FP&A at Early-Stage Companies & the Benefits of Using Python to Automate Data & Enhance Analysis

It's just fun to build something new. In my mind, FPA has a few components. You need to be able to explain and interpret the history of the actual results that your business is getting and then use that understanding to project and predict the future. I was part of a team that repurposed that bank. We have really good data on our customers. We do really advanced statistics and forecasting on that data to produce some very interesting results. What if you don't have any historical data?

What if Todd Niemann is currently the Treasurer of Varo Bank. His career has included extensive work in finance, bank and investing. He has been part of FP&A teams nearly all of his career at firms both large and small. He has helped build out the treasury and FP&A functions for three new banks. He is a CFA charter holder and earned an MBA from BYU and a BA from USU.

I was able to create these deposit decay curves that were very accurate. I mean, I was like 90% accurate on predicting these deposit balances. There's literally times when analysis that used to take me hours will take five to 10 minutes.

If Excel removed one feature tomorrow, which one would cause you the most panic?

Probably the keyboard shortcuts. Again, if I had to go back to using the mouse, I don't know.

If I FP and A Guy here and I'm excited to introduce you to our title sponsor. Meet Campfire, the AI native ERP built for modern finance and accounting teams. Automate revenue, streamline your clothes, and centralized accounting all on one unified platform. Learn more today by going to Campfire AI that is Campfire AI. Are you tired of being seen as just a spreadsheet person? Will others get a seat at the table? Well then, welcome to FP&A Unlocked, where finance meets strategy. I'm your host, Paul Barnhurst, the FP and A Guy. Each week we bring you conversations and practical advice from thought leaders, industry experts and practitioners who are reshaping the role of FP&A in today's business world. Together, we'll uncover the strategies and experiences that separate good FP&A professionals from great ones, helping you elevate your career and drive strategic impact. Speaking of strategic impact, our title sponsor is Campfire, the ERP that's helping modern finance teams close fast and scale faster. Today's guest is someone who's earned that coveted seat at the table. I'm excited to welcome Todd Niemann. Todd, welcome to the show. Hey Paul.

Thanks.

Yeah, really excited to have you. So let me give up just a brief background about Todd and then we'll jump into things, so. Todd is currently the treasurer of Varo Bank. His career has included extensive work in finance, banking, and investing. He has been part of FP&A teams nearly all of his career at firms both large and small. He has helped build out the treasury and FP&A functions for three new banks. He is a CFA charter holder and earned an MBA from BYU, my alma mater, so I'll give you that one, and a BA from usu, my wife's alma mater. So you've covered both of them. All right. We always like to start with this first question. If I was to ask you, what does great FP and A look like, how would you answer that?

In my mind, FP&A has a few components that you need to get right. It's that you need to be able to explain and interpret the history or the actual results that your business is getting and then use that understanding to project and predict the future. So if you have models and forecasts and structures that will both explain and predict, that's a good start. And then a few other things. I'm big on speed. You know, any FP and a process that takes, you know, a week to get through is probably not that good. So it needs to be as rapid as possible, while also, of course, ensuring timeliness and accuracy.

Speed, timeliness, accuracy, good data. You know, I agree with all those, you know, understanding the historical kind of running the future. Can you share a time where you've seen great FP and A in action and kind of what was the result of that?

From every, every organization I've been with, you know, they all have strengths and weaknesses. They all did aspects of FP&A. Well, but I don't know if I could say I've seen one company or at least one of the banks I've been with that nailed every component. So I can give you a few examples. The first bank I was with, you know, we had a model that we could iterate on very rapidly. So when it came to running stress scenarios or just alternative growth scenarios, we could perform that analysis because how the model was built, the frameworks and things in a very good and accurate way and very quickly. So that was a strength there. My current bank, varo, has by far the best data analytics that I've seen anywhere. We have really good data on our customers. We do really advanced statistics and forecasting on that data to produce some very interesting results there. But yeah, those are different. I don't see if I've seen one organization that's able to combine all of it. The results are typically, I mean, what FP&A should lead to at least are better decision making. Right. Which rapid iteration will of course lead to that because you can go through more scenarios, you can get a better handle for the business and then of course, just better. More accurate forecasting is better for everyone. It allows your plans to hold up better for sure.

And I really like how you emphasize, look, nobody was doing it perfectly. I've seen several maturity models and usually there's five, six, seven different elements they're grading you on almost. Nobody's a five on all of. Right. You have to pick and choose. It's really hard to be great at everything. Usually it's cost prohibitive, it's time prohibitive and it's not necessarily needed. You want to be at least be good. You have to be ldl, do the things that need to be done. But there are definitely areas where each FP and A is going to have its strength. Speaking of banks, you spent a lot of your career working for early stage banks. How did that end up happening? I don't imagine you went into your career thinking, I want to work for early stage banks.

