标题: Asset allocation process [打印本页] 作者: cfalevel2011 时间: 2011-7-11 19:09 标题: Asset allocation process
As opposed to Monte Carlo simulation (dynamic multi period), the resampled efficient frontier and Black-Litterman are both considered static (one period) asset allocation models right?
If so, what are other multi period methods? Any others we need to know?作者: mik82 时间: 2011-7-11 19:09
yeah except for monte carlo all other 3 are static approaches. I don't know of any others though.作者: Chuckrox 时间: 2011-7-11 19:09
pop, I just found it later in the notes. They confirm heer's understanding.作者: Analyze_This 时间: 2011-7-11 19:09
BL is tricky. I think for the exam, a basic understanding of what it does is ok (and why its better than MVO).
--<ignore this next part unless you're bored/drunk>--
From what I gather, the process basically examines market weights (based on market cap) and historic correlations between markets to determine total risk and other var/covar params for the 'traditional' model. It kind of 'backsolves' risk premia to determine what the market 'says' the optimal portfolio is. Then, you overlay your investor 'views' on that to weight more markets either heavier or lighter than what the 'backsolved' equilibrium allocations are.
I could be wrong. Peyote. Jack Bauer.作者: bboo 时间: 2011-7-11 19:09
Can we have a dynamic asset allocation without using the Monte Carlo simulation?作者: mp3bu 时间: 2011-7-11 19:09
deriv108 Wrote:
-------------------------------------------------------
> Can we have a dynamic asset allocation without
> using the Monte Carlo simulation?
It could be possible with a uninvented supersonic mathematic formula, but would take thousandfolds of time of ML, I think.
