Thursday, March 22, 2007

Week-Over-Week Changes

Sippn and others had an excellent question about whether recent median and average asking price drops were due to newer, lower-priced listings, or price reductions on existing inventory.

Although a close examination of the price inventory level graphs can give us some idea, it might be useful to know what kind of price changes the bulk of the inventory is seeing each week.

In that spirit, I've created and graphed four categories of listings: New, Unchanged, Decreased, and Increased. Note that these conditions are based on week-over-week changes. That means this data is more like a direction indicator, rather than an absolute number. (For you calculus nerds, this is a derivative or vector value.)

The first graph shows the market share each category claims:


The second shows the absolute number of listings in each category:


Although these graphs are fairly chaotic, some useful information can be discerned:

1. Roughly 10% of all listings have price reductions on a weekly basis.
2. Less than 2% of all listings have price increases on a weekly basis.
3. The number of new listings and the number of decreased listings follow a similar trend.

These graphs are just a start. (For example, it would also be interesting to know how many "new" listings are really relisted after a long period of dormancy.) So, let me know if you have any other ideas or suggestions on how to improve this.

No comments :