The post Microsoft Ignite 2018 Excel Session Summaries appeared first on .

]]>- Dynamic arrays to replace array-entered formulas ー a pretty big paradigm shift in how you will write formulas (start getting used to the #SPILL error message 🤣)
- To make this more complicated, cell reference in dynamic arrays will contain hashtags (#)

- Get and Transform your dat in Power Query is super powerful when combined with the Flash Fill feature ー predicts how you want to transform your data and let’s your
*preview*the transformation*before*you actually commit it - Reduced file sizes due to dynamic arrays (only write formula once) and optimized lookup formulas 🙌
- New data types seem interesting only if you’re a stock broker (tickers) or economist (geography/census data)

- Access anywhere, similar interface, working with the latest version
- Shareable link with permission setting → again, Google Sheets anyone?
- Threaded comments and @mentions
- Simplified ribbon toggle, thank God!
- Can create data validation and PivotTables now in Excel Online
**Bottom line:**Google Sheets had these features 5 years ago. seemed like it’s just to get parity with Google Sheets

- Introduction of dynamic arrays is a pretty big paradigm shift for Excel users
- Formulas can “spill” into other cells with a blue border around the spilled cells denoting which cells are in the “spill range”
- A little bit like array-entered arrays with CTRL+SHIFT+ENTER because you can’t overwrite a formula in the spill range → not all cells in the spill range have a formula
- Spill range referencing → instead of doing H5:H8, you do H5# to get a dynamic reference (crazy!)

- 7 new functions:
- FILTER – Take data from your table and filter on a certain dimension. Results in spills vertically and horizaontally
- UNIQUE – Unique data set from table
- SORT – Does what you expect it to do. FILTER and SORT part of calc engine.
- SORTBY – Used in conjunction with RANDARRAY to randomly sort a list of values
- RANDARRAY – Good for Monte Carlo simulations. Random values by x rows and y columns.
- SEQUENCE – Generate a sequence of numbers from 1 to n
- SINGLE – Didn’t cover in the video

TL;DR: If you are tied to a multi-year Microsoft license, makes sense to utilize all the tools in the ecosystem. Given modern ETL pipelines that rely on open-source tools, the Microsoft ETL ecosystem feels pretty closed and not interoperable with other platforms.

- The obligatory data-driven quote to tee things up: “Customers want to gain business insights to make better data-driven decisions.” 🙄
- Lifecycle of crunching data to sharing analysis with your team: Excel → Power BI → Office 365 (e.g. Teams – Microsoft’s version of Slack)
- Pretty cool to see your Excel files and Power BI reports inside of Teams
- If you see a chart from a Power BI report that you want to drill down into, you can open the source data model from the cloud inside Teams
- The Power BI report was created through Power Query Editor, feels like a more user-friendly version of writing macros
- As a data modeler, you have to know when seeing data in Power BI is better than Excel (and vice versa) → basically dependent on the number of rows of data
- Cleaning data in Power Query is similar to cleaning date in OpenRefine

- Good tip: “Keep data, calculations, and visualizations separate” → foundation for building any web app 👍
- Questionable tip: “Bing is your friend: search and you will find.” → I get it, it’s a Microsoft conference
- Flash Fill definitely gets rid of the need to do text manipulation involving CONCATENATE, MID, FIND, LEFT, and RIGHT
- New data types have been hyped for some time which pulls external data from the Microsoft Knowledge Graph
- While pulling data into Power Query looks easy, there are multiple join types which can be confusing if you don’t know SQL
- Once it’s set up though it’s a great substitute for VLOOKUP as it’s much easier to reconcile data and find rows that don’t match between two tables
- Probably makes sense when you have hundreds of thousands of rows. If it’s a small dataset, Power Query is overkill

- The Dynamic Calendar Pattern is pretty neat where you use Power Query.
- Allows you to create a self-updating table instead of creating a static table where you’re incrementing days one at a time with a formula
- Still a lot of data cleansing going on just to get a dynamic range of dates

- TL;DR: Transform and Flash Fill are probably the coolest features to help you cleanse your data and do some basic ETLing
- Excel is on 500 million devices worldwide each month, 1.5 billion file opes per day, 20,000 formula commits per second (!!!)
- Insert data from picture is a new paradigm where you don’t do data entry, but rather fix data issues with the table returned from the picture (since it won’t be perfect)
- Dynamic arrays actually reduce the file size since you’re only writing one formula that is “spilled” to other cells
- No more dragging and filling down your formulas to the last row in your data set

- New data types are geography and stocks; performance on loading 500+ stock tickers was pretty fast! You just enter in the value like any other value but as long as Excel knows the cell is one of the new data types, Excel “converts” the value to a rich data type
- Pulling external data via “Get Data” from a 3rd-party has increased the number of connections including MySQL db, Postgres db, JSON, SAP HANA, etc.
- The “Transform” feature combined with Flash Fill makes it easy to create new columns of merged data for novices with the “Column from Example> in Power Query. You can then see the actual Flash Fill formula that results from the smart autofill
- Performance improvement on lookup formulas! Showed an example of a spreadsheet that would take ~9 minutes to calc with all the lookup formulas that got cut down to 3-4 seconds with the new version of Excel
- Made a comment about how the saved time leads to better energy savings overall

- Working on private views!!!
- Discussed the “strong GitHub integration” where you can put in the URL of GitHub link to get a custom function from the repo
- You can stream real-time data into Excel 😃, showed a video of an Office customer in the inventory tracking business showing data getting streamed into Excel

- Really great point about tools adapting to humans instead of other way around;
- Today: Expect
**people**to learn and adapt to software and tools - Tomorrow: Expect
**software and tools**to adapt to people, and accelerate their job to mastery

- Today: Expect
- Interesting question about new data types: what does it mean for two data types to be equal to each other?
- The ticker data type is an object consisting of key → value pairs
- When comparing two MSFT ticker objects together, for instance, you’re comparing
*all*values of the objects together - Looks like the common pattern here is you’ll need to constantly refresh your data to ensure you’re not comparing old data with new

- Context for new data types is important for whether Excel will convert the value to a new data type
- Will Ideas make lives easier for analysts? Seems like this AI feature generates nice charts, but perhaps it will merely inspire analysts to create the charts they
*really*want to create? - Looks like formatting doesn’t “spill” over with dynamic arrays, need to tweak your formatting around the “largest expected size”
- No more A1:A20-like reference with dynamic arrays, you need to start using the hashtag (#) ー funny that hashtags typically denote errors in Excel

The post Microsoft Ignite 2018 Excel Session Summaries appeared first on .

