The FP&A Guy

View Original

AI vs. Human Insights in Financial Modeling to Build Accurate Models for Decision-Making

Paul Barnhurst

In the second part of "The Great Debate," expert financial modelers from around the globe discuss controversial and thought-provoking questions about the art and science of financial modeling. The discussion focuses on the enduring relevance of Excel, the role of AI and Power BI in shaping the future of modeling. This episode highlights the nuances, complexities, and future of financial modeling as a discipline.

This episode features Giles Male, co-founder of Full Stack Modeler and an Excel MVP, Will Wardle, an independent financial modeler with over two decades of experience, Carolina Lago, founder of Classic Financial, specializing in FP&A and financial restructuring for SMEs, Hedieh Kianyfard, a project finance expert from Iran, known for her practical solutions to complex modeling challenges, Oz du Soleil, a YouTube star and Excel MVP, renowned for his engaging tutorials and insights, Rishi Sapra, a former accountant turned Power BI advocate, and David Brown, a seasoned financial modeler with 27 years of experience, excelling in blending Excel with Power BI.

Expect to Learn:

  • The role of circular references in financial models.

  • Why named ranges in Excel can either simplify or complicate financial models.

  • The impact of AI on the future of financial modeling.

  • How Power BI complements financial modeling by enhancing data visualization and storytelling.

  • The debate on the most important financial statement and how it informs decision-making.


Here are a few quotes from the episode:

  • "Dynamic arrays offer huge improvements, and we should aspire to use them more effectively as a community." - Giles Male

  • "A three-statement model is the universal language of finance, connecting operations and strategy for decision-making." - Carolina Lago

  • "Power BI complements financial models by adding context, automation, and storytelling, making it a must-learn tool." - Rishi Sapra


This episode of The Great Debate offers a lively exploration of financial modeling. The expert panelists highlight how financial modeling combines art and science, requiring technical mastery, critical thinking, and creativity. The discussions underscored the importance of balancing simplicity with complexity, leveraging technology without losing sight of foundational principles. 


Follow Giles:


Follow Will:


Follow Carolina:


Follow Hedieh :


Follow Oz:


Follow Rishi:


Follow David:

Follow Paul: 


Follow Financial Modeler's Corner  

Newsletter - Subscribe on LinkedIn-https://www.linkedin.com/build-relation/newsletter-follow?entityUrn=7079020077076905984

Sign up for the Advanced Financial Modeler Accreditation Today and receive 15% off by using the special show code ‘Podcast’.

Visit https://bit.ly/3AK5dYo and use the code “Podcast” to save 15% when you register. 

In today’s episode:
[01:23] - Introduction to the Panel
[05:51] - Circular References in Models
[10:14] - The Use of VBA
[14:55] - Dynamic Arrays Debate
[19:57] - Named Ranges in Excel
[24:50] - Will Excel Ever Die?
[33:43] - Can AI Build Models?
[37:56] - Should Modelers Learn Power BI?
[53:28] - Most Important Skill for Modelers
[57:40] - Closing Remarks

Full Show Transcript
[00:01:23] Host: Paul Barnhurst: All right. Welcome out there everyone on LinkedIn and YouTube. We'll get started here in just a minute. We're just coming to the top of the hour. There we go. I can see David happen to join on his phone.


[00:01:36] Guest 7: David Brown: Hi, everybody. Hi, Oz.


[00:01:39] Host: Paul Barnhurst: Alrighty. Well, thank you, everybody, for joining us. Really excited. This is what I call round two of the great debate. We did this a few weeks ago. Several of you may have had the opportunity to attend that one. We had a bunch of people from Australia and the US. This time we had mostly Europe, a little bit of the US and also Africa. So we have a diverse group of expert modelers. And just in the interest of time, where we're going to start is we're going to start doing a quick introduction. How this is going to work today is we have about 13 questions we're going to run through. Each modeler is going to get a minute and a half or a minute 40 to argue their position. And then we'll move on to the next one. We'll share a link for you to vote. There's also a link in the LinkedIn live comments. For each question, there's a separate link to vote for who you thought did the best job arguing their side at the end. We'll show you how that went. So we're going to do a really quick lightning round of introductions 20 to 30s. Why don't we start with the humble MVP who's wearing my hat, Giles?


[00:02:43] Guest 1: Giles Male: That's because I'm trying to get in your good books. Paul. So, yeah. Pleasure to be here. Giles Male. I am co-founder of Full Stack Modeler. Humbled.


[00:02:51] Host: Paul Barnhurst: Like it, Will.


[00:02:53] Guest 2: Will Wardle: I'm Will Wardle. I am an independent financial modeler. I've been doing financial modeling for over 20 years. Cut my teeth in KPMG. I was there for five years and now I'm just an Excel geek, having used it since I was 15 years old. It's like Excel version 1.52, I think.


[00:03:13] Host: Paul Barnhurst: All right. Carolina.


[00:03:15] Guest 3: Carolina Lago: Hi I'm Carolina. I am from Brazil, originally. I've been working with financial modeling and FP&A for the last 20 years, I guess. And now I run Classic Financial, where I help small and medium companies to restructure their finances and their financial modeling as well.


[00:03:36] Host: Paul Barnhurst: Great. Thank you. Hedieh.


[00:03:38] Guest 4: Hedieh Kianyfard: Hi. My name is Hedieh. I'm originally from Iran. I made financial modeling, my profession and my passion and I'm very happy to be here today.


[00:03:49] Host: Paul Barnhurst: Great. Oz.


[00:03:50] Guest 5: Oz du Soleil: Oz here from Las Vegas and my living room. I'm an Excel MVP for ten years and I run the YouTube channel Excel on Fire, and I collect Kaleidoscopes and there's maybe a quarter of my collection behind me.


[00:04:09] Host: Paul Barnhurst: Love it. All right. Rishi.


[00:04:12] Guest 6: Rishi Sapra: Rishi Sapra from London. I've lived here my whole life. I'm not actually an Excel geek anymore. Or a financial modeler. I'm a former Excel addict, but I've moved on to the harder drugs now of power BI and fabric, a data platform. So my background is as an accountant, and I spent the first decade of my career working for the big four of the big banks in London. Then I moved into business intelligence. I worked for a company called Avanade, which is basically a joint venture between Accenture and Microsoft. So I still love engaging with the finance community, and it's been great to see how much Excel has moved on. So I thought I'd come in and share the love with them with the Excel community.


