UNLOCKING FULL POTENTIAL

OF EXCEL DATA TABLES

Part 1

By Andrei Okhlopkov

Published: December 2019

For the most recent version of this booklet,

accompanying Excel file (“Data tables_1.xlsx”)
and other publications please visit:

www.eloquens.com/channel/andrei-okhlopkov

Materials in this booklet are distributed freely and are not intended for sale. You are
authorized and encouraged to forward this booklet and accompanying Excel file to
everyone who might be interested.
Unlocking Full Potential of Excel Data Tables (Part 1)

Also in this series:

• Guiding Principles of Financial Modeling

• Dates and Timelines in Financial Models
• Sensitivity Analysis with Indifference Curves
• Charts and Dashboards in Excel (parts 1 and 2)
• Descriptive Statistics for Grouped (Weighted) Data
• Analyzing Data in Excel
• Monte Carlo Analysis without Macros
• Portfolio Analysis and Sales Forecasting
• Form Controls and Hyperlinks in Financial Models

The above list and the publications are constantly updated.

******

I appreciate your interest in my publications on financial modeling, Excel and VBA,

which I enjoy sharing, will keep coming and hope you will find useful.

Most of my publications are free but you can support my project through the PayPal
(ID: [email protected]).

And certainly feel free to let me know if you have any questions or comments.

Andrei Okhlopkov

[email protected]

Page 2
 Unlocking Full Potential of Excel Data Tables (Part 1)

TABLE OF CONTENTS

AN INTRODUCTION TO DATA TABLES 4

THE MODEL 5
KEY PROJECT PARAMETERS 7
CONDENSED FINANCIAL STATEMENTS 8
TORNADO CHART 10
WHAT’S NEXT? 11

Page 3
Unlocking Full Potential of Excel Data Tables (Part 1)

An Introduction to Data Tables

Data tables represent a powerful tool for sensitivity analysis in Excel. A conventional
data table is a one- or two-dimensional table showing the results of a calculation
(model outputs) at various model inputs in the headings of rows and/or columns of the
data table. Excel Help provides a detailed description of how data tables are created
and used (please check it out if you are new to data tables).

In this publication I will introduce a non-conventional way of using data tables which
allows to extend a traditional sensitivity analysis. This method relies on the ability of
data tables to evaluate a model many times at various input assumptions. Each list of
this file contains an illustration of a specific method of analysis with details and
comments on each of them further below.

The sensitivities are demonstrated through the following means:

a) a summary table showing the key metrics (IRR, money-on-money ratio, net
return, etc.)
b) a high-level cash flow summary for each scenario (showing cash flows from
investment, dividend income, debt service, exit proceeds)
c) a "tornado" chart which shows the effect of every input on the model's outcome

Data tables make the above visuals fully interactive - you can change one of the inputs
in any scenario and the effect of this change gets reflected immediately in the tables and
charts (all done without macros).

The method may seem complicated - but only until you look carefully at the setup and
the formulae.

Page 4
 Unlocking Full Potential of Excel Data Tables (Part 1)

The Model

Refers to sheets: “Key results”, “Executive summary”, “Tornado chart” (they all contain the
same model)

As an underlying example I have created a very simple demonstration model dealing

with a real estate investment project. This example is entirely fictitious; yet, I have tried
to make it close to reality. For the sake of simplicity, certain aspects of real estate
business are ignored - therefore, you should not rely on it as a practical tool for
evaluating investments. The model is the same for all demo cases on all worksheets and
is located underneath the data tables and related calculations.

Consider an investor wishing to purchase a fully completed and operational business

office. The building generates rental revenue, the amount of which is based on the
rentable area of the building, rental rate (indexed in the beginning of every year) and
vacancy percentage. Most of the operating expenses are compensated by the tenants (in
addition to the rental rate they pay), but there is a small "leakage" denoted as "non-
reimbursable opex".

The value of the building is determined by the net operating income it generates and a
rate of profitability ("capitalization rate" or "cap rate") which investors expect to earn
on this type of investment. Cap rates tend to get smaller over the years (the term "yield
compression" defines that), which creates the basis for capital gain upon disposal a few
years later. However, at rough times cap rates stay flat or even increase, demonstrating
higher expected returns due to increased cost of capital and the level of risk, bringing
the market value of real estate down.

The investor considers drawing a loan to finance part of the acquisition.

