Shaping Business Strategy: Ian Bennett's Insights on What Makes for a Great Financial Modeler

I'm sure you have a horror story from your career.

It's really an important conversation everybody wants to talk about at the moment. And as a financial modeller, you run that risk every day. They say best practices need to be followed.

Meet N. Bennett, the proud partner of Deals Modeling at PWC Australia, dedicated to revolutionizing financial modeling practices. Passionate about it and driven by commitment to create quality financial modeling. His approach combines technology, AI and discovering hidden ele of risk behind his career.

I'll go back maybe 10 years to a model which I was introduced to because it was being relied upon by quite a large organization. It was an extremely large model. When you rebuild a model like that, you go back to it with fresh eyes for today's purpose. I think two key things, but both around the purpose of a model.

How we can better manage that whole.

Every model tells a story and that story is known at the start of the project and the questions that the model will be able to answer and the model should be able to answer that Foreign.

Welcome to Financial Modelers Corner where we discuss the art and science of financial modeling with your host, Paul Barnhurst. Financial Modelers Corner is sponsored by Financial Modeling Institute. Welcome to Financial Modeler's Corner.

I am your host Paul Barnhurst. This is a podcast where we talk all about the art and science of financial modeling with distinguished guests from around the globe. The Financial Modelers Corner podcast is brought to you by Financial Modeling Institute. FMI offers the most respected accreditations in financial modeling. I'm thrilled to welcome our guests to the show this week. This week we have here with us on the Financial Modelers Corner in Bennett in welcome to the show.

Hi Paul, it really is great to be here.

Yeah, really excited to have you. So we're going to start off with our first question we ask every guest and I'm guessing you already know what this one is because I know you've listened to some episodes. Tell me about that worst model you've seen. I'm sure you have a horror story from your career.

So there are many. And I'm tempted just to talk about the one I saw yesterday when a client was demonstrating a challenge they were having and it broke every one of the best practices and I had to let them know that that was the case, but I think that was probably well understood by them already. But actually I'll go back maybe 10 years to a model which I was introduced to because it was being relied upon by quite a lot of large organization for valuation and such the like, and it was really struggling and they were really struggling. It was an extremely large model. I think it was 40 to 50 megabytes. Took about 10 minutes to open on a normal laptop. We ran our tools over it. It took over a day to get all of the analysis. It had had many, many owners, and each of them had left their fingerprints, their footprints, their personality in the model. And at that point, it had become a bit of a, I think a Frankenstein, if I've got that metaphor right. Of all the different people that had used it and the different things that had been built for it was relied upon very heavily. But no one person really understood how it worked or could trace through. And trying to give people confidence that it worked was extremely hard. In fact, I think we predicted that the cost of doing a full integrity audit would have been double the cost of rebuilding it from scratch. But there was no appetite or time for that. So we were just help out in any way we could with this beast of a model, and it just lived on. Roll forward eight years, I think it was. We did rebuild that model, but we learned a lot about models, and I think everybody involved in that model learned a lot.

I can imagine that's a long time before the model being rebuilt. I can only imagine the monster that it was. So is there one or two key takeaways from that experience that you've taken with you in your work?

I think the main thing is that models are built for a single primary purpose at the time that they are conceived. They may have some second, but there's a primary reason that model is built and this model. That purpose had evolved and new purposes had been added, and you could see that in the model. And that was its downfall. And of course, when you rebuild a model like that, you go back to it with fresh eyes with today's purpose. And the same model is significantly simpler, is significantly more easy to use. It's also built the way that the organization is currently thinking and the way that the people that are using that model or relying on that model are thinking. So it fits that purpose too. And that's a really important part of financial models and how they go on and be successful in an organization. So I think two key things, but both around the purpose of a model.

That's a great thing. And you know, as you were telling that story, I couldn't help but reminded of probably the most complex model I worked with. It was an FPA model where we had a ton of different products splitting out to two different businesses, all these cost centers, and just a monster of a model because There was a lot of timing things, mix of subscription and transactional. And when we first built built it, we were trying to understand the billing data. The billing data was a complete mess. None of the codes made sense. And so we built this model and it wasn't serving the businesses need. I couldn't hold them accountable. And, you know, we had it for about a year and at least had it working. And I finally said, I have to rebuild this thing. And it was so much better when I was done. But it was such a nightmare trying to untangle it and go through each line and just spend that time to go, what do we really need to be making smart decisions in a business and to use this to really forecast what's important. And yeah, I learned a ton from that one, but it was not fun by any, any stretch. So why don't you go ahead and just tell our audience a little bit about yourself. I know you've had quite the storied career, a lot of modeling and different things. So go ahead and introduce yourself to the audience.

