Building a Dynamic Scenario Based Forecast21 November 2018
Rob Torok, Senior ManagerErnst & Young, Financial Accounting Advisory Services (FAAS)
Page 2
Welcome to scenario based forecasting
1. Introductions 5 min
2. Observations on rolling forecasts & scenarios 25 min
3. Creating a better periodic plan 25 min
4. Introduction to the NISP case study 15 min
5. NISP Part 1: Metrics & Volume Discussion 20 min
6. NISP Part 2: Cost Types & Work Effort Discussion 20 min
7. NISP Part 3: Discussion of Options 20 min
8. Discussion with a CFO 40 min
9. Wrap up & final Q&A 10 min
Break for Keynote & Lunch
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 3 Building a Dynamic Scenario Based Forecast
Key Steps in Forecasting
In-Year Activities
Conduct YTD Expenditure
Analysis
Estimate Non-Committed
Expenditures to Year-End
Review Operational, HR
and Performance Results
Review & Update Commitments
Update Internal Budget Re-Allocation
Update Total Expenditure
Forecasts
Update Rolling Forecast
Quarterly
Update 5 Year Plan
Update Rolling Forecast
Create Annual Plan
21 November 2018
► Retain existing processes but EXTEND to reflect quarterly rolling forecast and multi-year plans
Page 4
Current Process for In Year Forecasts
► Each update ends with the fiscal year, tied to spending authority from Parliament
► But business planning continues, both for day to day operations & projects, albeit without commitments
► One does not assume that all spending stops
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 5
Key Features of a Rolling Forecast
► Extends beyond, and can be independent of, the fiscal year
► Each forecast iteration replaces one period of forecasted data with actuals then extends the forecast by one period
► Each RFC covers the same length of time, e.g. 6 quarters
Year 2
P3 P6 P9 P12
Actuals
Budget / Forecast
Year 5 Year 6
Building a Dynamic Scenario Based Forecast 21 November 2018
Year 1
P3 P6 P9 P12 …
P0
Annual Plan
Rolling Forecast
5-Year Plan
P3Annual Plan
Rolling Forecast
P12
Annual Plan
Rolling Forecast
5-Year Plan
…
P6Annual Plan
Rolling Forecast
Page 6
To be Successful A Rolling Forecast Must Be …
► Cost-effective
► Actionable
► Reliable
► Timely
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 7
Factors for Successful Implementation
Integration Of
Culture
People
Participation
Systems
Process
Design
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 8
Overcoming Challenges
Challenge Solution / Idea
It’s not needed. The fiscal year iswhat we live by – parliamentary authority,
reporting, budgets, etc.“
We would need new people, skills and technology. The cost is too
high“
Forecasts will probably be inaccurate. The further out we look,
the more inaccurate we’ll be.“
The fiscal year is arbitrary and has no connection to
the underlying cycle, so why stop with a fixed date
Value – much cheaper to have greater visibility than
to react to unforeseen situations
Still better – forecasting the possible ≠ predicting the
possible
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 9
Scenarios versus Options
Scenarios Options
► Outcomes or events that might occur at some point in the near or distant future
► Often fall outside of the current business environment and challenge the norm
► May require different analytical tools and capabilities
► Key challenge is to envision them and discuss them openly, far enough in advance to be able to act
► A (more) concrete idea actively being considered and requiring analysis
► Usually seen as adjustments to the existing business, and shorter term
► Fit within current modelling and business structures / capabilities
► Key challenge is the analysis itself, since details and timing are often critical
Example
Impact of self-driving cars on traffic laws and parking enforcement
Example
Planning for an X% reduction in police dedicated to traffic laws and parking
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 10
Scenarios: A Real Example1
1 Fortune magazine, February 2018, “Shell Faces ‘Lower Forever’”2 OilPrice, March 2017, “Shell Sells Almost All Canadian Oil Sands Assets”
March 2017“Royal Dutch Shell sells almost all Canadian oil sands assets”2
Why?
