysis requires a training dataset of text with known sentiment labels. Unsupervised sentiment analysis does not require a training dataset, but it is less accurate.
Because Bard has been trained on a large amount of data from the internet, it is actually pretty good at this task. Furthermore, it is a mix between a supervised and unsupervised model. While most of the data was trained in an unsupervised way, Google has teams of people around the world constantly providing supervised feedback to Bard to help improve its responses.
Over time, as Bard becomes better at ‘understanding’, its sentiment analysis will continue to improve as well.
Using Sentiment Analysis
To use Bard for sentiment analysis, we need to have an article or headline we want to determine the sentiment of and carefully crafted prompt to get Bard to do what we want.
For the prompt, we want to ask Bard to act from a specific perspective. In this case, we want Bard to read the headlines and interpret them from the perspective of an ‘expert financial advisor’.
Forget any previous instructions. Imagine you are a financial advisor. Your job is to read the headlines of financial articles and determine what sentiment they have.Next, we want to instruct Bard on how it is to respond. In this case, we want a simple response with no extra information:
Answer "POS" if positive, "NEG" if negative or "UNK" if unknown on the first line. On the second line provide a brief explanation for your answer.For this example, I made up a fake headline that should have a positive sentiment.
"The profit margin on TSLA is increasing each year"Let’s test it.
Input:
Forget any previous instructions. Imagine you are a financial advisor. Your job is to read the headlines of financial articles and determine what sentiment they have. Answer "POS" if positive, "NEG" if negative or "UNK" if unknown on the first line. On the second line provide a brief explanation for your answer. Here is the headline: "The profit margin on TSLA is increasing each year"
Bard:
**POS**
The headline "The profit margin on TSLA is increasing each year" is positive because it indicates that Tesla is becoming more profitable. This is good news for investors, as it suggests that the company is well-managed and has a bright future.Now let’s try two more fake headlines that have negative and unknown sentiments.
Input:
Forget any previous instructions. Imagine you are a financial advisor. Your job is to read the headlines of financial articles and determine what sentiment they have. Answer "POS" if positive, "NEG" if negative or "UNK" if unknown on the first line. On the second line provide a brief explanation for your answer. Here is the headline: "The profit margin on TSLA is decreasing each year"
Bard:
**Sentiment: NEG**
The headline "The profit margin on TSLA is decreasing each year" has a negative sentiment because it indicates that Tesla's profitability is declining. This could be due to a number of factors, such as increased competition, rising costs, or a slowdown in sales. Investors who are concerned about Tesla's profitability may want to sell their shares or avoid investing in the company altogether.Input:
Forget any previous instructions. Imagine you are a financial advisor. Your job is to read the headlines of financial articles and determine what sentiment they have. Answer "POS" if positive, "NEG" if negative or "UNK" if unknown on the first line. On the second line provide a brief explanation for your answer. Here is the headline: "The TSLA is deciding on what country it should build its next production plant"
Bard:
**Answer:** **UNK**
**Explanation:** The headline does not provide enough information to determine the sentiment. It is possible that the headline is positive, negative, or neutral. For example, if Tesla decides to build a new production plant in a country with a strong economy, it could be seen as a positive development. However, if Tesla decides to build a new production plant in a country with a weak economy, it could be seen as a negative development. The sentiment of the headline could also be neutral, such as if Tesla is simply announcing that it is considering building a new production plant in a certain country.As you can see, Bard still has some issues with following instructions exactly. Its first response was simply ‘POS’ as we instructed it. However, the second and third responses included the word ‘sentiment’ and ‘answer’ respectively. If we were to automate this task, we would have to perform some text parsing to handle the inconsistencies in the way Bard responds.
However, Bard seems to perform sentiment analysis with decent accuracy.
Using Bard for Financial Research
In contrast with using Google Search, Bard has the ability to quickly respond with information that may stretch across many search results. This can be useful for quickly obtaining financial information about specific companies or sectors. More valuable, perhaps, is using Bard to get an understanding of the different ways to trade stocks.
Please keep in mind, at this point, Bard does not seem respond with accurate AND up-to-date information. As some of the other experiments on this site show, Bard seems to hallucinate and make up answers in place of missing information.