No, it actually is something I kind of stumbled into. So I did study finance for my undergrad and I knew I wanted to work in finance and I did lean towards banks in the early days because, you know, I wanted to be somewhere where finance was the core of the business, not just a port function like HR or legal and as it is in so many others. Not that there's anything wrong with that. Of course. Every business needs a finance org. But I wanted to be here where finance was, was really a core part of the business. So I really just got lucky to start out with. I met some exceptionally good people in my first the first bank I worked with that had decades of experience in the industry. You know how it goes. Once you get in with a certain group of people, you network and you know, one thing leads to another. So yeah, I've been part of three or the launcher establishment of three new banks in my career. Yeah, so the first one was Green Dot bank and I joined just a few months after it had been acquired by so rather Green Dot had acquired a small community bank. So they wanted to get a banking license, but at the time it was almost impossible to get approval for a new license just given the regulatory environment. So instead they found a small community bank in Provo, Utah, of all places, and they bought it out. And then once they had the charter, they could completely repurpose and redesign it for their, you know, for their new strategy. So I joined again just shortly after that acquisition occurred and I was part of a team that know repurpose that bank. So we had to rewrite all the policies, we had to build all the new systems and of course the balance sheet and the profit streams and things changed completely over the course of those few years. So that was the first one and then the second borrow was truly a build from scratch. So at that time we were able to submit an application to the FDIC and the Federal Reserve and the occ, you know, all the major regulators to get a brand new banking charter license approved. Raised the capital from private equity and venture capitalists and we opened in 2020. And then the third one was Genius bank. And this one there was an existing charter but it was highly focused on, on like commercial lending or business banking. But the company wanted to launch a new branch or a new division of the bank focused on retail banking. So I was part of that team that launched that new division. But we had new systems, new branding, new marketing, I mean new everything. We were just legally part of that existing bank. So each time it's been a little bit different flavor or a little bit different spin on it, but nevertheless I still count it as three times creating a new bank.

Amazing. It sounds like a really great opportunity. All different. And funny enough, Green Dot, I remember reading when they went public their statements pretty regularly because I was working for American Express in their prepaid division at the time, which had Bluebird and Serve, which were competitors, you know, along with other products inside followed Green Dot pretty close. I knew you had the bank there in Provo and all those type of things. A small world at the end of the day. You mentioned, you know, you really like being finance, being a core to the business versus kind of that support. Is that what you find attractive about the banking industry or what are some of the things you find attractive now that you've been, you know, working in it mostly for the last decade or so. What are some things you really enjoy about the space?

That's certainly one of them. And I'd add I enjoy being at the new startup banks and they have their challenges. We'll get to that in a minute. But I enjoy those better than I think a large established bank. It's just fun to build something new. And then yeah, I like, like we said before that finance is the core business or the core element of it. You know, it's all, it's all numbers. Our products are Money, loans and deposits. So there's a lot of math and a lot of analysis and that, that's always been attractive to me. I also like the sophistication, I guess of the analysis. Right. We do a lot of pretty advanced statistics and modeling and things. That just makes it so you always have to keep learning and growing to perform the analysis. And just to be fair, there's a decent number of things I don't necessarily like about banking as well. One of them being the regulation. And it's just so burdensome. You spend so much of your time just reporting on stuff to regulators and answering questions to regulators, which of course we need that we need some element of regulation. It's just, it's just maybe heavier than I, I would like it to be in banking, but nevertheless, overall I, I enjoy it.

I, I can relate to the regulatory environment. I remember a lot of things with, you know, we invested all our travelers checks. That's one of the products I supported. And so there was a $2 billion liability that we were investing. You could imagine all the requirements around those investments of what they have to be, where you can put them. And because if they go bad, you are all of a sudden on the hook for $2 billion DIC.

You know, if they're going to grant you that insurance that all the customers and depositors want, you know, they're not going to do it, you know, without some expectation that you're going to manage and you know, safe and sound is the kind of the tag phrase that we use a lot in banking. So yeah, the regulation's heavy.

I'm curious your take just in general with banking, do you think, you know, we had the whole 2008 crisis, it's still kind of too big to fail. You think we're in the right spot? You mentioned which regulation just kind of know it's a little off. And you could say, I'd rather not comment it, but I'd love just high level. You're kind of one.

Banking in some ways is, is great because it's so relatable in that, you know, everyone has a bank account or almost everyone has a bank account. We should say it's, it's a business that everyone's familiar with. But at the same time, I don't know if it's a business that everyone like loves. You know, it's more a part of your life that, that you need. You know, we need payments, we need this save money and hopefully earn a little interest. So, so it's something we use, but it's pretty rare to find someone that truly loves or passionate about the bank that they use. So that's sometimes a struggle. But to get more back to your question, banks have been changing a lot in the last few decades. The number of banks has reduced dramatically. I don't have specific numbers, but I want to say loosely, there used to be maybe 8,000 banks in the 80s and 90s in the country, down to around 4,000 banks or so today. So that number is declining now. Most of that's been through mergers and acquisitions because total assets in the banking industry, they're actually increasing still. So it's gone up. You mentioned too big to fail. I mean some of the. I've never worked at one of those banks that's that big. I've always been at the smaller or mid sized banks. But certainly the major banks like JP Morgan or many others of that size, they, they do play a critical role in our economy. You know, they, they do at times get some special privileges for that. I don't know, it just kind of is what it is, I suppose.

