返回列表 发帖

Technology Efficient and Fatter Tails

Hi,

I have two separate questions (obviously).

First is about being technologically efficient. I don't understand this. I thought it just meant that you use the least amount of inputs.

Therefore, if you have 1 machine produce 500 gizmos or I have 5 workers producing 500 gizmos, you are more technologically efficient.

However, the schweser says a process that is not technologically efficient can't be economically efficient.

What if your machines costs 5,000,000 dollars and I pay my workers $2 an hour. Aren't you more technologically efficient but I am economically efficient?

And now for more question about fatter tails...

On page 23 of schweser secret sauce 2009 (sorry, it's what i've got) It says that leptokurtic distributions have fatter tails. However, on page 36 in a T-Distribution it says that the more degrees of freedom (i.e. more kurtosis) the thinner the tails b/c more of the observations are centered around the mean. Isn't this counterintuitive to what they said on page 23? Thanks in advance for any help you can provide!

@souljive99:

Q1.

Both the methods you described are technologically efficient. There can be more than 1 methods which are technologically efficient.

What technological efficiency means is that you use lowest amount of PARTICULAR inputs. So, the machine uses least amount of labour and the workers use least amount of machinery. Here each is using the lowest amount of particular inputs in either case.

Suppose we were to introduce 2 more cases, case C, with 250 hours of machine and 250 hours of labour, and case D with 250 hours of machines, 2500 hours of labour, Case D will be the only one which will be technologically inefficient.

Also, what Schweser says is correct. A technologically inefficient method can never be economically efficient. Besides, its just a matter of cost which is deciding which method can be economically efficient. So the costs can be structured to show each (A,B,C) as efficient but not D.



Q2.


T-Distribution and Normal Distribution are two different things.

A T-Distribution always REACHES towards Normal as degrees of freedom increase, but generally NEVER EXCEEDS it, while a leptokurtic distribution is more peaked (the tip of leptokurtic distribution EXCEEDS the tip of normal distribution)

Thus you always see statements like T-Distribution (or t-statistic) is more conservative than Normal Distribution (z-statistic) as t isn't exceeding z and thus is covering much more area under which your expected mean could lie.

----------------------------------------

smileyface is gonna clear the CFA L2 Exam on 5th June, 2011. And making this statement won't be a violation of Code of Ethics since its a fact. ;-)



Edited 1 time(s). Last edit at Saturday, May 29, 2010 at 09:42PM by smileyface.

TOP

Smileyface, thanks for the explanation on tech/econ efficient. I've been trying to figure this out forever, and you did a better job that the two prep providers I've been reading

TOP

返回列表