Financial modeling with GenAI

- [Instructor] Gen AI or generative artificial intelligence is one of the most important technology trends taking place right now. And it can have a major impact on you, in your role in finance and any financial modeling you're doing. So let's talk about at a high level what you can do with Generative AI. Now, there's a few major Gen AI sites that are out there. Certainly we're probably all familiar with ChatGPT. Anthropic is another good one. Google has their own AI tool, and Microsoft's Copilot as well. Now I wanna talk about each of these briefly in turn. So I'm here on ChatGPT, and I've pulled up ChatGPT 5, you know, which is only for professional users in general, and of course, they have the legacy models like 4. But you know, you can see that ChatGPT does do some pretty slick stuff, right? I asked it create an infographic for me, showing jobs lost and jobs created by various technological revolutions throughout the last century or so. And it created this, which I guess is pretty good, right? But there's a few issues here. It's spelled revolution's wrong. It's got kind of a weird thing going on with the H here, but most importantly, we didn't lose any jobs or create any jobs. So I say, hey, you know, you're gonna have to try harder. Revolutions is spelled wrong and I don't need throughout the last century or so in small print compared to the rest of it. And there's no numbers here. So I tell it to do it again, alright? And it says here, right, I missed my last pass, missing the mark. So it created this for me. I guess that's okay, at least it's got some numbers, but I wouldn't say this is amazing. It's a good start though. So clearly there's still some work to be done, at least for ChatGPT in terms of its ability to display data, right? The kind of output we might be interested in following a financial model. It's entirely reasonable that we'd build a financial model, looking at say, AI and what the impacts are expected to be over the next few years. ChatGPT might be able to help us with that, but it's clearly not ready to do it all on its own by any means. So let's turn to Google and we'll ask it for something else. So if you just go to google.com, you can click on AI mode and it'll bring you into AI mode. And I've done that and I asked to create a fictional set of financial statements for Big Bear Honey Inc. Now, this is not a real company, this is a company I imagined. I imagine it's run by bears and they're producing honey or something. Here we go, they have an income statement. Let's just take a look. All right, so revenue 2,500. Fair enough. Cost of good sold, 1200, gross profit is 1300. Okay, that's good. 2,500 minus 1200 is 1300. So it's doing some basic math right. Here's SG&A, it's getting the categories right and it looks reasonable. SG&A, R&D, I guess we're paying the bees or something. So that's 550 for these two. 1300 minus 550 now. Okay, so looks like it's about right. Operating income 750, my interest expense and income tax expense. Then, all right, that's another 230. So we get down to 520. Looks like the right level of net income. And now we've got the balance sheet. Balance sheet, 2550 in total assets, 2550 in total liabilities and equity. It balances, that's good. You know, overall looks reasonable. You know, it's only one year. I kind of wanted a few years. I could go back and ask it to add to that. But you know, looks reasonable. Statement of cash flows. Formulas all look reasonably correct. So, okay. And then I could go through and I could ask it, what are Big Bear Honeys primary business activities, or elaborate on the assumptions behind these financials. Now it's all fictional, so I'm not gonna bother with that. But you can see where, hey, even just for creating something simple and kind of silly like this, we're off to a good start. Now let's take a look at Copilot. So here I am in Excel. There's a web-based version of Copilot, but I've vastly prefer the Excel based version. They have it in all the Microsoft Office suite. It's nice just having it built in here right into the application. One big note, I would say, in order to get Copilot to work, you have to have auto save on. If this is off and you're trying to save it locally, it will not work. This only works when you're saving it to the cloud. But once you do that, you'll have the Copilot option right up here. You can click that, and I gave it the same prompt. Create a fictional set of financial statements for Big Bear Honey. And it goes through and says, yep, we've created it. I'm gonna put that info into this Excel sheet, and we'll see how it does. It says, I can't directly modify the Excel file you have open, but I can generate a version of the file that includes the fictional statements. Well, I've already got that. But you know, it's not bad. I can go ahead and I'll download this. So I've downloaded it and opened it up and we see, you know, it's okay. Doesn't necessarily look as nice graphically as Google's did. And the big downside, it's just hardcoded numbers. It's not formulas, right? It's not exactly what I'm looking for though, right? Probably wanna be able to go through and add in my formulas or have the formulas added for me, and then adjust from there. So I have to do this still manually on my own. Now we've got the balance sheet and we've got the statement of cash flows. Again, it's okay, it's not bad. It's a good start. At least it's in the format that I'm likely to be building financial statements in. But it's not perfect. Nonetheless, that is Generative AI and how it can start to help us when we're building financial models. Now, we're gonna go through the rest of this course and we're gonna be building in Excel. We'll get started on that coming up shortly. But at the end of this course, I'm gonna show you how we can use Generative AI to make ourselves faster and more productive. Gen AI is not ready to do all the financial modeling for us. We can't just pop in what we're looking for and it'll spit everything out. But it can make us faster and more productive and help in many important ways. So we're gonna talk about this later on in the course, and you'll see how we can do this. So stay tuned until the end when we get to that.