Constructing an econometric model

Discussion in 'Technical Analysis' started by joshua12, Apr 7, 2012.

  1. joshua12


    Hello, I am in need of some help regarding econometric construction and analysis. I need to construct a 'simple' econometric model of UK house prices, however all the reading that I have done around this area hasn't seemed to help me formulate a general idea of how this would be achieved.
    Can anyone help me/advise on how to go about it?

  2. Do you attend LSOE or South Hampton Institute of Technology? :confused: :D
  3. joshua12


    Thank you for your help, however I am more concerned with creating a model that isolates the effect of 'cost of credit' on house prices. I.e. the basic interest rate plus bank rate. Essentially I cannot find a single model that explains the relationship between the two, can anyone help?
  4. The econometrics projects I saw in my courses showed that people with insider knowledge in the sense of their even having the data was the only way to attack this problem.

    Basically you need to acquire a bank's proprietary macro datasets if you want to find this information, or create your own database which is much too time consuming.

    You can find these datasets if you know someone who has them, but you're really stuck if you don't. It is possible to create regional econometrics analyses if you understand what you're modelling, and it sounds like you don't, or else you would know more precisely what is required to know about the relationships of those two when most would be looking at the spreads.

    Your understanding further of interest rate derivatives and defining things like liquidity premiums in housing is essentially the same as a bond pricing formula to determine the attributes of the spread and the sizes of particular premiums.

    I wish I could help you, but you'll need to either make these datasets yourself or borrow someones' or some firms' data.

    To be more helpful, define your independent variable first, then figure out what quantitative or qualitative data you need to explain it. If this is a course project more than half of the struggle is getting data, about 1% is programming, and the other 49% is analysis.

  5. The difficulty is the word "isolates". There are many variables that determine house prices, most likely including nonlinear relationships and seasonal effects. However if one wanted to do a really simple (and bad) analysis, one could put average house price in column A, and interest rate in column B, and have Excel solve the equation.
  6. are you using Stata ?
  7. He's using STATA or SAS, and I'm pretty sure this wouldn't be excel.