[00:04:48] Host: Paul Barnhurst: And not sure I'd brag about moving on to harder drugs, but that's a separate discussion. Nice. All right. David Brown.


[00:04:55] Guest 7: David Brown: Hello, I'm David, I'm David Brown. Funnily enough, there's another David Brown who's a financial modeler, the other one from Nigeria. And I've been doing Excel. Excel has been doing me for the last 27 years. 


[00:05:06] Host: Paul Barnhurst: You're breaking up a little bit on us.


[00:05:08] Guest 7: David Brown: So I've been working with Excel for the last 27 years. I'm a financial modeler. Randy Brown Consulting and I work with power BI and Excel, so I kind of wrangle the two together.


[00:05:20] Host: Paul Barnhurst: All right. Great. So before we get started, one more thing. If you're out there, please let us know where you're coming from. We love participation in these events, so feel free to comment as much as you want on LinkedIn. We'll see those come through and just have fun with this. We're going to go ahead and get our first two guests up here. So I'm going to remove everybody else from the screen real quick and we'll get started. I should have kept myself on the stage. I removed me for a minute that I get. All right. So we have two guests here with us. Our first question and I'm just going to throw it up so everybody can see it in circular references in financial models, yes or no? I think we'll start on this one. Give you the opportunity to go first. I know she loves this question so much. So tell us why we shouldn't use circular references. Why is it a no for you?


[00:06:14] Guest 4: Hedieh Kianyfard: Well, I would say, you know, the question of a circular reference to be or not to be is like asking the question, or are you for or against cancer? And why is that? Because anybody who has ever worked on a spreadsheet that is iterative knows that it makes, you know, the spreadsheet unstable and very slow to run and all sorts of problems. And in the world where I am, which is project finance, it's just impossible to work with the iterative model. So I would say the better question to ask is how to deal with the circular reference. Not whether it should exist or not. And the conventional ways to deal with circular reference is copy and paste, which is just to, you know, freeze the problem. And this method of copy and paste to me is again makes an iterative process because you have to copy and paste. So and then as the model gets bigger and this is one of the main problems in project finance today is that these copy and paste becomes like blocks. And again the model, the same problems that we have with circular reference with an iterative model which is slow to run, it comes with the copy and paste method. So like any other issues we have in our personal life as well, when we are stuck in a loop, we need to get out of it. You know, either we find simple solution to get out of a loop, or we just go deeper and try to sort it out. And to me, and my challenge as a financial modeler has been to find a solution to really deal with the issue of circular reference, not to just, you know, deal with it with just a simple copy and paste. To me, that's a temporary solution, and there are better ways to deal with it. And one of them is to use this beautiful option that we have in Excel, which is VBA. Go behind and, you know, create a code, create a user defined function, and create a parallel model next to our model that can sort out the issue. And so that's my opinion about this issue.


[00:08:28] Host: Paul Barnhurst: All right great. We'll go to David. Now give us the other argument here.


[00:08:34] Guest 7: David Brown: So circular reference means that there is a circular in your model. It's a natural thing and there's a natural solution to it. It's just a simple tick. And we solve the problem. Now what that does is also enables us or forces us to document. All we need to do is have a place where we document and say, hey, there's a circular reference here and we've solved it. So when you have that documented, any other thing that's wrong with the model that caused by circular, you actually see why it is you see you see it clearly documented. So I think the problem people have with circular is they forget that there is a circular. It's not documented. So once it's documented it's a clean solution. It's pretty simple. VBA, well, I have my own misgivings about VBA. VBA is a no no. Most people even switch it off. So if your VBA is off as a company, you can't solve your circular because you don't use VBA. So I think circular references, yes, you can have them, but you just document and it's pretty good.


[00:09:36] Host: Paul Barnhurst: All right. Thank you David, and thank you. Hey, you can both vote for who you thought did a better job. So I'm going to put it up here. This is the link. You can find it on LinkedIn. Unfortunately I couldn't do a hyperlink here so you could just click on it. But if you go to LinkedIn in the comments for this event, you'll see every single question. You can go vote for who you thought did a better job arguing for or against circular references. I think this is one everybody has a strong opinion on, and there are. I've seen a lot of debate on this one. So thank you, both of you. We're going to bring our next two guests on for question number two. This is VBA. So here we go. All right. We're going to bring back David. He gets to argue one more time. And we're going to bring on Rishi here. So question number two that we have is use of VBA in financial models. So Rishi our wretched our resident power BI person I want to say financial modeler, but not not anymore. Our resident BI person who's going to be arguing the guest Yes, that we should be using VBA for models. So, Ricky, you're on the call. You got a minute? 40. Go ahead and tell us why our financial models should have VBA.


[00:10:58] Guest 6: Rishi Sapra: Well, I think as a financial model, before you want to learn VBA because it's going you're going to feel like the most powerful magician in the world in terms of being able to control over your own productivity with VBA as a financial modeler, you enjoy this experience of manipulating grids of numbers to achieve certain calculations and certain outputs. And because you have the Excel object model directly in VBA, that's exactly what you're able to use it to automate. And if you really want to experience the productivity gain you get from learning VBA, you could use it to leverage the micro automations that Eric has come up with, where you can literally just go in, put in your little functions in Excel, and create your own software application to work exactly how you want. And that's the power of VBA. The second reason is that the alternatives to VBA, things like Python and other scripts, they're just not as good. I have no idea why Microsoft decided to put Python into Excel cells, rather than have it as a version that talks directly to the Excel object model in a scripting environment, like VBA editor or something better.


