# What statistical tools does a analyst (financial) require?

I just graduated in engineering where I have a very strong background in pure maths, but unfortunately I didnt study any statistics in my course. I know calculus, probability theory, measure theory, analysis etc. But I have not done any courses on regression etc.

I am planning to apply for several "analyst" positions at various banks/investment firms in the coming months.

Amongst some of them, they specifically require the applicants to have a good knowledge in statistics.

What statistical techniques/concepts do I need and should I learn (self study), that an financial analyst would require? Basically I will be analysing and computing statistical and financial data, what techniques should I learn? Should I start with regression, time series etc.?

As taken from wikipedia, this is something I would probably be doing:

Financial analysts use spreadsheet and statistical software packages to analyze financial data, spot trends, and develop forecasts. On the basis of their results, they write reports and make presentations, usually making recommendations to buy or sell a particular investment or security. Senior analysts may actually make the decision to buy or sell for the company or client if they are the ones responsible for managing the assets. Other analysts use the data to measure the financial risks associated with making a particular investment decision.

So what statistical concepts/tools/techniques/methods do I need to study, in order to do the above (in the quotes)?

I just got the book George Snedecor: Statistical Methods, but what main concepts/techniques are most important/most used by analysts? I'd like to quickly start into those topics first.

• In a flag it has been suggested that this thread be made CW (community wiki), in analogy with the recent Elementary statistics for jurors. The analogy is good; but a subtle, important difference is that "juror" is an extremely diverse situation with an ill-defined relationship to statistics whereas "analyst" is much better defined. That suggests the present question need not be CW. For background and latest SE policy on CW, please visit The Future of Community Wiki. – whuber Aug 22 '12 at 14:44
• My students who have gone on to take positions like the one you seek almost universally report that the most useful statistical skill they learned was...how to use Excel effectively :-). – whuber Aug 22 '12 at 14:46
• Yeah, I was going to say that analyst normally means excel monkey. This is true if you're building DCF models on individual companies. However, there are quantitative roles at financial firms that require more statistical and programming skills. For instance, if you were building factor models than it would help to bone up on statistics. – John Aug 22 '12 at 16:31
• I am skeptical of your perception of "financial analyst". Are you going for (1) M&A, (2) FICC, (3) ECM/DCM or (4) back office? (1) and (3) need 0 statistics. (4) needs 0 statistics in most roles, unless you're specifically applying to a statistics/data analyst role. (2) ... well that's harder to say, but as an entry level role you're going to be asked 0 statistics questions unless they're specifically looking for someone in their central risk team (even then .. it's going to be brain teasers, nothing more than figuring out what the sample space is!). – user13253 Aug 22 '12 at 23:12
• Also consistent with your post is applications to (i) junior quantitative trader at a fund (ii) trader's assistant. But for (i) they want C++ gurus primarily, understanding of econometrics is highly unlikely to be required and for (ii) it's more varied. If it's a market maker they want you to know your derivatives and have good VBA. If it's a hedge fund it's harder to say, but statistics .. no. – user13253 Aug 22 '12 at 23:15

Maybe you want to take a look in Statistics and Finance: An Introduction by David Ruppert. It covers a wide range of topics but none in detail. The author actually written this book with engineers in mind and it's very application-oriented.

Table of Content:

1 Introduction
2 Probability and Statistical Models
3 Returns
4 Time Series Models
5 Portfolio Theory
6 Regression
7 The Capital Asset Pricing Model
8 Options Pricing
9 Fixed Income Securities
10 Resampling
11 Value-At-Risk
12 GARCH Models
13 Nonparametric Regression and Splines
14 Behavioral Finance

• That's the content of an undergraduate finance major in a nutshell. Portfolio theory (5 and 10), derivatives and bonds (8, 9), utility theory (14), econometrics (12,13,4,2,6), asset pricing frameworks (7). Nice find. – user14281 Sep 27 '12 at 8:21
• If you're applying to sales & trading at an investment bank you want to be expert on (8),(9) for general role and also splines (in (13)) for a quantitative role. You can ignore the rest except (2) for the brainteasers. For M&A/ECM/DCM you can ignore the entire list. All you will do is addition and division in Excel. For non-quantitative market making you want (2) for the interview and expert in (8). For statistician at a retail bank or other financial firm you want the econometrics. – user14281 Sep 27 '12 at 8:40

Take a look at these lecture notes for Financial Data Analysis from University of Chicago's Financial Math Program to get an idea of topics. Rene Carmona's Statistical Analysis of Financial Data or Ruey Tsay's Analysis of Financial Time Series or An Introduction to Analysis of Financial Data with R might be good as well.

There are a number of statistics books targeted for business and finance. You will find that they emphasize regression and time series. In the case of financial data volatility is an important concept and so the more recent GARCH models that tend to be covered in advanced courses may be important. When I taught business statistics I used an earlier version of this book by McClave and Benson. Its table of contents should give you an idea of what topics are important in business and finance.