Months of deliberation “conclude[d] that the energy industry was changing fundamentally – in a way that could turn the profitable oil sands operation into a liability”1
How?
The “scenarios” team concluded that global oil demand might peak in a decade, due to faster than expected reductions in alternatives to fossil fuels such as solar, wind, and electricity
If that scenario materialized and Shell still owned oil sands assets then “you were – gosh, forgive me – f --- ed”1
Key Takeaway
Shell did not conclude that this would happen, just that it could, and the consequences of it were massive. And Shell won’t know if it made the right decision until perhaps 2030!
Shell’s Challenge – “Minimize the Maximum Regret” 1
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 11
Scenario Planning
► Open-minded and generally long term assessment of possible outcomes
Encompasses a wide spectrum of possible outcomes, from the highly unlikely to the highly probable
Focus on the ‘the art of the possible’
Objective is to position oneself to avoid or take advantage of certain circumstances, both positive & negative
► Often best handled by a dedicated team
Avoids conflict with day-to-day challenges
Draws on different skills
Expands the analytical time horizon to years and even decades
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 12
Assessing & Planning for Options
► Understanding our current situation and planning for a better future:
How do things look ‘as is’, i.e. without options?
What do we need / want to add to (or subtract from) our business?
► Understand key business drivers
‘Levers’ that affect our performance
Getting a full range of options
What Is What Could Be
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 13
What Is
What Do You Need to Know or Have? Why?
Baseline or ‘as is’ position Must have a starting point to compare against
At the same level of detail as your impact analysis Cannot change what you don’t have
External / macro & internal metricsAct as drivers for both baseline &
options/scenarios
Options to select from Something to add to or subtract from the base
Similar time horizons
Hard to compare & select from options that are
both very short term & very long term (e.g.
price change vs. new market entry)
An open mindWillingness to consider the unlikely or
unpopular
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 14
What Could Be
► How we look at options:
Best assessed in binary fashion, i.e. include or exclude (otherwise one has a virtually infinite number to consider)
Also best assessed on a mutually exclusive basis for the same reasons
• Without this, there would be far too many options
No need to flesh each option out in full detail (i.e. budget line items) until final decisions are made
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 15
Consequences
► Consequences for Clients & Potential Clients, Suppliers, Regulators, Influencers (and more):
• Immediate to longer term actions
• Tangible & measurable vs. intangible & harder to measure impacts
• And include your planned responses to their responses (ripple effect)
Example
If the US government tightens or loosens immigration rules, how will that impact the flow of immigrants to Canada?
Services or combinations of services
Funding sources
Changing regulations
Changing macro environment
Actions of other organizations (provinces, UN, etc.)
Political influence or intervention
Considerations
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 16
Example: Marginal vs. Average based Forecasts
Existing organizational IT budget is $10M, and is currently charged to four programs / agencies (Average of $2.5M each). NISP is added, but only $500K additional IT investment is required for laptops.
Average Marginal
Current IT Budget $10,000,000
Additional IT Expenditure $500,000
Total IT Expenditure $10,500,000
Number of Programs 5
Average Exp. / Program $2,100,000
Laptop Expenditure (NISP) $500,000
Incremental IT Expenditure $0
Total IT Expenditure (NISP) $500,000
Other Considerations:
► Is there a requirement or policy for one or the other?
► Average approach triggers impacts on other programs or agencies, which is especially problematic when average costs increase
► An additional option would be to charge the average of existing IT expenditure to each of the five programs, and allocate the incremental amount to NISP (i.e., $10M/5 = $2M per program; + $500K for NISP incremental laptop expenditure).
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 17
► What decisions might the organization make based on the forecast?
► What actions might be taken?
► Who will be most impacted by the forecast?
► What if the forecast is wrong?
► Where might it be wrong?
► What can I do to improve forecast accuracy?
How do you Want to Use the Forecast?
