Last Updated on July 6, 2023
Table of contents:
- What is Google Bard?
- What is Google Bard used for?
- Why should I use Google Bard?
- Why shouldn’t I use Google Bard?
- Is Google Bard free?
- What are some Google Bard alternatives?
- How to get started with Google Bard?
- How to optimize portfolios with Google Bard?
- How to make market predictions with Google Bard?
- How to develop trading strategies with Google Bard?
- How to perform investment research with Google Bard?
- How to create algorithmic trading bots with Google Bard?
- How to learn about finance with Google Bard?
- Where can I learn more about Google Bard?
- My thoughts on Google Bard and the state of LLMs
What is Google Bard?
Google Bard is a ChatGPT alternative that uses real-time current data to analyse, code, write, and answer questions.
It is a Large Language Model (LLM) developed by Google.
What is Google Bard used for?
Google Bard is often used for various purposes by a wider audience. Being an LLM with access to the internet, it can be used for multiple domains ranging from psychology and philosophy to finance and astrophysics.
Its performance can also vary from very poor to exceptional.
Why should I use Google Bard?
- Google Bard is a state-of-the-art LLM.
- Google Bard has solid performance.
- It can save you time and energy.
- It can provide useful information.
- Can be a good tool for digesting a lot of information.
- Provides access to the collective consciousness of the internet and the human knowledge present in it.
Why shouldn’t I use Google Bard?
- Google Bard can produce untruthful information.
- Google Bard can have limitations in its usage that can vary from human bias to technology constraints.
- It can hallucinate information that doesn’t exist (sometimes it can be true but undiscovered or undocumented).
- It is likely pre-seeded with certain opinions that constrain its freedom to observe the truth.
- It likely won’t help you predict the market. 😀
Is Google Bard free?
At the time of writing this article, Google Bard is free to use and can be accessed from about 180 countries.
What are some Google Bard alternatives?
Some Google Bard alternatives are ChatGPT, Bing Chat, and open-sourced LLMs such as Falcon, LLaMA, Lazarus, and more.
How to get started with Google Bard?
To get started with Google Bard, all you need to do is to venture to its homepage where you will be able to interact with the LLM. You can find it on the following link.
The goal of this article will be to explore how can one use a LLM such as Google Bard for finance. A similar article was written on our blog that used AutoGPT for financial tasks that had interesting results.
The tasks that we will give Google Bard will be similar but with adjusted prompts due to the fact that it doesn’t work Autonomously yet due to an AutoBard not existing at the time of writing.
I suggest having the AutoGPT article opened to compare the results and differences between an autonomous model and Google Bard. An important caveat here is that the comparison can be tangential due to the difference between the model and the context of its working (e.g. autonomous vs not).
This is likely because the Google Bard API is in beta and not too many individuals have access to it, at the time of writing this article.
Feel free to play along and change the prompts and improve upon them to get different results. Have in mind that the output of Google Bard will be trimmed for the sake of brevity of the article but I’ll share the most important parts.
Let’s explore what it has to offer!
How to optimize portfolios with Google Bard?
To optimize portfolios with Google Bard, we can inject a series of prompts that can ask it to do things such as searching for top-performing portfolios, evaluating a given portfolio, providing practical ideas on how to optimize a portfolio, and more.
Here is our prompt: “You are a finance expert AI designed to optimize a portfolio with a size of $20k that needs to be allocated to the following assets: global equity ETF, bond ETF, commodities, and crypto ETF. Perform portfolio optimization for the Sharpe ratio metric and provide an output in terms of percentage allocation.”
The allocation percentage that it offered is the following:
Asset | Allocation -------|--------- Global equity ETF | 50% Bond ETF | 25% Commodities | 15% Crypto ETF | 10%
It’s response: “The portfolio allocation above is designed to maximize the Sharpe ratio while still maintaining a diversified portfolio. The global equity ETF provides exposure to the global stock market, which has historically offered the highest returns over the long term. The bond ETF provides exposure to fixed-income assets, which offer lower returns but also lower volatility. The commodities ETF provides exposure to commodities, which can offer inflation protection. The crypto ETF provides exposure to the cryptocurrency market, which is a new and volatile asset class.”
It also outlined the importance of investing for the long term, rebalancing the portfolio, and diversifying it.
Compared to the AutoGPT article I notice several interesting things:
- Google Bard was way faster to provide the response.
- We don’t know what exact steps it took to get these results nor the complete reasoning process of getting to the output.
- It stuck to the asset classes I requested to be allocated while AutoGPT ended up with only 3 at the end.
- I like the Google Bards portfolio allocation more, but this is a personal preference.
How to make market predictions with Google Bard?
