Quantrix Approach to Simplify Complex Financial Problems with Gilbert Hangartner

What were the benefits you were seeing in Quantrix? Why do you prefer that over modeling in Excel?

It has two features which are crucial to modeling, which Excel doesn't have. First is.

And second is Excel gives you pretty much unlimited flexibility in the sense you can build whatever you want, getting navigating around.

And don't lose your overview. That's key to working efficiently with Excel.

Is modeling more of an art or a science?

It's an art. I cannot think that you can really learn this or have a systematic process. It's a creation. It's really like writing a piece of art. Don't be afraid to appear stupid. Simulation is used because it's so simple. Most people who are doing quantitative risk analyzers using statistical methods, not all of them, but very often they get the wrong results.

Will AI build models for us in the future?

No. If AI can do that, then everything else is replaced before.

It's hard to know what you don't know, so to speak.

You really have to understand how the business functions. If you want to make a model, it has to be 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. The Financial Modelers Corner podcast is brought to you by Financial Modeling Institute. FMI offers the most respected accreditations in financial modeling. Traveling this week, we are thrilled to have with us Gilbert Hengartner. Gilbert, welcome to the show.

Hi, Paul, nice to meet you.

Yeah, really nice to to meet you as well. Where are we today? I know you're out on your boat. So where are we at?

I'm currently in Canarian Islands in the middle of the Atlantic Ocean on a little island called La Gomera.

Well, enjoy it. I hope you get to go out and enjoy the beach and spend some time on the island. That sounds like a fun trip.

I will do that.

All right. As you know, I always like to start the show with the same question, and it's tell me about that worst financial model you ever seen, that horror story.

There are obviously a lot of very bad models, especially complicated models, which are difficult to understand and which give the user a pain to work on. But the story I want to tell you is the model that had the most disastrous effect, and that was new coming, new to a company. I was just hired and the CFO came up to me and said, gilbert, we just finished this project. Can you look at this model before we go to the board for the final decision? And I took a look at the model. And they worked half a year on it, 10 people. So it was a lot of work. Dozens of taps and thousands of lines. And, you know, once you wanted to understand the calculation, you had to follow a formula over 10 steps to find the right number that gets into it. That was horrible. In the end, first they just modeled the new business was a business extension, about 20% of the business they wanted to invest in. And then obviously the board asked, well, let's compare this with our traditional business. If we won't do the extensions or that we compare, how is it with the investment and without the investment? And there were 18 times that it just took the current business we already have, they took other numbers, and then it made the difference. And these other numbers weren't calculated the same way in the same model. So the difference gave a huge positive increase of net revenue for the investment. So the board said, yeah, cool, let's do that. And then when I calculated, I realized that if you would calculate it correctly with the same assumptions and the same logic, the difference would be very small. So going from a very good business case, 20% invest, 50% increase in net income, it left to 5% of net income. So it was really not worth doing it. And it increased a lot of risk and a lot of work. And I went to the CEO and said, I'm sorry, this is all bullshit, you have to cancel the project. And said, well, we told the board now for months and we can't come up now two weeks before the final decision and tell them it's all wrong. So they went ahead and they did the investment. And two years later, it turned out that, well, the numbers weren't what they had expected.

That is a really bad one. And it's unfortunate you weren't there a little earlier, before they gone to the board. It's unfortunate that they weren't willing to go back to the board and say, look, this isn't the idea we thought it was. We reran the numbers, there were some errors, but nobody wants to look bad. But at the end of the day, it's like, okay, look bad at the beginning and save the company money or look bad at the end. You know, I. I would much prefer just owning it up front. Feels like the better way to handle that. But what was the key takeaway from that experience for you? What. What did that teach you? Going through that model and that whole experience?

It again shows that. That understanding the question and thinking thoroughly about the question in the beginning is much more important than making a complicated model. Going to all the details, I mean, they, they really calculate. I always, I mean, the first science thing a bad model is if the, the amounts are in dollars and cents. Okay, they calculate everything to descent. And that's not the point. You should not seek precision. You should really seek to answer the question. In this case, it was, is it A or B? Should we invest? Should we not invest? And then you have to build a model that gives confidence in giving you the information to make the good decision and not having something complicated, something precise. That's not the point. So understanding the question and think, how can I build a model that reliably answers this question?

