Anthony Percaccio, research manager, product development and property values at MPAC, said MPAC updated its SAS software from version 6.12 last year. It built an automated valuation model (AVM) to run on the software as a means of unearthing more up-to-date information than it had previously had.
The tool is used mostly by the financial sector for purposes such as processing mortgage applications. For property assessments, MPAC estimates Current Value Assessment at a point in time according to the Assessment Act.
“The upcoming assessments will be based on the Jan. 1, 2005 valuation date, but its AVM is updated monthly and reflects the first of the month,” he said. That means the current AVM value of a residential property is Nov. 1, 2005, which is different from the assessment valuation date.
Percaccio explained that the software uses multiple regression analysis to arrive at the real-time value. It looks at all the characteristics of a property, such as the location, square footage, quality of construction, depreciation or renovation and lot area, and applies a general weight to each.
“There can be an addition to the property where they’ve renovated the interior — for example, the basement,” he said.
MPAC collects that information through building permits, sales investigations and the reconsideration process in which there is a review with the homeowner. “The other way is when we inspect properties, that’s when we find out the changes that have occurred,” he added.
MPAC, a non-profit organization established by the province, has been using SAS for six years for the development of the AVM product, said Percaccio.
“We scoped all the products out there,” he said. “SAS was only option that would allow us to create the product in a monthly fashion. Some of the other products took too long to operate, and SAS offered more efficiency.”
The upgrade, which took about six months, has also allowed the organization to apply ARIMA (auto regressive integrated moving averages) modelling to look at data over time and try to come up with patterns that predict the future.
Michael Turney, manager for strategy and market development at SAS, said organizations such as MPAC need to be able to access data from many sources, formats, servers and locations, which then has to be merged into a single version of “the truth” by discarding duplicate information. After that, it has to be cleansed before it can be used for the purposes of analysis.
“Those reports present information we refer to as actionable intelligence that allows you to make decisions,” he said.
The real value, he said, comes from the predictive intelligence built into the software. In MPAC’s case, it provides the organization with an idea of what any particular property is going to be worth in a year or two.
“That type of additional forecasted view puts us in a different arena,” he says. “Where MPAC is looking for value is what’s going to happen next.”