]]>The post Modeloff 2018, The One and Only Financial Modeling Competition appeared first on .

]]>*Source: Modeloff*

The first round is October 6th, but unlike previous years, the competition will be much more accessible and enjoyable. Here’s what I like about this year’s changes:

- Access Round 1 questions anytime within a 24-hour period, and you have 12 hours to complete the questions
- There is an “open” Round 2 where anyone can try and answer the questions, but you won’t be considered for the Modeloff Finals
- Enter as a team (up to 6) in either Round 1 or Round 2 (but teams aren’t eligible for the Finals)If you want to have a chance to make it to the Finals in London, you still have to go through the Round 1 and Round 2 “official” path, but for newbies to Modeloff, the Open rounds are a good way to go. See this link for the full details about these new formats. These new formats should make you feel:

In the past, Modeloff definitely felt a bit exclusive, meaning only the best financial modelers should be able to enter and compete. At the core of the competition, this makes total sense, and the best financial modelers in the world *should* still compete since there is still a path to the finals. As an observer and learner, the opportunities to participate in Modeloff is at the very end during the Finals event and GTC (Global Training Camp). The Global Training Camp has become a production in itself, giving up and coming financial modelers an opportunity to learn from the best and brightest in the industry.

Observers also have the opportunity to review previous questions on the Modeloff website, but this is usually a few months after the competition takes place. To capitalize on the excitement around the actual competition starting next week, I believe being inclusive of all skill levels is the right way to go and should encourage more participation in GTC and future Modeloffs. So if you are just learning about financial modeling or on a team of analysts, enter the Open rounds and see how you stack up with the rest of the community!

Although I’m working at a place that is re-thinking the building blocks of spreadsheets and docs, my first love will always be Excel. How often do you see “first love” and “Excel” in the same sentence? Some of you Excel nerds out there feel what I’m saying :).

I’ll be entering the competition this year for the very first time, and looking forward to flexing my financial modeling muscles which I haven’t done in *years*. It should be a very humbling experience, but I think it’s good to stay on top of your Excel skills no matter how much or little you use the tool at work or in your personal life. What I’ve gained using Excel has carried over into work with databases, web development, and programming. Hope you join me next week!

The post Modeloff 2018, The One and Only Financial Modeling Competition appeared first on .

]]>The post International Financial Modeling Competition Modeloff 2017 Recap appeared first on .

]]>That was then, this is 2017. This time around, I had the opportunity to attend the Global Training Camp in NYC which is a 2-day camp done in conjunction with the Modeloff finals. The topics covered during the training camp included Modeling Best Practices, LBO Modeling, Data Visualization, and more. There was also a track of classes taught by professionals from the Microsoft team. All in all, it was great to be a part of the community of Excel enthusiasts of all ages learning together.

NYC Global Training Camp begins!

Andrew Grigolyunovichs class on corporate budgeting techniques.

Awards given out during the finals

Everyone getting ready to see who wins Modeloff 2017.

ModelOff Finals testing under way at Microsoft New York! pic.twitter.com/eN3eF69S6r

— Fin Mod World Champs (@ModelOff) December 3, 2017

The coffees are keeping Finalists focused during the gruelling 4-hour Finals testing pic.twitter.com/Su0ua5nJTf

— Fin Mod World Champs (@ModelOff) December 3, 2017

I didn’t get a chance to see the winner and runner-ups for Modeloff this year, but the winner was Alvin Yoon from Bank of America! You can read more about Alvin and the other finalists here.

To celebrate the holidays, we are offering a 40% discount off your entire order through December 12th! Use the code **YAYHOLIDAYS2017 **during checkout to take advantage of this discount which includes shipping. Give the gift of productivity this holiday :).

The post International Financial Modeling Competition Modeloff 2017 Recap appeared first on .

]]>The post Why You Need To Enter The Ultimate 2014 Excel Financial Modeling Competition appeared first on .

]]>That video is pretty bad ass huh?

There are only a few days before the first round starts, so check out some of the old questions from previous competitions here. The questions also include the actual Excel workbooks the contestants from previous years saw as well. Everything from Data Analysis to Energy Operations, these previous questions will give you an idea of what to expect if you make it to the final rounds of the Modeloff competition.

We are a huge support of Modeloff, rarely do you see the analysts, bankers, and consultants of the world heralded as world champions for their line of work. We attended the Modeloff 2013 awards ceremony last year and was blown away by the number of Excel enthusiasts waiting to see Excel celebrities like Bill Jelen from Mr. Excel. We also did a thorough write-up of the four challenges the Modeloff 2013 finalists were faced with during the awards ceremony. Check out the challenges and solutions here: 1, 2, 3, and 4. There’s also a lively discussion over at Mr. Excel on these questions as well.

Modeling (financial, that is) is an art and a science. Those who have been in the trenches staying up until 2AM figuring out why there’s a #REF error in your spreadsheet know that creating models involves a lot of trial and error, debugging, and just plain banging your head against the screen until you’ve tried every formula until it just *works*.

Embrace your Excel super power. Don’t be afraid to call yourself the king of spreadsheets. Show disdain towards those who still hard code numbers into their spreadsheets and don’t format their models correctly. Through practice, repetition, and grit, you are the guy or gal at your office that everyone looks to when they can’t understand why their VLOOKUP() doesn’t work. *That*, is power my friends.

Can you compare being an Excel champion to one of the greatest of all time? Hell yeah you can. MJ failed over and over again, which is what makes him great. You can be great too. Join the Modeloff this Saturday and show why your failures have led you to succeed.

The post Why You Need To Enter The Ultimate 2014 Excel Financial Modeling Competition appeared first on .

]]>The post The Best Of The 2014 Accounting Technology Show NYC appeared first on .

