Why I’m taking a break from teaching Financial Modelling
Excel is a beautiful tool but has it’s natural limitations. This is a message for anyone thinking about signing up for my financial modelling course at IBA CEE scheduled for April 15 2016
Financial Modelling and the usage of excel has defined my career to a great extent, I have taken pride in being able to use more keyboard shortcuts than most and I have thoroughly enjoyed building beautiful looking insightful dashboards that tell a tale which numbers can only dream of. Even on Linkedin, Financial Modelling remains the number one skill I am most endorsed for.
2 years ago I started thinking about models at a more deeper level, what are models after all? My conclusion was that a good financial model answers queries, but a great financial model guides the human brain towards more compelling and actionable insights. This series of questioning led me to meeting a guy named Osama Khan with whom I started discussing how the process of building financial models can be automated so that humans spend more time thinking and less time punching keys to make excel do their bidding.
This conversation led to Osama describing a world beyond traditional efficiencies and he wrote a little concept paper for me in which he described a story I really like, it goes like this:
“John reaches office and his attendance is marked automatically as soon as any of his wireless devices connects to the company network. The Autoflows Central Service running on John’s MacBook launches upon login to a computer giving him a summary of unread emails and tasks/objectives for the day flowing in from multiple sources.
John sees that his objective for today is to ‘make an initial draft for project titan deal slide deck’. An overlay checklist opens in front of John as soon as he marks the objective as ‘commenced’. The checklist ensures that John follows all company guidelines developed in line with best practices to conduct research on a given Company/Industry. Now John’s exact time spent on this task can also be logged. In addition he can access the company’s Knowledge Base powered by Autoflows which will also auto-suggest articles related to the current task at hand. Moreover, John also has the capability to post any insights he might’ve discovered during his research to the Knowledge Base.
John posts an updated version of his pack to the company’s Document Management System and shares it with his team for feedback via Slack from within Autoflows. John’s team can discuss and collaborate over the work and put together the final piece over Slack. Once the deck is finalized and posted John ticks it off his list, his supervisor gets an a notification in his Autoflows Central menu bar to review and appraises his work in the Performance Management system.
My conclusion from spending almost a year with Osama was that virtually all jobs will get automated. Nearly half of U.S. jobs could be susceptible to computerization over the next two decades, a study from theOxford Martin Programme on the Impacts of Future Technology suggests.
Therefore the days of Excel are numbered and thus Excel should be used as a tool to understand logic, to do some basic arithmetic and calculations, but for the really neat stuff, learning about SQL databases, programming languages like Python and even machine learning is the way to go. To me it’s VERY clear now that financial modelling or for that matter a lot of work that human analysts do will be done by computers within the next 3–5 years, and therefore in this world, teaching just the mechanics of financial modelling using excel is insufficient.
See this video below about Google’s AI, Deepmind beating the game of Go, a game that has more possible permutations than there are atoms in the universe!
Therefore, I have decided to take a break from teaching financial modelling using MS Excel. I will be teaching my last course at the IBA CEE on April 15th, I am hoping to go a bit beyond traditional modelling though. This is going to be tough because most of my paying customers are expecting a certain product i.e. learning how to build spreadsheets. Despite my numerous pleas, most of my past students still insist on getting “templates” or “examples” as opposed to learning the logic or the thought process behind building models.
Some useful links for this upcoming course are given below:
- Notes to my course and samples: These materials are dated and at best can be used as a point of reference
- Observatory of Economic Complexity this is sort of an economic model but very intuitive
- Modeloff: These guys organise the world championships of Financial Modelling!! I’ve competed in the first ever competition and lost. It’s got GREAT resources, especially the Find That Error problem is lovely
So the big message for those who are to pay me and IBA hard earned money to learn financial modelling:
“Financial modelling using Excel is a great tool, use it as an easy way to build logic, but PLEASE do not count on it and PLEASE keep evolving and PLEASE keep learning about how Artificial Intelligence, Machine Learning, Deep Learning and generally Data Science will effect the work you do today”
Looking forward to seeing you all and I hope I can teach you something useful. The mechanics of a model in excel or memorisation of formulae won’t be useful to you. Please look at the video above and think about how you will contribute to a world in which AI is widely prevalent.
Some random clippings below to support my spiel above.
