Artificial intelligence has brought new efficiency and productivity to the valuation field, particularly where it concerns data entry, processing and analysis, which is a major part of the process, as well as generating the reports to explain what the data says. Lari Masten, who heads valuation advisory firm Masten Valuation, said that working the technology into her own processes has saved countless hours, turning tasks that were once an interminable slog into relatively quick jobs that can be completed in minutes.
One of the biggest use cases in the valuation world, said Masten, is data entry and processing. Valuation tends to require a lot of data inputs that can take hours or even days to complete, but recent AI advances have allowed her to sort through literally thousands of documents to find data patterns. This not only aids her own insights but also thwarts those who want to hide those insights in massive piles of unstructured data.
“Recognizing patterns [means] you can dump a ton of PDFs into a model and it can quickly summarize what is going on and put things in order a lot of times. I do a ton of litigation, and it’s so helpful there because for valuation purposes generally it’s not somebody who wants somebody looking at all their financial records and it comes in a big old pile, and so it helps organize it — ‘Here’s 4,000 pages; organize it by date and by financial record first and then underlying quantitative data second, etc.,'” said Masten.
Her firm also uses AI for coding support when creating new macros, some of them quite complex and developed for specific clients for evaluation purposes. Instead of having to constantly test and retest an ineffective macro, professionals now can describe what it is they need to do and why, and the AI can provide assistance.
“We’re not having to go out to Microsoft or Google and ask how to do a particular task. We can just go to one place and not be distracted by it,” she said.
But while these might seem like purely quantitative processes, Masten said there are uniquely qualitative elements that make it difficult to imagine AI ever completely taking her place, saying professional valuation is both a science and an art. While valuation involves a lot of calculations, what exactly is calculated, and how, can come down to the holistic judgment of the professional. This has been the case, she said, even before AI came onto the scene.
(See our feature story, “
“For decades, valuation has had software out there. You can dump a bunch of numbers in there and churn, but it doesn’t mean that it gives you a good output. I don’t have confidence that if I put in a bunch of information into a tool that it’s going to be able to understand how Lari Masten would reason through and solve that problem. It can do the math, but did it do the math in the way that when I do valuations? When anybody does a valuation, there’s three or four dozen decisions that they make and each decision gets them to a different place,” she said.
There are reasons she does the things she does that a computer can’t necessarily understand at this point in time. While the calculations can be automated, according to Masten, the thinking behind them cannot, particularly where it concerns the ultimate conclusion of a valuation engagement. She has yet to find an AI model that can do that.
“They’ll say, ‘Here’s three ways to solve it’ but they don’t understand if this one is more reasonable than that one. If you work on it you can automate a lot of portions … but the bigger picture of, what is the problem I’m having to solve? What standard of value do I have to follow based on that problem? How is the valuation date going to make a difference? What was known or knowable on a date? AI is not going to be able to really sift through that depending on what the inputs have been already. It’ll just pull information and it may not know when something became known or knowable, so there’s that professional judgment. The good reasoning, the backbone really behind any valuation, can’t be automated,” she said.
Overall, she is supportive of AI and believes it will be of great benefit to the profession, but noted that there are some cons, particularly the need to vet information and not automatically trust what the AI says. Further, she expressed concern that even though AI cannot replace a valuation expert, professionals over time might lose some of those inherent human qualities that prevent this from happening now.
(Read more: “
“There’s some measure of responsibility on the valuation people to go in and make sure that they’re vetting that information, that they’re still applying their logic, overlaying their skills, knowledge, expertise, training, that kind of stuff because [no matter] how great it is as a tool, it can’t use that logic that we have, it’s not a replacement for the interactive qualitative piece that the valuation analyst knows and tie that necessarily to the quantitative art that it does,” she said.
This ties into communicating to clients the value of a human valuation expert: Yes, a computer can crunch the numbers, but that’s not all there is to an engagement, and not even necessarily why someone might hire a human professional in the first place.
“It’s not [just] how to do valuation. The value ad is that you understand the problem,” she said.
(See how AI is impacting firm services in
Credit: Source link