I love how you said that reliably answers the question. And then also, you know, one of the terms I've often heard is the false idea that increased precision increases accuracy. You know, like going to the sense like you don't need to. Forecasting every line at a detailed level doesn't make you more accurate. It usually makes you less accurate. Knowing your question and focusing on the big picture is so important. So I appreciate you sharing that. Can you tell our audience about your background and a little bit of your career story and how you ended up where you're at today?

That's a great story. I started in science, so I did chemistry and physics. And I mean, physics is all about modeling. So I built crazy models, quantum chemistry with electrons and atoms and so on. So I was always very puzzled by having a problem and not knowing how to solve it and thinking about the problem and then go from there and from there. Later in my career, I came to an energy company and there we had obviously technical risks. So it was always, well, we don't know how to handle this. Gilbert, can you take a look at that? And so I came really from the technical side to do modeling of physical processes, of investments, of infrastructure, infrastructure stuff. And then all the more it connected to science because knowing what your risks you have always then goes to money. Should I invest to reduce the risk? How much would it cost if my plant goes offline? So it goes all back to money. And then I realized having to present my results to boards and to finance the guys, I realized that you can apply the same principle from a physical model to a financial model. It's just the problem and modeling the question. And this, I think this modeling part and understanding that the model is not the reality, it's a difference between reality and a model. And this difference doesn't harm. You just have to know what is important, which behavior you want to make. And a model, as we said, before should answer your question and not to try to be the reality.

I, I like, I really try to answer your question. So I, I have a question and I think we know the answer. You started your degree in physics today you do a lot of financial and business models. It sounds like you had a role where they kind of had you start doing that. Did you find you liked it and can wanted to continue to do that or was there a specific moment when you said hey, I really want to be more of a modeler than focusing on kind of that physics and chemistry side. Like how did that transition come about?

It was just interesting. It's such a creative work for me. For me, making a model is like for others, maybe doing a painting or writing a song. It's a kind of a creative act. And each time I did that I liked that and was very rewarding. And I have seen that you can really bring a benefit to the company. People like them because you can come up with insights they didn't have. That's one thing. And the other thing I discovered when I was starting to move from the technical modeling more to the financial modeling, that a good model can bridge a gap. Because often in companies you have different world, different interest zones, and especially between technical people and financial people, there is often not that much communication going on. And I really learned that if you build a good model, you can take everything that's important to one guy and his assumptions into the model, from the technical guy into the financial guy, and make them agree on their inputs. And then you have a systematic method to get to the results and then they believe in the results. And I often had the situation where finance wanted to do A and Technics wanted to be B. And in the end, thanks to my model, they could talk to each other and they understood each other's arguments and they found the common solution which was backed up by the entire company. So a good model is a tool to think. A good model is a tool to communicate mainly. And it's not at all about the numbers.

I really like you said a good model is to think. A good model is communicate. And we'll definitely dig more into that. But one thing I really liked is you said you love the creative process. So I'm just curious, your take is modeling more of an art or a science? Do you have an opinion on that?

It's an art. I think it's an art, definitely. I don't know, maybe I'm too stupid or maybe I don't know the right people.

There's not A right answer to this.

I cannot think that you can really learn this or have a systematic process. It's a creation. It's really like writing a piece of art or creating a piece of art.

I like, I like it. I figured that's what you say, but I just kind of wanted to get your take. You know, some people think of it more of a science, some think of it more of an art, some are in between. And so, yeah, a lot of different opinions and I think it's a little bit of both. But it definitely is a big part of art to it. No question in the. In the creative. I just worry sometimes when people don't follow a little bit of science in the structure. Right. But that's a different story. When you and I chatted, and you've already hinted at this quite a bit, we talked about quite a bit before we met here about the importance of making sure you understand the question before building the model. Why is it so important that you really understand the question and what you're trying to solve before building the model?

When we agree that the model can't be a complete exact reality, then you always make some simplifications and there are some simplifications that hurts and others don't. And if you don't understand what is the definite goal, the ultimate goal of the model, you cannot choose the right simplifications. You cannot make it reactive. Very often people think on a deterministic equilibrium point, okay, they know their business and they say, well, this is like it has ever been. And we just calculate this. And I want maybe to know If I move 10 cm to the right or 10 cm to the left, what happens? But this is not. Especially when you're going to risk management and things, then you have to have big movements. What happens if Covid happens? What happens if Ukraine in war happens? What happens to my system if the energy prices are not 5% higher, but if they're times 10? Okay? So you must really make a model that acts more or less correctly. If you make big movements on your drivers. I mean, if you just start and do anything, then you have a model that moves in all directions. So you have really to understand what is constant and what will stay and what can move. And how are these interactions and are there interactions? Can I for. I mean, you can make a linear extrapolation and if this happens that the price changes, then I will produce the double, but you can't. So you have really to understand what is the capacity, where can you shrink, where can you expand and you really have to understand how the business functions. And the trouble is when it comes to a company, often you talk to specialists and they know 5% or 10% of the company, but you very rarely find people that understand how this little thing on this edge of the company interacts with something else somewhere completely in another part of the company. But this has to be in a model. If you want to make a model, it has to be an end to end model which connects all the pieces of the company. So you have to talk to all the people and always try to get to this higher level to have the overview of the company to make a model that in this big picture works the right way.