Yeah. Hi everyone. So I live in Sydney with my wife and my two daughters. Been here for 14 years. I was 10 years in London before that. And I'm a professional financial modeler and I have been for 24 years now. I started my career as a financial auditor, was doing audits of companies in Manchester and England. And it very quickly became apparent to me and the people I was working with that that was not the right career for me because I really enjoyed spending my time and I'm going to show my age now. Paul. But there was one laptop in the middle of the audit room and it was primarily there for printing. It connected to a physical printer and there was this thing called Excel on it, and I'd never seen that before. And I realized that I could sort of test working capital with it. And so I built some little calculations and I used that for the next audit that I did, and I expanded it and made it more complex and again for the next one. And other people heard about this and wanted to use it for their audits. And I realized I much more enjoyed building these little tools that people could use to solve their problems than ticking cash or whatever it was I was doing in my audit world. And at around about that time, I heard about this team in London that had just started, was doing financial modeling. I moved down to London and the rest is history. Professionally, I'm really lucky to lead a team here in Australia. Team of about 50 people based around Australia and an amazing team in our Offshore Center, Kolkata. And I am advisor to a number of different areas of financial modeling. And I guess one of my proudest things is that I became a master financial modeler in MFM in the last 12 months with FMI. And that was a really wonderful moment.

Thanks for that background. You know, since you mentioned that, let's talk a little bit about that before we get into the rest of the questions. Can you just tell our audience a little bit of what the MFM is, why you did that, kind of why it was such an accomplishment for you.

So the first two levels I sat and really enjoyed, they're highly technical. The learning process for most is as valuable as sitting the exam itself. And that was a great experience in the financial modeling skills and that greater awareness of what it is to be a good financial modeler. The MFM is a little different. It asks you to prepare a proposal, if you like, of your career, your cv, your history, but most importantly, your contribution to the financial modeling industry. So I was encouraged to prepare that document and a little humbled to be accepted.

That's great. Now, congratulations on that. And Speaking of the AFM, I posted this on LinkedIn today. The show will come out after that, but I'm taking the first level test Saturday, so I've been trying to spend some time studying for that to make sure I can do all my building and linking and schedules and all those things. And it's been really good. It's helped me, you know, power shop some of my skills and learn some things because I haven't been much of a three statement modeler. I work big companies, a lot of P and L, so it's been really good exercise for me. So thank you. All right, so recently you had the opportunity to attend the Global Excel Summit that just ended in London. Tell us a little bit about what you were doing there and what that was like.

So it was a great time. I guess there were three things I took away from that. So the first one was just how wonderful it was to be in a room full of people who love the things that I love. These were my people. They were celebrating things that, that I super enjoy that. People that I've followed from afar, heroes if you like. But also people that I've been working with or known for years, but have never met in person, which was a really lovely time. The second thing is they were talking about things which are new. So I think Excel has evolved more in the last 18 months than it's evolved in the last, well, certainly the last 24 years. That I've been using it on a daily basis and there are some super exciting things in there. Now we might come back to what some of those things are just to get a sense of that evolution, but you could feel that in the room. This is new stuff. But and this is my third takeaway for is it's not just shiny new toys that we can't find a purpose for. It's like Microsoft has been listening to us and they're building the things that we need. And I can already see, and the global financial modeling leaders at PwC can see how these can be used to make our clients lives easier, to be more efficient, to solve problems quicker, better, a more stable solution that will live in a business longer with less risk. This is real stuff now. And it's a really exciting time to be a financial modeler.

I agree it's an exciting time for Excel. And subject we chose for this episode is around, you know, what makes a good financial modeler. And we're going to dive into that here in just a minute. But before we do, kind of two fun questions I'd like to ask you. You've spent pretty much your entire career modeling, outside of the brief time in auditing when you realize, yeah, this isn't what I want to do all day. Ticking and tying is not the best option for me. What is it that caused you to fall in love? What is your love about modeling? About building something, helping people with financial models. What created such a passion for that?