FP and A related. But as we got talking. So I'll get back to FP&A here. I'd love to know, and it may be different for each bank, but what are the key metrics that you really like to look at when you're assessing kind of the health of the bank and how you're performing? What are some of those metrics that are common?

Banking is a little different in that banking is a balance sheet business. Most businesses are primarily focused on the income statement or the P and L. Right. And of course we look at the P and L also. But banking gets just the balance sheet in banking gets a lot more attention and a lot more scrutiny than it does at any other bank. And there's a lot more regulation around it. We have to maintain certain capital ratios, certain liquidity ratios, certain coverage ratios. There's, there's a whole lot of analysis that goes in there. So anyway, so if you're analyzing a bank, place to start is of course the balance sheet. So I would look at the capital levels first off. And the other good thing about banking is there's standardized ratios that are consistent for all the banks and all of the data is publicly available. The fdic, they actually collect data every quarter from banks. And you can go pull up financial statements on any bank out there. Even if it's not a publicly traded bank, they still have to disclose and report all their financial statements. So it's very easy to get information about the Financial health of any bank. So anyway, I would of course start with the balance sheet, look at the leverage, look at the liquidity, and then of course look at the profitability. And so in banking in particular, we look at things like the net interest margin and then something called the efficiency ratio, which is a measure of cost relative to income or like a ratio of your margins effectively. And then we look at the standard return on assets, return on equity, those sort of things. So those would be the basics if you wanted to go a little more in depth. I'd start looking at the quality of the loan book, the charge offs and delinquencies on the loan book, how they're provisioning for the loans and things like that to get a health of credit quality or health of the bank.

I always find it interesting, right, because very few business, especially for FP and A are balance sheet heavy. Finance industry is the one. That's how it was when I supported Travelers Checks is it was a balance.

Sheet business, to be fair. We look at all of them. Again, we're looking at the income statement, we're looking at the balance sheet. We're also looking a lot at cash flows. So I look at all three statements quite regularly or at least some version. When we look at cash flows, it's not always in the accounting format of a cash flow statement, but it's in the spirit of where's the cash moving in and out of the business still.

Yeah, you still have to understand the cash position. Whatever method you're using to do it, it doesn't necessarily have to be a formal cash flow statement. It can be whatever makes sense for the business. I'm curious. So we talked a little bit about the metrics. What have you found? You mentioned regulation. I'm sure that's one. But what are some of the challenges? You know, continually working in an early Stage bank and FP&A and treasury, you know, kind of helping a build up. What have been the kind of themes as far as those challenges you've seen?

Yeah, one of the major challenges is lack of data. And this isn't just in banking, but this would really be in any new business that you're starting from scratch. You know, so much of standard FP and A, once you get up and running is, you know, looking at what happened last month or last quarter, last year and kind of using that as a guide for next month, next quarter or next year. Right. You based a lot of your forecasting on the past. Well, what if you don't have any historical data? What if It's a brand new product or, you know, a brand new institution that you're dealing with. How do you forecast in the absence of historical data? That's a challenge. So there's a few workarounds. You know, one, because banking is a fairly standardized business and like I said, financials from other banks are so readily available, you can often borrow some, some analysis from other banks and, and say, well, we think we'll perform similarly or a little better or a little worse, depending on what the metric or, or area might be. So you can kind of benchmark that way. It's possible to buy data. Actually you can get like anonymized loan data if you wanted to say, you know, we're thinking, just hypothetically if, if someone was thinking about launching say a, you know, a subprime credit card product, well, you could, you could get data from like TransUnion or, or some of the other, you know, credit agencies. They, they have data on, on how those loans have been performing and so you can use that also. And then lastly, you have to rely a little bit on management assumptions and experience. That's why, you know, having experience in the industry and hiring a team that, you know, has done it before and they're familiar with the space, it goes a long way. So I'd say that lack of data and solving around that is probably one of the biggest challenges when you're at a new bank, but not far behind. The second one I'd mention is just lack of infrastructure and lack of established methods to do stuff. There's often, even with the same data, there's often five different approaches that someone could take to analyze that data and to present that data. And often, especially with a new management team, you know, maybe you haven't, you haven't come together on an agreed upon framework for how you want to analyze and evaluate the business. So inevitably, always you'll hear comments from people like, well, at my last bank we did this, and at my last bank we did this from someone else. And you'd be surprised. They're not the same. They're often quite different approaches. So you have to think about it and say, well, you know, maybe that made sense for your last bank. You know, maybe it also makes sense for us or maybe because of XYZ differences in the customer or in the, you know, current economic environment or whatever it might be, we need to adapt that and establish a new framework or a new structure that we're going to use to evaluate this bank and kind of, you have to lay all that out in the beginning. So that when you start getting the data and it starts flowing in, you have a framework to evaluate it under. So those are both, those are both challenges of a, you know, working at.