So it sounds like if I'm hearing right, I mean, a big part of it is you really got to understand the business and the right people to talk to in the business. And so how do you ensure you're kind of reaching the right people? Especially if you're working on a project, maybe you're not super familiar with the industry, you're new to the company, how do you make sure you're asking the right problems and you know when you've really defined the problem to ensure you're able to build that end to end model, how do you go about that, making sure that you've kind of covered everything? Because it's hard to know what you don't know, so to speak.

Don't be afraid to appear stupid. I always say, I'm sorry, this is too complicated for me, I didn't understand. And very often, you know, I'm in a meeting with three, four, five guys of the company and one of the guys explained, well, you know, it's very easy. It's like this and this and this and this and sorry, maybe I'm stupid, but I didn't understand. Can you explain me this in an easier way? And then the four other people after the meeting tell me, oh, that's the first time that I understood this. So even in the company people are afraid to understand that everybody's a specialist and nobody wants to show that he has no clue about how this working. And often I'm trying to all this complex stuff to reduce it to the bare minimum. What is the essence, what is the worry? All details apart, what is A plus B equals C? What's the basic equation? And I'm digging as long as I have it. And if he can't explain me in simple words, then I'm not satisfied and only when I'm there. And then in the end it gets very Simple. And once it gets simple, then I'm sure that I've understood.

I really like that. So basically keep asking questions. Kel, it feels like you can explain it in a simple way. Yeah, it's like the whole. I think Reddit has a site in other places where you hear explain it to me like I'm five. Yeah, right. You know how you explain it to a little kid? Yeah. Because if you explain it in simple terms, it's going to be much easier to understand. No reason to use big words. But I, I love the just keep asking questions. I had someone on a show that said one of the best signs of a good CFO is they keep asking. They ask a lot of questions. So I think there's some truth to that in a modeler is don't be afraid to keep asking till you're comfortable. Otherwise you never know if something you're modeling is really that important or if you understood the other implications, your assumptions. It's much easier to make a mistake.

And as long as you so complicated you can't even understand it, it will be too complicated to model. You will lost. You will model to precision which is not worth doing it. And the most questions are rather simple and, and, and the goal is to, to make it simple.

Yeah, I, I've always said and I'd love your thoughts on this. One of my favorite quotes, I use it a lot in training is simple is hard, complex is easy. It's often easier to build something that's complex. What's your thoughts on that?

I couldn't agree more.

That's what I figured you say just from the way you're explaining it. Right. You really know. You, you really know. You understand something when you have to explain it to others and you have to make it simple enough so they'll understand it. If they don't understand it, there's probably a question of how well you understand it. Not always. Sometimes you're just using big words. But I know for me training people, when they get it, then I'm like, all right, I really know this, this subject. So I want to move on and talk a little bit about risk because I know when you and I chatted we talk quite about risk. I know it's a big part of what you're doing, big part of your work. First, let's start with what tools do you use to model risk. I know if you're doing statistical modeling, that can be a real challenge in Excel. So if you used add ins, you use a different tool. How do you go about Modeling risk, especially if you're doing any kind of Monte Carlo or other statistical methods, which I'm sure coming from a physics chemistry background, you probably have a very strong math in there around risk modeling.