[00:12:01] Guest 6: Rishi Sapra: And you know, I do feel bad that the VB editor hasn't had an upgrade in the last 20 years. It's probably the only component of Microsoft I've seen that hasn't had copilot integrated into it, so it hasn't had much love, but it should have loved it. Once Microsoft realized that, and maybe at some point in the future they will, they'll get their heads screwed on, and then it will be a million times better than Python in cells or F scripts or Taskbars. And the last reason is what we were talking about the other day. Paul, as a consultant, as a financial modeler, the key to success is niching down right and being one of those few people with unique skills that are highly in demand and financial modelers with VBA, that's exactly what they are. You could do fewer things with VBA than you could do with Python, but it doesn't have as rich a library, but sometimes less is more, and combining financial modeling skills with VBA gives you that edge that most people just don't have.


[00:12:54] Host: Paul Barnhurst: All right. Thank you Rishi. Appreciate that. To our audience. Please provide comments throughout. We're going to give David his opportunity to respond, but keep the comments going. The guests that are in the background, feel free to comment. Let's have fun with this. And David, your counterargument. Time for you to tell us why we should not be using VBA.


[00:13:13] Guest 7: David Brown: So first of all, Excel is king. You have to understand that VBA is like another. It's like a software that kind of sits on top of Excel. Why not use Excel directly? If you can use Excel directly to solve your problems, then you should. And I can tell you 99.99999% of the time. Excel can do it perfectly on its own. You don't need VBA. And guess what? Why? Why is Microsoft not doing anything with VBA for 20 years? Because frankly speaking, there are other things, the other tools that VBA can do that actually you don't need VBA and Excel has gone. I don't know if you've heard of dynamic arrays in the audience. Dynamic array is lambdas. They're stuff that VBA could only VBA could do. But guess what? Excel can do it now. So you stick with Excel. Stick with Excel. That's number one. Stick with Excel. Because even the investments in VBA is not happening, because we know there are other solutions that are better in native Excel. And then VBA gets switched off by most companies. Why? Because it can take viruses. You're not too sure. And guess what? Modelers were not building models for ourselves. We're building it for our companies. We're building it for people that don't know excel as well as we do. So they don't know Excel, forget VBA, they don't even know Excel. So you need to build the stuff that people can actually open. And the risk is people can open your models full of VBA. VBA is not enabled and your model doesn't work. Models not working is a no no. So VBA is also a no no.


[00:14:38] Host: Paul Barnhurst: All right thank you David. We'll let everybody vote which argument they think they like bath where they stand on this one. So go to the link there and we're going to bring our next two guests on here to keep this debate going. So thank you both of you. All right. We got question number three. This one we have Carolina coming on the stage. And we have ah where is he. There he is the guy known as Giles. He's trying to get some brownie points from me from wearing that. We'll see how it works.


[00:15:14] Guest 3: Carolina Lago: It's not gonna work.


[00:15:16] Host: Paul Barnhurst: See, there you go. You've already been told it's not going to work. So question number three. Should we build fully dynamic models using dynamic arrays so they fully adjust in Excel? This is the debate I've had with several people. Question. We ask our guests now because it's becoming a bigger issue. And what we're going to do is I think it's ladies first. Right. So we're going to let Carolina go. She's going to be arguing no, we shouldn't be building fully dynamic remodels. And after Giles will give us the counterargument. So, Carolina, you're on the clock. Go ahead.


[00:15:51] Guest 3: Carolina Lago: Okay, so what I think about the dynamic arrays, it's awesome that Excel can do it. And it's very nice to see it. Just like the circle reference. I have the same opinion about it, but it brings up an extra layer of complexity to financial models. It's fine. It works great, but fully dynamic array financial models. I don't believe it will work because it's just too much for just one single simple model that could be simpler. And I am totally for simplicity all the time. I try not to put any kind of complexity that it's not necessary. It will bring auditing challenges. It's much harder to audit a financial model that is built with a full dynamic array itself, and especially if it is fully dynamic array, then it's almost impossible. It's very hard to be compatible with all types of Excel. So not everybody has the type of Excel that will support dynamic array. Not yet. I'm not going to say that in the future it's not going to be more fully used, more widely used. But for now, I think it brings a complexity that it's probably not necessary. So I will vote for no fully dynamic array financial models.


[00:17:21] Host: Paul Barnhurst: All right. Thank you for your argument. And I saw a comment come in. It was Jack Kennedy asked in the voting, are we voting for our opinion or who we thought made the best argument? Go with who you thought made the best argument. Hence, it's a debate. That's kind of the idea here. Not just voting for what you think for the actual live vote, but what you think in the comments. Totally fine with that. But for the voting, go with what you like best. All right, now we're going to hear the counterargument of why we should do this. So Giles, you're on the clock.


[00:17:51] Guest 1: Giles Male: Awesome. Well, firstly, Carolina and I were joking before that. We pretty much agree with everything. So this is just a strong debate position. Look, if the question said specifically, should all financial modelers be forced to convert to full dynamic array modeling today, my answer would be categorically no, because we're not ready for it. So I think you have to answer this a little bit aspirationally and a little bit technically. So technically today, dynamic arrays offer huge improvements to modelers. The fact you can't go into period five in your array output and tweak the number to make your balance sheet balanced is a huge step in the direction of best practice, modeling the fact that dynamic arrays deal with data more effectively, more dynamically, more dynamically is a huge improvement for modelers. But we have one big challenge, which is the circularity that dynamic arrays create when you're dealing with three statement models and balances. We have the best and brightest minds looking at this challenge. We will solve it or Microsoft will evolve the tool so that it's solved for us. So you have to answer this question aspirationally. And I think the answer has to be yes. We should be looking at this positively as a community. How do we solve these problems as quickly as possible, so that we are all using dynamic array modeling to the best effect in the future?


[00:19:11] Host: Paul Barnhurst: All right. Thank you. Giles. One thing you're going to notice, I think you'll notice with all our guests, obviously these are nuanced. We say yes or no to have a debate here. But you're going to see different comments. Hey, is it the future? Is it aspirational? Everybody's got to take their different approach to argue their side. But thank you both of you. I'm just real quick while I have you, we'll see what a couple people said. I love what you said. The best mantra is simplicity. So there's one for Carolina And another one that really liked simplicity. Yes. Blessed simplicity. Say that again. So I'll say it. Keep it simple. But dynamic arrays are wonderful. So you pick which side you want to vote on, because there's a little bit of a dichotomy sometimes, but we're going to go ahead and move to our next guest. So thank you both of you. And we'll go to the next question here. Alright. So next up we're going to have will our self-proclaimed geek from the age of 15. And we get Carolina again on this one. So welcome back Carolina. This is going to be question number four. So this is should modelers use named ranges in models. So here we're going to again let Carolina go first. She's going to be arguing. Yes we should. Will's going to argue the opposite that we shouldn't. So Carolina you're back on the clock. Go ahead and tell us why we should use named ranges.


