The oversimplification of models by keeping them as AR1 at
many places may cause many errors in the projections; these projections are
then cause wrong policy decision. The errors are accumulated in the residuals
for the projection & hence making sure that the past residuals and the
future residuals of AR1 are same which could greatly enhance our projections. To
calculate the Cholesky Residuals, we can create relation between stock price
and other things failing AR1 and then use it. In this case we can use stock
price as related with S&P and exchange rates / bond returns or others. The
residuals would then help to right make the mix between AR 1 models.
It is not just about implementation of residuals but also
about the effect of using different residual policy and back testing the data.
This should become an important area of the research.
|
Sectors
|
Historic Data (Multiple, Slope,
Volatility)
|
Consensus Projections
|
|
Technology
|
||
|
Defense
|
||
|
Health
|
||
|
Retail
|
No comments:
Post a Comment