We asked ChatCPT to Create an Algorithm from our Data fields

If you are anything like us, you are spending a lot of time in front of ChatGPT and its equivalents. The use cases are countless and with our unique 100% complete dataset of every Regulation Crowdfunding offering, the ability to analyze what is happening in this industry just exploded. Particularly when you consider layering on our dataset with datasets comparing states, economic output, other parts of the private capital markets, etc.

Back in 2016, we launched the industry’s first data aggregator. We collect 125 data points on every company that is raising capital online from who they are, when they  incorporated, what industry they are in, to where they are located, their valuation, their financials, daily amount committed and daily investor count. This data has been available on Bloomberg’s Terminal since 2020.

We track not only each of these deals while they are live but track their successive rounds and their progress via their annual reports. Needless to say our tiny dataset when we started is quite huge today. You’ve seen much of our analysis in our monthly recaps and our annual report.

We were fascinated with what ChatGPT can do so we asked it the following: “Create an algorithm that reviews investment crowdfunding offerings and rates them from good to bad based on when a business was incorporated, where it is located, if it has sales, how fair its valuation is compared to others in its 3-digit NAICS code, how much capital has been invested, how many investors in the round, and other factors like cash, accounts receivables, short-term debt, long-term debt, and income.”

Here’s what it had to say:

Certainly, here is an algorithm that reviews an investment crowdfunding offering and rates it from good to bad based on various factors:

  1. Gather information about the business being offered for investment, including its incorporation date, location, sales, valuation compared to others in its 3-digit NAICS code, cash, accounts receivables, short-term debt, long-term debt, income, how much money they have raised to date and how many investors have participated.
  2. Calculate a score for each of these factors based on the following criteria:
  • Incorporation date: The longer the business has been incorporated, the higher the score. For example, a business that has been incorporated for over 5 years might receive a score of 5, while a business that has been incorporated for less than a year might receive a score of 1.
  • Location: The score for location will depend on factors such as the state’s tax policies, labor laws, and the state’s business-friendly index. For example, a business located in a state with low taxes, business-friendly policies and labor laws might receive a score of 5, while a business located in a state with high taxes, stringent labor laws and a low business-friendly index might receive a score of 1.
  • Sales: A business that has a high amount of sales will receive a higher score. For example, a business with over $1 million in sales might receive a score of 5, while a business with no sales might receive a score of 1.
  • Valuation compared to others in its 3-digit NAICS code: A business that has a fair valuation compared to others in its 3-digit NAICS code will receive a higher score. For example, a business that has a valuation within 10% of the median valuation for its 3-digit NAICS code might receive a score of 5, while a business that has a valuation that is more than 50% higher than the median valuation for its 3-digit NAICS code might receive a score of 1.
  • Funding amount and number of investors: A business that has raised a high amount of money and has a large number of investors will receive a higher score. For example, a business that has raised over $1 million and has over 50 investors might receive a score of 5, while a business that has raised less than $100,000 and has only a few investors might receive a score of 1.
  • Cash, accounts receivables, short-term debt, long-term debt, and income: Each of these factors will be evaluated individually and will receive a score based on the specific circumstances of the business. For example, a business with high cash reserves might receive a score of 5, while a business with a high amount of long-term debt might receive a score of 1.
  1. Calculate a weighted average score based on the importance of each factor to determine the overall score for the business. For example, the overall score might be calculated as follows:

Overall score = (Incorporation date score x 0.1) + (Location score x 0.2) + (Sales score x 0.2) + (Valuation score x 0.2) + (Funding score x 0.2) + (Cash score x 0.05) + (Accounts receivables score x 0.05) + (Short-term debt score x 0.05) + (Long-term debt score x 0.05) + (Income score x 0.05)

  1. Rank the business based on its overall score, with the highest-scoring businesses being rated as good and the lowest-scoring businesses being rated as bad. For example, businesses with an overall score of 4.5 or higher might be rated as good, while businesses with an overall score

Pretty interesting right? While this doesn’t take into account necessary diligence needed on founders, their industry experience, their startup experience, or the experience at raising capital. It does provide a way to narrow down the focus.

What do you think? Are you looking for data for machine learning or AI that you can use in conjunction with other data sets? If so, contact us for more information!

Contact us