How do you evaluate a stock?
|
The naive, Category 1 Trader simply looks at the stock's price and change in price. The stock is abstracted away from the company it represents. They know that some stocks go up and some stocks go down, but they don't have a true understanding for why that's the case.
|
The Category 2 Trader knows a little bit about fundamentals. They read articles and blogs and figured that stocks represent the companies behind them. Businesses that earn more money consistently should be valued better than businesses that don't. They learn about the PE (Price to Earnings) ratio, a ratio of the company's share price to its per-share earnings. They understand that stocks with a low PE ratio have a low share price compared to the amount of money they earn per year. Thus, they've concluded that you should only look to buy stocks that have a low PE ratio. Seems smart at first glance, right?
|
Then we have the category 3 trader. These traders use data to formulate hypothesis, test different ideas, and back their results up objectively. They understand why PE ratio could be a good measure, but they want to validate this explicitly. They do not trust traditional investment "wisdom".
|
Spoiler alert: It is not.
|
Finding our stocks with low PE ratios
|
Query for the top 10 stocks that had a PE ratio above 0 and below 10 on Jan 1st 2023. Sort by market cap descending.
|
The full response from the model is as follows:
|
Here is a summary of the data:
1. Company: Exxon Mobil Corporation (Symbol: XOM) - Description: Engages in the exploration and production of crude oil and natural gas globally.
- P/E Ratio: 9.04616
- Market Cap: $471,642,800,000
2. Company: Pfizer Inc. (Symbol: PFE)
- Description: Discovers, develops, and sells biopharmaceutical products worldwide.
- P/E Ratio: 9.65214
- Market Cap: $287,302,680,000
3. Company: Toyota Motor Corporation (Symbol: TM)
- Description: Designs, manufactures, and sells passenger vehicles and related parts globally.
- P/E Ratio: 7.47751
- Market Cap: $189,102,358,004
4. Company: Verizon Communications Inc. (Symbol: VZ)
- Description: Provides communications, technology, and entertainment products and services worldwide.
- P/E Ratio: 8.5848
- Market Cap: $165,519,400,000
5. Company: ConocoPhillips (Symbol: COP)
- Description: Explores for, produces, and markets crude oil, natural gas, and LNG globally.
- P/E Ratio: 8.56037
- Market Cap: $152,195,338,000
6. Company: AT&T Inc. (Symbol: T) - Description: Provides telecommunications and technology services worldwide.
- P/E Ratio: 6.76593
- Market Cap: $140,781,270,000
7. Company: QUALCOMM Incorporated (Symbol: QCOM) - Description: Develops foundational technologies for the wireless industry globally.
- P/E Ratio: 9.54411
- Market Cap: $123,352,680,000
8. Company: The Goldman Sachs Group, Inc. (Symbol: GS)
- Description: Provides financial services for corporations, institutions, and individuals worldwide.
- P/E Ratio: 9.02693
- Market Cap: $121,899,900,000
9. Company: Intel Corporation (Symbol: INTC)
- Description: Designs, develops, and sells computing products and services globally.
- P/E Ratio: 8.12612
- Market Cap: $108,838,740,000
10. Company: Citigroup Inc. (Symbol: C)
- Description: Provides financial products and services globally. - P/E Ratio: 5.71683
- Market Cap: $87,814,045,000
These entries offer insights into the descriptions, P/E ratios, and market capitalizations of the mentioned companies.
|
This list looks to be about accurate. Being the developer of this platform, I know that the model isn't hallucinating and that the results come from a sophisticated and advanced process in the backend. Unfortunately, validating the results for the end user is not easy. This limitation is discussed in more detail at the end of this article, but for now, trust me when I say these results are backed by objective data.
|
Next, we will create a portfolio using these assets. Our goal is to buy and hold each of these stocks.
|
Create a portfolio. Call it PE Test Trade. I want 10 strategies. Buy and hold each one of these assets
|
Next, we'll save this portfolio to our profile. We'll navigate to it, and then click the "Backtest" button.
|
A backtest simulates the historical performance of our portfolio. It executes the strategies that we generated on past data to see how our portfolio would have performed if we had held during the specified time period.
|
The result for stocks with low PE ratios? Highly disappointing.
|
This strategy performed terribly! Buying and holding the S&P500 (SPY) would have given you MUCH better performance. It would also be easier, less to think about, and requires absolutely no research. Wow!
|
Let's just sit back and think. PE ratio is public knowledge. Calculating it is trivial, and in fact, the most popular brokerages actually show it to you directly.
|
The PE ratio being low litterally means that the stock price is cheap on an earnings per share basis. Thus if it's cheap, there must be a reason. What it is?
|
It's because investors are selling.
|
They don't want the stock, for whatever reason. Either they're not growing, they don't have great free cash flows, or they are in a dying/dead-end industry. Investors don't believe the company will grow significantly beyond its current stock price.
|
So they're selling and buying stocks with a higher PE ratio like NVIDIA.
|
In fact, what would be another interesting analysis is seeing how well high PE ratio stocks perform in the market. Do we see the opposite? 🤔
|
Limitations and Drawbacks to this Approach
|
In this article, we've shown that AI was able to automate an extremely complicated workflow and give us new insights on a type of fundamental metric. Doing so was as easy as talking to a person. However, there are a number of downsides with this approach.
|
For one, it only works with the GPT-3 powered chat. This infuriates me. This wasn't a complicated request, and the insights were extremely valuable. Thus, it should work well with GPT-3, but no matter what I do for the prompt engineering, it simply fails to return the correct results. This should not be the case.
|
Moreover, another limitation of using AI was that it, unless it failed to work at all, it wasn't entirely clear on how to audit these results. The model didn't return any articles or SEC statements to confirm the validity of the numbers in the system. While I know what's happening under the hood, a Category 3 trader would want to validate the source data to ensure the insights actually make sense. This isn't currently possible, and is a HUGE limitation for this type of research.
|
The perfectionist in me HAS to fix this chat. It is one of NexusTrade's best features and biggest differentiators; it should be able to execute this exact same complex functionality using GPT-3. I won't sleep well at night until it does, so I WILL fix this.
|
Also, the chat should return sources for its data. Nobody should be forced to blindly believe the output of a large language model, something that tends to literally make things up on the fly. Investors should have confidence that the insights from the model are accurate without having to worry about finding another way to validate the results.
|
We cannot understate the value of using AI and automation for financial research. The Category 2 trader lost money this past year because they blindly believed traditional investing wisdom. The Category 3 trader, however, uses data to back up their result. At the very least, they don't use a low PE ratio to justify buying a stock. Their using metrics that actually matter, like how much a company's cash flow is changing over time.
|
This demonstrates why NexusTrade is so powerful. In literally minutes, we were able to run a complex experiment and gain new insights on the "PE ratio" metric. We've shown that traditional wisdom turned out to be extremely wrong, and that investors are better off buying and holding assets that represent the S&P500, such as SPY or VOO.
|
AI and automation will transform how informed investors approach the market. NexusTrade makes this process fast, easy, and straightforward. I've learned something about investing that most people do not know, and this was with one of the simplest possible examples. I hope this demonstrates the importance of learning how to use these AI tools to maximize your future potential gains.
|
Disclaimer: The content provided in this post is for informational purposes only and is not intended as financial advice or a recommendation to buy or sell any securities. I am not a financial advisor. The insights and analysis shared are meant to demonstrate the capabilities of NexusTrade in automating financial research. It's important to conduct your own due diligence and consult with a professional advisor before making any investment decisions.
|
|