]]>On a rainy Wednesday morning, we headed to herald square where a bunch of disappointed tourists were finding shelter under the Pennsylvania Hotel’s awning. Nestled in the middle of the hallway with tourists walking by left and right was the sign for the New York Accounting Technology show. After going up to the 18th floor, we caught the end of Dana Richardson’s (Richardson Media & Technologies) opening keynote about “What’s New In Technology For Your Accounting Practice.” The room was packed with tax professionals and hung onto every single world Mr. Richardson had to say. He even made his predictions about where technology is going. The most startling prediction, in our opinion, was that the U.S. Post Office delivery people will become obsolete in a few years to be replaced by community mail centers that you pay for and go to to pick up your mail.

Back in the day, ING used to have have the best online savings account and that’s how people in the personal finance space came to know the brand. After speaking with Mr. Akhil Kumar from Arch Global Advisors who was representing the booth, we learned that ING has some pretty decent wealth management services. For those of you thinking about rolling over your 401(k), the ING Select Advantage IRA may be a good option as it only requires a $5,000 initial investment. Seems like a pretty good deal and knowing ING is an innovator in the personal finance space, this could be worth some more research if you are looking around for IRAs.

We met with the CEO of HotHeadTech which provides IT outsourcing to businesses among other services. Mr. Joseph Ferdinando’s story about how he started HotHeadTech shows that with hard work, you can do anything you want. He mentioned he started off making money literally cleaning toilets, and now today he owns a whole portfolio of businesses ranging from pest control to construction to his IT outsourcing firm. The name of Mr. Ferdinando’s company, interestingly enough, apparently came from Mr. Ferdinando’s own characteristic of being a hothead.

We spoke with Ms. Fei Dollinger from CPA.com and discovered that the AICPA (American Institute of CPAs) is one of the most involved and active associations in the U.S. The team here is responsible for finding the right vendors and solutions that add value for CPA professionals and educating the association on these new solutions. If you are a CPA, chances are you’ve come across AICPA and CPA.com.

Here is another example of where an old association partners with a leading technology firm to develop solutions to help tax professionals. Ms. Nancy Kasten works at the NATP (National Association of Tax Professionals) and they needed a solution to help solve the problem of helping tax professionals with getting their clients’ taxes done. NATP partnered with app development/marketing firm App Giraffe to create TaxPocket, a tax recordkeeping app that has a mileage tracker, expense tracker, and other features to make the tax filing process easier. The app first was created to help just the NATP to assist the NATP’s tax professionals with their own tax filing needs, but now the app is open to all tax professionals, hence the reason why TaxPocket was at NYAT.

We spoke with Ms. Diana Waterman from Sage, one of the bigger companies at NYAT. While a much bigger company compared to some of the other exhibitors, Ms. Waterman made it clear that Sage is committed to serving the small to medium-sized business segment as well and helping these businesses succeed. Very simply put, Sage helps your business manage your finances, people and payroll, and operations through a suite of products. We got to see quick demo of Sage One, one of Sage’s flagship products, and liked how the product was geared towards entrepreneurs and small business owners who don’t time to manage these back office operations. Surrounding the exhibit were thought leadership pamphlets and guides on how to run your business, using technology to automate your operations, and low and behold, a guide called 101 Microsoft Excel tips and tricks! Here is the full PDF if you’re interested, definitely a pretty comprehensive guide on using Excel 2007 and even provides some history about spreadsheets!

Working at a corporation requires detailed expense tracking and submitting receipts. Even at Fortune 500 companies, employees are still stuck photo-copying receipts and submitting their receipts through Concur or some other expense tracking platform. Certify aims to simplify the entire expense report process from submitting receipts to handing reimbursements. One thing thing we liked about their solution is the focus on the employee who is constantly traveling and needs to be able to manage his or her expenses on the fly.

Everyone has heard of Thomson Reuters, but did you know they offer various accounting software as well? CS Professional Suite is a line of software applications that helps your business with all accounting operations. We spoke with Ms. Christina Wiseman, a product manager on CS Professional Suite, and found that she has been with the company for 15 years. We really appreciated her stories about the fun events they would have in their Dexter, MI office like a prom event that helped bring the employees together. It sounds like it’s a great place to work and it’s clear everyone at the Dexter, MI office is one big family. It’s also good to see companies out of Michigan succeeding and Ms. Wiseman is a testament to the awesome people working in the accounting technology field.

Here is a company straight out of Brooklyn, NY. We spoke with Mr. Vincent Valenti from TGI and learned how TGI helps businesses with streamlining anything dealing with documents, scanning, and printing. Want to scan something and save to PDF? Looking to print from a certain device? TGI has developed software to solve these common problems. Beyond the services. Mr. Valenti described the green efforts by TGI such as recycling the thousands of ink cartridges and actually turning the material into building materials for homes. The company also had very humble beginnings but now is one of the most renowned companies in terms of helping businesses streamline operations.

We are big fans of Accounting Today since they provide really interesting content around accounting but also Excel as well. Their articles are helpful for any business professional looking to understand how to utilize technology to improve their business. Articles like “Should I Ditch the PC for an iPad” provide professional advice on how to adopt new trends in the marketplace around business. We are looking forward to working with Accounting Today on future content collaborations!

Last but not least, Microsoft Store was in the house showcasing some of their new laptops and tablets running Windows 8. Mr. Devon Cooper from the White Plains Microsoft Store showed us some of the new features in Windows 8 and we were definitely impressed. Without Microsoft, KeyCuts would not even be here today so we definitely had to stop by the booth and pay our respects. We didn’t know this, but Microsoft Stores, similar to Apple stores, offer tons of training to customers who have bought Microsoft products in the store. Mr. Cooper explained how he is currently training a woman one hour every week for the next year since she wants to learn how to use Microsoft’s products beyond the basics. While Apple stores tend to get the attention of most passerbys at malls and outlets, don’t overlook Microsoft Stores in terms of playing great product advisors and testing new laptops and tablets.

We went to the conference knowing we would learn some interesting technologies for accounting and financial analysts, and we definitely found some cool stuff as mentioned above. More importantly, the people we met along the way didn’t strike us as the stereotypical straight-edge and pedantic types. Everyone had an interesting story to tell about themselves and their experience at the company they work for. It’s great to see these niche communities of professional get together and geek out about the next big thing coming out for accounting professionals. We were so privileged to be an attendee this year! We leave you with some final words of wisdom from Sage’s 101 Excel Tips and Tricks book:

Using Excel shortcuts can increase your overall efficiency by saving you time, hence saving you money.

The post The Best Of The 2014 Accounting Technology Show NYC appeared first on .

