“Data doesn’t exist, it’s the insights and actions that do.”


- Ken Cukier, author of “Big Data”

Why Such High Demand for Quants and Data Scientists


The simple answer is the explosion of data, combined with the reduction in the cost of computing. Together these have created an enormous opportunity to mine and present the underlying data (be it quantitative, or behavioral) for insights and actionable ideas. Wall Street is at the beginning stages of this shift, but significant momentum is occurring. For example, the majority of trading volumes today are quantitatively driven and low-cost style based funds (smart beta, factor oriented) have started to encroach on active managers.


At EDS, we are taking a leading position at generating robust, insightful and actionable analytics on immense data sets, including fundamental information on thousands of global companies, behavioral data, such as credit-card transactions, and clickstream data. Our process relies on four main types of analytics, which we believe quants/data scientists perform from the lenses of an investment-oriented firm. These are described in detail below and are utilized to both filter for and enhance the predictive probability of a stated outcome.

  1. Mean Reversion

  2. Cross Sectional Analysis

  3. Linear Regression

  4. Correlation Analysis

A cross-sectional study, in particular, is ideal in the construction of Price Targets, as it presents a non-biased, real-time view of the upside or downside valuation risk of a particular stock.

Mean Reversion


Mean reversion is the theory suggesting that prices and returns eventually move back toward the mean or average. This mean or average can be the historical average of the price or return, or another relevant average such as the growth in the economy or the average return of an industry.


Corporate performance rarely moves in only one direction, as performance could adjust relative to the health of the economy, the industry it competes in, as well as other company specific or macro factors.  Since there is some degree of cyclicality in most companies, it means they will oscillate around a median or mean level over time. Mean-reversion analysis seeks to identify how a company is currently valued relative to its mean and/or median over a specified period of analysis. At EDS, we perform this calculation on thousands of companies and sectors in seconds, meaning you can spend less time searching for ideas and more time researching the ones that have the greatest upside or downside.

Linear Regression


A linear regression is constructed by fitting a line through a scatter plot of paired observations between two variables. Traditionally, the main application on Wall Street has been geared towards performance correlation to the market (beta or the CAPM model). However, to a quant, and at EDS, we are taking the next step, and using our deep fundamental experience (using linear regression) to pair up variables across multiple fundamental and sector based metrics (margins, valuation, ROE, P/B, Sales, etc.). For example, within the Financials sector, we believe that Price/Book (P/B) and Return on Equity (ROE) are closely related.

Now, in just a few seconds, and across the entire universe, we can test the validity of these (and any) relationships to hone in on the most important fundamental metrics that drive stock returns.

Correlation Analysis


Correlation quantifies the degree to which two variables are related. Correlation does not fit a line through the data points, the same way a regression does. Simply, you are computing a correlation coefficient (r) that tells you how much one variable tends to change when the other one does. For example, if one of your themes relates to rising interest rates, at EDS, we provide precise guidance on which stocks correlate most closely. In this example, you can see that there is a 69% correlation to change in the 10-year yield for Spirit Realty.

Cross-Sectional Analysis

A cross-sectional study is a type of observational study that analyzes data collected from a wide universe, such as the entire US Market or a representative subset, at a specific point in time (cross-sectional data). This differs from traditional comparisons that are biased by history and are generally focused within a peer group.


There are many advantages to a cross-sectional study, which is why it is so popular among quants and data scientists:


  • Used to prove and/or disprove assumptions.

  • Captures a specific point in time with current vs. historical data.

  • Contains multiple variables at the time of the data snapshot.

  • Outcomes often lead to new ideas and opportunities.


With EDS, a cross-sectional study puts valuation parameters into perspective across a wider universe; based on metrics, such as margins, market capitalization and growth.


This analysis, in particular, is ideal in the construction of Price Targets, as it presents a non-biased, real-time view of the upside or downside valuation risk of a particular stock.