During the holding period, our investor receives dividends which represent 100%
distribution of the net profit of the asset. The investor uses those dividend proceeds to
make loan interest and scheduled principal repayments. If these are insufficient, the
investor must cover the gap with his/her own money.

In the end of the project, the building is sold at a price determined by its net operating
income and the current cap rate. Proceeds from sale are partly used to repay the
remaining loan balance (a so-called "balloon payment").

The investor is concerned about the main profitability parameters (IRR, Money-on-
Money ratio, net return) of the project and needs to make sure the financial covenants,
such as loan-to-value and debt service coverage ratios (typically established by the
banks and monitored over the loan period) do not fall below the prescribed hurdles.
And he/she certainly needs a full cash flow model underlying these calculations.

The model considers three scenarios (Downside, Base and Upside). These scenarios
differ in the expected future vacancy and indexation rates, the amount of non-
reimbursable opex, exit timing, loan interest rate and yield compression. The scenarios
can be switched from and to with a drop-down list next to the input section.

Page 5
Unlocking Full Potential of Excel Data Tables (Part 1)

Just as that, a typical model usually contains several sets of parameters, or scenarios.
The user switches between these scenarios and each scenario's outcome is displayed
one by one, which is not quite convenient. The data table tool, as I have designed it,
allows seeing the outcome of all scenarios at once without having to switch between
them, and any change in inputs is reflected immediately in the outcomes. This method
does not involve macros. Here is how it works.

Page 6
 Unlocking Full Potential of Excel Data Tables (Part 1)

Key Project Parameters

Refers to sheet: ”Key results”

This is an "extended" version of a simple one-dimensional data table. The data table is
located in the range K1:N17 (highlighted in light brown) with output formulae in
column K (they calculate key results, financial covenants and overall cash flow figures
for the currently selected scenario; I have left these outputs in light gray but you can
white them out completely in your model). The data table populates the outputs
through columns L, M and N for the Downside, Base and Upside scenario respectively.

The data table allows seeing the key results of all the scenarios at once. Try changing
the inputs (in the cells highlighted in blue) for a particular scenario and see how this
affects the results of that scenario in the data table.

The cells of the data table can be formatted as amounts or percentages, can be blanked
out where needed and even conditionally formatted (I have done this for the traffic
lights checking the compliance with financial covenant; the hurdles can be
seen/changed in cells F5 and F6). I have highlighted the data table here for educational
purposes, but you obviously can format yours as you wish.

Cell C17 is a row input cell for the data table and is also a manual switch between the
above scenarios in the model (located underneath). You can select a scenario you
would like the model to display and see it in detail. This does not affect the data table -
it remains unchanged showing the key outcomes of the model for all the three
scenarios.

Page 7
Unlocking Full Potential of Excel Data Tables (Part 1)

Condensed Financial Statements

Refers to sheet: ”Executive summary”

If you have a big and complex model, you will normally make a snapshot of it with the
main results (IRR, Net return) and a short cash flow summary. With the data table
functionality, you can create such summaries one per each scenario, displaying the
results at once independently from which scenario is currently selected for display by
your model.

An example of such a summary is given in range A18:G44. For the output formula to
populate the data table it needs to determine what to show in this or that cell, and
under which scenario to show this.

The answer to the first question is contained in the data selection table to the right
from the data table (range I19:N44). The format of this table mirrors that of the data
table. It contains the calculations and hard-coded value we want to display in the data
table. As we have three scenarios, they are repeated here in three sections.

The second table (scenario selection table, range P19:U44) tells under which scenario
each particular cell needs to be calculated (Downside, Base and Upside). If scenario
does not matter, the cell is left empty.

As the data table calculation goes over its cells, it puts values 1 to 6 (columns) into cell
K11 and values 1 to 26 (rows) into cell K12. Notice that the input cells can have either
formulae or hard numbers - the data table algorithm will just temporarily replace them
with the values from the row and column heading of the data table. Formula in cell K13
finds out which scenario is put by looking it up at the intersection of the given row and
column in the scenario selection table (if the cell is blank, it puts "Base" - this does not
really matter but the model still needs some scenario to run under). Then the model is
recalculated at this scenario, the current results of the model are reflected in the data
template and are picked up by the data table output formula (cell K14), which puts
them into corresponding cells of the data table. As the formula will return zero for any
empty cell in the data template, the formula replaces such zeroes with an empty space
for the sake of aesthetics.