]]>The post Excel Formula Final Question #4 from Modeloff 2013 Released with Explanation (4 of 4) appeared first on .

]]>Challenge number 4 was the most difficult challenge at Modeloff 2013. So hard, in fact, none of the contestants were able to solve the challenge in the allotted time. The teams were furiously experimenting with different formulas but as the clock started ticking down, a few simply gave up with exasperated looks on their faces; fingers resting on the keyboard instead of tirelessly typing new formulas to solve the challenge. If you can solve this challenge, you are in the top .1% of modelers out there, and you should definitely pre-register for Modeloff 2014! If you want to take a look at the first 3 challenges from 2013 Modeloff series, here are the questions along with the explanations:

As always, thanks to Dan Mayoh, the creator of all these challenges for providing us with the answers. Dan’s consulting practice is online now, so go check out Fintega Pty Ltd for more info. Now, onto the challenge!

Take a look at the first post in this series to familiarize yourself with the rules of the challenge. No VBA, Defined Names, or any other Excel black magic. Your formula must be entered into the grid of blue cells below:

There is also a 5 X 2 table of dates and duration periods that will help you with writing the formula:

Don’t forget! There is also a table of answers in the Excel file (download) that will turn green if the formula you entered into the blue answer cells is correct:

Write a formula that counts the number of cumulative flags to date that have been “set off” for each period end date (above the blue answer cells) relative to the first flag date. For instance, if the period end date is December 31st, 2014, and the first flag date is December 31st, 2013, and the duration of the flag is 1 year, the answer for the number of flags set off is 2. The answer is 2 because December 31st, 2013 is in the past (relative to the period end date), and a 1 year has passed since December 31st, 2013, so the flag was set off again on December 31st, 2014.

As another example, consider the period end date to be June 30th, 2015. The first flag date is June 30th, 2014, but the duration is 0.5 years. This means every 6 months the flag gets set off. The answer in this scenario is 3 since the flag was set off on June 30th, 2014, December 29th, 2014, and finally June 30th, 2014. In cases where the period end date is BEFORE the first flag date, the answer would simply be 0 since the flag would never be set off in the first place.

Ready to get your ass kicked? Or become a legendary financial modeler? Download the Excel file here that the contestants were given and try to solve the challenge yourself! The goal is to find the formula that satisfies the conditions of the challenge using the least number of characters possible in the formula.

Knowing that none of the contestants were able to solve this challenge, I was easily discouraged from trying to spend too much time banging my head to figure out the solution. I did take a few stabs at writing the formula as I knew it would involve dividing the difference between the period end date and the first flag date by 365. For instance, in cell O19 I started with this:

=(O$19-$E19)/365

You need to add the “$” to the row in O19 and to the column in E19 so that when you copy the formula across the blue cells, the reference to the period end date and first flag dates to not move around in the formula and stay fixed. I also knew that the formula would probably involve dividing the number of days by the duration cell, so the formula would look something like this in O19:

=(O$19-$E19)/365/$F19

This is pretty much where I got stuck as there are a few things you need to account for in the formula which remained unanswered to me:

- How do you account for the edge cases where the flag needs to increment by 1 given that subtracting the period end date from the flag date and then dividing by 365 still results in 1?
- How do you ensure that the answer is 0 for the period end dates that are before the flag date (columns H-J)?
- Piggy backing off of that last question, do we use the MIN() function somewhere to make sure we get the right answer for when the period end date is before the flag date?

There are still many unknowns, and the main problem I had was handling the edge cases (as is the case with most of the challenges in this series). Just finding a solution that works is difficult enough, forget trying to find the formula with the least number of characters.

The final solution contains 34 characters and only uses the INT() and MIN() functions:

=-INT(MIN($E19-H$17-9,1)/365/$F19)

This formula is entered into cell H19 (not array-entered) and copied across and down to show all the flags in the blue answer cells:

I was surprised myself that the final formula only utilized the INT() and MIN() functions, but through the clever use of these two functions we are able to see the edge cases handled elegantly.

Let’s take a look at each segment of the final formula to see how this formula is able to handle all the edge cases. Let’s take the formula in cell Q19 as an example:

=-INT(MIN($E19-Q$17-9,1)/365/$F19)

$E19 equals December 31st, 2013 and Q$17 equals the period end date, June 30th, 2014. Let’s forget the “-9” for now and see what happens when we just evaluate =MIN($E19-Q$17,1). The result is -546, which is essentially the difference in days between December 31st, 2013 and June 30th, 2015. The reason why MIN() evaluates to -546 is because this function takes in the two arguments, -546 and 1 and returns the lesser of the two arguments. Now let’s divide by 365 and we see the answer is is -1, or -1.50 if you show two decimal places.

This is starting to make sense now. The difference between December 31st, 2013 and June 30th, 2015 is indeed 1.50 years. If we divide the -1.50 by the duration, we will still get an answer of -1.5 since the flag duration is 1 year. Now let’s add in the INT() function and see how it affects the formula:

=INT(MIN($E19-Q$17-9,1)/365/$F19)

The result is -2, so we are definitely on the right track here. Why does =INT(-1.5) evaluate to -2? You usually use the INT() function to get the integer of a decimal number. For instance, the INT() function would return 5 for the number 5.25. However, for *negative *numbers, the INT() function returns the first negative number that is less than or equal to the expression. You would think that =INT(-1.5) would result in -1 (the integer of -1.5) but the closest integer to -1.5 that is less than -1.5 is -2. Even if the expression is -1.001, =INT(-1.001) still evaluates to -2.

The final solution has a “-” in front of it so this would turn the -2 to a 2 which is the correct value for cell Q19.

This leads us to the question about the “-9” which we omitted in our explanation above. It looks like this formula yields the correct answer for us:

=-INT(MIN($E19-Q$17,1)/365/$F19)

If you copy this formula across the rest of the blue answer cells, however, you’ll start to notice some errors in the table of answers. The reason is because the -9 helps us with edge cases when the period end date is the same as the flag date or one or two years ahead of the flag date. For instance, let’s look at the formula in cell Q20 *without* the -9:

=-INT(MIN($E20-Q$17,1)/365/$F20)

The result is 2 but this is incorrect since the actual answer is 3. The flag date is June 30th, 2014 and the period end date is June 30th, 2015, exactly one year ahead of the flag date. The duration is only 0.50 years in this case. This means that the flag has been set off 3 times: once on June 30th, 2014, again on December 31st, 2014, and one final time on June 30th, 2015. The issue with the formula above is that $E20-Q$17 results in -365. When we take -365 and divide it by 365 and then by the duration, 0.50, we will get the -2 answer which we currently see. What we need is a number to decrease the -365 so that when we divide by 365 the answer is not exactly -1, but rather a little less than -1. Then, when we incorporate the INT() function with the given expression, the result will will be the lesser negative number, or -3 in this case.