So I think it's the thrill of the problem solving and then the pride of building something which is loved and embraced and enjoyed by the people that are going to use it. That problem solving piece, that's core to financial modeling and it's definitely something that sits really deep inside me. And I'm relishing that opportunity to understand the problem and then using the sort of wealth of tools and experience at my disposal to solve that problem. And these problems, they're commercial, you know, they are business, but they're also really human and trying to understand how that problem's evolved in the mind of the people that you're solving it for. Not just one, but also a complicated relationship of humans? Probably. I really enjoy that component.

And so speaking of modeling, I'm curious, do you have a favorite industry that you like to model? And if you do, why is that? Why is it your favorite?

So I have. There's so many industries that I've worked in over the years and I am really enjoying working in renewables because it has. So it's got some great complex modeling in it. Project finance is that in many ways the pinnacle of complex modeling in the way that produces the numbers and solves complex commercial problems, but also because it's part of that energy transition. And I, and I really do identify with the importance of that. But I think maybe I'm going to say hotels and I haven't built a hotel model for a while, but I just love how driver based and vertical they are. You know, many businesses are very boring. They're like this thing, this will grow by 3%. This thing, it'll grow by 2%. And you just sort of push all these lines out over the timeline. Hotels just, you start with a number of rooms and then sort of occupancy and it just drops down the P and L. It drops through the balance sheet, it drops through the cash flow in a really elegant way. And there's just something about that that.

I really like fine. And I could see, yes, very much driver based and, and how that would flow. So fun on that one. All right, so let's jump into the kind of focus here of the interview. Let's start with the question here. From your perspective, what makes a good financial modeler? You know, what is someone needing to do to be a good financial modeler? Let's start there.

So I guess there's a few foundations if you aspire to be a great financial model. So you do need financial modeling skills, the technical skills. You need curiosity. So you need a desire to understand the power of the things you have in front of you and not be scared, in fact, actually being used by the new things that turn up the new ways, the alternatives that are presented. You're going to be solving problems, so you need to be able to think about the best ways that you can solve those problems. So I think curiosity and problem solving are those core things. And then of course, really good financial modeling skills.

That makes sense. So I'm curious, you know, we talked about importance of those financial modeling skills and other things. What makes a good modeler from the client's perspective? What is the client looking for when they're trying to find someone to build that model?

That's a tricky question as well. I guess a key part of being a great financial modeler is understanding that for each client that you come across, I think so first thing is if they're going to kind of outsource that, if they're not going to build it themselves, then there's a trepidation about asking someone else to do it. And the first Thing they're looking for is someone who's not going to let them down and build something that's not right, or somebody who's not going to listen to them, or it's going to build it in their own way or it's going to be super. You know, financial modelers have a reputation of building things very complex, aspiring even to build things very complex, which I think is unfair, but that is a reputation. So they're nervous about that. So they need to be allowed to understand that you're going to have their best interests at heart. After that. They really just want to know that you've got an approach and a methodology, as well as the skills and experience that's going to get them to the place at the end where they need to be that outcome. A definition of success is, is the most important thing to them and they're looking for someone who's going to understand that.

So when you talk about understand, I mean, it sounds to me like a big part of that is really being able to listen to the client. Right. How. How do we become better listeners as modelers? What does that mean to be a kind of a great listener of your client? Any thoughts on that?

Yes, I talked about this at the summit in London. So, as a financial modeler, listening skills are universal listening skills. And so the key thing is you're not waiting to speak, you're not waiting to tell them how you've done something before five times, and this is how you did it. You're looking for those and listening for those cues around what's really important and leaving spaces for the client to fill. If there's something on their mind, which is the most important thing to them, but isn't about the model. If you ask the question, what does this model need to look like, they're going to go and tell you about the Excel spreadsheet. But if you ask them what's the definition of success for this project? You might hear something about the spreadsheet and something about the business. But if you leave a space silence, then they will hopefully fill that with something a little bit more meaningful for them. And that's the gold. That's the stuff that really allows you to create a successful model and a successful project.

I like that. Can you maybe share an example from your career where, you know, kind of that listening really helped you realize something beyond just a model that allowed you to be successful in a project? Can you think of maybe an example there?