[00:20:34] Guest 3: Carolina Lago: Yes yes yes. For named ranges I've been using more and more lately because I think it brings a readability to the model. It makes the model more easy to read, easy to understand the formulas, and much easier to audit because it's all laid out there. It's much easier to understand a formula that is named. For example, if I say tax rate sum. Something times tax rate. It's easier to see, to understand. Then something times c five for example. So it's much easier to understand a formula to explain a formula to audit a formula. And you make less errors. And it's easier also when you're using the prompts or anything that you're using that you're going to copy from one sheet to another one like the form control, for example, you can copy and all of the characteristics of the form control will go with it if it's name ranged, Granted, if you put a named range that will go across the full workbook, then you can copy around everywhere and it doesn't mess up your formulas, it doesn't mess up your form controls or anything else that you're using. So it makes it simpler. It makes it simpler and makes it easier to understand and easier to explain as well. And I'm always for simplicity, you're going to see on all my, my, my questions. I mean, I am the person for simplicity. So that's one more thing that will help with that.


[00:22:07] Host: Paul Barnhurst: All right. Thank you. Appreciate the arguments. You get it just in the nick of time. Will, are you ready to tell us why the opposite is true?


[00:22:16] Guest 2: Will Wardle: Well, I think I do actually have to agree with Carolina about this. The simplicity thing. I mean, I am not against using named ranges in a financial model. I use them for constants and for ranges when I'm doing VBA code. So they are truly dynamic. So as you said about simplicity coding, you can actually do a tax free. She can do all of this. Now. In my experience, over the last 20 years, I've had some horrific models which have had lots of named ranges in them. And one of the worst one was they had a telephone directory of all the name ranges in that particular model. So they had it wasn't simple. You actually had to go and you could never trace anything back. Now in the simplistic model, if you've got a range called revenue and cost, you can actually multiply the two together and you can see, oh, that's the revenue range, that's the cost. And then you can work these things out. Now when the model steps up to a much higher level of complexity, you start developing your own cottage industry of maintaining your named ranges. Then they manager isn't up to speed. It is. It hasn't changed in the last 20 years since I've actually used it. Eric in A thing About Robots said to me, he said it's as if Microsoft has actually lost the code for name manager. It is not fit for purpose. Then you've got the problems of named ranges being deleted. You end up with hash refs, and there's lots, I mean, and then and now this is the best one to say. It is hidden named ranges audit nightmare. They can appear out of nowhere. Links to all sorts of different things. I just don't like name ranges because of that one.


[00:24:11] Host: Paul Barnhurst: All right, so go ahead and pick which argument you like better there. I think there were some great things brought up. You know name ranges. Yes. We I think we can all agree whether you're using them right now or not, we'd love to see an update by Microsoft content to named ranges, because who's ever had that file with 10,000 references with all About broken links and things to named references. So I think there's some. Opportunities to improve it. But you guys decide. Use them yes or no. Thank you. Both of us, both of you for joining. We're going to bring our next people up on the stage now. All right. So that was question four. Question five. This is a fun one. Please put your thoughts in the comments. We got Oz and Rishi to battle this one out. It's the will excel ever die debate. We've never had that one before right.


[00:25:04] Guest 5: Oz du Soleil: Many books and keep it fair.


[00:25:07] Host: Paul Barnhurst: Yep. No kidding. You know what? We're going to let you go first Oz, and tell us why Excel is eventually going to die.


[00:25:16] Guest 5: Oz du Soleil: All right. There's two ways that we can think about this obsolescence. No, it won't die from obsolescence. No software is going to come along. And then we put Excel in a paper sack and take it out to the alley for the garbage man. Instead, it is going to die from irrelevance, because right now we have technology run amok and we are staring at the Four Horsemen of the apocalypse. We have engineers who have forgotten that they live in a society with the rest of us. We have slimy data brokers and the wanton desires of venture capitalists and soulless digital marketers. They will usher in the death of humanity. And once society, once regular people who've lost their jobs to AI, who've gotten fed up, who are tired of hearing, well, have you cleared your cookies yet? They are going to take to the streets with table legs and rocks and that ride that you have in the middle of artificial Christmas trees. And that will be the last gasp of human dignity. And when we have no more dignity as a society, mon frère. We're in trouble. We got bigger fish to fry than thinking about Excel. And the last formula that will be written in Excel will be equals. Survival. Open parentheses and the battery will die. And the user will look out across his shack and out the window and see the sun glistening on our shame.


[00:27:00] Host: Paul Barnhurst: Alright. Thank you Oz. Go ahead and put it in the comments. What did you think of Oz's argument there? We're going to bring Rishi up now and give him his counterargument.


[00:27:10] Guest 6: Rishi Sapra: So here's a little thought experiment that I want everyone here to do with me. What would an immortal software product actually look like? So think about this. There's an immortal software product would be software that a significant proportion of people, not just thousands, but millions, hundreds of millions or even billions of people use and love on a daily basis. It would be that kind of software product that's ingrained into society, ingrained into communities like sports, like professional sports. People would compete using the software products in the entertainment capital of the world, Las Vegas, where it was his job to win cash battles and fierce competitions that are aired on ESPN. It'd be a software product that allows so many types of people to express themselves in different ways, from marketers and a fortune 500 company to heads of households trying to work out if they have enough cash at the end of each month to cover their kids school fees. And because of all these different use cases, it would be a software product that's accessible to everyone in an environment that gives them the opportunity to be creative without being constrained by complex UI or fixed set of features that a software vendor has decided to implement, and one that people could customize themselves. They could build their own automations in and customize exactly how they want. It would be a software product that would generate the world's largest software firm billions of dollars of revenue a year, so that it's got the funding to continuously innovate and be the best in class. And it would be a software product that is so entrenched in corporate workflows that it would just be a part of almost every process that everyone in the entire organization is involved in, and one that the entire banking system and even the entire world runs on it. That would be the definition of an immortal software product, and that is Excel today.