Let’s add in the -9 to the formula and see what happens:

=MIN($E20-Q$17-9,1)/365/$F20)

The result is -2.05, which is slightly less than -2 but makes a world of difference in terms of what the INT() function evaluates the expression to. The lesser number of -2.05 is -3, and when we multiply this number with the minus sign at the beginning of the formula, we will get the 3 that we need in cell Q20. There is no specific reason why the solution uses a “-9,” we could also use “-1” and the answers in the blue cells will be correct. The important concept here is to decrease the negative result of the the flag date minus the period end date by a little bit for the INT() function to evaluate the expression correctly.

This challenge shows how a formula may appear to work, but once you analyze the edge cases, you’ll begin to see how the formula breaks down and you need to incorporate other functions or elements to make the formula work. The INT() function wouldn’t even be incorporate into the formula if it wasn’t for the “-9” since the purpose of the INT() is to find the lesser integer from the expression. If we assumed the formula worked without the “-9,” the resulting formula would’ve been a lot shorter and wouldn’t have caused so much anguish among the Modeloff contestants.

Through all these Modeloff challenges, the one common theme I see is to observe the details about how data that is fed into the formulas. What may seem like a straightforward formula may become very complicated as you deal with the edge cases, as we saw with the final challenge. Hopefully these challenges helped you learn a little more about building advanced formulas in Excel, and some of the real-life issues analysts, consultants, and modelers face on their job.

The post Excel Formula Final Question #4 from Modeloff 2013 Released with Explanation (4 of 4) appeared first on .

]]>The post Excel Formula Question #3 from Modeloff 2013 Released with Explanation (3 of 4) appeared first on .

]]>Ground rules are the same as before, create a formula that solves the challenge in the least number of characters without using VBA and all that fancy stuff. Read the first post in this series if you want the full rules. This week, the challenge is only one blue cell. Yup. That’s it. Just one cell to enter your kick ass formula.

There is also a large 40X15 block of numbers that you will utilize to solve the challenge:

The numbers in this block are a bunch of random integers.

Write a formula that counts how many cells in the number block are less than -10 or greater than +10. Ready, set, gooooo! During the Modeloff 2013 award ceremonies, the contestants easily created formulas to solve this challenge, but the problem was trying to find the *shortest *formula. I’ll give you a hint, the number of cells that have values less than -10 or greater than +10 is 158.

Give it a shot! Download this file that represents the data discussed above and see if you can create the formula.

I’m going to do things a little differently for this post, and walk you through my thought process for finding the solution before posting the actual solution as per Dan Mayoh. If you simply want to see the 22-character solution and don’t give a crap about my meandering thoughts, then scroll to the bottom and get instantly gratified.

As stated earlier, the solution definitely involves the use of array-entered formulas to test all the numbers in the block against the condition greater than +10 or less than -10. It also involves the SUM() function in order to add up the number of cells that meet the condition. Ok let’s get started.

The first formula I wrote was this:

=SUM(N((E20:S59>10)+(E20:S59<-10)))

This formula is array-entered (CTRL+SHIFT+ENTER) into cell E17 on the worksheet and it will result in the correct answer of 158. Why does this work? Well let’s first look at what the two conditions we are testing for: E20:S59>10 and E20:S59<-10. Nothing too crazy here, we are simply checking the entire range of random integers to see which ones are less than -10 and which ones are greater than 10.

The interesting thing about array formulas is that you can put OR() and AND() operators to test your conditions. For array formulas, the syntax to use is the “+” sign to tell Excel that we want to see which numbers are less than -10 OR greater than 10. If we used the “*” sign in between our two conditions, this would tell Excel to look for values that are less than -10 AND greater than 10. This conditional makes no sense because it is impossible for a number to be both greater than 10 and less than -10. Duh.

If we array-enter this formula =SUM((E20:S59>10)+(E20:S59<-10)), we will get the correct result of 158. However, the number of characters in this formula is 32, a whopping 10 more characters than the actual solution. There must be a better way to optimize, so I thought about using the ABS() function which returns the absolute value of whatever number you give it. This would allow me to not compare the range of values to both 10 and -10, but rather just one number.

The next iteration looks like this: =SUM(ABS(E20:S59)>10). I got really excited because this formula only contains *21 *characters, which is one less than the correct solution. When I array-entered the formula, the result made me realize I am still just a regular Excel geek compared the top financial modelers out there. The result of this array-entered formula is 0, and I struggled to figure out why. Here’s what I discovered.

ABS(E20:S59)>10) yields an array of TRUE and FALSE values indicating which values are indeed less than -10 or greater then 10. The array would look something like this (abbreviated to save space):

{FALSE, TRUE, FALSE, TRUE, TRUE, TRUE, [...]}

Which is based on the first few values in the number block:

{9, 22, 0, 15, -14, -18, [...]}

While we know which values are indeed less than -10 and greater than 10, we cannot simply sum up the resulting array of TRUE and FALSE values. I am not completely sure about this next statement, but the reason why I think =SUM((E20:S59>10)+(E20:S59<-10)) works is because when you introduce a conditional on top of two other conditions (using the “+” sign), the result is the numeric representation of TRUE and FALSE. You’ll get an array of 0s and 1s which you can easily sum up. If you are only working with one condition, as in the above formula where I used the ABS() function, the resulting array is the non-number values of TRUE and FALSE. If you have a better explanation for why this occurs, please leave a comment below!

Ok, so I know I need to somehow convert these TRUE and FALSE values to numbers. I cheated a little here and Googled for the answer and found a lesser known function called N() which converts non-number values to a number, dates to serials, etc. Exactly what we need! Now the formula looks like this with the N() function doing its job:

=SUM(N(ABS(E20:S59)>10))

Array-enter that bad boy and you get 158. However, the character count is at 24 (including the equals “=” sign). Still. Need. To. Optimize. At this point I was getting restless and just wanted to see the solution. I think if I spent another 10 minutes on this sucker I could have come up with the solution.