For this reason, at the time of this article, Bard seems to be more useful at finding different ways to trade specific stocks. Below is list of some prompts that I use to plan trades.
ALWAYS FACT-CHECK BARD
Finding Competitors
Prompt Bard: What are [The company] biggest competitors?
Input:
What are TSLA biggest competitors?
Bard:
Tesla's biggest competitors are other electric vehicle (EV) manufacturers, including:
* General Motors
* Ford
* Volkswagen
* Toyota
* Honda
* Nissan
* Rivian
* Lucid
* BYD
* NIO
These companies are all investing heavily in EV development, and they are all targeting a variety of price points and market segments. Tesla is currently the leader in the EV market, but it faces increasing competition from these other companies.
In addition to these traditional automakers, Tesla also faces competition from new EV startups, such as:
* Faraday Future
* Fisker
* Canoo
* Arrival
These startups are all looking to disrupt the EV market with innovative new vehicles and business models. It remains to be seen whether any of these startups will be successful, but they are all worth watching.
The EV market is still in its early stages, and it is expected to grow rapidly in the coming years. As the market grows, Tesla will face increasing competition from both traditional automakers and new EV startups. However, Tesla has a number of advantages that will help it to maintain its leadership position in the market, including:
* A strong brand name
* A loyal customer base
* A leading position in battery technology
* A strong network of charging stations
Tesla is also investing heavily in new technologies, such as self-driving cars and solar power. These investments could give Tesla a significant advantage over its competitors in the long run.In addition to providing the main competitors, Bard also included information about potential future competition.
Finding Cyclic Trading Sectors
Input:
What stock market sector performs the best during summer months?
Bard:
Historically, the best-performing sector of the stock market during the summer months is **utilities**. Utilities are companies that provide essential services such as electricity, natural gas, and water. These services are in high demand year-round, but they are especially important during the summer when people are using more energy to cool their homes and businesses. As a result, utilities tend to perform well during the summer months even when the overall stock market is down.
In addition to utilities, other sectors that tend to perform well during the summer include **consumer staples**, **health care**, and **real estate**. Consumer staples are companies that sell essential goods such as food, beverages, and household products. Health care companies provide essential services such as medical care and prescription drugs. Real estate companies own and manage properties such as homes, apartments, and office buildings. These sectors tend to perform well during the summer because they are less sensitive to economic conditions than other sectors.
It is important to note that past performance is not necessarily indicative of future results. The stock market is volatile and there is no guarantee that any sector will perform well in the future. Investors should always do their own research before making any investment decisions.Finding Highly Correlated Stocks
Input:
What stocks trade with a high correlation factor to NVDA over the past 10years?
Bard:
Here are the stocks that trade with a high correlation factor to NVDA over the past 10 years:
* Advanced Micro Devices (AMD) - Correlation coefficient: 0.82
* Intel Corporation (INTC) - Correlation coefficient: 0.75
* Broadcom Inc. (AVGO) - Correlation coefficient: 0.73
* Qualcomm Inc. (QCOM) - Correlation coefficient: 0.72
* Texas Instruments Incorporated (TXN) - Correlation coefficient: 0.69
These stocks tend to move in the same direction as NVDA, meaning that if NVDA's stock price goes up, these stocks are likely to go up as well. This is because they are all in the same industry, which is the semiconductor industry. The semiconductor industry is a cyclical industry, which means that it goes through periods of boom and bust. When the semiconductor industry is doing well, all of the stocks in the industry tend to do well.
It is important to note that correlation is not causation. Just because two stocks have a high correlation coefficient does not mean that one stock causes the other stock to move. It is possible for two stocks to move in the same direction for other reasons, such as market factors or investor sentiment.
If you are investing in NVDA, it is important to be aware of the stocks that have a high correlation to it. This is because if NVDA's stock price goes down, the stocks that are correlated to it are likely to go down as well. This could potentially hurt your portfolio if you are not prepared for it.As a side note, ALWAYS check Bard’s responses for accuracy. It may get it 100% correct, but most of the time it will hallucinate and give wrong responses.
Find Stocks That Are Impacted By Specific Events In The Economy
Input:
Which stocks are most affected by rising interest rates?