Yeah, I can think of some. I can think of a lot of examples I think the main one, and I'll characterize it as the secondary objective, is where maybe I'll share an example of one that didn't go so well actually quite a long time ago, but a model that was being built with a very specific purpose in mind. We understood about the model really, really well what it had to do. And there were some complexities around the stakeholder group that we didn't understand and hadn't taken the time to understand. When the model was built, it was presented to our primary stakeholder, who was a very senior executive, and the rest of the senior executive. And we got about five to 10 minutes into that presentation where the CEO said, Hang on, does the model do? We said no, but we said no in a way of, well, no, because no, it doesn't do that. We weren't asked to do that. That wasn't important. That wasn't mentioned. It's almost no use then. And we really misunderstood. Primary stakeholder was the CEO, and that he had a very clear purpose. And if I'm honest, and then with the benefit of hindsight, you go back over that project and the clues were all there, the hints were there, but we were probably so focused on the spreadsheet at that time, I think I'm going back 20 years now, that we didn't stop to ask those questions. And it has a very real impact in my business, Paul, because that client didn't pay for that model. So that's a really important lesson I think I learned that day.

Yeah, that's a kind of a painful lesson to learn when you don't get paid for that model. Because, I mean, at the end of the day, as much as we love doing this, there's a component that needs to be met and that's the, you know, financial needs. Right. And so I can see where that would be a hard lesson to learn.

I'm sure you felt that too, though, Paul, with the models that you've built, it's that payment is not so much the money, it's the fact that it validates. Done for somebody. They value what you have done, which was the idea at the outset. So for them not to value it, that is really painful.

I agree. I still remember, you know, kind of two experiences with that. I'll share real quick. One, I had a boss, he had worked in investment banking and was a very good modeler, had spent years building models. You know, I hadn't done that, and I wouldn't consider myself a good modeler at this point. I remember I built one model this new idea and just really didn't know how to do it. You know, you're trying to figure it out, what does it look like? It's a new idea. I get it all done, and we're reviewing it with my boss, and he pauses at one time and he kind of makes a comment. He goes, this looks like government work. Which was his way of saying, this is not a good model. You know, he's like, you should have done this and this and this. And, you know, we finally got there. But at that moment, it was like, all right, I blew it. He obviously doesn't think this model is worth much of anything. It was a good lesson to learn, but at that moment, there definitely wasn't validation. There was anything but. So, you know, that's one experience I had. And that same boss, I can still remember when he first took over, he had all of us build brand new models. And you want them built a certain way. And one of the things you want done, I just didn't even know how to do. I'd never built a model that way and brought in someone else to help work with me. And I learned a ton over that next year around modeling and building more robust models. And it was a good learning experience. But there are definitely some humbling moments where there was not validation.

And as a financial model, you run that risk every day because every person you work with has a different perspective on what is a good financial model, and they're entitled to that view. Is it an issue? There's a prevailing view sometimes in our industry that when you are in our industry, you know the best way to model. And then collectively, we know the best way to model. So it's a bit of an echo chamber. But really, the best way to model any particular problem is how the individuals who are going to receive that model feel about it. They have all sorts of different opinions. And if they say best practices need to be followed and they are this and they're different to the ones that you're used to, you can have a conversation about that and you might be able to sort of find a happy medium. But ultimately their view on best practices is the thing that you're going to have to adopt for that project.

Yeah. And that makes a lot of sense. Right. Because there's a lot of different opinions. Obviously, there's a few things of like, hey, I want you to do this. You can be like. And that's. That's a real problem. But for the most part, there's a lot of subjectivity in the sense of hey, do I build the depreciation this way or that way? Do I use this formula? Do I do a waterfall or do something else? There's plenty of areas where judgment can be used on what's the best way to do it, right beyond the standards, like don't hard code. Right. Obviously that's one where it's just like, yeah, no, that's just a bad idea, don't do it. But beyond that, there really is a lot of subjectivity there and different ways you can build a great model while still following similar procedures. So, you know, next question here. I want to ask you, you know, for a great modeler, well, in your opinion, how much of the process is actually spent in Excel? Like how much time do you think they should be spending in Excel?

So I think there's two answers to that question. So the first one is about approach and methodology. So in that regard we have a very well defined and deployed use and experience methodology in our team, in fact in our global firm. And that takes you through a process of scope, specify, design, build, test, and then finally handover. Those phases need to all occur within the time that you've got to build that model and hand it over. We would allow 40% of the project time to those first two or three phases, scope, specify and design. So one answer to Your question is 40% of the time before we get into Excel. And if you have two weeks to build a model, which is sometimes a luxury, but let's say you have two weeks to build a model and let's say you don't work the weekends, that means that Friday morning of week one is when you're picking up the Excel file and starting to code. Now that's a hard thing for people to understand and, and adopt because you've been asked, build a model. And so on day one, you're kind of inclined to do that, but if you can resist, there are many benefits, mostly around the fact that you won't have to make changes to the model later. And it's making changes to a model which causes errors. That's the primary benefit. But the second answer to that question is that I guess we cheat a bit. Paul, during that scope and specification phase, we're building artifacts which we share with our clients to get their buy in to that specification. So it's going to be an input template and that'll be in Excel and it's going to be an output template and that's going to be in Excel too. And we'll build a prototype model which is our Little calculation of something particularly complex. And that prototype model might go down the page, whereas the final model will go across the page. But it's just there to help people understand and agree to the complexity. And of course, that's in Excel too, and that has formulae which the client is over the moon about, of course, but we're still specifying at that stage. But we are in Excel using Excel to do that specification.