A new bank FP and a guy here. And I want to share a brief message from our title sponsor, Campfire, the ERP that's helping modern finance teams close fast and scale faster. Finance is changing fast. Are you keeping up? Campfire is hosting the first ever AI Summit for finance leaders in San Francisco. And you're invited. Finance Forward will bring together 150 of the brightest minds in finance and operations, featuring finance powerhouse speakers from Snowflake, Vercel, Anthropic and more. You'll get expert led discussions, hands on sessions and real world playbooks designed to help you put AI to work immediately. Can't make it to San Francisco? No problem. You can sign up to receive exclusive tailored post conference content aimed at equipping you for the future of finance. Learn more about this can't miss event at Campfire Summit. AI, I think that's a common one you see in different, any industry that's similar, where you look at similar metrics, everybody does it a little different. And so when you have to align, it can be a challenge because, well, I like this way and I like that way. And if you get into the meeting and everybody's like, well, what's your assumptions? You know, you're not getting anything else done for the rest of the meeting.

And no, and to your point, that's often multiple meetings and you know, weeks of effort just getting people to understand these are the terms we're going to use, the acronyms, these are the formulas we're going to use, this is the output and you know, make sure everyone's okay with it. That takes a long time.

Right? It's like, you know, with Gap, we pretty much all know it's aligned. You can read the statement, yes, there's a few little different things you can do. Lifo, fifo and you know, but it's all there, it's laid out, there's a very strict framework. But when you get into non GAAP metrics, it's sometimes just the wild, wild west. And so you have to figure it out as a company.

Yep, exactly.

Before we move on, I want to talk a little bit about Python. I know you're a big, big believer in Python. You used it quite a bit in your work and I want to spend some time there. But before we do that, any, any thoughts you could offer? If we have somebody listening that's kind of interested in the banking industry or thinks they might want to, you know, work in it or learn more. Any advice you'd offer them?

Just general advice for getting started in banking. Yeah. Well, in particular in finance. Obviously. Banking like every other company has many divisions and I, I couldn't, I couldn't share much about how to get a marketing job at a bank. So I'll in fpa, you know, there's a few like core skills that I'd say are very helpful in banking and especially so one would be some accounting knowledge. And that's, that's really true across the board.

Right.

The more you understand the historical accounting numbers, the better you'll be at forecasting them in the future. So that, that's a, a basic principle. Another one would then be data. We often have to deal with very large data sets, whether it's, you know, customer data or economic data or you know, whatever else. So being, you know, adept at using your statistics, running regressions, understanding quality of data and how to, you know, how to clean it if it's not good, all of those things so that you're getting good data and analysis. So, and I'll put all of that into probably like statistics category. So accounting would be one, statistics another and then finances is another category. And I separate that from accounting because what I mean by finance is like finance theory, things like time value of money, duration of assets and liabilities, you know, the MPVs and IRR CLCs and things like a lot of us, sure, you know, we take corporate finance in school and you know, we hear these ideas but a lot of us maybe aren't using much in our day to day and especially for a lot of income statement focused bank or other companies, they might not be as heavy into that sort of analysis. But again in banking you are so really understanding the time value of money and those ideas and how they apply to your bank. That's critical. So that's someone starting out their career or if you're still a student, make sure you get a good grounding in all of those disciplines and that would lay the groundwork. And after that of course, it's just networking. Fortunately with banks, there's a lot of banks, a lot of people that work at banks, so it should be pretty easy to find some people that you can talk to that will kind of show you the robes, help with an internship or you know, entry level job or something. And from there of course like anything else, it's just all performance based. If you do well, people recognize the Value you bring, you'll, you'll likely succeed.

I like that part about, you know, the performance based. Right, you got to have the skills to get the job. You get in there and then you have to show that you're, you can do it and the rest will take care of itself if you're a, if you're a good team player and you work hard and do the right things. All right, let's talk a little bit about Python. I know when you and I grabbed lunch here a few weeks ago, you talk quite a bit about, you know, you're using on Python a lot to work with and analyze your data. So how, first let's start. How did you get started using Python? I know that's something most people don't gain coming out of school. Python experience. How'd you get started?