[00:28:54] Host: Paul Barnhurst: Perfect. Right on time. Thank you Rishi. Thank you Oz. Go ahead and vote for who you think won that argument there. Loved it. Some great points. Everybody should learn the equals survival formula. Try it in Excel today. Let us know how it works. All right. Let's bring up our next guest. So this is the surprise question that they. I'm going to throw out here. So Carolina gets to come on stage for a little surprise. She doesn't know what the question is, but I bet she might have some ideas. So tell our audience you get one minute here. Why? Understanding three statement models is the most important skill for FP&A professionals.


[00:29:39] Guest 3: Carolina Lago: Oh, that's a good one. Well, I think it's the most important skill that a finance professional can have. Because when you start understanding the three statement model, first of all, the three statement is the language of communication in finance. You have to agree with that every time you're going to report numbers, either on the stock market or to compare companies or to buy companies, you're using the three statements. So it's the language of communication in finance. Why it's so important to understand the three statement model. Because as you are working with the three statement model over and over and over again, you start to build a model in your head and that will help you understand what every business decision, how every business decision flows through your finances as it flows through your finances. It's going to show in the three statements. So if you are very familiar with the three statements, we are going to be able to understand how each business decision will impact the business, in the finances, in the numbers. So that's why I always tell all my students and my team and everybody that works with me, build a financial statement, a financial model in your head so you are able to take decisions quickly and understand how any, any business decision will impact in the business. That's fine.


[00:31:07] Host: Paul Barnhurst: Thank you. And arguing no on this side is going to be the FP&A guy. So I'm gonna put on my hat here for my argument. So I'm different from the host. And first I'll say I'm the FP&A guy and I say no. So you should vote for me. The real reason I say that is in 12 years in FP&A and corporate fact, I never once built a three statement model. I built unless I understood the basics, but I don't think it's most important. It's important. I think people should understand it, but I would much rather have someone who's a critical thinker, strong Excel skills, communication skills. I'd probably put it fifth or sixth on the list of skills I want when I'm hiring somebody. I can't remember ever hiring someone and giving them a test to see if they can build a three statement model. And I had some great employees, so I think it's important. Do I think it's the most important? I don't, but you guys get to decide. That's the beauty of this. So go ahead and vote for whose argument you like best. And thank you for coming up and humoring me on the spot there, Caroline. I had to have a little fun. We got one more surprise question a little later. That will be for Giles. So you'll get to enjoy him being on the spot. All right. So keep the comments coming. We have a lot of them coming in. Let's see what we have.


[00:32:28] Host: Paul Barnhurst: FP&A guy here and as you know, I am very passionate about financial modeling and the Financial Modeling Institute's mission. I have been a huge fan of the FMI for years, and I was super excited when they decided to sponsor the Financial Modelers Corner. I recently completed the Advanced Financial Modeler certification and love the entire experience. It was top notch from start to finish. I am a better modeler today for having completed the certification. I strongly believe every modeler needs to demonstrate they are a qualified financial modeler, and one of the best ways to do that is through the FMI's program. Earning the accreditation will demonstrate to your current and future employers that you are serious about financial modeling. What are you waiting for? Visit www.fminstitute.com/podcast and use Code Podcast to save 15% when you enroll in an accreditation today.


[00:33:35] Host: Paul Barnhurst: Yeah, we got some about I going work in P and it's cashed and the causes of cash. So we got a lot of different opinions. We're going to go ahead and go to our next question. So now we're going to bring up on the stage for question number seven here. This is going to be Will and Hedieh. So we got them both on the stage. Let me pull up the question. And so we have number seven. Will AI eventually build the models. For us. This has been a debate that I think a lot of people are having a lot of works being done, a lot of different opinions here. So we're going to let Will kick us off by telling us why I will not build the models for us.


[00:34:22] Guest 2: Will Wardle: Well as a trained economist, Keynes said, in the long term, we're all dead. Eventually they may do, but at the moment AI is nowhere near capable of developing a financial model. I mean, for me, financial modeling is an art as much as it's a science. You can put 2010 financial models in a room together and you may get 20 completely different answers. The art in financial modeling is that you will find one of those 20 answers. That is the best for that particular situation that you're actually dealing with. So what I'm trying to say is financial modeling. Well, AI in financial modeling, it is a large language model. There's no intelligence in that. There's no spark. There's no creativity. There is Nothing that. I mean, the whole concept of light is the fact that we can actually think independently. Large language models are just regression models of a various different sort. So what I'm actually trying to say is you have the ability to have inspiration, the thinking and creativity when you're building financial models, which a computer doesn't have to do. Now, I'm not saying that Excel has or AI hasn't got the ability to do elements of automation to help you actually do things, or with the analysis of data or the analysis of outputs. But in the ability to in the creation of a financial model, I do not think that I can ever replicate the ability of a human being with this high seconds amazingness that we can actually do.


[00:36:13] Host: Paul Barnhurst: All right. Great. We're now going to hear the counterargument from Hedieh. We're going to get her up on stage here and let her tell us why is AI going to build the models for us?


[00:36:26] Guest 4: Hedieh Kianyfard: I would definitely think it will. I'm not saying that it will be a success, but I already see the trend, you know? I see now most companies, the way they are dealing with the financial modeling is that they have separated the work of building the model and doing the analysis. They have a bunch of people sitting in an office. You know, building the model. These people, they have no idea about the deal. They are completely isolated from the transaction. They don't participate in the project calls. They don't go to site visits. And so these are the people who are building the model. And then some other people are using it. And this has caused us to deal with models that are not overly complex. And I deal with them on a daily basis. So I think what's going to happen is that AI is going to replace those spreadsheet engineers, and I hope that AI will be more intelligent and can simplify the task, and not so that we don't have this complexity in our models what we see today.