In today's business world, financial modeling skills are more important than ever. With Financial Modeling Institute's Advanced Financial Modeler Accreditation program, you can become recognized as an expert in the field by validating your financial modeling skills. Join the Financial Modeling Institute's community of top financial modelers, gain access to extensive learning resources and attain the prestigious Advanced Financial Modeler accreditation. Visit www.fminstitute.com podcast and use code podcast to save 15%. When you register.

Modelers, obviously it's a real challenge to find that kind of balance between making sure you spend the time on scope and process and design versus building. You know, we, we like to build. We want to get our hands in there and get started. What does a good modeler do to find that right balance and make sure they're doing their homework and doing those things that will pay dividends in the end in the whole modeling process? Any thoughts on that?

Yeah, that's a, that's a tricky one because you've kind of got to go through that experience a few times to really understand that. So it's all very well me saying to my team, here's the methodology, but unless they've been through that, it's hard for them to really buy into that approach. So I, I understand your question. So how do we learn to actually do that? Well, you could start by putting post it notes on your monitor that say don't start building. But I think because we all have that temptation, don't we? We just want to dive in and start, start building that model. We wanted to bring it to life. We want to show that we've understood there's all sorts of really good reasons. I think the measure is that the end of the, at the end of the specification phase, you have an unambiguous understanding of the model, of its purpose, of how it will be used by the 1, 2, 3, 4 users of that model, what that outputs will produce, the story that that model will tell. Every model tells a story and that story is known at the start of the project and the questions that the model will be able to answer. So when someone interacts with it and maybe, maybe this will be the future. Paul, we won't open up the Excel so much as talk to it, but we'll say to it, I really need to understand what happens if we do this in the business. And the model should be able to answer that because you've understood at the start what those questions are going to be. So that unambiguous understanding tells you you're ready to go, time to start building.

So that, that leads to two questions there. You kind of piqued me thinking when you mentioned hey, we'll be talking to the model. So first let's talk about what are some of the tools beyond Excel? Right. We always talk about Excel and everybody loves Excel, but we know it's not always the best tool for everything. And there are other tools we can use in modeling. So what's maybe some of those other tools that a good modeler should be aware of or consider using that helps them in this modeling process from your career?

Okay, I'm going to put that into three categories. So the first one is the non Microsoft cloud based modeling tools and such. The like that came about around eight, nine years ago, grew to quite a lot of popularity. We invested a lot in those, we trained a lot of our people in those. They haven't taken off as much as we'd hoped. But there are good alternatives out there for cloud based multidimensional modeling. They're not going to replace Excel in most use cases, unfortunately. Maybe we'd hoped they would. Secondly, there are all the wonderful add ins that you can find which other people that have had the same problem as you have put into an add in that you can then deploy and there's more and more places that you can find those because there's more and more ways that you can deploy them. But there's some amazing third party add ins out there that you can use and we use a number in our team every single day. And I think the third category, and probably the most important one is to now understand the full potential of the Microsoft Sweet. That power. Well, I'll call it the Power stack, which is not on all of our laptops yet. But if I describe what it's going to look like relatively soon, it's going to be Excel as the control center, it's the cockpit, it's our entry point because we're familiar with it. We love the grid, we love the way it looks when we have a confidence level when we open it up first time. From there we can already access power query and PowerPivot and we can access M and Dax and we can access of the languages and then we can access Power BI to dashboard and display those things. We can then start to use Power Automate to make things move and make things happen overnight when we're not there. To automate things which have been very manual in the past, and Power apps to make the way that we do that easier, more meaningful on the run, quick and easy. And then overlaying all of that is fabric and I honestly believe that is going to be a massive game changer. So in answer to your question, and it's hard because some of these things aren't really readily available yet, is to be curious about those. There's enough around, enough content available for those sorts of things. If you can understand how that that stack works and how it can be deployed to solve problems, I think that's a great way of thinking about the breadth of offerings and but different ways to solve problems that are now available.