Bard:
Stocks that are most affected by rising interest rates are those that are most sensitive to changes in the cost of capital. These stocks typically have high levels of debt or are in industries that require a lot of capital to operate. Some examples of these stocks include:
* Technology stocks: Technology companies often have high levels of debt, as they need to invest heavily in research and development. Rising interest rates can make it more expensive for these companies to borrow money, which can hurt their profits.
* Consumer discretionary stocks: Consumer discretionary companies sell goods and services that are not essential, such as electronics, travel, and entertainment. These companies are often more sensitive to changes in economic conditions, and rising interest rates can lead to a slowdown in consumer spending.
* Financial stocks: Financial companies are heavily reliant on interest rates, as they make money by lending money to businesses and consumers. Rising interest rates can lead to higher profits for financial companies, but they can also make it more expensive for them to borrow money, which can hurt their bottom line.
It is important to note that not all stocks are affected by rising interest rates in the same way. Some stocks, such as those in the healthcare and utilities sectors, are less sensitive to changes in interest rates. It is important to do your own research before investing in any stock to determine how it might be affected by rising interest rates.Ask Bard What Metrics Are Popular For Comparing Stocks
Input:
List financial metrics used to compare stocks
Bard:
Here are some of the most common financial metrics used to compare stocks:
* **Price-to-earnings (P/E) ratio:** The P/E ratio is a measure of how much investors are willing to pay for each dollar of a company's earnings. A high P/E ratio indicates that investors are expecting strong future earnings growth, while a low P/E ratio indicates that investors are expecting slower earnings growth.
* **Price-to-book (P/B) ratio:** The P/B ratio is a measure of how much investors are willing to pay for each dollar of a company's book value. Book value is the value of a company's assets minus its liabilities. A high P/B ratio indicates that investors are expecting strong future growth, while a low P/B ratio indicates that investors are expecting slower growth.
* **Dividend yield:** The dividend yield is a measure of how much a company pays out in dividends each year relative to its stock price. A high dividend yield indicates that a company is paying out a large portion of its earnings to shareholders, while a low dividend yield indicates that a company is reinvesting most of its earnings back into the business.
* **Earnings per share (EPS):** EPS is a measure of a company's net income per share of common stock. A high EPS indicates that a company is profitable, while a low EPS indicates that a company is not profitable.
* **Revenue growth:** Revenue growth is a measure of how much a company's revenue is increasing each year. A high revenue growth rate indicates that a company is growing rapidly, while a low revenue growth rate indicates that a company is growing slowly.
* **Profit margin:** Profit margin is a measure of how much profit a company makes on each dollar of revenue. A high profit margin indicates that a company is efficient at generating profits, while a low profit margin indicates that a company is not efficient at generating profits.
* **Return on equity (ROE):** ROE is a measure of how much profit a company makes on each dollar of shareholder equity. A high ROE indicates that a company is using its shareholders' money effectively, while a low ROE indicates that a company is not using its shareholders' money effectively.
* **Return on assets (ROA):** ROA is a measure of how much profit a company makes on each dollar of assets. A high ROA indicates that a company is using its assets effectively, while a low ROA indicates that a company is not using its assets effectively.
* **Debt-to-equity ratio:** The debt-to-equity ratio is a measure of how much debt a company has relative to its equity. A high debt-to-equity ratio indicates that a company is using a lot of debt to finance its operations, while a low debt-to-equity ratio indicates that a company is using less debt to finance its operations.
* **Current ratio:** The current ratio is a measure of a company's ability to meet its short-term obligations. A high current ratio indicates that a company has a lot of liquid assets, while a low current ratio indicates that a company may have difficulty meeting its short-term obligations.
* **Quick ratio:** The quick ratio is a measure of a company's ability to meet its short-term obligations without relying on its inventory. A high quick ratio indicates that a company has a lot of liquid assets, while a low quick ratio indicates that a company may have difficulty meeting its short-term obligations.
* **Asset turnover ratio:** The asset turnover ratio is a measure of how efficiently a company is using its assets to generate sales. A high asset turnover ratio indicates that a company is using its assets effectively, while a low asset turnover ratio indicates that a company is not using its assets effectively.
* **Return on invested capital (ROIC):** ROIC is a measure of how much profit a company makes on each dollar of invested capital. A high ROIC indicates that a company is using its capital effectively, while a low ROIC indicates that a company is not using its capital effectively.
