Why do PropertyRadar's market analysis numbers differ from other resources?

Why do PropertyRadar’s numbers different from other real estate websites that publish market analysis?

Our market analysis numbers may be different for one or more subtle reasons:

  1. Many analysts base their reports on MLS data. MLS data while valuable for some analysis does not necessarily reflect what actually occurred in the market. First, it is self-reported by the agents who enter it. Second, it does not include off-MLS transactions, which in some markets can be significant.

  2. Some analysts base their reports on when the data is received into their system, often called “published date” and not the official recording date at the county. Think “cash” vs. “accrual” accounting. For us, a January sale is a January sale if the sale took place in January as recorded by the county, not data from a sale in December recorded in December and populated as a record into a database in January. We believe that any analysis based on “published date” isn’t even worthy of discussion, and should simply be ignored.

  3. "No wine before it's time." Many analysts face publication deadlines for content and may work with incomplete or smaller than desirable samples of the market data to rush an analysis to publication. The "plumbing" between public record offices and the data aggregators frequently breaks down. County recorder offices get behind for a myriad of reasons. Data aggregators face staffing and format consistency issues. There are many and frequent reasons why it can take a longer than expected for all the data for a month to come in. PropertyRadar derives no economic benefit from publishing our reports, so we don’t rush them. We take our time and only analyze and publish a report when the quantity and quality of data is appropriate. We're quite happy to be the last to comment.

  4. Seasonal adjustments. Many analysts use seasonal adjustments to make their reporting more consistent. There is value to this, but those adjustments are more art than science, and as a result, always include some degree of bias. Instead, we report on the actual numbers and make a point of noting these seasonal factors so that users can make their own conclusions on the impacts.

  5. Finally, public records are messy. As a result, each analyst has to implement some degree of modeling, for example, to determine whether a grant deed represents a market sale, or non-market transfer between related parties, like moving a property into a trust. Over the last ten years, we’ve developed our own algorithms and models to determine these differences and to correct common public records data errors. We also have a data research team employing real live humans that handle subjective data issues that algorithms cannot correct instead requiring research to rectify.
Have more questions? Submit a request

Comments

Powered by Zendesk