And I like how you broke that down into three categories, right? Those cloud based multi dimensional modeling tools, we have a lot of them. We've seen that grow. They're more common in FPA than probably other places, but they definitely have their benefits. I've seen a lot of them. And then the add ins, right, sometimes can be overwhelming, but I've heard a lot of people swear by add ins and there's ones that they love. And then I really like the third category because Microsoft has done a lot on the tech stack front. Like you mentioned, Power Automate, Power Apps, Power Query, Power bi. Right. There's a trend there. Power, power, power, power. Except for design, fabric. Fabric went a little different, but that's all right. So let's go back to Excel within that tech stack. You mentioned a couple times when we've chatted that, you know, what's been done in the last 18 months in Excel is probably greater than the entire time of your career. Why is that? And what has you so excited about the future of Excel? What are they doing?

What are they doing? So I think they're listening. I think they've realized how important Excel is. Like we've always known it, right, Paul? Like it's the most important software in the world. Like it runs every organization. We don't like to admit this, right, but it runs every finance department, it runs every government. If we woke up in the morning and Excel wasn't on our laptops, I don't know, would the world stop? Maybe it's an extraordinary dependence we have on this wonderful little grid and it has Been mostly ignored and now it has all these new things coming in and the people that use Excel have all these new tools that they can use and Microsoft are adding them all the time and they're integrated and they work together. Good example of that. Dynamic arrays came out actually quite a while ago. We weren't quite sure about it. Adding Lambdas, it becomes extraordinarily powerful. Dynamic arrays and with Lambdas is a really important new feature. We have Copilot, which at the moment is relatively limited in Excel, but I'm convinced it will expand and become more useful. But it's also really useful for other things. Might come back to some of that. We've got Python in the grid and we've got Python in Labs as well. Different way of accessing Python. We've got the advanced Formula Explorer which is in Labs, which is really neat way of exploring formally like Lambdas. Office Scripts is coming. Well, it's here, but it's coming and it's coming to take your vba and I think another one is, and it was demoed on LinkedIn, I think by Joe McDade from Microsoft is the stale formulas concept. Have you seen this?

Yeah, I did see something about that. I've heard, I think it might be in early beta or whatever. I have heard a little bit about it. I'm not real familiar but I saw something on LinkedIn this past week on it.

So it's, it's so simple but it basically shows you which formulae have been calculated and which haven't. So if you're, if you've got a model in manual mode and that's probably not best practices but for any reason and it has other benefits but just for the, for the moment, if you're using it in manual mode, some formally would have calculated, some haven't. The ones that haven't have got a little strike through through them so you can see immediately which form. I love it. What a beautiful addition to Excel. It's things like that, Paul, that make me super hopeful about the future.

Yeah, I know it is exciting. And you know, as we're talking about Excel, I'm going to ask you a question here. I know this might be like picking between your kids, but I'm going to make you pick here. What's your favorite shortcut in Excel? Do you have a favorite?

24 years ago, Paul, the shortcuts were quite different. I am somewhat embarrassed to say that I still use Alt esv. And one of the things I love about Excel and Microsoft more than anything else is that they preserved the shortcuts that technically shouldn't work anymore because of the ribbon, but they still do. I still use that. Okay, I've been weaning myself off that. I now use Control Shift V and that's coming in most Microsoft platforms. It's not an all yet, so it's a bit confusing. I'm weaning, but alt est for formats, alt ESF for formulae for people. Just to be clear, that's paste special values, paste special formats, paste special formulae. I still use those keyboard shortcuts.

So funny. The last interview I did, I asked what their favorite was and they said Control Shift V. And I mentioned I'm still using Alt esv. And so as I've been prepping for this test just today, I was like, well, let's try starting to use Control Shift V a little bit more. And I can see where it's a little quicker, but old habits die hard, so I'm with you. I'm not there yet, but I'm dabbling in Control Shift V, so I can appreciate that one. So I thought that was great. You chose that one because. Yeah, just today I was thinking about that.

So yeah, I feel, I feel your pain trying to re. Teach fingers to do something they've been doing for decades.

Right. So one last question here before we move to rapid fire. When you're, you know, when building a model, what's a technical nugget you could maybe give our audience that you wish more people incorporated when building models. Like, you know, could be something around the waterfall schedule or the debt and the revolver, whatever that may be, but something you wish more people would use. Kind of a technical nugget that could help them when building models.