[00:37:28] Host: Paul Barnhurst: All right. Thank you. Appreciate that. Thank you both for arguing that you can go ahead and vote for who you thought the winner was. We're going to keep going as we have a number of questions to go through. I love what Davidson said to that one real quick. Wow, that is a tough question. To be honest, I don't know how to answer it. So I think, yeah, there's definitely some people struggling with it. Right? It's an art and a science. But go ahead and vote for which one you like best. And we'll go to the next question. All right. So we're on question number eight. Now this one is should financial modelers learn power BI. And we have two people coming up on stage. I think they've had to argue against each other once before. So we're going to give them a second round of going at each other. I think this time since last time Oz got to go first. We'll let Rishi go first here this time. And it should. Financial modelers learn power BI. Rishi is going to be arguing. Yes. So you're on the clock. Go ahead and tell us why.


[00:38:30] Guest 6: Rishi Sapra: Thank you. Well, so yes, I think financial modelers should learn power BI not as a replacement for financial models, but having both together. That duo of financial models and power BI reports are the future of data decision making because they feel different types of data decisions. Power BI is so much more than data visualization. There's three things I think it gives you on top of the top of your data. The first is context. And if I could sum up Will's arguments of why I won't build financial models with one word, it's context. It's because AI doesn't have enough context to turn your spreadsheet of numbers into a financial model, and that's what you get with power BI. It's a semantic model. It's a structure that captures relationships to your data set financial logic with Dax and even row level access permissions. And it's that context which gives your data meaning. It's actually a cube, a SQL Server Analysis Services cube, which is ideal for calculating ratios, performing variance analysis, and understanding what's driving your numbers. And this intelligence applied on top of your data is what both AI and humans can use to make those data driven decisions. The second is automation. So hopefully everyone here knows Power Query. You can use it to clean transform shape data with the click Unpivoting split text from numbers, merge datasets, merge data sets together all in this transparent, repeatable workflow. You combine that with Power Automate and with Microsoft Fabric, and you have a really powerful data pipeline where you can integrate Python notebooks and AI for data validation or anomaly detection. And yes, you have Power Query in Excel, but pulling it into power BI and into fabric, it makes that automation so much more robust and so much more supportable. And the third thing is storytelling. So finance data isn't about the numbers, it's about the insights. So this is interactive features. Cross filtering drill through from a high level member of the income statement to underline journals. All these kind of things through to copilot to answer natural language questions. That's powered by AI. Thank you.


[00:40:28] Host: Paul Barnhurst: Now we're going to hear the counterargument of the no. So Oz give us the no.


[00:40:36] Guest 5: Oz du Soleil: Power by interactive dashboards. Data visualization is sexy. It lures people in. They go around power by licking, licking, licking like flies in a Venus fly trap, licking that sweet nectar until it just snaps them in because they've fallen for the puffery. Because when have you seen power BI used in the financial modeling competitions? You've not. Because you have to get to the hard core stuff, not puffery. Don't get caught in the forks of a Venus flytrap.


[00:41:23] Host: Paul Barnhurst: All right, so Paul is telling us, don't get caught in that Venus fly trap. Thank you. Both of you love the arguments. We're going to give you that chance to vote. Remember, you can also find that on LinkedIn in the comments for this event. Vote for who you like there. Let us know in the comments. We're going to move on to question number nine here. So we're going to bring two more people up on the stage this time. We're going to welcome back Carolina and Will for this argument. So our question is are financial models the number one corporate decision making tool. We have Carolina arguing yes. Will arguing no. So we're going to give the floor to Carolina first and go ahead. You're on the clock.


[00:42:17] Guest 3: Carolina Lago: Okay, so this pretty much ties with what I just said before about this financial modeling, especially the three statement modeling being the number one skill that a finance professional should have a financial model. And now I'm not just talking about the three statement. And actually I'm not talking about Excel financial modeling. I'm talking about financial modeling in general. This is everywhere in the company. So there are financial models everywhere in any anywhere that you see numbers, you see a financial model behind it. So that's why I think it's the number one decision making tool. And also when you see there is a thing, a thin line between the connects between the operations and the strategies, and there is a financial modeling connecting it. So it's very important to understand Explain what is this model that will connect operations with strategy. So every decision needs to tell operations what to do in order to achieve strategy. So that's why the financial model would be the first tool that people will take for decision making.


[00:43:32] Host: Paul Barnhurst: All right. Thank you. Appreciate that. We're going to now let Will tell us why it is not so will go ahead. Let us know why it is not.


[00:43:43] Guest 2: Will Wardle: This is a very difficult argument to hold against. I mean if you take a financial model in its purest sense, then financial models exists within organizations. If you are taking a pure financial model, as in a three statement financial model with that's fully integrated, then no, it is not the first go to thing. There are so many more areas that you can actually look at in some sort of business, you'd be looking at your. As Paul is not a guy, you'd be looking at your budget, your actuals. You'd be actually doing your variance analysis to see where you're actually are. Now, that is much more important than a financial model that's looking 4 or 5 years ahead. That is why it's not really the go to model. But there's also other models that exist in the business. There are metrics about sales. There are metrics about customer satisfaction. You do not, I mean, if you go into banking, you've got the whole which I've worked in product. You will have this whole subclass of models that look at your balance sheets. So how but then again, would you classify those as a financial model in what we're trying to look at. So it's the semantics of what is a financial model. A financial model can actually be a spreadsheet with two numbers that adds things together that. But then that's. We lead on to a whole different subject.


[00:45:23] Host: Paul Barnhurst: Sure, we could start to touch me in what qualifies as a model.


[00:45:28] Guest 2: Will Wardle: Yeah, exactly.


[00:45:30] Host: Paul Barnhurst: Thank you. Will. My favorite answer on the show so far is I asked someone. Our corporate financial model is the number one decision making tool. The response I got was no, it's politics. And I was okay. That's hard to argue with. So that would be my counterargument if I was one of these here. But go ahead and vote for these two. Which one you like best? I think they enjoyed that one. And you know what? Funny enough, we're getting comments that we need to bring Giles up. He's being asked for. He's humbly listening. Let's bring on Giles. So guess who's next. Where is the most humble MVP? So we're going to meet the demands of our audience here. The crowd. We're going to bring Giles on for a surprise question that he has not had time to prepare for. So, Giles, there's nobody arguing the opposite side of this. You're on your own here. You get one minute, not a minute. 45. Please explain to the audience why you are the self-proclaimed most humble MVP in the world.


