Our company has a strong real estate soul, and like all operators in this market, our division has approached the proptech world with curiosity (acronym for Property Technology, i.e. the application of information technology to the real estate world).
The platforms created with this model offer among other things: innovative real estate brokerage services, real estate property management according to automated models, crowd funding, remote property evaluation.
It is precisely this latter application that has particularly aroused our interest; it is significant for our business to be able to evaluate the greatest number of properties in the shortest possible time. The huge amount of data available to the suppliers of these services (so-called Big Data) is a promising source of accuracy.
According to a recent survey conducted by Crif (source “Il Sole 24 Ore”) this is not exactly the case. More than two thirds of these reports report values that diverge from the reality of the facts.
Is it therefore possible to rely (only) on these assessments to make decisions?
With this degree of uncertainty, who can these reports serve?
And above all, what does “value” mean?
The game is played basically on the details. There are variables that affect the property evaluation that cannot be measured remotely, or only partially.
A few examples:
- the views
- the neighborhood
- the context
- the ratio between m2 and number of rooms
- the quality of the comparables
- the relationship between the status of places, intended use and user
Starting from a static value, these parameters have a significant impact on the appraisal price. However, it is complex to attribute the impact of these variables a priori, the only possibility is to question the market of intermediaries and potential buyers.
And at this point we come to the initial question.
We tend to forget that the economic “value” of a property (and any good) is nothing more than a conventional, floating value, continuously redesigned over time according to countless variables, some of which are not directly related to the real estate market (e.g. fiscal, political and social issues).
It is therefore already complex to assign a value in the presence of all the variable data, let alone if we want to do it in the absence of some of them and remotely.
In the NPL market, the technological tool at the service of real estate valuations can be a valid aid for evaluating very large portfolios; the data of thousands of positions fed to the electronic brain will return a total number, which will only work if considered as a whole.
In the case of servicer buy-side such as GMA, this type of evaluation (at least with the current technology available) can only provide indicative data. In the presence of a concentration of risk, a specific assessment, drive-by and with territorial investigation, is required.
It would be interesting to see the deviations of the error curve in relation to the number of practices taken into consideration; until then, we combine proptech with tradition.