So, so many. Let's go with error checks. So error checks are often a last minute thought and addition. But in reality, error checks should be there from the start as you build the model. You should be building them as you go. Use an absolute function. So an error check is checking that two things are the same. So use an absolute function to take the difference between the two and then check it against an error tolerance just in case you've got one of Those crazy floating point 10 to the minus 15 errors in your model. Use the absolute with the error tolerance and then give it conditional formatting. If it turns true, then it's green. If it turns false, then it's red. And add those everywhere in your model where there's two numbers which should be the same. So the most obvious one being the balance sheet, does it balance? But let's check that the P and L equals the movement in reserves. That that cash flow equals a moving in cash. That the sum of my four quarters equals my annual. That the sum of all my cash flows across all of my business units equals my consolidated. Anywhere where you can do a check, put the check in. And it gives such a lot of confidence. There's green lights all over the model and when something goes wrong, there's red lights to checks.

Good advice. Error checks is definitely an area I could get better at my models. I sometimes neglect it. I was just thinking about the other day on something I built going, there's no really good error checks here. Outside of that the balance sheet ties. Right? You know, but that doesn't mean your model's right there. There could be plenty of issues with your model and your balance sheet ties. So I think that's really good advice. All right, so I know this is going to be your favorite section because I get to put you on the spot, right? You're probably like, ah, I'm dreading this. So I think you know the rules, but I'll repeat them. I got about 10 questions here. You get 10 seconds to answer. You can't tell me it depends. So you got to pick a side and then at the end I'll lab allow you to elaborate on one or two why you picked an answer. Because I get there's nuance to all these, but the fun is just kind of seeing whether you. Where you think about them, where you sit, if you have to pick. So when building models, circular or no circular references.

No, no, no, no, no.

Okay. Where were you again? Was that no?

I'm pretty sure that's a no.

All right. VBA or no vba?

Yes for now.

All right. Horizontal or vertical?

Horizontal.

Should have been using dynamic arrays in your models. Yes or no?

Yeah, embrace. I wasn't sure, but yes.

All right. External workbook links, yes or no?

No.

Yeah, I could tell which ones you have a strong opinion on. You get more than one. No named ranges versus no name ranges. Where'd you sit on that? In a model?

Yeah, Name ranges. I like named ranges.

Great. And so when building your models, you follow the formal standards of one of the boards out there smart or, you know, some of the others, they're out there fast, etc.

We use PwC's global financial modeling guidelines that we authored just a few years ago. It's publicly available. It's a really good set of guidelines which you can dissolve to a standard if you want to look at it. That way. But it's a really great way of designing a model.

Great. Yeah, no, that's, that's good. Now we'll call that a yes. Will Excel ever die? Yes or no?

No.

All right. We've, we've had a few different ones on that one. Always fun to see the opinion. I know this will be a little bit of a controversial one. We talked about it. Will AI build the models for us in the future?

So it is a difficult one to answer because it might not be the right question, but yeah, it almost certainly will build models for us at some point in the future.

Okay, fair enough. So sheet sell protection in your model. Should you use it? Yes or no?

Oh, yeah. If everybody's happy with it, yes.

Makes sense. And then this is a question I like to ask people. Do you believe financial models are the number one corporate decision making tool? If no, what do you think is.

Categorically yes, Excel is the number one decision tool.

I like it. All right, so last question. This is just kind of a fun one. We asked everybody. What's your lookup function of choice? Do you like choose Vlookup, xlookup index match or something? I didn't mention.

Plenty of. Plenty of good ones there. I try to default to xlookup these days. I think it hasn't taken off as much as I thought it would. I wrote an article on that some years ago and it. There are some use cases where it doesn't quite work for, but I think we should default to xlookup Now.

I would agree with you. I generally tend to default to xlookup. I think there's use cases you can make and arguments for all of them. But I do like xlookup. I think it's very flexible. So which one would you like to elaborate on? Go ahead and pick one there.

