Typo in GarleanuPanageasModel?

[MaximilianJHuber]

I am working through your GarleanuPanageasModel and I think I found a typo, but I cannot be sure, because I do not know with certainty how your pA and mcA maps into the paper's g^A and n^A. The former is suggested by the definition of κ, the latter by the definition of r.

Anyway, the denominator of equation 9 has the terms concerning A with a plus sign and the terms with B with a minus sign. This is an error, unless pAx/pA = - gAx/gA.

Furthermore, equation A.16 in the appendix defines n^A which resembles your mcA almost perfectly. But, the coefficient in the second line agrees with you definition of mcA, only if γA/(1-γA) = 1, which is not true.

Is there a documentation that could accompany the example?

[MaximilianJHuber]

Yeah, that was a quick shot. Indeed, your pA is 1/gA and hence pAx/pA = - gAx/gA, so there is no typo, the second issue remains, though.

[matthieugomez]

Ok. What is the second issue?

[MaximilianJHuber]

A.16 in the appendix: is your mcA the paper's n^A. If yes, I cannot match second line in A.16 with your definition of mcA.

[matthieugomez]

Good catch. I think it is a typo in the Garleanu Panageas appendix. Can you try to rederive A.16 from A.14 and tell me whether you obtain their formula vs my formula?

[MaximilianJHuber]

You are right. I collected all terms with a κ from A.14 and found your expression, not the one in the paper. How did you know it was a mistake? Again, is there a supporting document you have for your example?

[matthieugomez]

Cool, thanks. I’ll upload some notes.

[matthieugomez]

Actually just have a look at slides 9 here