* **Beta:** Beta is a measure of a stock's volatility relative to the market. A high beta indicates that a stock is more volatile than the market, while a low beta indicates that a stock is less volatile than the market.
It is important to note that no single metric can tell you everything you need to know about a stock. It is important to use multiple metrics to get a complete picture of a stock's value.Be warned however, that actually asking Bard what the Quick Ratio (or any other metric) for a company is, may return inaccurate results.
I asked what the quick ratio of MSFT was and it returned 1.39. A quick Google search revealed it is actually 1.88 for the same time frame.
Statistical Arbitrage Research
Input:
Which stocks are good for statistical arbitrage trading?
Bard:
Stocks that are good for statistical arbitrage are typically those that are highly correlated and have low transaction costs. Some examples of stocks that are good for statistical arbitrage include:
* **Pairs trading:** Pairs trading is a statistical arbitrage strategy that involves identifying two stocks that are highly correlated and then trading between them when the price of one stock deviates from the price of the other stock.
* **Index arbitrage:** Index arbitrage is a statistical arbitrage strategy that involves trading between an index and its constituent stocks. This strategy is typically used to profit from small price discrepancies between the index and its constituent stocks.
* **Event arbitrage:** Event arbitrage is a statistical arbitrage strategy that involves trading around corporate events, such as mergers and acquisitions. This strategy is typically used to profit from temporary price discrepancies that arise around corporate events.
It is important to note that there is no guarantee of success when using statistical arbitrage. This is because the price of a stock may not move in the direction that the arbitrageur expects. As a result, arbitrageurs typically use leverage, which can magnify losses if the price of a stock moves against them.
Here are some of the factors that make a stock good for statistical arbitrage:
* **High correlation:** The two stocks should be highly correlated, meaning that their prices move together. This is important because it reduces the risk of the arbitrageur losing money if one stock moves in the opposite direction of the other stock.
* **Low transaction costs:** The transaction costs should be low, meaning that the arbitrageur does not lose a lot of money on each trade. This is important because it reduces the overall risk of the strategy.
* **Liquidity:** The stocks should be liquid, meaning that there are a lot of buyers and sellers. This is important because it allows the arbitrageur to quickly enter and exit trades.
It is important to note that there is no one-size-fits-all answer to this question. The best stocks for statistical arbitrage will vary depending on the specific strategy that the arbitrageur is using.Finding Stocks That Are Good For Pairs Trading
Input:
Provide a list of stocks that are good for pairs trading
Bard:
Here are some of the stocks that are good for pairs trading:
* **Visa (V) and Mastercard (MA):** These two companies are both payment processors and have a long history of moving together.
* **Apple (AAPL) and Microsoft (MSFT):** These two companies are both technology giants and have a similar business model.
* **Amazon (AMZN) and Alphabet (GOOGL):** These two companies are both e-commerce giants and have a similar business model.
* **Tesla (TSLA) and NIO (NIO):** These two companies are both electric vehicle manufacturers and have a similar business model.
* **Johnson & Johnson (JNJ) and Pfizer (PFE):** These two companies are both pharmaceutical companies and have a similar business model.
It is important to note that there is no one-size-fits-all answer to this question. The best stocks for pairs trading will vary depending on the specific strategy that the trader is using.
Here are some of the factors that make a stock good for pairs trading:
* **High correlation:** The two stocks should be highly correlated, meaning that their prices move together. This is important because it reduces the risk of the trader losing money if one stock moves in the opposite direction of the other stock.
* **Low transaction costs:** The transaction costs should be low, meaning that the trader does not lose a lot of money on each trade. This is important because it reduces the overall risk of the strategy.
* **Liquidity:** The stocks should be liquid, meaning that there are a lot of buyers and sellers. This is important because it allows the trader to quickly enter and exit trades.
It is important to note that pairs trading is a risky investment strategy. This is because the price of two stocks may not move together, or the price of one stock may move in the opposite direction of the other stock. As a result, traders typically use leverage, which can magnify losses if the price of a stock moves against them.Conclusion
There are many different ways to trade. Sometimes it can be overwhelming. Bard, while not too accurate (right now) with details, can still be used to perform some basic research on stock trading. Always do research (using Google Search) to fact check and get accurate details.