As most of you know, valuation has always been a core focus of efforts and a core area of expertise given Areté’s value-oriented investment philosophy. As a result, it was a big decision to terminate a relationship with a vendor last year that provided a valuation model and to bring that functionality in-house.
As with anything that plays such central role in activities, it was a hard decision at first. Many things would have to change and it would take work to evaluate and incorporate new alternatives.
While there is still work to be done, things are progressing nicely. One of the most valuable aspects of the old relationship was the access to data – not just for valuation but for portfolio analysis and reporting as well. This is one, area, however where tremendous progress is being made by a number of startups working with XBRL data.
For those not familiar with XBRL, it is the data tagging protocol implemented by the SEC in 2009 for public companies. One of the most important aspects of XBRL is that it is cheap. The actual data is free to consumers since reporting companies produce it. The only cost is to access the raw data in a user-friendly format which a number of providers are starting to do.
As a result, the total cost to access corporate financial data through XBRL is almost an order of magnitude lower than it is through historical methods from legacy providers. In addition, data accessed through XBRL is typically available within minutes of its being filed with the SEC.
When I first started researching different providers, my biggest concern was that data quality may not be as good. What I have found through my own tests is that there were initially some data issues, but that they have been substantially cleared up. The data now is actually quite good.
A bit of investigative work revealed why the high data quality shouldn’t be much of a surprise. Since a large chunk of companies outsource the tagging to the big acccounting firms, the numbers are subject to an array of standardized quality checking processes that highlight anomalies and correct errors.
Areté’s current provider of financial data is Calcbench and they provide yet another benefit; they make sure each of the data items it reports are traceable to the SEC document where it was reported. As a result, there is absolutely no mystery about the numbers.
By contrast, some other providers are not nearly so transparent. While some of those providers try to “add value” to the underlying data by making adjustments to better reflect operating numbers, it can be difficult or impossible to reconcile “operating numbers” to anything in the reported financial statements.
While such efforts are standard analytical fare and laudable as such, they also expose consumers of the data to risks. Without being able to easily and regularly audit the data, it is virtually impossible to monitor the quality of the key input (i.e., data) to any valuation model. The risk is, “garbage in, garbage out.”
Regardless, now that I have much more control over the data, and as the XBRL eco-system continues to get built out, it is now also becoming clear that there is a long runway for continued improvement. One effort in particular will be exploring ways to combine XBRL’s cheap access to data with our analytical expertise in order to provide useful, data-driven insights for investors.
Finally, with the valuation model work in progress and as a purveyor of good old fashioned security analysis, I am encouraged by what seems to be a shift in the market back to a valuation focus after several years. If that holds, it may be a nearly perfect time to be stocking up Areté’s analytical toolkit!