New Project: Buy a House

I've been informed by my better half that I cannot start the next book until I buy a house. "No pressure", she says, "but June/July 2011 is the deadline".

I figure that the only way I can get through this terribly boring process is to crunch the data (hopefully on this blog). I also figure the best way to start the journey is to define the acceptance criteria and primary information sources. I'm sure some of this is taboo to declare, but I figure any provoked discussion can help me to learn faster (worth it). Also, public declarations can be blunt instrument for motivation if all else fails.

Generally: I'm constrained to Melbourne, Australia. Common sense suggests the house must be treated as an investment, requiring an estimate on the return on that investment.

Acceptance criteria:

• Buy a house to live in:
• 3+ Bedrooms
• Within the "zone one" public transport boundary
• With ~15 minutes walk of a tram or train to the CBD
• Granny flat or very separated bedroom (for wife-parent-visits)
• East or North of the CBD (in that order of preference)
• Buy a house for investment
• Within the "zone one" public transport boundary
• Car park or garage

Primary sources of information:

• http://www.realestate.com.au (has a better rank on Alexa)
• http://www.domain.com.au
• http://www.canstar.com.au (loan interest rate comparison)

Areas of interest seem to include properties of the loan (interest rate, constraints, etc) and properties of the prospective house (properties of location, section 32, etc). Early discussions with a mortgage broker has suggested NAB Homeside with an interest rate of 6.90% (which meet our needs and appears reasonable). The loan side seems to be the easy part, for now.

At least one sticking point with the house side is private sales which will likely not be listed on aggregates sites (requiring direct contact with real estate organizations?). This may or may not matter, like looking for a job and relying on the single most popular job board as the source of advertisements. On the 'properties of the house' side, some specific studies I am going to crunch include:

• Graph of country, state, city, suburb mean house price (the Bureau of Statistics was useful here in the past).
• Regression of house price in city and suburbs of interest (something like square suburb/footage/price).
• What are the asking price bands by suburb?
• What is a good suburb to keep an eye on?
• What is a fair price for an area?
• Automated notification of anomalous asking prices
• Automated notification of downward per/suburb clearance/asking price trends.

Some of these studies I have run ad hoc before, and obviously the extent of the work I can do here is contingent on data availability.