Question

5. Suppose you want to estimate by MCO the following function:

? = ?0?1?1?2?2?

Using the proposed function, suggest a model that allows econometric estimation by MCO. What benefit does the transformation you proposed offer? Explain the assumptions that the error term must meet to obtain an estimate by MCO that can be validated as adjusted to reality.

5. Suponga que desea estimar por MCO la siguiente función: y = Box, B1xxBze Utilizando la función propuesta, sugiera un modelo que permita la estimación econométrica mediante MCO. ¿qué beneficio ofrece la transformación que usted propuso? Explique los supuestos que debe cumplir el término de error para obtener una estimación mediante MCO que pueda ser validada como ajustada a la realidad.

Answer 1

y = Bo X, MX, se This model needs to be linearized in terms of parameters to allow us to do estematedor by meo. To lenearize, we need to apply begarithmic beans forema- tions. Taking logarithms on both sides: logy = log Bo + ß, Loga, + Pe loga + loge > Y = x + B, X, +B₂X₂ + € , where Y=logy, a = logßo, log u=X, Log X = X₂ and lege=t. If we transform the variables (r. x, and x), then we can estimate the above linear function to e obtain estimates of Bio P2 and Bo= exp(2). To estimate the above function by Mco, the exercon () must meet the following conditions. - E(E)= 0 i.e. mean (expected encoR must be zero. - ECE²) = 8².e. variance of ernare in constant (homoskedastic). - Fort CarCE, 2,) and car CE, Nn) 20. inl. ZeRo covariance between error and independent variables. - E(Ei,G)=0 i.e. no autocorrelation.