So why don't we go? Well, there's a couple. But we should probably talk about AI because it's really an important conversation to be having and everybody wants to talk about it at the moment. And I think I'd summarize the situation by saying that and a number of your guests over the last few months have said similar things. It's here to help us. It's a friend. It's someone you can delegate things to. It's someone you can ask advice from. It's going to help you at every stage of a model build those. Scope, specifier, design, build, test and handover. It's got a role to play today in all of those. We have a chat genai tool within PwC called Chat PwC, and the expectation is you'll be using that multiple times a day to support you in the work that you do inside the grid, inside Excel, it is currently relatively limited in terms of what Copilot can do. It's fun, it's exciting to see it doing those things, but it hasn't really changed the way we do things in the future. Will it have a bigger role as its use case expands, as the technology expand, as we teach it things about financial models? And it can teach us more about the financial models that it sees? Yeah, it will. It'll help us to review models, it'll help us to understand them more quickly, it'll help us to build the driver trees, it'll help us to build the input templates, the output templates, it'll help us build the prototype models, and maybe in the end, it'll build the model for us. But when you reflect on everything we've talked about today, the human needs, the priority objectives, the secondary objectives, the people that are. How can IT know all of those things that are actually critical to a successful project? That is the bit I'm still trying to understand.

Yeah, I know. I agree with you. I think there is a lot you said there. And one of the ways I've heard it said is it's not human versus technology. The way we need to look at this AI and other things is humans with technology. It's there to make us more effective. And the example I give from just this week of generative AI, I think last week, but I had to do this indirect. I just don't write indirect formulas very often. When you're referencing sheets and yeah, I know, you're like, good. Different rows and columns. It can get really complex in the syntax to get it right. And I'm like, I. I'm going to be sitting here forever figuring out the syntax. So I just told Chat GPT what I wanted and just kept asking, did write, okay, give me the next part, give me the next part. And 10, 15 minutes later, I had it working beautifully. And I probably would have spent an hour going, all right, why? I forgot the stupid quote here or the ampersand and the quote aren't in the right order, Whatever. And so it's just an example of where generative AI can help us in that process, whatever it may be. And it's that much more powerful if we understand what we're doing, if we have a good grasp of things, where we can really ask the right questions or be smart enough to validate what it's giving us, I think AI is most deadly in the hands of the competent, the ones who know their career. Like, we get this fear that we don't need to learn anything now that we have AI. I think those that have learned and put in the time, if they're using AI will be the most effective people with it.

There's no doubt about it. And there are very real use cases of that right now happening in our team where people are using it to do exactly those sorts of things. A couple of examples that people could start doing right now, if you've got Copilot, it's in Power Pivot, it's writing Dax, that's a really useful place to have it. It can have a pretty good go at coding vba, and it's got a pretty good go at while it's learning to code VBA into Office Scripts. And as I sort of hinted at, I'm going back to one of your quickfire questions. I apologize, but I hinted at VBA will, will be replaced, I believe, and it'll be replaced because it doesn't work in the cloud. And the cloud is going to become the most important place for our financial models. And we need a language that does. And that's Office Scripts, Office JavaScript. It's relatively new. The Gen AI tools are still learning how to code in that language. But you can see a world where we upload our VBA macros and they turn up in a whole new language and we can just copy and paste and that'll be a really important part of the transition, I think, from desktop Excel to the cloud Excel.

I agree with you. And I think we could spend a whole episode on that subject and BBA and the cloud and how we make that transition. Because there's no way around it when that transition happens. And if they do decide, look, we're gonna really not support VBA and slowly take that away. That's going to be a painful, long process. There's no way around that. So that will definitely be interesting to watch. But maybe we'll hold that one for episode two to get into a little bit more, because I think there's a lot we could unpack there. So, you know, we're coming at the end of our time. I've thoroughly enjoyed having you. We just have two questions left, as we conclude here. So the first is, we ask everybody this. If you could give one piece of advice for our audience to be a better financial modeler, what is that one thing you would tell, would tell them or remind them to do.

Be curious as you're facing problems, as you're seeing things of content on YouTube, you've got a new problem to solve. Be curious about how you might solve that in new and interesting ways. And learn. Learn the trade through that curiosity.

If our audience wants to learn more about you or get in touch with you, what's the best way for them to do that?

I'll just reach out to me in LinkedIn. I'd be happy to hear from people.

Great. So we'll put that in the show notes and thanks so much for joining me. I know it took us a little while to get this scheduled. We've had some great conversations and I'm really excited to share this with our audience. So keep doing the great work you're doing in modeling. And thanks for joining me.

Paul. Thank you. Genuinely. Thank you for everything you've done for this industry, for profiling it. I so enjoy listening to your podcast. Thank you very much.

Thanks.

Financial Modelers Corner was brought to you by Financial modeling institute. Visit fmi@www.fminstitute.com podcast and use code podcast to save 15% when you enroll in one of their accreditations today.

It.