There is a very big battle about statistics and simulation, especially with one client I have, they have a very big, a lot of people working on very good statistical models and they're afraid of simulation. I'm completely against using statistics for risk management because statistics would assume that you know the result and normally a normal person, a simple person, doesn't know the result. And that's the advantage of Monte Carlo methods. You just build easy bits one by one and you let the simulation figure it out. What then the statistics is and the statistics is not the input, but it's the result. So that's the first thing which is very complicated and often people are very afraid of it because they think it's complicated and that's not true. Simulation is used because it's so simple and you should always make it simple and you shouldn't be afraid, oh, I don't know the z distribution, I don't know the probabilities and so doesn't matter. If you understand the problem right, it will be very easy, just formulate it. Very stupid, make the most simple assumption and it would always be good enough. So what tools do I use for Monte Carlo? I did it in Excel, I did it in Excel using At Risk, which I liked quite a bit for time. And then people didn't want to buy these add ins, so I did it in Excel. With a little vba it's possible. You can do quite a lot of Monte Carlo simulation just in Excel. That's maybe not the good tool, but it's possible because it's so easy. You can do it in Excel. Now I'm using it in Quantrix. So I built a little plugin, very stupid little plugin to extend Quantrix to be able to do Monte Carlo simulation. But there are obviously other tools, Gold, Sim or real simulation tools. But for the normal business modeling, it's not worth it, it doesn't matter. You can use anything you want. It's the simple. The Monte Carlo is the simple part. If you have a business model, understanding the business is the hard part. And doing the Monte Carlo and the risk on it is the easy part.

And what advice would you offer if you have someone listening, they're like, I haven't done Monte Carlo. Do you have a resource you'd recommend they go to? Or a way to get comfortable in learning how Monte Carlo Simulations work.

Take dices or a deck of cards and play and try to model this game afterwards. Start with simple questions and go until you really understand it.

I like it. So I need to get a deck of cards. I think I got one back there. We can play cards because I might need to understand it better. So basically just starting to understand the whole idea of risk. But any, like, as far as setting it up, is there any math stuff they need to know of how to do a Monte Carlo simulation? Do you recommend kind of reading about it if they're not familiar with the concept at all? Not, not so much the idea of simulation, but just how to structure one. Is there like, you need to know certain statistics or anything like that? Like, is there any different beyond, you know, kind of your typical modeling for risk that you need to understand?

No, obviously there, there is a lot of theory and, and, and a lot of, a lot of statistics, but you don't really need most of that stuff. You really need the basic thing. Okay, I, I, I have a random variable dice and I throw the dice. I mean, just do the basic exercise. Take two dice, throw them and write down the numbers and make the statistics out of it. And if you really understand this, then you have understood everything. And then go from there and do a budget little company. The revenue, you throw the dice and the cost, you throw the dice and then you add it together. And once you did that and you get the distribution of the net income, then you're done. Then you have understood what, what simulation means to, to financial modeling. It's not more complicated than that really.

Like the way you put that, right? If you take a dice, we all know what six sided or whatever it is. You roll it. You got 12 different, you got six, you got six, you got, you know, 12. Whatever the outcomes are, you roll it each time you get your different numbers and you kind of figure out what the probability is going to be. And then you're, so you're just running those simulations and figuring out the probability and what that looks like. So you have an idea of, all right, I think I'm gonna do 5 million, but my range looks like 3 to 8 and it skews toward the 8 or it skews toward the 3. Right? Isn't that kind of the thinking? So you have an idea to say there's really, If I think fives, where I'm gonna add it, there's probably about a 60% chance that I end in that range.

Don't even think about probabilities. I mean, it's okay, just. Just throw the dice.

I like it. I might have to do that on the next show. Bring some dice and I'll throw them.

I always bring some dice. When, when people, when I'm teaching a course on Monteculism, Monte Carlo simulation, or when I want to convince people, you know, they say, well, let's make the model simple. Let's make it deterministic. And I say, well, if you make it deterministic, it's wrong. Yeah, but Monte Carlo is complicated. So then I bring them some dice and then we do a little game. And then they understand it's not more complicated. The only thing that is complicated, I must be honest. There's one thing which really bothers people. The calculation is straightforward, but you get a stupid result. You don't get one result, but you get a lot of results. And this is where people have problems with. Because they're used to accuracy. They want to say, my net revenue is X million 75 cents. Okay? And if the 75 cents or 76 cents, then they have a problem because they believe in this number. But once you accept, I mean take your last year's budget and compare it to what really happened. Take your budget you made in 20 before COVID and take the result you got the year after. I mean, it's nonsense. The budget has no connection to what happens in reality. And when you accept that your planning is wrong anyway, then you can make the next step and accept. I'm doing a model and I'm not getting one number, but I get in some kind of a distribution, I get an uncertain result. Most probably here a little bit up to there can be very worse. I get something. Okay, it's about that. And this is the only thing which is difficult to accept. You don't get an exact number, but numbers are anyway wrong.

