What is it about Financial Modeling?

When I was first exposed to financial modeling in the late 1970's, I thought I had died and gone to heaven.  What could be more interesting and fun than something involving business and overflowing with math?  That involved light programming of a computer.  Where you could represent a business by a series of equations.  Load up history to provide a solid basis to demonstrate that the model works.  Make assumptions about the future and examine the impact instantly.  Provide real insight into the drivers of profits and cash flow.  Improve decision making when the stakes are high.  This is the financial modeling I grew to love and adore. 

Back then there was a problem, though, and that was the expense of developing and running financial models.  Financial modeling "languages" had just come into being, and you could rent time on big mainframe computers to use them.  Popular modeling languages back then included Profit II from Service Bureau Corp, and IFPS from Execucom.  (I learned a product called FCS while working in the Washington D.C. sales office of Comshare, one of the smaller computer timesharing firms).  Customers would reach what we called the "ouch point" of timesharing usage.  This is when the model is so successful that the bill grows large enough to be noticed by senior management.  I had customers spending as much as $10,000 per month for corporate planning models, or investment decision models.  We used to joke about the $1,000 carriage return for customers who ran Monte Carlo simulations which ate up CPU cycles then faster than a millennial changing radio stations during a short drive to the store.

This was the idea behind Ferox Microsystems: that financial modeling is great, but there had to be a cheaper way to get it done.  Well - that's the idea I like to tell people.  The reality is that when I left Comshare in late 1979, and bought my first personal computers - a Radio Shack TRS 80 that depended on cassette tape for storage, and then luckily an Apple II with real floppy disk storage - I didn't think the PC's were powerful enough to handle a fully featured financial modeling language.  I just wanted to fulfill my lifelong dream of starting my own company, and being my own boss.  I figured maybe I could do something simple like accounting software for small companies.  I was 25 years old at the time, and of course didn't know diddly-squat about what accounting software needed to do.  There were lots of things I didn't know, but I had a newly arranged $6,000 second mortgage, and a working wife - so at least I had a little runway.  I figured it was about 6 months when I first started off.  (It never got up that high ever again!  Such is the life of an entrepreneur.)  Almost 40 years later, I still have (the same) working wife, and (a different) second mortgage.   But what a journey in between!

What happened in the early days is something I know now to call a "Pivot" in  modern day startup parlance.  I was doing some financial modeling on a consulting basis to help keep the cash flowing in the early days, and kept hearing from customers, "can't you find a cheaper way to do this?"  I had been digging into the Apple II, and learned to program in Pascal.  A stroke of brilliance on Apple's part was the way they integrated the language, Pascal, with its own operating system that took care of memory management.  Without a lot of technical skill, you could write programs too big to fit in memory, and Apple Pascal would take care of loading into memory only that code needed to run at the time.  My early foray into earning money from my Apple II was to write a Pascal program to do cash flow projections and calculate an internal rate of return for a dog kennel expansion.  This was for a CPA friend of mine (and his dog kennel client) who worked for a Comshare customer, but did book-keeping and tax work on the side.  He knew about the value of financial modeling to do budgeting and evaluate investments, but also knew the cost was way out of reach for his client.  I took on the project to learn and earn at the same time.

The nice thing about using a financial modeling language vs. a general purpose programming language is that it is much easier to make changes.  Also, the number of lines of code you need to write to accomplish a given task is much less with a financial modeling language.  Back in the 70's and 80's, a financial modeling language was considered a 'higher level" language.  I think we called it a 3rd generation language - with the first generation being assembly language, and the second generation being general purpose programming languages such as Fortran, COBOL, or Pascal.

So to use a general purpose programming language on a "micro-computer" to do financial modeling meant that you saved a lot in computer costs, but at the expense of your time.  This was driven home to me about the 14th time my CPA friend asked me to add just one more column to the final report.  I figure I earned all of 50 cents per hour on this particular project.

But the real benefit was that I had to code up many elements needed by a modeling language - such as the ability to format numbers with commas, and decimal points;  use brackets around negative numbers; calculate financial functions such as the Internal Rate of Return.  I had created some of the building blocks of a financial modeling language.  I reached a point where I decided to go full bore, and see if I could prove to myself that it was feasible to develop a modeling language to run on the Apple II computer.  I was off to the races to develop what I now know to call an MVP, or Minimum Viable Product.  This was a race against time.  Every hour spent developing the product was an hour I couldn't sell, and chipped away at my shrinking runway of small business survival.

In the Spring of 1980, I got to the point where I felt I had something to show, and decided to put on an "open house" in my fancy new digs - the basement of a four story office building in Falls Church, Va.  We learned to keep flashlights  in our desk drawers at all times in the event of a power outage.  When the lights went out, it went pitch dark. I was down to my last $500 by the time the event was scheduled.  I invited my Comshare friends as well as several Comshare customers.  Even the bank manager from next door attended.  I forget his name, but he figures prominently in this startup story before too long.  I enlisted my mother and father to help with the preparations as well as the event itself.

And no, we did not have to deal with a power outage.  In fact, the open house was a resounding success.  I had three very strong expressions of interest from Comshare customers, and a bank manager who assured me his bank could provide financing.  What could possibly go wrong?

It turns out "a lot."  But a lot went right as well. 

And I found that financial modeling has been a constant in my career.  A tool for delivering high value to customers who need help understanding the complexities of their companies, and exploring tough decisions.  Who plan to  use the models over a long period of time to make plans, and monitor performance against plan. 

For me personally, financial modeling has been a godsend.  Financial models have helped me guide my enterprises through tough times and good.  I've been able to get the peace of mind of looking forward with confidence, and not just backward with fright.

Fast forward to the present.  I've landed as the CFO of Issuetrak, a company of 40 people my brother, Hank, founded 25 years ago.  I still can't seem to get away from financial modeling.  I'm getting ready to update the Issuetrak Corporate Planning model with May actuals, and re-forecast the rest of the year.  It's kind of poetic justice that I now am the customer and not the vendor.  I can see what it is like to use a complicated model over a long period time - and how important it is for the model to change and grow with the business.  I plan to use this deep well of real world experience in the latest incarnation of SurvivalWare, my financial modeling tool.

OK, back to the Ferox open house in 1980, and the bank manager from next door.  Not only did I get strong expressions of interest from prospective customers, but the Marriott Corporation committed to buying the software if I could provide them the hardware for 30 days to prove it worked.  With the printer, monitor, and extra disk drives, we were talking $5,000 or to make it happen.  The bank manager said no problem, and I arranged to deliver the setup to Marriott on a Friday.  On a Thursday, I got a call from a sheepish bank manager.  The loan committee had turned down my lousy $5,000 loan.  I had to scramble big time, and ended up convincing a computer dealer to sell me the equipment with 30 day terms.  The software worked, Marriott became my first customer, and life was good.  But to this day, I have an attitude about bankers.