John sat at his regular Starbucks table sipping his tall pike. It was his Saturday morning routine to spend a few hours every Saturday morning, turning pages of the Wall Street Journal and thinking about the activities of his week and the upcoming week. As a Quant, trained in Mathematical Finance, John had worked at a hedge Fund for the last 5 years and had a keen interest in technologies and how the Quant Finance profession was transforming with the advent of newer technologies. No longer were asset managers relying just on traditional methodologies and funds were seeking an edge by leveraging newer data-driven technologies. Every conference John went to, he would hear panels and speeches on how AI, Machine Learning and Data science is transforming the investment profession. Though John had heard about Data science techniques, he had never used them and was intrigued by how Data science applications would be implemented. The company John worked at predominantly used Excel for analysis, but they were hitting the limits on what they could do in Excel with the huge increase in the quantity of data they were dealing with the last few years.
Earlier in the week, at the team meeting, there was discussion on how the competition is using Data science techniques and the company was worried about customers moving to other funds seeking alternate strategies. "We should be looking into Data Science", Anna, John's manager who had a PhD in Computational Physics remarked. "I think it is high time, we develop competence in this field and evaluate if this can help us address some of the data challenges we are facing", another colleague added. "Many firms are using Python and Machine Learning techniques nowadays and that is something we should look into", Jacob, the summer intern who just started at the company remarked. John was assigned to do research about Data science techniques and how it could be applicable to their firm. He was asked to study how Data science and Python and present to the team a strategy towards adoption of Data science techniques.
With the Boston Fintech Week coming up, John had signed up to the QuantUniversity's Data Science in Finance Crash course and was excited about learning how data science applications are built in Python. He was also interested in learning how Jupyter notebooks were used and was intrigued about QuSandbox, the tool used to run the labs. As he finished his coffee, he started reading tutorials on Python and Machine Learning gearing up for the crash course coming up the next week!
This case has 4 parts. Each Day, we will cover one part with a lecture and a QuSandbox Lab.
Instructions to use the ResearchHub and QuSandbox will be provided. Contact us if you need additional information on running the labs.
Get data fromhttps://www.kaggle.com/wendykan/lending-club-loan-data
Exploratory Data analysis: Write a Jupyter notebook using Python to graphically represent different summaries of data. Summarize your findings in this notebook.
Summarize your key insights about different user profiles, states, loan amounts etc.
Your next goal is to build a model to predict interest rates. You will get leads from people with different profiles and you must decide if you will give loans or not and if you will give a loan, how much interest you would charge for those loans.
1. Speaking Data Science with an investment accent - CFA Magazine, Cynthia Harrington
2. Machine Learning, an intuitive foundation, Sri Krishnamurthy