# How to define vectors when used in equation

So in short I want to have something that looks like this:

So far I managed to get to the following output:

with the code as follows:

\documentclass[12pt, a4paper] {article}
\usepackage{amsmath}
\DeclareMathOperator{\var}{var}
\usepackage{mathdesign}
\renewcommand{\vec}[1]{\mathbf{#1}}

\begin{document}

$$\begin{gathered} R_{it} = \alpha_{i} + \beta_{i} R_{mt} + \varepsilon_{it} \\ E(\varepsilon_{it} = 0) \qquad \var(\varepsilon_{it}) = \sigma_{\varepsilon_{it}}^2 \end{gathered} \label{eq: marketmodel}$$

$$\vec{R}_i = \vec{X}_i\vec{\theta}_i + \boldsymbol{\vec{\varepsilon}}_{it} \label{eq: mmvectors}$$
where $\vec{R}_i = [R_{iT_0} \dotsm R_{iT_1}]'$ is an $(L_1 x 1)$ vector of estimation-window returns,

\end{document}


Note please that especially my in-text definition of the variables in the equation are messed up. Also further help for the following equations (4.4.3 - 4.4.6) would be great! Thanks!

• Isn't that a \tau (a stopping time), not t, in the subscript to R? – A.Ellett Dec 12 '14 at 0:49
• Don't use \mathbf (for \vec), but \boldsymbol instead. See How can I get bold math symbols? – Werner Dec 12 '14 at 0:52
• Better use \usepackage{bm} and \renewcommand{\vec}[1]{\bm{#1}} for example – user11232 Dec 12 '14 at 0:55
• @Werner (and Harish) the image shows bold roman for vec R (so \mathbf rather than \boldsymbol or \bm, but then somewhat inconsistently uses bold italic greek so \bm{\theta} – David Carlisle Dec 12 '14 at 0:58
• @A.Ellett - Actually, t is much more natural than \tau. It's an application of the Capital Asset Pricing Model (CAPM) -- no stopping time arguments arise. The only reason I can think why the book's author is using \tau is because of a possible distinction between "event time" (denoted by \tau, I gather) and "calendar time", which presumably would be denoted by t. – Mico Dec 12 '14 at 2:50

I can't read the image very well. This uses bm and \tau. I've also replaced x with \times. I've also added some potentially useful packages - remove any you don't want. I've used Latin Modern just for the example because mathdesign didn't display well in my viewer with the default settings and I didn't want to mess around with it.

\documentclass[12pt, a4paper] {article}
\usepackage[T1]{fontenc}
\usepackage{mathtools,bm,amsfonts,amssymb,lmodern}
\DeclareMathOperator{\var}{var}
% \usepackage{mathdesign}
\renewcommand{\vec}[1]{\bm{#1}}

\begin{document}

$$\begin{gathered} R_{i\tau} = \alpha_{i} + \beta_{i} R_{m\tau} + \epsilon_{i\tau} \\ E(\epsilon_{i\tau} = 0) \qquad \var(\epsilon_{i\tau}) = \sigma_{\epsilon_{i\tau}}^2 \end{gathered} \label{eq: marketmodel}$$

$$\vec{R}_i = \vec{X}_i\vec{\theta}_i + \vec{\epsilon}_{i\tau} \label{eq: mmvectors}$$
where $\vec{R}_i = [R_{iT_{0+1}} \dotsm R_{iT_1}]'$ is an $(L_1 \times 1)$ vector of estimation-window returns,

\begin{align}
\vec{\hat{\theta}}_i &= (\vec{X}_i'\vec{X}_i)^{-1} \vec{X}_i'\vec{R}_i\\
{\hat{\sigma}}^2_{\vec{\epsilon}_i} &= \frac{1}{L_1 - 2} \hat{\vec{\epsilon}}_i'\hat{\vec{\epsilon}}_i\\
\hat{\vec{\epsilon}}_i &= \vec{R}_i - \vec{X_i}\vec{\hat{\theta}}_i\\
\var[\vec{\hat{\theta}}_i] &= (\vec{X}_i'\vec{X}_i)^{-1}\sigma^2_{\vec{\epsilon}_i}
\end{align}

\end{document}


• anybody knows what the first symbol in the definion of X_i is??? – mt1111 Dec 12 '14 at 16:06
• @marinko Is it $\ell$? – cfr Dec 12 '14 at 16:10
• ... found out it is \iota, if that's a help for anybody – mt1111 Jan 12 '15 at 22:43

You are better off with bm package and

\renewcommand{\vec}[1]{\bm{#1}}


Code:

\documentclass[12pt, a4paper] {article}
\usepackage{amsmath}
\usepackage{bm}
\DeclareMathOperator{\var}{var}
\usepackage{mathdesign}
\renewcommand{\vec}[1]{\bm{#1}}

\begin{document}

$$\begin{gathered} R_{i\tau} = \alpha_{i} + \beta_{i} R_{m\tau} + \varepsilon_{i\tau} \\ E(\varepsilon_{i\tau} = 0) \qquad \var(\varepsilon_{i\tau}) = \sigma_{\varepsilon_{i\tau}}^2 \end{gathered} \label{eq: marketmodel}$$

$$\vec{R}_i = \vec{X}_i\vec{\theta}_i + \boldsymbol{\vec{\varepsilon}}_{i\tau} \label{eq: mmvectors}$$
where $\vec{R}_i = [R_{iT_0} \dotsm R_{iT_1}]'$ is an $(L_1 x 1)$ vector of estimation-window returns,
\end{document}


I have replaced t with \tau, revert if I am wrong.

the image shows a rather inconsistent font choice using bold upright roman and bold slanted greek for vectors. Perhaps the simplest way to achieve that is using bm package and

\newcommand\vec[1]{\bm{\mathrm{#1}}


then \vec{R} and \vec{\theta} should do the right thing