0
$\begingroup$

I have data science projects coming up in a university course in the upcoming quarters, and I was interested in using financial/economic data since there is so much of it. Also, I do have a hobby for trading, so I felt that this would be a good area to try and practice data science in.

So far I have collected a large amount of stock market data from the Americas, Europe, and Asia. Additionally, I've been collecting foreign exchange data, GDP, interest rates, and yields. I'm interested in other possible data sets that are worth looking into as well.

However, I'm quite lost as to how to approach these large data sets. I feel that in the professional sphere, the first step would be to analyze the data to try and get an idea of what the global picture is like. I'm not sure exactly how professionals would examine the data, despite having taken some micro/macroeconomics courses in university.

Can someone give me some tips?

$\endgroup$
  • $\begingroup$ There is a quantitative finance stack exchange, I believe. There might be more useful responses there. The short answer from economics is that “markets are efficient”; there’s no easy way to make money by looking at patterns in existing data. Patterns do exist, but they tend to be things that happen simultaneously (so they offer no predictive power). If basic number crunching could uncover money-making patterns, there would be a lot more rich mathematicians out there. $\endgroup$ – Brian Romanchuk Nov 20 '18 at 12:15
  • $\begingroup$ So when an economist looks at data, they are looking for fundamental cues as to what is going on? I suppose they look at business confidence, interest rates, and GDP to see whether or not it is time to invest or not? $\endgroup$ – yqz09 Nov 20 '18 at 20:08
  • $\begingroup$ There’s a whole industry devoted to trying to find the best time to buy/sell. It’s going to be very difficult to research something that hasn’t been investigated already if you aren’t familiar with the literature. I think you’d need to narrow the scope, and then see if there’s a book on investing that has some discussion that looks similar to what you want to do, $\endgroup$ – Brian Romanchuk Nov 21 '18 at 23:31
0
$\begingroup$

As an answer already indicated there are a lot of papers on this subject. However if the project is about data handling and not so much about economic theory, I would suggest you to have a look at the correlation heatmap(basically covariance matrices illustrated) and maybe do an factor analysis to find the driving forces within the dataset. Both are built-in in most statistical software, and are very intuitive to interpret.

From this you might find yourself asking questions, and looking for answers and all of sudden you have a topic !

$\endgroup$
  • $\begingroup$ Hi: Also, note that having vast amount o data isn't necessarily that useful because the processes in finance and economics are constantly undergoing structural changes anyway. So it won't necessarily be helpful to have a lot of data. Data science in finance and economics might be more hype than reality. Be careful of hype. It's all over the place, mainly due to marketers, salespeople and media. $\endgroup$ – mark leeds Dec 22 '18 at 16:43

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.