what about the capital and financial account balance? anyone remember thier answers?
Dsylexic Wrote: ------------------------------------------------------- > 4.9 = 0.84 +4.06 thought it was 0.8 (1.66*0.5) everything else held constant…maybe I’m wrong
^^ That’s what I did. They asked for change so I didn’t include the intercept as that is constant.
mcpass Wrote: ------------------------------------------------------- > ^^ That’s what I did. They asked for change so I > didn’t include the intercept as that is constant. Yep
The model as a whole (so including intercept) predicts the change in exchange rate. So the intercept needs to be included.
Wasn’t the dependent variable itself (the FX rate) stated as %change itself?
that is what i thought as well…did it say “everything else held constant”? i know it said what is the change, implying the values were already set, which is why i did 0.5 * 1.7 and not add the intercept
MarkvanOmmen Wrote: ------------------------------------------------------- > The model as a whole (so including intercept) > predicts the change in exchange rate. > > So the intercept needs to be included. i had a glance at this in Schweser last night - they indicate that it would be the coeff * the change when holding everything constant - intercept not included. i better not have this wrong as easy marks are slipping by
But this model does not predict the exchange rate, but the change in exchange rate. Therefore the intercept also contributes to the change.
I actually meant that we do had to include the constant to get the change…if the dependent variable itself is the change, what would we get if we just multiplied regression coefficient x value? But I’m totally not sure anymore.
My head is still mush, but, for mult regression this is how you describe the impact of a change in an independent variable on the dependent variable. Say y = b0 +b1 (my change) + b2 (your change) + e if b1 is 0.80 and my change is 5% then holding all else constant… y will change by 4%…
slouiscar Wrote: ------------------------------------------------------- > My head is still mush, but, for mult regression > this is how you describe the impact of a change in > an independent variable on the dependent > variable. > > Say y = b0 +b1 (my change) + b2 (your change) + e > > if b1 is 0.80 and my change is 5% then holding all > else constant… y will change by 4%… yep and b1 was 1.66 (0.5) = 0.8 but who knows
balls, no intercept? I might have to sign off AF untill after the results, I’m bleeding.
if you have an equation, and then you change one variable by 0.5 times the coefficient, that should be the only change. nothing else should be changed. that is my interpretation
completely agree with the no intercept concept…even though ive seen questions where the intercept was included. I think this question was addressing % change in Y and not the level in Y. In Schweser, there was a question concerning the change in the level in Y, and that included the intercept.
Not when the dependant variable is a %change. The the coefficient only describes the change in change. You do need to add the intercept.
No intercept needed. They did not ask to predict the value from the change.
over05 Wrote: ------------------------------------------------------- > No intercept needed. They did not ask to predict > the value from the change. Agree. If you ask what is the change in the dependent given a change in X1 of 5% holding all else constant, then the change in the dependent Y is b1 * (X1). if dependent Y is the change of ratio Z then the end result is that the change changes by b1X1 holding all else constant If you ask what is the value of the dependent given a change in X1 of 5% holding all else constant then you plug into the whole y = b0 + b1X1+ b2X2 and determine what is the dependent. change, changes, changing changes, all fair and fun ways to test my knowledge of quant. I am going to be irritated if I fail. Take that stress, pressure, comprehension nonsense, the LI calcs and the what is the best Ho: out and ask 2-3 q’s about time series or model mispec, or DW or any of the other 20 quant topics that could separate the hard work we spent to understand the curriculum vs the glance over quant crowd.
Once more, If you ask what the change in exchange rate is and the dependant variable is defined as “change in exchange rate”, then you need to run the entire model. The model states that when both coefficients are zero, the change in exchange rates is still equal to the intercept.
I’m pulling for you Mark.