In all it’s 22 character glory:

=SUM(N(E20:S59^2>100))

So simple and elegant. It’s very similar to my solution except the condition is cleverly written to cut down on characters. The E20:S59^2 portion of the formula tells Excel to raise each value to the power of 2 which basically eliminates the need for the ABS() function since any integer raised to the power of 2 will yield a positive number. Of course, we still need our trusty single-character N() function to convert the array of TRUE and FALSE values to 0s and 1s. After you array-enter this guy, you’ll get your solution of 158.

In the real world, rarely will you encounter the need to optimize your formula to this level of detail. On a project you are usually working under a deadline and whatever formula you come up with that simply works and gets the job done will suffice. There are still two points I want to finish on that this challenge can teach us:

**Speed up large Excel files**– You will work on a file that has thousands or millions of rows, and writing an efficient formula will actually*decrease*your productivity since Excel will take a long time to think and spit out the answer. This can also lead to your file crashing and data getting lost and you end up playing the whole AutoSave game to recover your file. When you are optimizing a MySQL query, these shortcuts and optimizations can mean the difference between a successful and unsuccessful product.**Formula sex factor**– There’s no other way to say it. Seeing a really simply formula that can break down a complex problem and spit out the answer is just*sexy*. I may be taking this too far, but understanding why a formula works allows you to think about other data problems differently. In computer programming parlance, this is like learning a new module and being able to use it over and over again in another part of your program.

Whew, that was a lot of talking from me for a 22-character formula. Did you come up with an alternative solution that is less than 22 characters?

The post Excel Formula Question #3 from Modeloff 2013 Released with Explanation (3 of 4) appeared first on .

]]>The post Excel Formula Question #2 from Modeloff 2013 Released with Explanation (2 of 4) appeared first on .

]]>In case you missed the ground rules for the group challenge, read the first post in this series of Excel formula challenges. Similar to last week, you have a range of blue cells where you need to enter in the solution formula:

There is also a range of numbers next to the blue cells which you must use in the challenge:

**The Objective:** Write a formula to rotate the values in the 4*4 square through 180 degrees.

All the numbers need to be “flipped” along both the X-axis and Y-axis. Here is what the final table of answers should look like in the blue cells:

Here is a side-by-side comparison between the original range of numbers and what the final output should look like in the blue cells:

Give it a shot! Download this file that represents the data discussed above and see if you can create the formula!

When I first looked at this challenge, I knew it involved using array formulas. For those not familiar with array formulas, here is a good guide from Mr. Excel that explains how array formulas work and why they are useful. I was seriously stumped by the question, and struggled to figure out how to “flip” the numbers in an automated way.

The solution involves using the simple SUM() function in an array formula in an elegant fashion. The final formula consists of **33 characters**:

SUM($E$18:$H$21*(B15:E18=$E$18))

This is the formula you enter in the first cell of the blue cells in the array format (by pressing CTRL+SHIFT+ENTER) after you’ve written the formula. You then copy over to all the cells in the range to yield the final table of answers.

This formula compares two ranges of values through an array formula. If you simply entered the formula as is and pressed ENTER, you would get a #VALUE error in the cell since you are supplying the SUM() function with a conditional statement.

Let’s back up and see how array formulas work in the context of this solution. Let’s analyze the $E$18:$H$21 portion of the formula. This is simply the range of numbers we need to flip to get our solution. By entering this formula as an array formula, this specific portion of the formula returns an *array* of values like this:

{100, 111, 112, 114, 118, 126, 135, 136, 142, 147, 149, 151, 174, 186, 193, 197}

The B15:E18=$E$18 portion of the formula is a little different since we are testing a condition against an array of values. In this case, we are testing each value in the B15:E18 range to see if that value matches the value in $E$18. The B15:E18 range of values returns an array that is mostly empty except for the last value:

{"", "", "", "", "", "", "", "", "", "", "", "", "", "", "", 100}

The reason why there are so many blanks is simple. The range B15:E18 is what I like to call a “dummy” range since it’s created mostly to help with making the final formula work. Cells B15, C15, D15, etc. are all empty in the worksheet. The B15:E18=$E$18 condition actually returns an array of TRUE and FALSE values, or 0s and 1st based on whether or not the values in the B15:E18 range equal cell $E$18, or 100. Thus, the array of values returned from testing B15:E18 with the value 100 is this:

{FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, TRUE}

Mathematically, this array actually looks like this:

{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1}

So now we have two arrays: one with the values in $E$18:$E$21 and one with the 0s and 1s from testing range B15:E18 with the value 100. The “*” in the formula multiplies the first array by the second array. The product of this is…you guessed it! Another array of 16 values.

{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 197}

This array above is what’s returned from $E$18:$H$21*(B15:E18=$E$18) when entered as an array formula. Why is the last value 197? Remember the first array $E$18:$H$21 consists of all the numbers in the original square. When we multiply all those numbers with the second array with the 0s and 1s, the first 15 numbers in the resulting array are 0 (due to the FALSE values) and the last number 197 gets multiplied by 1 (the only TRUE value in the second array). The SUM() function simply adds everything up in the array together which results the value 197. This is the correct number in the top left of the blue answer cells.

The real magic occurs when you copy this formula to all the cells in the blue answer cells. Let’s look at the formula in cell K19 of the answer cells:

SUM($E$18:$H$21*(C16:F19=$E$18))

You’ll notice that the only thing that changed about this array formula is the relative range C16:F19. All the other referenced cells are absolute references. As you copy the formulas over to the other cells in the answer table, the “dummy” range of cells begins to shift to help create this “flipped” axis of numbers. If we break down the two arrays that are multiplied together, you’ll see how this works. $E$18:$H$21 remains the same:

{100, 111, 112, 114, 118, 126, 135, 136, 142, 147, 149, 151, 174, 186, 193, 197}

When we test C16:F19 against $E$19 now, the C16:F19 dummy range starts to include some of the numbers in the original table of numbers. This effectively helps us “count” which position in the second array contains the TRUE value. The resulting array of C16:F19=$E$19 is:

{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0}

The 11th value in the above array is the TRUE value because that’s when the value in range C16:F19 equals 100. When we multiply this array of 0s and 1s with the original table of numbers, the resulting array is this:

{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 149, 0, 0, 0, 0, 0}

Of course, when we sum up the numbers in this array, the result is 149 which is the correct number in cell K149 of the blue answer cells.