FP and 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 certificate application 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 FMIS 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. Yeah, as I've always said, if I could forecast the right number I would have invested in the stock market and be sitting on a beach somewhere not building models. Because you're right. Anytime you rely on an individual number, it's never going to be right. Having a range and understanding kind of where that range falls. Like you said, being able to see that distribution gives you a, a comfort level to say okay, we think we'll be between here and here that risk is acceptable. And yeah, you may take the midpoint and say okay, that's the goal of where we want to be. But if you have the understanding that there's a range around that and that that that's just a, a point not really that you're, you're not going to hit. I mean nobody ever hits the exact number. At least any complicated model, maybe something super simple. But as you know in general it, you're just, you're it's estimates. So I, I agree with you. Definitely an area I need to get better. I've, I've played a little bit with Monte Carlo but I never really used it in my career career which I think is a common challenge. Why do you. So you say most people don't use it for their complicated. In your experience what percentage of companies and people are modeling do you think are actually using Monte Carlo? You have any fill on that?

Nobody.

Very low. Interesting. I mean I've seen it in a few areas.

Mo most people who are doing quantitative risk analyzer using statistical methods and not all of them but very often they get the wrong results because it's too complicated and because you're basically knowing the result before and then you're doing the statistics about it and then you have some nice looking graphs. But if you're doing Monte Carlo modeling, you're doing the other way around. You don't know the result but you know what happens in the real world and you put that to the model and the model explains you then what will be the result? And maybe I'm wrong but I know very, very, very few decisions really been taken there more in technical domain. But in the financial business decisions, not that much.

I would agree with you especially in the financial. I think there's some other areas that use more but I've seen at least in corporate finance it's rarely used there. I think there may be some, you know, risk project modeling and some areas where it gets used but I would agree with you.

Project risk. Yes. Project is often. And I don't understand why, because it's so easy and it brings so much benefit.

Like I could tell you in fpa, in my career, I never used Monte Carlo.

We could use it a little demo model together once when you have time.

I might have to do that. It'd be good for me. So I'll take you up on that sometime. We'll do a little demo model. That'd be fun. We'll. When I get back from vacation, I'll.

Reach out in two hours. In two hours we have the first model.

All right, sounds good. I'll. We'll, we'll schedule that guests will know coming in the future. A demo model for your Monte Carlo. All right, so I want to move on to a little bit. I know you spent a lot of time Excel in your career. You still use Excel some. But recently you were introduced to Quantrix and I know you're a big fan of Quantrix. You really liked it for modeling. You mentioned even built a little Monte Carlo simulation. What were the benefits you were seeing in Quantrix? Why do you prefer that over modeling.

In Excel, it has two features which are crucial, crucial to modeling which Excel doesn't have. The first is multidimensionality. It's just stupid having columns and rows. I mean, the world is not two dimensional. Never. No problem. I mean, every problem has 5, 10, 15 dimensions. Okay? So you cannot do a two dimensional model. And if you want to model it in Excel, you have to do a lot of work to cut your problem in slices and put it on this flat paper. So this is the first multidimensionality, definitely. And second is flexibility and scalability. I mean, in Quantrix, I first solve the problem very easy with two examples or three. I make everything works right, the formulas work right, the structure works right. And then I pump in 100,000, 10,000 items, doesn't matter and it will all be the same. There will not be an error, there will no formula to be adopted. So the scalability, the security, the stability of the model is just another word. And I was very, very funny because I built a rather good model and stable model which was used for production and simulation model in Excel. And I showed this to somebody when I worked with Quant Extreme and I created little VBAs to create this multidimensionality. And I created a formal parser so that I could write the real language formula into Excel, which then were translated into the cell reference formula of Excel. So I basically built my little Quantrix in Excel without knowing that quantrix exists. So these things are really, and again quantrix is not a good modeling tool. Quantrix is a good business, a finance modeling tool, but not Monte Carlo. It can't do that. It's not very good on this. So I added this. But again the business model is the complicated part and this is so easy, so clean in Quantrix and then just the Monte Carlo stuff is easy peasy to, to add on top of it.

Got it. So it sounds like one of the things you really appreciate. I mean you mentioned a couple there but the multidimensional modeling is a really big one and I would say that's one of the biggest strength of when you model outside of tools like Excel is that multi dimension, whether it's Quantrix causal, there are others out there, quite a few that I can list that allow the multi dimensional modeling. That is one of the biggest limitations. I love Excel. I use it for just about everything. But I also recognize you're, there's definitely some trade offs you're making when you choose to use it. And you've liked quantrix so much. You mentioned that your youngest child, you taught them Quantrix vs Excel. Talk about that for a minute. I, I, I found that very interesting.