Yeah, no, exactly, I didn't, I didn't learn any programming at school. In fact, of course I learned a little Excel during school, but really I learned a lot of Excel once I started working in banking. And what I found is, you know, Excel is, Excel is amazing. I don't want to say anything negative against it. It's certainly the workhorse. And you know, 90% of what we do in FBA still is excel based and probably will be for a while. I mean, we'll see how some of these new AI tools come up, but for now it's the workhorse. But what I found is there was a lot of analysis that either I couldn't do or that I just kept thinking there's gotta be a better way. You know, reading around and talking to people. Python quickly stood out as something that I felt would be very powerful and I just honestly started learning on my own. I'm totally self taught. There's a few good resources I could recommend that I use. So there's some online courses, just simple stuff, you know, from Coursera, it's free, or a good website called Kaggle. It's K A G G L E Kaggle. It's actually owned by Google, but it has some free online courses or some sample data sets and things and you just start, you know, start experimenting to learn the very, very basics. I actually learned pretty well from reading textbooks. I developed that, that skill during, during my educational days. So I also just got a few books and, and you can again just search for that, say, you know, Intro to Python coding books and, and you know, free PDFs or something. It'll, you'll get some resources that come up and. Well, none of them were exactly what I needed to say, do you know the financial analysis for Banking particular? None of them were that close. But nevertheless, they taught me the very basic skills that I needed to understand the syntax and some of the libraries and kind of the structure of the Python code. And once you get the very basics, which again I think someone could do in a couple weeks if they focused on it, you get those basics and then it's just trial and error. At least that's how it was for me. I'd think of what project I thought would be the most beneficial or what I was most excited about and I'd try, I'd get started and try to figure it out. I had a lot of Googling answers and asking questions and maybe AI can help there. I've used a few of the AI tools now, like Copilot from GitHub and things they're very good at helping you correct your code and especially finding small little syntax errors where you miss a comma or something tiny like that that we've all done well, instead of having to read over it, the AI can identify and correct that for you really, really quickly. So it's probably much easier to learn now than when I was trying because I started learning back in 20, probably 2017 or so, so about eight years ago now anyway, so that's kind of how I got started. But then what, what happened is I kept finding more and more use cases for it and it's something that I'm using daily now for multiple different analyses. And personally I feel it's been just tremendously beneficial with the time savings and the accuracy and in some cases there's like new analysis or doors it's open to us that we really couldn't even do before. We couldn't even perform because of the limitations of Excel or other data systems. Whereas before we might have had to bring in someone from your data science team or someone like that that specializes in that. Well, there's a lot of that work that I can just do on my own now or, you know, or at least get a start from them, but continued on my own. So it's increased autonomy and speed tremendously for me.

Thank you for sharing, that's great. And when you mentioned, you know, AI helps a ton with syntax, right? Yeah, I've found that true. Whether it's Excel, you know, SQL, whatever you're doing, it's great with syntax. And one of my favorite shirts I saw and I, I wrote code for about two years, SQL script, so, you know, not really code, but ask similar type. I did that for about two years in a report writing job and Slasher at one time said, roses are red, violets are blue, you have a left bracket on line 42. And it's just like, yes, that I think I've spent hours trying to find that one thing. And then you want to kick yourself when you finally find it. In your view, what, what are some of the tasks that you think are kind of really tailor made for Python that we're often using spreadsheets for? What are some of those things that you think have really, you know, been beneficial? Not so much the things then we could talk a little bit about some of those things that you couldn't do with a spreadsheet, but what are some things that maybe are traditionally done in a spreadsheet? You're like, yeah, I really should use Python for that.

From your perspective, a few different areas. First would be anything that's really routine and you know, there's a lot of FP and a work that can get pretty routine where you're downloading from a couple data sources, crunching a few numbers and merging them into, you know, different Excel workbooks to, to create some output, whether it be a chart or table, you know, whatever it may be. And Python can automate most of, of that sort of thing. So a lot of the data sources that we use have API calls that you can, you can tap into. So to be more specific, like we use NetSuite and there's a NetSuite API. So I can download any accounting information I want directly in, into Python. So that's one example. So like for your, your monthly update, it's common for people to download the latest financial results. Then you gotta, you know, just summarize and format them in the right way because you're not, you know, you're not forecasting every single line item on the income statement and balance sheet. But there's, you know, there's categories that get summarized. So you do that and then you copy and paste it usually into your, you know, your forecast and it updates well, because that's so like routine. You could have Python download all the, the, you know, all of the accounting information. Then you can have Python summarize it in the exact right groupings that you want and you can have Python output it in a table that is very clean and tidy exactly how you want it. And it'll do all of that in about, you know, a minute or two because it's so rapid. So instead of, you know, whatever, maybe an hour update process, you're down to three Minutes and one copy paste and that's it, and your monthly update is done. So things like that, it's super, super valuable. The other one would be combining multiple data sources. So there's times where, you know, you got to connect to netsuite to get the accounting information. But, you know, we also want to connect to our customer database to, you know, get what, whatever metrics we're interested in there. Maybe we want to pull in some economic data from, you know, the Federal Reserve or something, or some competitor data. And I can have Python connect to all of those data sources, merge them all together and even perform a lot of the analysis that I need. So instead of again having to manually go through each of those steps, it's automated and there's literally times when analysis that used to take me hours will take five to 10 minutes. Now, of course, it did take me a couple days to build out the code initially to do that. But the payback on some of this I found is super short. I mean, within a month or two, you're time ahead.

So it's amazing how often people we've all been guilty of. You hold off automating a process because it's going to take you a long time to do it.

Yeah.