Cell K11 contains a bit tricky a formula the purpose of which is to allow the user to
switch between scenarios to display in the model without interfering into the table
supporting formulae. If, for instance, the user chooses "Downside" in cell K8, it will
cause "2" to appear in cell K11, which, together with 22 entered into cell K12 will result
in "Downside" appearing in cell K13 (derived from cell Q40 which corresponds to
column 2 and row 22 of the scenario selection table) which feeds the model. Same, if
the user chooses "Base" or "Upside", the formula in cell K11 returns 3 or 4 respectively,
which will result in "Base" or "Upside" (contained in columns 3 and 4 and row 22 of the
scenario selection table) in cell K13.

Page 8
 Unlocking Full Potential of Excel Data Tables (Part 1)

Data selection and scenario selection tables are helper tables and you obviously don't
need to show them in the presentation version of your model - you can hide them or
move to a different spreadsheet.

Page 9
Unlocking Full Potential of Excel Data Tables (Part 1)

Tornado Chart

Refers to sheet: ”Tornado chart”

The model you have constructed shows the total IRR, net return etc. parameters at
various scenarios. The difference between net return in e.g. the Base case and the
Upside case comprises the effect of change of all sensitivities (vacancy rate, indexation
date, exit date etc.)

However, you might be curious what the individual impact of each input is on the
change of net return. The tornado chart (built with the help of the data table
functionality) can help you to find this out.

In order to figure out the individual effect on e.g. the net return from change of an input
(e.g., vacancy rate assumption), we need to calculate first the scenario under the base
case assumptions (for all inputs). Then we will change our vacancy rate assumption
(only this one assumption) from the Base case (5%) to the Downside case (10%) and
note the resulting net return. After that we will change it to Upside case (2%; again, just
that one input) and note the result again. The differences between the Base case result
and the two results we have just received will be the impact of change of Vacancy rate
on the net return (favorable variance and unfavorable variance). After that we need to
change vacancy rate back to the Base case assumption and do the same test for the next
input (e.g. indexation rate), and so on.

The data table functionality allows doing this calculation automatically for any number
of inputs. In our example the data table (range A28:D35) starts its procedure by
evaluating the Vacancy rate at the Downside case scenario (putting those into cells B39
(column input cell) and cell B38 (row input cell) respectively). Cell E29 returns 10%
(vacancy rate under the Downside scenario) and cells E30:E34 return their
corresponding assumptions under the Base case scenario. These assumptions feed the
model which calculates the net return under the Base case assumptions except vacancy
rate calculated under the Downside scenario. The model returns 12.6 million USD, after
which the data table moves on to the next combination of row and column inputs. If the
data table is evaluated at the Base case or at the Full scenario (for all cases)
assumption, the model returns the result of the Base case, the Downside case or the
Upside case net return.

Table to the right (F29:H34) calculates by how much each of the figures we have
obtained deviates from the Base case. Based on the total swing (from the Downside to
the Upside), the inputs are ranked in descending order, and are used to build a tornado
chart. The name of this chart becomes clear once you look at it. Two vertical lines
represent the net return amount under the full Downside scenario (to the left) and the
full Upside scenario (to the right).

The chart demonstrates that under the current assumptions the yield compression has
the biggest effect on the net return among other inputs. The model is less sensitive to
the exit date, indexation rate and loan interest rate, and changes in vacancy rate and
non-reimbursable opex have the smallest effect on the net return of the project.

Page 10
 Unlocking Full Potential of Excel Data Tables (Part 1)

What’s Next?

Refers to sheet: “Next”

This concludes my introduction to advanced usage of data tables - but there is more!

The second part of this file expands this material further as follows:

– In the examples we have just looked at every data table is linked to a separate
model (we had to repeat the model on every sheet). The second part explains
how to link several data table to the same model without conflicting with each
other and interfering with the inputs of the model.

– By default, data tables require that input cells are located on the same sheet as
the data table. I am showing a way around it.

– There is a built-in functionality in Excel which allows to turn off automatic

update of data tables. This speeds up the calculations but then by looking at a
data table you cannot figure out if it is updated or not. To turn data table
recalculation off and then back on you have to do a few manipulations with your
mouse. My solution is more simple: a check box located next to data tables
which turns automatic updates on and off with one mouse click.

– Creating a waterfall chart using data tables:

Check out my Eloquens page at the link below for the second part of this publication:

https://www.eloquens.com/tool/QkbvsGeb/finance/sensitivity-analysis-excel-
templates-and-methods/unlocking-full-potential-of-excel-data-tables-part-2

Page 11