[00:46:30] Guest 1: Giles Male: I was born this way, Paul. I mean, this is a mix of perfect genes and a lifetime of great decision making. I carry this burden with me every day. You shouldn't really need to ask. I am just the humblest MVP. The next time you have a bad day, whether it's at work or with the kids, just remember the burden that I carry around remaining this humble every day. When I look in the mirror and I see this staring back at me, there is no counter-argument. But thank you very much.


[00:46:59] Host: Paul Barnhurst: Thank you. We appreciate that argument, Giles, and we will bring him back on again to make a real argument. But he did a good job there. Loved Love the coat. Love the glasses. All right. Next we're going to bring back on Oz for our next question here. So we're going to kick Giles off the stage. He can bring back his hat and try to score some points. You know funny enough this question is should financial modelers learn Power Query of everybody that answered for this session? We only had one. No. And he wasn't able to make it. It was Chris Riley. So feel free to comment on one of his things. Dm him and let him know that he's out on a wilderness. Everybody else thinks we should learn Power Query, and Oz gets a minute here to tell us why we should learn Power Query as a modeler. So Oz, take it away.


[00:47:49] Guest 5: Oz du Soleil: Imagine the olden days of being required to use VBA to bring data in from multiple folders, multiple sources. Those were the bad old days when people had a life expectancy of 35 and they were dying in mine collapses. If you aren't using Power Query, that's where you are. You're out there in the wilderness with this fellow. You don't want to be there because Power Query makes this easy. Beginners can use Power Query. If all you need to do is split a column, you can do that. You don't have to be bones in the bottom of a mineshaft because you are hard headed.


[00:48:40] Host: Paul Barnhurst: All right, love it. Thank you. Oz. So don't be bones in that mineshaft. All right, we're gonna move on to the next one here. We just have two questions left, and then we'll quickly show. I do not believe there is a link for question 11. Jack. I see that coming through. So we gave him the win. There was no need to vote. There is the opportunity to vote on Giles, though, so you'll see what's there. Right. We're going to move on to question 12. See if I can keep this all straight. Which financial statement is most important for a modeler to understand. So we're going to bring on our humble MVP. And we're going to bring on Hedieh to argue this one. We had one person who voted for PNL. They couldn't make it. So it got kicked out PNL. Who cares. It's not important because what we're saying. So now Giles is going to tell us why it's all about the balance sheet. So Giles, you're on the clock.


[00:49:42] Guest 1: Giles Male: So I think you can answer this in a couple of ways. I would look at it one from like a training and skills perspective. Which of the three statements would it be the biggest problem for a modeler not to understand? I think there's more to the balance sheet than any of the other statements. So to not have a modeler comfortable with the technicalities and the principles would be a massive problem for the balance sheet in particular. But then I think when you look at the three statements, look not having the three statements at all is a problem. I think a lot of people will look at the cash flow and yes, cash is king and you have to understand your cash generation over multiple years. I totally get it. But the question I ask myself is, is it possible for something fundamental about the health of a business to be hidden from your view? If you just looked at the cash flow statement or the income statement, and I think the answer is yes, is it possible with the balance sheet, I think it's as close to no as possible. And the real kicker for me is then if you assume that you're getting multiple years worth of numbers and you've got your balance sheet in front of you, you can actually sort of back calculate quite a lot of the information that you want to get from the bottom of the cash flow statement from the bottom of the income statement. You can't get the detail, you need the income statement and the balance sheet to do that. But you can reverse engineer quite a lot just from looking at a balance sheet over multiple years. You cannot do that with a cash flow on its own or with an income statement on its own. So I think it's a tight call. But for me, I was looking at investment. I would choose the balance sheet over multiple years.


[00:51:16] Host: Paul Barnhurst: All right. Thank you. Did that just in the nick of time. We will now hear the counterargument or not counterargument. We'll hear why cash flow is most important. So tell us about the cash flow.


[00:51:29] Guest 4: Hedieh Kianyfard: Yes judge. You have already lost the debate. You already declared that cash is king. And I would add the financial model is the Queen. And in my world, which is project finance. I'm so sorry, Giles, but we don't even care about your balance sheet. We just feel the balance sheet because it helps us to make sure that the model mechanics is correct. But I mean, I would even argue that, you know, what we have in the project finance, which is the cash flow waterfall, is even, you know, the most important. It's the most important statement in a project finance. But I would even, you know, go further. And when I work, even in corporate modeling, I always also ask them to build that statement because it's such a beautiful statement. It tells exactly who gets what and at what time. So it tells you the priority of each cash flow statement. It tells you it reveals to you all the risks of the project. It tells you if things go wrong, what can happen, who can get what. So I think there is no argument against cash flow being the most important statement.


[00:52:37] Host: Paul Barnhurst: All right. So there you have it. We've had two people share their thoughts. Let us know. Nobody wanted to argue for the PNL. So it obviously it's not going to win. It comes down to that cash statement and cash being queen or you know, balance sheet is, as Giles argued, is that really the most important one? So let us know. Give us your opinions. Let's see what we have in the comments. Here we go. This is Will must be a consultant. It depends on the business. Come on. It depends. No, I'm just kidding. We all know there's nuanced politics. All right. Last question, and then we'll be wrapping up. So we're going to bring few people up on the stage for this last question. We have David Brown. We have Giles and we have Oz. So we got three people. And this last question we're going to give you each about a minute and a half here, because we're running close to the end of our time. So it can be a little shorter here. But the question is which is the most important skill for financial modelers? So they had a list of ones to choose from and from that list David Brown chose critical thinking. So we're going to put him on the clock to tell us why critical thinking is the most important skill. Take it away, David.