**Conclusion**

There are two other formulas that deserve honorable mentions which we won’t go into detail about how they work, but they show how you can get creative in Excel to solve problems with different functions. Both of these formulas clock in at **35 characters**:

INDEX(E18:H21,{4;3;2;1},{4,3,2,1})

OFFSET($E$18,21-ROW(),13-COLUMN())

The first formula is array-entered into all 16 cells in the blue answer cells simultaneously. The second formula is entered into the first cell of the answer table (Cell J18) and copied across and down to the rest of the cells in the range.

Were you able to come up with other solutions? How long did it take you to answer this question?

The post Excel Formula Question #2 from Modeloff 2013 Released with Explanation (2 of 4) appeared first on .

]]>The post Excel Formula Challenge – Question #1 from Modeloff 2013 Released (with explanation) appeared first on .

]]>The goal of the challenge is to write the most “efficient” Excel formula that solves the problem, e.g. the formula with the least number of characters. The teams were generally given 10 minutes or less to create their final formulas. The formula must be copied across all the cells in the answer range to solve the objective. No VBA, Defined Names, Helper Cells, or references to the Table of Answers or Check Sum cells may be used.

** **You have a bunch of empty cells as shown below with these start dates and end dates above them:

You also have these Start and End dates in yellow directly to the left of those blue outlined cells:

**The Objective:** Write a formula that counts the number of days between the Start and End dates in the above table that fall within the Start and End dates listed above the first table.

For instance, in the first cell in the blue-outlined table, the number **69** should be the output, like this:

Why is that first cell 69? The period in the blue-outlined table is January 1^{st} to March 31^{st}, 2013 (one quarter). The Start date in the yellow table is January 22^{nd}, 2013, which is greater than the period start date (January 1^{st}, 2013). The End date from the yellow table, however, is greater than the period end date at January 21^{st}, 2014. Therefore, the only overlap would be from January 22^{nd} to March 31^{st} 2013 in this case (69 days total).

The correct number in the second column for the blue-outlined table is **91**. Why? The period start and end date for this cell is April 1^{st}, 2013 and June 30^{th}, 2013. The Start and End dates in the yellow table are April 1^{st}, 2013 and June 25, 2030. This entire date range of 17 years overlaps with the entire period from the blue-outlined table, so the resulting answer is the entire quarter, or 91 days. If there is no overlap, then the formula should return nothing in the blue-outlined cells.

The Table of Answers shown in the Excel file tells you if the output in the blue-outlined cells are indeed the correct calculation (Green = correct, Red= incorrect):

The Table of Answers quickly helps you figure out if the formula you are copying across and down in the blue-outlined table are correct. After creating the correct formula and copying it across all the cells in the range, your table should look like this:

Give it a shot! Download this file that represents the data discussed above and see if you can create the formula!

The very first thought that comes to my mind is to use a bunch of IF statements to solve the problem. While this brute force method will work, it’s not the most efficient solution since the challenge calls for writing the shortest formula.

The solution involves using the MAX() and MIN() formulas in a creative way. The final formula only uses **38 characters **as reported by Dan:

MAX(,MIN(H$17,$F18)-MAX(H$16,$E18)+1)

This is the formula you enter in the first cell of the blue-outlined cells and then copy over to all the cells in the range.

Why does this work? Let’s first focus on the MIN(H$17,$F18) part of the formula. MIN() simply returns the smallest value you provide the function—in this case it simply returns H$17 or $F18 depending on which is smaller. H$17 refers to March 31^{st}, 2013 in the blue-outlined cells, and $F18 refers to January 21^{st}, 2014 in the yellow cells. Ok, so we know that this MIN() formula will return **March 31 ^{st}, 2014 **(cell H$17).

For the MAX(H$16,$E18) function, MAX() does the opposite of MIN(), and returns the greater value of these two inputs. H$16 refers to January 1^{st}, 2013, and $E18 refers to January 22^{nd}, 2013. The greater of these two dates is, of course, **January 22 ^{nd}, 2013**.

Now, let’s step back and look at what this does: MIN(H$17,$F18)-MAX(H$16,$E18)+1

We are taking March 31^{st}, 2014 and subtracting from it January 22^{nd}, 2013. The result is 69 days. The reason we add the 1 at the end is to make sure the result is inclusive of the start and end dates.

If we simply copy this formula MIN(H$17,$F18)-MAX(H$16,$E18)+1 across the range of blue-outlined cells, we’ll get something like this:

This is obviously incorrect since there negative numbers in parentheses that doesn’t make sense if we are trying to find the overlap. These negative numbers should in fact be empty cells since the dates from the yellow table to not overlap with the period dates from the blue-outlined table.

Here is where the MAX() function from the outside of the formula we found comes into play: MAX(,MIN(H$17,$F18)-MAX(H$16,$E18)+1)

We are asking Excel to give us the greater of MIN(H$17,$F18)-MAX(H$16,$E18)+1 or 0. If there is indeed an overlap, then we will always get a positive number from the MIN(H$17,$F18)-MAX(H$16,$E18)+1 part of the formula. When this formula returns a negative number, 0 will be the result of this outside MAX() function, so therefore the resulting output will be the 0 or the blank we are looking for in the answer set.

Interestingly, the formula could have been written like this: MAX(0,MIN(H$17,$F18)-MAX(H$16,$E18)+1).

You see that extra 0 (zero)? That can simply be deleted and Excel automatically knows to compare against the number 0 as stated in the final solution. This also allows us to save one character from formula!

This is the first question the teams answered and most teams finished it in less than 10 minutes. How long did it take for you? Were you able to come up with a shorter formula?

The post Excel Formula Challenge – Question #1 from Modeloff 2013 Released (with explanation) appeared first on .

]]>The post The Hunger Games Meets Excel Financial Modeling – Modeloff 2013 Recap appeared first on .

]]>There was already a big line forming in the Microsoft lobby in midtown as finalists arrived a bit late and one security guard working the desk. He was probably not used to all the brain power and Excel analysts carrying their desktop keyboards in their backpacks (F1, CAPS LOCK, and INSERT keys removed, naturally). Check out the live strong Modeloff bracelets all the participants received!

