Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I'm working through an econometrics textbook and came upon this supposedly simple problem early on.

Suppose you win $\$1$ if a fair coin shows heads and lose $\$1$ if it shows tails. Denote the outcome on toss $t$ by $ε_t$. Assume you want to calculate your average winnings on the last four tosses. For each coin toss $t$, your average payoff on the last four tosses is $w_t= 0.25ε_t + 0.25ε_{t-1} + 0.25ε_{t-2} + 0.25ε_{t-3}$

Find var($w_t$), and find var($w_t$) conditional on $ε_{t-3} = ε_{t-2} = 1$

For the first part of the question (finding the variance without any conditions on $ε$), I know that:

Var($w_t$) = Var(0.25ε_t + 0.25ε_{t-1} + 0.25ε_{t-2} + 0.25ε_{t-3}

= Var(0.25ε_t) + Var(0.25ε_{t-1}) + Var(0.25ε_{t-2}) + Var(0.25ε_{t-3})

=0.0625 Var($ε_t$) + 0.0625 Var($ε_{t-1}$) + 0.0625 Var($ε_{t-2}$) + 0.0625 Var($ε_{t-3}$)

This is the point in the problem that confuses me. I know the variance for the entire time series is 1, because each possible result of the coin flip (-1 or 1) is $\$1$ away from series' expected value of 0. Does this mean that Var($ε_t$) = 1, $∀ t$ as well? Based on that logic, I could expect Var($w_t$) = 0.0625 (1+1+1+1) = 0.25, but this textbook does not provide any solutions, so I have no way of checking my logic.

I'm confident that once I figure out this part of the question, the part of the question that assumes $ε_{t-3} = ε_{t-2} = 1$, but I added that part to my initial question in case there are other nuances I'm missing, in which case I can use said nuances to form a new question.

share|improve this question
2  
The variance of w_t is 0.25, for the reasons you explained. –  Did Oct 7 '12 at 20:27
    
@did Then the condition that $ε_{t-3} = ε_{t-2} = 1$ doesn't change the problem in the slightest, correct? –  Ricardo Altamirano Oct 7 '12 at 20:28
1  
It does change the problem since now, $w_t=0.25ε_t+0.25ε_{t−1}+$some constant, hence the conditional variance is only 0.125. –  Did Oct 7 '12 at 20:33
    
@did Ah, ok. Post your comments as an answer and I'll accept it immediately. Thank you for the help. –  Ricardo Altamirano Oct 7 '12 at 20:39

1 Answer 1

up vote 2 down vote accepted

The variance of $w_t$ is 0.25, for the reasons you explained. But conditioning on $ε_{t−3}=ε_{t−2}=1$ changes the problem since now, $w_t=$ 0.25$ε_t+$ 0.25$ε_{t−1}+$ some constant, hence the conditional variance is only 0.125.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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