# Study Session 3: Quantitative Methods for Valuation

## Mean Reversion

lets say assume that we have Xt=0.5+0.25t-1 and and the mean reversion is 0.666

does the 0.666 mean that on the long run, the mean of the above equation will be 0.666?

## Multiple Linear Regression Simplified

Hey Guys,

Hope you guys are preparing well for CFA. Here is my take on simplifying Multiple Linear Regression

Multiple Linear Regression Simplified : https://youtu.be/fGTAcPup6UE

## Autocorrelation and Graphs

If we are asked to identify by looking at a graph if a time series model has the problem of autocorrelation, how do we identify that?

By definition autocorrelation means that todays value depends on yesterdays value. So lets say that the trend is persistently going up or down, does it mean we have the issue of autocorrelation?

And how a graph without autocorrelation supposed to look like?

## Random Walk and Drift

In quants, we speak about random walk with drift and that if we have a drift, b0 is different than 0, but what exactly is the drift supposed to mean?

## Analysis of Variance (ANOVA) table simplified

Quants being one of the hardest topics in CFA level 2. ANOVA table is one of the important topics and a lot of students find it hard to understand. I have tried my best to simplify in this video.

CFA level 2: Quantitative Methods 2020: Analysis of Variance (ANOVA) Table Explained

https://youtu.be/YehQrDdDwEs

Am willing to hear your thoughts.

## Heteroskedacity and R squared

can we say that if we are trying to understand heteroskedacity one thing that may appear is that our R squared may result higher than it should be?

## Heteroskedacity, serial correlation and multicollinearity

Do we need to test for heteroskedacity, serial correlation and multicollinearity in level II?

Based on the learning objectives, they have only asked to”explain the types of heteroskedasticity and how heteroskedasticity
and serial correlation affect statistical inference” and to  “describe multicollinearity and explain its causes and effects in
regression analysis;”.

But then i see problems at the end the chapters asking to test each of this problem

## Heteroskedacity & Standard Errors

Why do we say that with conditional heteroskedacity, the standard error is underestimated?

Technically, i think standard errors can be overestimated, and our T and F values will result smaller. As a consequence, we wouldn’t reject the H0 when we would should have and that would cause a type II error

## Hypothesis Testing in Regression Analysis

If the problem asks to determine whether S&P 500 returns (being the independent coefficient) affect ADM’s returns( our dependent variable), that means that when we set up the hypothesis, we would have H0:B1=0 vs B1 different than 0. Am I correct with how I have built my hypothesis ​​​​​​?