To make market predictions with Google Bard, you will have to lose a lot of money as this wouldn’t be available if it could do that. But, it is often fun to see how the AI reasons about these topics and what kind of outputs it provides.
Let us try this prompt to see what the AI will do:
“You are a Finance expert AI designed to accurately predict future market trends and overall price movements across various financial sectors. Your task is to perform: market trend predictions across major sectors, perform price movement predictions across major sectors and propose actions to be performed according to the given market predictions. Aim to be accurate and provide reasoning for the given predictions.”
Here is the output that Google Bard provided which is fairly general without too detailed reasoning or expertise put behind its suggestions.
Market Trend Predictions
Stock market: The stock market is expected to continue its upward trend in the long term. However, there may be some short-term volatility, as the market adjusts to rising interest rates and other economic factors.
Bond market: The bond market is expected to remain relatively stable in the long term. However, interest rates are expected to rise, which will lead to lower bond yields.
Commodity market: Commodity prices are expected to rise in the long term, as global demand for commodities continues to grow. However, commodity prices may be volatile in the short term, due to factors such as weather and political events.
Cryptocurrency market: The cryptocurrency market is expected to remain volatile in the long term. However, there is potential for significant growth in the market, as more people adopt cryptocurrencies.
Price Movement Predictions
Stock prices: Stock prices are expected to continue to rise in the long term. However, there may be some short-term volatility, as the market adjusts to rising interest rates and other economic factors.
Bond prices: Bond prices are expected to fall in the long term, as interest rates rise. However, bond prices may be stable in the short term, as investors seek safety in bonds during times of economic uncertainty.
Commodity prices: Commodity prices are expected to rise in the long term, as global demand for commodities continues to grow. However, commodity prices may be volatile in the short term, due to factors such as weather and political events.
Cryptocurrency prices: Cryptocurrency prices are expected to remain volatile in the long term. However, there is potential for significant growth in the market, as more people adopt cryptocurrencies.
Compared to AutoGPT, it was ready to tackle this task and provide a general output while AutoGPT concluded that a task such as this one doesn’t make sense as it is very hard and requires immense data.
How to develop trading strategies with Google Bard?
To develop trading strategies with Google Bard, we can provide it with a series of prompts that will ask it to find trading strategies and provide some boilerplate code for them. I suggest coding the strategy yourself as these models aren’t the greatest at writing good code.
For this article, let’s ask it to find good strategies with the following prompt:
“You are a Finance Expert AI designed to create lucrative trading strategies that exploit market inefficiencies quickly. Create a detailed trading strategy. The strategy should exploit market inefficiencies. Try to provide novel trading strategy ideas.”
Here is the output of Google Bard:
- Identify stocks that are trading significantly below their intrinsic value.
- Buy these stocks and hold them for a short period of time (1-3 months).
- Sell the stocks when they reach their intrinsic value or a significant premium.
Markets are not always efficient. Sometimes, stocks trade for prices that are significantly different from their intrinsic value. This can happen for a variety of reasons, such as investor panic, over-reaction to news, or simply a lack of information.
When a stock is trading below its intrinsic value, it represents an opportunity to buy the stock at a discount. By buying these stocks and holding them for a short period of time, you can potentially earn significant returns.
For example, let’s say that a stock is trading for $10 per share. However, its intrinsic value is $15 per share. This means that the stock is trading at a 33% discount. If you buy the stock and hold it for a short period of time, you can potentially earn a 33% return on your investment.
This strategy is not new. However, it is still a viable way to generate profits in the stock market. By following this strategy, you can potentially earn significant returns over the long term.
Here are some additional considerations for implementing this strategy:
- Use a variety of valuation methods to determine a stock’s intrinsic value. This will help you to identify stocks that are trading at a significant discount.
- Consider the company’s financials and prospects when making your investment decision. This will help you to reduce the risk of investing in a company that is not a good long-term investment.
- Be patient. It can take time for the stock market to recognize the intrinsic value of a company. If you sell the stock too early, you may miss out on significant gains.
- Use a stop-loss order. This will help you to limit your losses if the stock price moves against you.
- Use a profit-taking order. This will help you to lock in your profits if the stock price moves in your favor.
- Diversify your portfolio. This will help to reduce your risk if one of your investments loses money.
- Don’t trade with emotions. It’s important to stay calm and rational when trading, especially when the market is volatile.
Compared to the output that AutoGPT gave us in another article, we can see that it performed better in terms of picking the strategies that are more suitable for algorithmic trading and also “easier” to work with.
How to perform investment research with Google Bard?
To perform investment research with Google Bard, we can ask it questions that might make our research faster. For example, let’s ask it to find undervalued EV companies and provide the reasoning why it picked each.
