# Tag Info

4

If you give a covariance matrix an inverse Wishart prior, then it simplifies a lot of math in the calculations. This is called a conjugate prior. If you don't understand conjugate priors, you might want to work through the math on the univariate normal case with an inverse gamma or chi square prior for the variance. The Wishart distribution is just a ...

3

Of course you can choose the prior. As far as I understand the literature, the BL-model is characterized by using the equilibrium implied returns. Otherwise it would just be a Bayesian model. If you estimate the returns in a different way (not taking implied returns from the market portfolio), you could lose the stabilizing inverse optimization step ...

1

Use Mean Squared Forecast Error (or any other forecast evaluation metric). Your question appears to complicate the problem: If your goal is to forecast a given parameter you can test the rolling forecast against the actual observed values. This will also give you a metric of uncertainty as you can then create confidence intervals around your forecasts based ...

1

I know kalman filters and have ... If this knowledge extends to Unscented filters, UKF, you can think of the UKF being a sparse particle filter. With a UKF you have a few sigma points which are propagated forward via your model function and then after the measurement update these sigma points are updated via covariance estimation. With a particle filter, ...

1

Since you add the bayes-theory tag here I'm gonna speak in bayesian interpretation; I'd say it's just because this is the simplest way to obtain the distribution of prior; A better way to do this is by finding a prior optimal (essentially finding best mean and variance that fits our assumption of return distribution based on the data you have; usually done ...

1

There appears to be some confusion on how Bayesian methodologies work and their differences to null hypothesis methods such as Pearson and Neyman Frequentist, Fisher's Likelihoodist or Chebyshev's method of moments. Having old data is not a prior distribution. That is the data. It is the same for all modeling methodologies. The prior distribution ...

1

You have a correct understanding. There is a subtle difference in that Knight effectively distinguishes uncertainty from chance. It could be argued that there is no such thing as chance in the Bayesian posterior density. To understand why, imagine that you were holding a strictly fair coin and you were going to gamble with an unknown stranger who would ...

Only top voted, non community-wiki answers of a minimum length are eligible