I, I have four childs so the, the older three are a little bit older and the younger is the 10 years younger. So I always try to teach them Excel. You know, I teach you Excel so you have to learn something for the world and don't even if it's a stupid little calculation you have to do, do it correctly because in all your life you will be using Excel. That was what I always told my kids. And now my youngest I was working with quantrix said I just tell the teacher Quantrix and so everything she has to calculate something or doing some little example for her math classes or something. She always uses Quantic. So she's, she's 11 now but she started using Quantic with nine, with nine years.

I love it. That, that's, that's great. Thank you for sharing that. And she'll be the Quantrix expert when she goes to college.

Hope so.

All right, so we're going to come back to Excel for a minute because toward the end we like to ask this question to people just for fun. Within Excel, do you have a favorite shortcut you like? Is there one that you can't live without if you had to pick one?

I'm very old fashioned mouse guy. I grew up with Macintosh so I'm very tied to my mouse. But the Most important thing, somebody once told me, control arrow down to go. Especially when you have sheets of other People, which has 10,000 lines. This is great. So I mean, instead of scrolling down 10 minutes and it builds up screen after screens of this, this is really. Wow, that saves 50% of the time.

Yes. Those control arrows, the end, the home. If you're sitting there scrolling more than a second, there's a better way to do it. And I've been guilty of it. There's times when I start scrolling, I was like, what's the column that doesn't have a bunch of blanks? So I can get to the end here really quick. And how do I get this where I'm not? But yeah, control the, control down, arrow the right, left. All those are fabulous.

I think navigation takes the most time in Excel. And if you master navigation, how to get from one point to the other, how to jump to the sheet where the data is and come back without getting lost. I mean, if you manage that, then you have really half of the.

I think that's huge. Like, I know. Control, left bracket. If you're in a formula and you want to go where it's referencing, I think it's alt F5 to go back to where you were. There's lots of little things like that, that along with those control arrows that could just save you a ton of time. Not even learning any of the others of performing certain functions, but just starting with increasing your navigation speed around a spreadsheet.

Navigation. Yeah, I think navigation. Because actually it's so difficult to get the whole picture. It's so unstructured and normally it's so big. So getting navigating around and don't lose your overview, that's key to working efficiently with Excel.

Yeah, and that, that's. That's the greatest strength and weakness. What you mentioned is unstructured. Right. Excel gives you pretty much unlimited flexibility in the sense you can build whatever you want. It's just a blank canvas. At the same time, you don't have the structure. You get a lot of flexibility, but with other tools, you often get structure and some benefits that when you learn them, can make a lot of things easier. But you almost always give up at least a little bit of flexibility because you have a database on the end. You're all sudden using structure. And that's the biggest thing I ask. I have people ask me all the time, well, I want a tool that's just as flexible. Excel. If all you care about is the flexibility, then just stay in the spreadsheet. You're going to trade off some of that flexibility for a lot of other things that may make your life a lot easier depending on what you're doing. And you have to decide if those benefits make sense. That's usually kind of how I frame it with people.

Completely agree, completely. I'm just not sure if flexibility is really what you want because, sorry to say, it's good for people who are afraid to think about the problem. And if you have quantrix is really annoying because you have to think and it forces you to really make a clean, clear structure. But this clean, clear structure then helps you not only because you're more efficient, but also you have a better understanding of your problem. And you can build anything in Excel which doesn't make any sense and you wouldn't realize, but in Quantrix or in any other tool which is more structured, you have to think about this structure. You have to make decisions and think things then are possible or not possible because you defined it this way. And this really forces you and I think that's a good thing. It forces you to be clear and to have a clear structure, clear data and you know what you're doing.

I, I think we're gonna, I'm gonna contact all the multi dimensional tools so see if we can make you the spokesman. You definitely, you have it down and what you see as the advantages and it's clear you really appreciate those and I like that. I love a strong opinion and coming with your reason. You know, I'd say most of people we have, and most people obviously model in Excel. I'm no exception. And that's their tool of choice and getting them to look at other tools can be difficult. I've looked at a lot of tools and there's definitely some benefits. One, like I said, one of the biggest things I like, whether it's Quantrix or dozens of others I can think of, is one of the biggest benefits is that multi dimensional modeling. I am with you. That's a huge thing. And definitely there has to be more structure. And it's a good question. Flexibility is what you can to do.