But once you do it, the savings quickly add up. Yeah, I've seen that in my own work more than once where it's like, why didn't I just fix this? Those are some great examples. I appreciate, appreciate you sharing that. What are maybe some of the things you've seen where you just couldn't do it in a spreadsheet? Some analysis really help you, I imagine. Big data is one area, but are there others?

Yeah, so the big data is clearly the, the most obvious one. You know, Excel has a limit of about a million rows, which is, you know, great for most things. But, you know, I've done analyses with up to a billion records or 2 billion records in Python and it'll, it'll handle all of it, no problem. So, yeah, anytime you're dealing with very large data sets, and I have an example of that I can go into later if you'd like. But. So that's one. The other one is like doing, you know, Excel is great for, for simple things. You know, if you're adding and averaging and making little charts and stuff, it's great. But Excel is sometimes limited with the robustness of some of the statistical analysis that you can do. So I find there's some of that that you can only do in Python or only do it good in Python. And then third, kind of going back to the automation, I guess, when you automate things, it's amazing how you can combine things in ways that maybe didn't occur to you before, or just simplify analysis, that you thought you had to take 10 steps. And it turns out if you put in Python, it's five steps and it performs it automatically for you. So it doesn't, you know, it doesn't take near the time that it used to.

Got it. Yeah, makes a lot of sense. You said you had an example you could share, kind of big data you did with Python. I'd love to hear one in again.

The bank, we're focused on our balance sheet and one of the most important things are all of our deposit or information. And there's an analysis that all banks have to perform that we try to estimate the length of time that the average depositor keeps their, their money with, with the bank. So, you know, you open an account, let's say you put $1,000 in or something, how long on average can we expect that thousand dollars to be there? Because you know, the bank is going to use that to fund the loans or whatever other operations that, that we're doing. So like when I got to a few of the banks I've been at, actually they just weren't doing this or they, they outsourced it to maybe a third party, a consultant, or they had the data team doing something. But what the data team was doing, it wasn't exactly what we wanted, but it was kind of just what we thought was feasible. So they were doing a lot of compromising and really cutting corners on that analysis just because they didn't feel like we could do it or they didn't know how at least. So anyways, I was able to come in and because of Python and some of these APIs, I could connect to our customer data set and download every single customer's balance every single day. And that's what I was talking gets to several billion records pretty, pretty quickly. But then using Python, I could group it by certain types of customers. I could take averages for the month and create kind of cohorts and curves and things to really reduce that data. And then we augment it with economic data, whether it's economic interest rates or unemployment or whatever the variables might be. And once all that's in Python, with those variables added, I can fit a regression. It was a log linear regression that we used to estimate what that decay curve looks like. And it's similar to like a, a half life of, of you know, you know, anything in the, in the natural.

World, you're like a chemical or whatever. A half life. Yeah, yeah, yeah.

The math is similar to this and that goes to the finance theory I was talking about before. You actually have to understand the statistics and the financial theory, like why do we want it to work out this way? But anyways, at the end of the day I was able to, to create these deposit decay curves that were very accurate. I mean I was like 90% accurate on predicting these deposit balances, at least on the, on the historic data. There's always the caveat that historic data and future data are not going to always align. And again, best of all, it was extremely fast. So this is an analysis that we really couldn't even do properly before that. Now I can update anytime with near real time data because it will pull in up to yesterday's data from the, you know, the customer database there's. And the whole thing can be run in about 15 minutes. So it gave us much better quality of data analytics for estimating those, again the average life of those deposits, which was a key aspect of our balance sheet that was really helpful for us to understand.

That's a great example. And yeah, billions of records, all those curves. Statistical modeling.

Right.

That's just not something that Excel, a traditional spreadsheet is designed for it. And the time it would take, you could imagine writing all the formulas. You're trying to run a formula on a billion rows, it'll probably crash.

But yeah, if it does run, it's going to take a long time.

Agree. Right. It's not a, not just not a viable solution. So I appreciate you sharing that one. I'm curious, as you look back, you mentioned automation, different things. What would you say has been the one or two most significant benefits in your career for using Python?

Well, the most significant would probably be the time savings. Again, like imagine a process that takes, you know, two people, three days to update. Well, if you could automate very large chunks of that, and now it takes one person half a day to update. Just again making this up. Like those are some huge savings both in, you know, the team's time, but then also the cost, you might be able to reduce the size of the team or expand the scope of work that the team is doing. So I think the biggest one is the time savings, but pretty close behind it again I'd say is the quality of the analysis. In many cases it's analysis that we couldn't do or we couldn't do as well before. So you're getting both better analysis for less time and less cost. So that's a no brainer, I'd say.

And you mentioned this a little bit before for, I think you said Coursera, Kaggle Books. Any other resources you'd recommend for somebody listening that wants to get started with Python?

Nah, those are, I mean that's how I learned. So that's, you know, where I go first. I would recommend like GitHub Copilot that we talked about before because that will really simplify a lot of those rookie mistakes that you make early on. It'll ease the learning curve. The other one though, I'd say is, you know, a good friend or colleague that knows this. There are times when there's features or functionality to Python that you know, I didn't even know existed or I didn't even know were possible. But in having a conversation with someone on our data science team, they would just kind of off the cuff say, well, have you tried this or have you tried that other thing? And you know, upon trying it, it turned out to be the perfect solution. But you know, you don't know what you don't know so much of the time. So an experienced guide really goes a.