The hallway to the main room was decked out with Modeloff’s various sponsors like Capital IQ and Kaplan. The most impressive sponsor to grace the Modeloff poster is, of course, KeyCuts! We are sandwiched in between Mergers & Inquisitions and Mr. Excel.

The crowd is intently listening to John Persico, Executive Director for Modeloff, discuss the history of the event and introducing all the judges and sponsors. Mr. Excel himself (Bill Jelen) is one of the judges for Modeloff as well as Professor Simon Benninga, author of Principles of Finance with Excel. Kiel LaFrance is the final judge for the event. I had a chance to listen to Professor Benninga speak at the pre-Modeloff Innovation in Financial Modeling event Friday at Star Mountain Capital. He told an interesting story about how one of the reviewers for his books had only reviewed one other book on Amazon–and that book was about sex positions. That’s right. Sex and Excel complement each other very well.

To get things started, there would be group challenges between groups of finalists competing for cash prizes. The room was organized so that two tables were set up in the middle of the room where the finalists would sit head-to-head death-match style. All the challenges involve writing formulas with the least number of characters to solve the problem. It’s kind of like “Excel Golf,” as Dan Mayoh put it (2nd place 2012 Modeloff winner and 2013 Modeloff questions designer). The least number of strokes (in this case, keystrokes) the better. The finalists are grouped together by regions: London, Australia/New Zealand, USA, and rest of the world.

People in the crowd naturally gathered around the finalists as they started testing and experimenting with different formulas. The Excel files the finalists were working on were projected so that audience members can easily see what Excel monsters the Question Design team had created. Bill Jelen channeled his inner Bob Costas and provided the play-by-play commentary as INDEX, MATCH, and OFFSET were being tossed into the proverbial spreadsheet ring.

For Challenge 1 there was a tie between the teams, as the winning formula from both team had exactly 38 characters. Challenge 2 involved a block of 4 X 4 numbers and you need to write a formula that reverses or reflects the numbers across the X and Y-axis (see picture 3 above this text). The winning answer used the INDEX function instead of the OFFSET function since it accepts less arguments. Challenge 3 is involved a big block of numbers and the contestants needed to count the number of cells that are greater than 10 or less than 10. The 4th challenge was the most difficult and based on a debt facility model. Finalists needed to count the number of flags to find, for any given period, how many times the flags were triggered. Finalists needed to create a formula just to quickly sum up the occurrences without creating the flags themselves. No team solved this challenge, hats off to Dan for a tough question.

After 4 challenges, the **Australia/New Zealand** team came out on top!

During one of the challenges, Bill made a rhetorical statement that summed up the passionate and self-deprecating love everyone in the room has for Excel:

I just want you to consider this for a second: you are here in this room on a Sunday afternoon…watching other people enter formulas into Excel. – Bill Jelen

Very eloquent, Mr. Excel.

Mr. Rosenbaum, author of Investment Banking: Valuation, Leveraged Buyouts, and Mergers and Acquisitions, gave a short keynote following the challenges and discussed the state of financial modeling education today. Mr. Rosenbaum made some very salient points about how most modelers learn their craft. The best modelers are self-taught which means there’s a huge opportunity (or deficiency) for universities to train the next wave of financial modelers. Just because you received a 4.0 from Harvard Business School doesn’t automatically make you an expert modeler, Mr. Rosenbaum observed.

Leontios Toumpouris and David Loong were the undergrad and graduate student winners. For the moment you have all been waiting for: the winners!

2nd Place – Alex Gordon (Modeloff 2012 Winner)

Hilary was the only female finalist this year and won the grand price of $30K. This was a big step for female financial modelers, as female entrants to Modeloff increased 10% to 19% year over year. John then invited up all the volunteers of the event and it was pretty amazing to see all the people involved behind the scenes.

I waited patiently to talk with the Justin Bieber of the Excel world, Bill Jelen aka Mr. Excel aka The King. If you have ever had a problem with Excel or needed help on writing a formula, chances are you have landed on one of Mr. Excel’s forums. I asked Mr. Excel what his favorite keyboard shortcuts were, and here they are (with Mac equivalent in parentheses):

**1) ALT, E, S (COMMAND+CONTROL+V)** – Paste Special

** 2) F2 and ESCAPE (CONTROL+U, Esc)** – Quickly view the formula and get out of the formula edit mode

** 3) CONTROL + F5 (FN+F5) **– Display the Go To dialog box

A shot of all the 2013 finalists.

Finally, I had the pleasure to talk with the excited yet poised Modeloff 2013 winner, Hilary Smart. She also gave her top 3 favorite Excel keyboard shortcuts (with Mac equivalent in parentheses):

**1) CONTROL (COMMAND) + ARROW KEYS** – Zip around an Excel sheet fast

** 2) CONTROL (COMMAND) + [ ** – Trace formula dependents

** 3) CONTROL + PAGE UP/PAGE DOWN (FN+CTRL+UP/DOWN) **– Scroll through worksheets in a workbook fast

Everyone hiked a few blocks through the blizzard that had been raging on outside while the heat was turned up at the event. The post-event drinks were held at 48 Lounge sponsored by EY.

Pre-registration has already begun for next year’s event! Check this out if you think you have what it takes to be the 2014 world financial modeler. Thanks to Johann Odou and the rest of the Modeloff team for having KeyCuts as a community sponsor for an awesome event. Watch out for great things coming from the Modeloff team and the Excel community as a whole!

I would like to make a few corrections and clarifications to this post. The Modeloff event referenced in this blog post refers to the award ceremony held at Microsoft. The challenges did not determine the 1st, 2nd, and 3rd place winners for Modeloff 2013, they were simply challenges between different teams of finalists at the ceremony.

The 16 finalists sat through **4 grueling hours** of testing earlier that day which ultimately determined the winners for Modeloff 2013 (congrats again to Hilary Smart!). The testing was done behind closed doors and the World Champion truly deserves all the respect and approbation for beating out 15 other finalists.

Finally, Challenge 3 referenced in this post required counting cells that were *greater than 10 or less than negative 10* (not less than positive 10).

Thanks to Dan Mayoh for the corrections!

The post The Hunger Games Meets Excel Financial Modeling – Modeloff 2013 Recap appeared first on .

]]>