Structured modeling in Excel. Yes, but Excel doesn't force you, you have to force you yourself.

Correct. And that's why there's so many guides and so much discipline is, you know, like investment banking. One of the first things they teach is standards and here's how you model because they're trying to force a certain discipline and I never learned that earlier in career. My, my first few models were terrible because it's like nobody talking I'm just trying to figure it out and you're just throwing it all on the paper and then you get done and you're like, what have I built? This makes no sense. Whereas if you start with something that forces you for structure from the beginning, there is definitely some benefit. And so everybody has to decide, where's that trade off? What makes the most sense. Obviously, Excel is extremely inexpensive. We can do just about anything in it. That's probably one of my favorite things about it. But you just have to recognize, like anything, it comes with trade offs. I, I, I love how you shared that there. So I want to ask one more question. Then we're going to move into our rapid fire section, which is kind of our fun section, where you have to choose a side and then at the end you can elaborate. So what is the one thing, if you had to list one thing that you've learned during your career that's helped you the most, what is it?

I don't know. Not being afraid. Not being afraid. Asking questions. Again, that's, I think that's, and, and the confidence that everything that's important can be easily modeled for, for, for everything. It's an easy, there is an easy answer. There is a nice little neat formula. If you understand the problem, it's always possible. Don't be afraid.

I love that. So basically, don't be afraid, ask questions, keep digging till you're able to get to where you can answer it in a simple way. I'm going to have to do some more digging on some things now. Does that apply to relationships? Because I can't understand women at all. All right, we'll move into our rapid fire section. And this is going to be related to Excel. I know you use quantric sum, but we'll run through this for Excel. And this is, you have to pick a side. You can't say, it depends. And at the end, if you want to elaborate on one or two of your answers, you can so kind of make it fun. It's supposed to go quick, you know, just a quick one side or the other, 10, 15 seconds at most. So in your models, should you use circular or no circular reference? No circular VBA or no vba.

Yes, but not for calculation. Only for graphics user interface, this, that kind of stuff. But not, you should not, not never anything calculate in vba.

Okay, I've heard that before. You're not the first that has that opinion. Do you think the model horizontal? So basically, right, lots of sheets or vertical where you kind of build all your schedules for the most part, on one sheet, you have a preference between horizontal or vertical.

The question doesn't matter. Structure and display something completely different. You have to separate that. Even in Excel, if you don't do that, you're lost. Don't build your model on the layout.

Dynamic arrays in your model. Yes or no?

Yes. Yes.

External workbook links, yes or no?

No. No. No.

I usually get multiple no's on that one, named ranges. Yes or no?

Name? Yes, absolutely. That's the key to make an Excel model work.

I like it. Formal standards, do you follow them for your modeling? Do you like Fast or Smart or any of those others? No, and that's fine. You're not alone on that. Do you think Excel will ever die? Do you think it's with us forever?

It will die, absolutely. And that's what I want to elaborate on. Afterwards, please.

All right, great. We'll look forward to that. Will AI build models for us in the future?

No. It will help, but it will not. It will not. Definitely not.

All right.

I mean, if, if that's the last thing that will be replaced by AI. If, if, if, if AI can do that, then everything else is replaced before we're in trouble.

If it's doing that, is what you're saying. Do you think you should use sheet cell protection in your models?

No.

All right. Do you believe financial models are the number one corporate decision making tool?

I'm sorry, no.

What is.

How do you call this? I think in English you say napkin. In German we say the thing, the carton thing under the beer glass. The corporate decisions are taken by gut or by very stupid calculation on a napkin. And people are afraid of models, so they don't use them.

So you think back of the napkin gut feel. You don't think most decisions are really made with data and a model.

People are afraid. Even for a lot of clients, they always have a tendency. They make the decision, they tell me what the decision is, and then they tell me, can you please make a model that confirms that my decision is right?

I'm laughing because I know what you're talking about.

Then there is the process. The education starts and the successful processes end with a company which two years after they do it the other way around, they model first and take the decision afterwards. And projects end with the company wants to take the decision like they always took. And then I say, well then it's not worth it, I'm wasting my time.

If they're just looking for the model to confirm what they've already decided, then they're Just playing kind of corporate politics and a sense versus really trying to make good decisions. And we've all seen it.

An open mind and, and, and the research attitude. I don't know what the good decision is and I want to find it out. And then it's worth to build a model.

