Subscribe to DSC Newsletter

Have you looked at the syllabus of the Finance class on financial modeling at the University? It is outdated: It teaches you how to do modeling on one single company, if possible multiple years of the same company. To any Analytics professional, it is a joke. It's like having a table with a few records - one record for each year. Let's wake up and go for a change:

1. Pour the whole world of companies into research: instead of putting the projection of financial data in different columns, and different financial items such as sales, cost of good sold in rows, can we flip it the other way? having the years in rows and financial items in column will allow us to replicate the same calculation steps across years. Let's do it once for all companies in the interested space.

2. Load company data from market intelligence provider via a data-feed. M&A rumors used to be reported as a news. With the advancement in text mining, some third party provider can add an additional field to indicate if the news is a M&A rumor. With this additional field, we can screen all of the rumors and pick the interesting company for modeling. To be more teckie, we could even write a function to simulate the combination of companies and project the outcome, with all the necessary accounting adjustment.

3. Use volume to supplement precision. Research guys used to be smart in tapping the right contact for "insight". If you are not able to get one target company's assumptions, can those number be extracted from world wide average or same/comparable country(s) average?

4. Private companies and mirco-economic data become more transparent. We used to be blind about the performance of a single company's overseas unit due to its unlisted/private nature. With the activity seen with the data collectors, we are gaining more information about one's company in structured data format. Similarly, the Economic data are collected and analyzed more detailed than before. e.g. world input output table depicted how the different countries industry import and export their products - it is tracking how each dollar is spent and earned.

5. Do back-test. Can equity research guys use predictive modeling to predict what economic factors/ financial factors impact the company's performance? 

I learnt all of these from my experience of building credit risk model. It works for risk, why not equity research?

Views: 2251

Tags: big, data, equity, research

Comment

You need to be a member of AnalyticBridge to add comments!

Join AnalyticBridge

Comment by Jeffrey Ng on July 30, 2013 at 10:13am

You are right David, there is the sea of data. What is needed is the ingredient to mix them up with Analytical skills, Accounting and Economics.

Comment by Jeffrey Ng on July 27, 2013 at 7:30pm
yes. what you have written is interesting but they ate more related to equity strategy, not fundamental research.
Comment by Vincent Granville on July 27, 2013 at 12:47pm

On Data Science Central

© 2019   AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service