Long way too, I could imagine. Especially, I know there's a lot of different libraries. Sometimes you may not be aware of what a certain library can do. I can try downloading this library that's better for statistical than what you're doing or whatever it might be.

That's exactly right. The libraries are always shifting and then there's, there's new libraries being created. I mean, and Python is not my full time job, right. I'm, I'm running the other treasury functions. I just use Python as a tool to make my life easier. But it's very possible that they've, you know, released a new version of the code or they've added a new feature or a function that, you know, I didn't know about. But once you learn about it, it's hugely valuable. So yeah, someone that can help you kind of stay abreast of the latest updates in that world is very helpful.

Always good to have somebody. It's like everybody goes to the Excel person in the office, right?

Similar type thing, the same thing. And you know, you mentioned that I found that once I started learning Python, you'd be amazed at how many requests you get from people. I get more questions on Python than Excel these days because people are all the time asking, well, could it do this, could it do that? Can you help me with this? And I'VE become someone that's teaching the rest of the finance team now often how to, you know, how to use it.

It's funny. Yeah. Once you become the expert, all of a sudden, everybody comes to you, which can be a blessing and a curse. But that's a separate discussion.

Yeah, that is a separate one.

So I know you're an avid reader and I know you've often shared books on LinkedIn that you've read. Do you have a favorite business book or one you'd maybe recommend to our audience? You know, kind of being an FBA finance audience, that's hard.

I do read a lot and I read widely. I mean, I read, of course, business finance books, but I'm reading history, I'm reading politics, I'm reading religion, even a fair amount of fiction, like, I read extensively. I don't know if I could narrow it down to just one, but I could maybe give, like, categories or authors. So when it comes to just great stories of finance, you can't beat Michael Lewis. Of course, everyone knows Michael Lewis, but he's got some great stories, whether it's the Big Short or Flash Boys, these are just classic stories that are so well written and fun to read. So the. He's. He's, of course, one of my favorites. Another one would be Nassim Nicholas Taleb, if you've read his stuff. So Fooled by Randomness or Antifragile, for example, though he has others. Taleb is very contrarian in his thinking and it bent my brain to read his stuff because it is the opposite of what you hear in so many places. But the way he explains it, you're like, well, of course. That's right. And it really opens your brain and help makes you think ways that you've never thought before. So I highly recommend his stuff as well. One of my old classics is always Stephen Covey. I thought his stuff is. And it's not finance specific, of course. It's more just strategy or even life in general. But his stuff I found was just really, really helpful. It keeps you grounded. Oh, I will put a plug in for one other that I read recently. It's called the Five Types of Wealth by Sahil Bloom. And again, it's not a finance book. It's more just a life strategy book, like how to Be Happy in Life. But it was really good.

It's a good one. Thank you. Appreciate that. All right, so we're going to move into kind of a standard FP and a section where I ask a few different questions. We ask similar ones to every guest. First one is, what is the number one technical skill that you think FPA professionals need to master?

You know, earlier we spoke of what I feel the. The core ingredients are. You got to have accounting, you got to have statistics, you got to have finance. So assuming all of those are. Are behind us, I'm going to go back to where I started with speed It. You can often give two analysts or two people on the team the exact same task, and it takes one person, say, eight hours and the other person two hours. That's a big difference. If you can be the guy that can get it done in two hours, like that is a huge value add to the organization. And there's a few tricks to that is one is just practice getting familiar with it. But, you know, there's several of us don't use the mouse. In Excel, if you can learn the keyboard shortcuts to just rapidly perform analysis that I found is really, really, really beneficial. So that's probably it. Be fast. Use keyboard shortcuts.

What about that soft or human skill?

Soft skill. So frankly, this is not my strength. I often struggle with these soft skills as well. So I have sympathy to those who struggle here as well. But, you know, to that point, maybe, maybe patience. It a lot of the time in finance, you know, we look at the numbers. We're good with numbers, we understand numbers, and we look at the results of the analysis, and it's just clear as day, you know, black and white, we understand it. But you present that to someone else on the executive team, and often they don't get it or they're not seeing it clearly. And so being able to explain things and educate people and kind of bring them along the analysis in a way that they understand it and they're supportive and you're building that cohesive team that really is a talent. It's something I wish I did better.

It's definitely one I've had to work on developing myself over my career. So I can relate to that of having that patience be able to explain things simply and help influence the outcome. Right. They really kind of all go together and you need all three to accomplish. Ability to simplify, ability to influence. And really having that patience, which I've often not been patient. And I remember more than once where it showed on my face and I didn't realize it. I've had to work at like, okay, all right, if Excel removed one feature tomorrow, which one would cause you the most panic?

