Frequency conversion - Low to High - Quadratic match method

[niluki]

Hi,

I want to know the method underlying the Quadratic Match sum.

It says in the userguide that "The quadratic polynomial is formed by taking sets of three adjacent points from the source series and fitting a quadratic so that either the average or the sum of the high frequency points matches the low frequency data actually observed"

After fitting the polynomial, what method is used to calculate the High frequency points?
If I were to replicate this in Excel how can I do this?

Any help is greatly appreciated.

Thank you!

[EViews Gareth]

Say you have the following quarterly data:

Code: Select all

2002Q3 150 2002Q4 105 2003Q1 100

And you want to know the monthly data for Oct, Nov, Dec 2002. You need to find the values of A, B, and C such that the average of At^2+Bt+C (where t is a trend) over the three months in each quarter is equal to the quarterly values above:

Code: Select all

Month Trend at^2+bt+c Average Jul-02 0 172.4074089 Aug-02 1 148.5185195 Sep-02 2 129.0740746 150 Oct-02 3 114.0740743 Nov-02 4 103.5185185 Dec-02 5 97.40740727 105 Jan-03 6 95.74074059 Feb-03 7 98.51851846 Mar-03 8 105.7407409 100

Here the 3rd column is calculated as A*trend^2 + B*trend + C. You need to find A, B, C such that the average of the first three rows is 150, the average of the 2nd three is 105, and the average of the last three is 100. The solution is A=2.22222, B=-26.11111, C=172.4074.

[niluki]

Hi Gareth,

Thank you for the prompt reply.
How do you find the solutions to A,B & C? Is an equivalent of a 'Solver' being used?
And will the trend always start from 0?

Thank you,

[EViews Gareth]

Yes and yes.

[lmvk99]

hi,

I am trying to replicate this but I cannot use solver because there are multiple target criteria (ie. that each of the quarterly sums equal the quarterly number). Does any one have a solution?

Thanks