[00:53:56] Guest 7: David Brown: Yeah. So, financial modeling is a multidisciplinary. I mean, we need excel. We need financial modeling. I mean, we need accounting, finance, quite a lot of stuff together. And if we check all universities in the world, you will not see financial modeling. Why? Because it needs so much critical thinking to be an excellent modeler. And critical thinking too, is not really in most universities' curriculum. Right. But critical thinking is extremely important because you need to take those assumptions and need to make sense of it. I is going to do the mechanics for us, but critical thinking is what the humans are really going to do very well. So critical thinking is number one. If you go and check our recruitment and see what's the number one skill most people need from recruiting new people, it's critical thinking. And I think even to be the best modeler ever, you want to be the best in Excel. You need a lot of critical thinking. So critical thinking I think is number one skill. And of course many modeling projects have gone through. It just takes a little bit of, well, not a little bit, but a lot of thinking, a lot of understanding. How does that balance sheet actually how does this line item make sense to the business? How does this particular delay ten days? The delay? What will really happen? You have to critically think through that process. And then you then do the mechanics. So critical thinking comes first, then comes second comes last. Because at the end of the day, that's what brings value. Critical thinking.


[00:55:17] Host: Paul Barnhurst: All right. Thank you. So we're now going to go over to Giles to argue for understanding financial statements. Now Giles don't forget about that cash flow statement. It's important I've heard P and L you can forget about. But go ahead. You're on the clock.


[00:55:33] Guest 1: Giles Male: They're all quite important. So? So, look, this is not an ideal situation at all. We would want to have a modeler that understands financial statements, has world class critical thinking skills and is obviously using Excel. We don't want them using Google Sheets. Let me blow Oz's position out of the water straight away. If Excel disappeared today, we would find an alternative. So you cannot argue that Excel is the most important thing out of these three. So it comes down to critical thinking and understanding financial statements. Now, I agree with everything David said, but I don't think you can have a financial model that doesn't have financial statement financial modeling knowledge. You could have a situation where a modeler has zero critical thinking skills, but if they understand the financial statements, they can be given a specification document. They are told what to do and they have the skills to build it. You can have the best critical thinking skills in the world. If you don't understand financial statements, you are not technically going to be able to do the primary thing you are supposed to do for your role. So it has to be understanding financial statements out of these three.


[00:56:43] Host: Paul Barnhurst: All right. Thank you. Giles, we're now going to get the last word before we wrap up here to Oz. Tell us why it is Excel, that tool that will die.


[00:56:54] Guest 5: Oz du Soleil: I have built a career knowing excel and I don't know, financial statements. I didn't know the business of the marijuana dispensary that I did a project for. I don't know, insurance, I don't know being a chef or running a restaurant. But I have built a career knowing Excel under the guidance of chefs and marijuana shop owners and insurance agents and project managers, all kinds of people. And I'm going to tell you what this kaleidoscope cost like $700. I bought it because I know Excel.


[00:57:36] Host: Paul Barnhurst: All right. Thank you. Oz. We're going to bring everybody back on stage here. And just first I want to thank everybody. You know, the idea here was to have fun. I hope everybody had fun with this. We all know the reality is every one of these is nuanced right. There are some more so than others, but you can make an argument. You could argue it depends. Definitely your role and what you're doing matters. But we're going to run through. I want to say something.


[00:58:08] Guest 5: Oz du Soleil: I want to say something because you said David Brown chose critical thinking. I don't know if he chose it. I know I was assigned every one of my positions.


[00:58:19] Host: Paul Barnhurst: You were. Yes. But you all filled out a form. I could go show you the form. I asked you all questions I based it on. 


[00:58:26] Guest 5: Oz du Soleil: Oh, here's what the guy said, you know. Oh, yeah, but you signed a waiver, right okay. Yeah, right. Paul Barnhurst. Go ahead.


[00:58:32] Host: Paul Barnhurst: Give me a hard time because I'm ready for it. Oh, I forgot to bring Giles on the stage. Sorry, Giles. You were so humble. The background. I just did notice.


[00:58:41] Guest 1: Giles Male: I noticed, trust me.


[00:58:43] Guest 2: Will Wardle: Not that valued.


[00:58:45] Host: Paul Barnhurst: I was keeping them off screen so he would remain humble. All right. 


[00:58:50] Guest 1: Giles Male: Happens naturally. It's fine.


[00:58:51] Host: Paul Barnhurst: So real quick, we're just going to run through and not in any specific order because it's just the way it came up. Fully dynamic arrays are winner based on votes was Giles. So congratulations Giles named ranges are winner. They're based on votes. Carolina. Congratulations. Power by Oz is our winner. Will AI build the models for us? This one was a landslide. We had more votes for this one than any of them. And it was all One Direction. Hedieh. Congratulations. All right. On the most humble. We did get a vote. The options were. Yes, of course. And why is this a question? And the answer we got was of course. So congratulations, Giles.


[00:59:39] Guest 1: Giles Male: I voted. That was me.


[00:59:40] Host: Paul Barnhurst: All right. He's the only one that voted. Surprised he voted for himself. Right. All right, so let's keep going here real quick. Next one we have up is three statement models that ended up as a tie on it being most important for FP&A, and a one vote for me and one vote for Carolina, so it's important. All right, will excel ever die. Oz won the vote there. So someday Excel is going to die. Good job on the financial statements. Which one's most important? Turns out it's cash flow. So, Giles, you're going to have to give it up to our queen happily. All right. Most important skill again. Oz is our winner. They said excel. Congratulations, Oz. All right. I think we have two more left. Should we learn VBA? The vote here was two thirds for Rishi, one third for David on that one. And circular references. We ended up with a tie. I see that look. So like how do we tie what, 50 over 50 there. So thank you everyone for joining us. We're right at the top of the hour. We hope you had fun with our panelists and go forth and learn modeling, Excel, Power Query, all that stuff. Thank you everyone.


[01:01:07] Guest 5: Oz du Soleil: This was fun. Thank you. Thanks for having me here.


[01:01:10] Guest 4: Hedieh Kianyfard: Thanks everyone.


[01:01:11] Host: Paul Barnhurst: Thanks, everyone. Thanks, guys.


[01:01:13] Host: Paul Barnhurst: Financial Modelers Corner was brought to you by the Financial Modeling Institute. This year. I completed the Advanced Financial Modeler certification and it made me a better financial modeler. What are you waiting for? Visit FMI at www.fminstitute.com/podcast and use Code Podcast to save 15% when you enroll in one of the accreditations today.