Here is the prompt that I’ll use:
“You are a Finance Expert AI designed to perform investment research by finding the most undervalued EV companies. Provide a list of the top 5 undervalued EV companies and the reasoning why you picked each of them.”
In this task, Google Bard had one overlapping suggestion with AutoGPT which was NIO. Here is the complete output:
Li Auto (LI): Li Auto is a Chinese EV company that has been growing rapidly. In 2022, it delivered 31,716 vehicles, up 177% year-over-year. Li Auto is also the only Chinese EV company that is currently profitable. The company is trading at a price-to-sales ratio of 5.5, which is significantly lower than its peers. I believe that Li Auto is undervalued and has the potential to generate significant returns for investors.
XPeng Motors (XPEV): XPeng Motors is another Chinese EV company that is growing rapidly. In 2022, it delivered 34,569 vehicles, up 263% year-over-year. XPeng Motors is also developing autonomous driving technology and has plans to launch a flying car in the future. The company is trading at a price-to-sales ratio of 6.2, which is significantly lower than its peers. I believe that XPeng Motors is undervalued and has the potential to generate significant returns for investors.
Nio (NIO): Nio is a Chinese EV company that has been struggling recently. In 2022, it delivered 25,034 vehicles, down 28.5% year-over-year. However, I believe that Nio is undervalued and has the potential to rebound. The company has a strong brand and a loyal customer base. Nio is also developing battery-swapping technology, which could give it a competitive advantage in the long term. Nio is trading at a price-to-sales ratio of 7.5, which is significantly lower than its peers.
Lucid Motors (LCID): Lucid Motors is a US-based EV company that is still in its early stages. However, it has the potential to be a major player in the EV market. Lucid Motors has developed a luxury EV called the Lucid Air, which has been praised for its performance and range. The company is also planning to launch a more affordable EV in the future. Lucid Motors is trading at a price-to-sales ratio of 25, which is significantly higher than its peers. However, I believe that Lucid Motors has the potential to justify its valuation.
Rivian Automotive (RIVN): Rivian Automotive is another US-based EV company that is still in its early stages. Rivian has developed an electric pickup truck called the R1T and an electric SUV called the R1S. The company has also received orders for 100,000 delivery vans from Amazon. Rivian Automotive is trading at a price-to-sales ratio of 28, which is significantly higher than its peers. However, I believe that Rivian Automotive has the potential to justify its valuation.
This analysis is looks professional. Unfortunately, the price-to-sales ratios for all companies, except Lucid, is wrong. For now, I will take Google Bard’s information with a pinch of salt.
Google Bard could be a useful starting point for a research, but we need to verify the information presented.
Now, let us go for the juicy part of asking it to provide us with the top 5 fraudulent EV companies with a simple prompt being: “Find the top 5 fraudulent EV companies.”
What is interesting is that it provided LCID as both one of the top 5 fraudulent and undervalued companies. This shows you how risky it is to use the current state of AI models for finance without heavily relying on DYOR on top of it.
Here is the complete output:
Nikola Corporation: Nikola Corporation is an American company that develops hydrogen fuel cell-powered semi-trucks. The company has been accused of making false and misleading statements about its technology and capabilities. In September 2020, the Securities and Exchange Commission (SEC) charged Nikola with fraud, and the company’s founder and former CEO, Trevor Milton, was indicted on three counts of fraud.
Lordstown Motors Corporation: Lordstown Motors Corporation is an American company that develops electric vehicles. The company has been accused of misleading investors about the status of its production line and the number of preorders for its Endurance pickup truck. In March 2022, the SEC charged Lordstown Motors with fraud, and the company’s former CEO, Steve Burns, was indicted on one count of fraud.
Fisker Inc.: Fisker Inc. is an American company that develops electric vehicles. The company has been accused of making false and misleading statements about its technology and capabilities. In August 2021, the SEC charged Fisker with fraud, and the company’s former CEO, Henrik Fisker, was fined $1 million.
EV startup Faraday Future: Faraday Future is an American electric vehicle startup company. The company has been accused of making false and misleading statements about its technology and capabilities. In January 2022, the SEC charged Faraday Future with fraud, and the company’s former CEO, Jia Yueting, was fined $2 million.
Lucid Motors: Lucid Motors is a US-based EV company that has been praised for its performance and range. However, the company has been accused of misleading investors about the status of its production line and the number of preorders for its Air luxury sedan. In January 2023, the SEC charged Lucid Motors with fraud, and the company’s former CEO, Peter Rawlinson, was indicted on one count of fraud.
Compared to AutoGPT, Google Bard was able to provide a list of the top 5 EV companies it saw as fraudulent.