Probably the keyboard shortcuts. Again, if I had to go back to using the mouse. I, I don't know if I could.

And I think there's a few people like you. I, I still use the mouse. I'm not, I, I can use a keyboard and I use it quite a bit, but I think I'd be okay. But I can think you and a lot of other people would panic if that happened, right? If you could wave a magic wand and change one thing about fpa. So you're, you're keen for a day, whatever you want to call it, what would you change?

Variance analysis is one of my biggest pet peeves. I would do away with it. I mean, mostly. So I've been at some organizations where, you know, of course you have to do your standard month over month and maybe your, you know, actual versus budget, of course. But then I've been at some places where they want quarter over quarter and they want year to year over year and they want year to date versus year to date, and then everything versus the plan and then everything versus the prior plan and you end up doing like 12 different variance analyses. And I, I don't think they add value. I mean, again, the simple month over month and versus plan, actual versus plan or budget. Some simple variance analysis, sure, beneficial, but most of the time I feel people take it way too far. So I would, I would eliminate, reduce, or, or change that.

Funny story, my favorite on a variance analysis. One time we had a budget that our corporate decided to completely change it in the middle of the year and we didn't know what they did. I couldn't understand the numbers. I finally just commented in my variance commentary. I have no idea what's driving it because I have no idea what you did. I don't understand the numbers.

Did they respond well to that?

I was, I had a very good reputation and I had worked hard on the rest of it, so they just let it slide. They just live with it because they, they knew I wasn't slacking. And it was just an honest, like, I, I can't make heads or tails of how you came up with this number. So I can't tell you what the variance is. So they, I'm sure they weren't happy. I can't imagine when they first read it, whoever read that report was like, oh, thank you, this is helpful.

Again, one, one or two variances I, I get. But when people are trying to perform 8 to 10 or 12 or whatever number of variances, it's like, what, what is that going to show you that, you know, you didn't get from the first one or Two. It's just, I. I'd view it a total waste of time.

We're going to move into our get to know you section. I just have a couple questions to get to know you a little bit better. So what's kind of favorite hobby or passion? What do you like to do with your free time?

Well, we spoke already about reading. I. I do a lot of reading and I. I love it. That's. Although most to be, you know, transparent. It is mostly audiobooks these days. I listen while I'm at the gym or driving and things, but I still count it as reading. But beyond that, I've. I've recently gotten into mountain biking. So I bought a bike last year and started exploring the local trails. It's been. It's been a lot of fun. And then this year, my. My son actually joined the high school team. They have a mountain bike team, and I signed up to be an assistant coach or ride leader with them. So I spend, you know, at least two, sometimes four or five days a week mountain biking, either in the mornings or in the evenings. But it's. It's become a big, big part of my life, and I. I'm enjoying it.

Fun. So we got the reading and the mountain biking. All right. If you could have any one person's job in the world for a week, who you picking and why?

I'm gonna go two ways on it. One, I would say, like, you know, Trump, just to see what it's like, you know, to be there. But I also don't think I'd actually enjoy it. I'd just be more curious to see how it goes. But if I look at who probably has the most fun in their job, it would be like Elon Musk. I would love to just not necessarily have his job, but just follow the guy around. I mean, from cars to rockets to, you know, brain implants to AI. I don't know what a day in the life of Elon would look like. So that would be fascinating to.

With all the different businesses he has, or even like a Jeff Bezos with all the different things he's. He's had going on with. It would be really interesting to see what's like, you're like, I can barely manage my one job, and they're like CEO of like, you know, especially Elon, like, four companies or whatever. Like, am I missing something?

Yeah.

So, last question. If you could have dinner with one person, dead or alive, who would you. Who would you pick?

It's hard. All your questions limit me to just one. Right I'm gonna, I'm gonna go with two again. But it's the old partnership of Warren Buffett and Charlie Munger. You know, Munger passed a few years ago. Warren's still around. But, you know, I think like most of us, I've, I've. Anyone in finance, of course, has heard of Berkshire Hathaway and Buffett and Munger, and I've been a follower of theirs for years, have read their books and things and just been so impressed. So I would love to sit down and have dinner with those two if it were possible.

Not surprised to hear you say that. I think a lot of people have a ton of respect for, for them. How can you not?

Yeah, of course.

All right, last question. If anyone wants to contact you, learn more about you, you have kind of things you do. Best way for them to do that.

Probably just LinkedIn, you know, I'm there. Reach out. Happy to chat.

Well, thank you so much for joining me today, Todd. I've really enjoyed chatting. It's been a great conversation and appreciate coming on the show.

Oh, thanks, Paul. It's been fun. I appreciate you making time for me as well.

If you enjoyed the show, please leave us a five star rating and a review on your podcast platform of choice. This allows us to continue to bring you great guests from around the globe. As a reminder, you can earn CPE credit by going to earmarkcpe.com, downloading the app, taking a short quiz, and getting your CPE certificate to earn continuing education credits for the FPAC certification. Take the quiz on earmark and contact me, the show host, for further details.

It.