I'm, I'm fully with you. You really have to. You need a question and you need to be open. Got a problem? You need to be open to figure out what the answer is, what the right answer is or the best answer or at least the risks around the different options. Sometimes there may not, often there's not a best, but at least having an idea the risk being open minded because there will clearly be times that it will show you whatever you think is a good idea is a bad idea.

I mean you can still take the decision you want, but take please an informed decision. Seeing all options, seeing the advantages and disadvantages of the options and understand what your decision is about. And then you can still, I mean you're still the boss. It's not the model that will take the decision for you, but the model makes you understand and it makes the decision transparent. The problem is not everyone wants a transparent decision.

Good point. And I have one more question, then we'll go back to the Excel ever die? Because I know you wanted to elaborate on that one. Do you have a favorite lookup function you like to use in Excel? Like Vlookup Index, Match xlookup, choose something else. Is there one? You tended to tend to use some ifs. Okay, yeah, there's definitely some that use that one. That's a good one.

I mean normally I separate the model in two parts. Calculating, pushing everything in a flat database, you know, value and attributes and then making the analyzer section which is a dashboard that just uses sumifs and grabs the right results out of the database. So I mean you can build a database in Excel if you structured enough.

Oh yeah, no question you can. I agree with that.

And then some ifs works, works nicely.

Yeah, some of this is a fabulous formula especially for financial modeling. It's a, it's a favorite. So back to Excel because you know we got to start and end there. What's elaborate on when it's going to die or why you think it's going to die?

My year of birth is 70. Okay. So my father was a carpenter and he had little business, two people and he had this famous big paper book where he wrote everything down. Okay. Every, every transaction he wrote in this book with a pencil and made Little notes to it. And in the end of the year, there was the bookkeeping to be made for the taxes. So he gave the book to me and said, when I was six years old, seven years old, gilbert, can you please do the profit and loss statement for me? So I took the calculator and calculated everything, wrote it down and calculated again. And the trouble was, each time you type the thing, the numbers are gone. So this was a stupid way. And then some years later, I went to school and I made the first Apple computers, Macintosh with multiplan, which is like X, a little bit more stupid. And I discovered this, and it was, wow, a new world. Because you can write it down and you see the numbers and you have the formula. And for the first month and the second month, you can use the same formula. You can copy, paste. It's so nice, it's so clean. It's not the same thing with another tool. It opens a different world. And I think if you never did the profit and loss statement of an entire company by hand and then move to multiplan, you don't understand this step. And for me, going from Excel to a real modeling tool, it's the same step. It's another world. It's other possibilities. You can do the same thing. You can do it in Excel like you can do it by hand, but it opens up so more other questions and other possibilities. And I think the times where you have just, you know, to add up numbers to get the sum of this number there are gone. We are living in a much more complex world. We have data analytics, we have scenarios, we have data we want to cross with. We have risks. And this is another world, and it needs another world of tools. And I mean, I always say Excel is good for writing a shopping list, and everything that's more complicated than a shopping list you shouldn't do with Excel. You can, but you're really wasting your time and you're passing about a side of a new universe. You could open your mind much more in going to other tools.

Love it. I love that we get a contrarian view. Most our guests love Excel. I would definitely say I go beyond a shopping list, but I love getting to hear different people's opinions. And it will be fun to watch because we're definitely in a changing world where technology continues to evolve, and we'll. We'll watch it unfold. It's exciting. So I appreciate you taking the time to join me. And we'll go ahead, and I think this is a good spot to end and wrap up so just one last question for you. If our audience wants to learn more about you, get in touch with you, what's the best way for them to learn more about you?

Drop me a note on LinkedIn or mail and let's chat. 10 minutes. Chat. It's worth much more than private face to face chat. It's what I prefer.

Perfect. All right. Well, we'll put your LinkedIn information in the show notes and I'll have to reach out at you sometime and maybe we can do some kind of little demo on Monte Carlo. That'd be fun. But thank you so much for joining me. I really enjoyed it. Have fun on the island this weekend and happy boating.

Thank you very much. Paul, very interesting question. I really like to be on the show. Thank you. All the best. Bye bye.

Financial Modelers Corner was brought to you by the Financial Modeling Institute. This year I completed the Advanced Advanced Financial Modeler certification and it made me a better financial modeler. What are you waiting for? Visit fmi@www.fminstitute.com podcast and use code podcast to save 15% when you enroll in one of the accreditations. Sam.