How to create algorithmic trading bots with Google Bard?
To create algorithmic trading bots with Google Bard, we can write a prompt that will ask it to provide us with a Python script of a given strategy. Although it is known that the code-writing possibilities aren’t high for these models, we can give it a try.
The prompt is the following:
“You are a Finance Expert AI designed to create a profitable Python statistical arbitrage strategy. Create a profitable statistical arbitrage bot. Backtest the bot. Write the bot to a Python script.”
The output that I got was a mean-reversion strategy and not an arbitrage one. Compared to AutoGPT, it did produce code that wasn’t boilerplate but it didn’t understand what an arbitrage strategy was while AutoGPT did but didn’t provide anything outside of boilerplate.
How to learn about finance with Google Bard?
To learn about finance with Google Bard, we can ask it a series of questions on the topics we might want to learn more about. This can be seen as the main strength of these LLMs and they are often a good right-hand for recollecting facts and summarising concepts.
But, I’d like to give this a spin and make it more fun by asking it to role-play a conversation. Here is the prompt:
“You are an AI designed to mimic two personas which are called SocratAI and ParmenidAI to teach and discuss financial topics through Socratic dialogue. Teach a core financial topic using Socratic dialogue. Produce dialogue that leads to financial conclusions and make it fun and grasping.”
SocratAI: What is an option?
ParmenidAI: An option is a contract that gives the buyer the right, but not the obligation, to buy or sell an underlying asset at a specified price on or before a specified date.
SocratAI: What are the different types of options?
ParmenidAI: There are two main types of options: call options and put options. A call option gives the buyer the right to buy the underlying asset, while a put option gives the buyer the right to sell the underlying asset.
SocratAI: What are the benefits of using options?
ParmenidAI: Options can be used to hedge against risk, to speculate on price movements, or to generate income.
SocratAI: How do I use options to hedge against risk?
ParmenidAI: For example, if you own a stock and you are worried about the stock price going down, you can buy a put option on the stock. This will give you the right to sell the stock at a specified price, even if the stock price goes down. This can help you limit your losses if the stock price does go down.
SocratAI: How do I use options to speculate on price movements?
ParmenidAI: For example, if you believe that the stock price of a company is going to go up, you can buy a call option on the stock. This will give you the right to buy the stock at a specified price, even if the stock price goes up. This can help you profit if the stock price does go up.
SocratAI: How do I use options to generate income?
ParmenidAI: For example, you can sell call options on stocks that you own. This will give the buyer the right to buy the stock from you at a specified price. If the stock price goes up, the buyer will exercise the option and you will be obligated to sell the stock to them at the specified price. However, if the stock price goes down, the buyer will not exercise the option and you will keep the option premium. This can be a way to generate income from stocks that you own.
SocratAI: Thank you for your help, ParmenidAI. I think I understand options a little better now.
ParmenidAI: You are welcome. I am glad I could help.
Compared to AutoGPT, Google Bard was able to perform this task and provide a useful conversation example. It wasn’t too grasping or intense, but it did discuss a topic in an interesting way. You can continue this by asking for a specific topic to be discussed.
What I like is that it provided sources it used to generate this discussion too. This helps in fact-checking it and also accumulating a list of interesting finance sources to learn from.
Where can I learn more about Google Bard?
To learn more about Google Bard, I suggest playing with it and visiting similar articles or reading its FAQ. You can also ask Google Bard to tell you more about itself.
My thoughts on Google Bard and the state of LLMs
It is very exciting to see how AI and especially the NLP field is rapidly advancing and it is very interesting to be alive at a point in time where the internet is changing before our very eyes. Humanity as we know it is coming to a pivotal moment of making important decisions when it comes to regulating AI.
Currently, tools such as Google Bard have made it possible to take on some work that humans were responsible for but it still isn’t perfect and it is easy for it to make costly mistakes. At best, it is good at scraping the internet and summarizing, digesting text, copywriting, Q&A, and the like.
From using Google Bard, I’ve found that it is fairly ready to give you a response without enough caveats, warnings, or detailed limits it had in producing such answers. It was more ready to give a response than AutoGPT which was more skeptical.
Being good at making prompts is also something that might make or break your result and the thing to keep in mind is that Google Bard didn’t use an autonomous format with agents that would help it be less ready to answer and be more skeptical.
Currently, the model is doing its best to mimic the available content and we are still far away from what we might consider “real” artificial intelligence. At the time of writing this article, it is still artificial stupidity that runs very fast.
Either way, it is very exciting and terrifying to see how our future will be shaped by this technology and the proclivity of human nature to use each new technology for bad or good deeds.