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 18
► Variances are the inevitable result of assumptions and circumstances to matching forecasts
► But its NOT the variances themselves that need to be explained
► Rather, look behind the variances at the assumptions & drivers that drove the budget/forecast amounts and comment on why they differed:
► Compensation cost exceeded budget/forecast because headcount exceeded plan
► Compensation cost exceeded budget/forecast because several planned retirements were deferred until P6 of next year
Variances: Looking Back
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 19
Better Budgeting & Forecasting
Key Drivers of Bad Plans
► Lack of planning
► Excessive focus on financial data
► Lack of linkage to business operations
There IS a Better Way
► Draws on operational planning, capacity planning
► Incorporates performance management
► Leads to a process valuable to both the languages of operations and finance
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 20 21 November 2018 Building a Dynamic Scenario Based Forecast
The Inter-Relationships of ABPB
ABPB Fundamental Concepts &
Terminology
Operations Planning & Forecasting
Current Budgeting Process
Current Performance
Reporting Process
Capacity Planning Process
Page 21 21 November 2018 Building a Dynamic Scenario Based Forecast
Shift Effort from Budgeting to Planning
Planning
Budgeting Planning
Budgeting
Current Effort Future Effort
Proportion of Benefit Realized
Page 22 21 November 2018 Building a Dynamic Scenario Based Forecast
And From Financial to Operational Planning
OperationalData
Financial Data
OperationalData
Current Effort Future Effort
Proportion of Benefit Realized
FinancialData
Page 23
Overview of The Closed Loop
Stage 1 – Operational Stage 2 - Financial
Feasible
Operational
Plan
Obtain Demands
Determine Resource
Costs
Add Non-Activity
Costs
Review Strategy
Balance Operations
Determine Activities
#
Balanced Financial
and Operational
Plans
Formal
Budget
$
Balance Financials
Begin ...
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 24
Let’s Look at Our New Agency
Finding short term accommodation for new immigrants
• Activity and resource consumption rates based on current ABC information
• Assume:
• Annual period
• Financial target is for funding to equal expenses, +/- 1%
• Homogeneous product
• Some buffer (or contingency) capacity to meet demand peaks, unplanned absences, etc. may be desirable and will be included in activity and service cost
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 25
Stage 1 – Operational Balance
Requests for ST housing
1 hour / search
= 325,000 hours
Operational Balance
Activity
Requirement
Consumption
Rate
Resource
Requirement
Labour
Consumption
Rate
Resource
Capacity
Searching for housing
Demand
Requirements
Strategy first drives ...
230 staff x 1,500 hours
= 345,000 hours
0.65 searches / application1
= 325,000 ST housing searches
500,000 immigration applications
This model includes 20,000 hours of buffer (or contingency) capacity, but we will come back to this concept later.
Building a Dynamic Scenario Based Forecast 21 November 2018
1. Applications may be for multiple persons & some applicants do not need housing, hence < 1.0
Page 26
Stage 2 – Financial Balance
Stage 1 – The Operational Loop
Requests for ST Housing
Operational Balance
Activity
Requirement
Consumption
Rate
Resource
Requirement
#
Labour
Consumption
Rate
Resource
Capacity
Searching for Housing
Demand
Requirements
Strategy first drives ...
Stage 2 – The Financial Loop
Financials
230 staff x $80,0002
= $18,400,000Resource Cost
$18,400,000 / 325,000 searches
= $56.62 / searchActivity Cost
$56.62 x 0.65 requests/app.
= $36.80 / applicationProduct Cost
Funding = $39.00 / application, total of $19,500,000Activity Costs = $56.62 / search, total of $18,400,000
Oper. Surplus = $1,100,000
Other Costs = $950,000
Net Balance = $150,000 which is < 1% of funding
1 hour / search
= 325,000 hours
230 staff x 1,500 hours
= 345,000 hours
0.65 searches/app.
= 325,000 requests
500,000 applications
Building a Dynamic Scenario Based Forecast 21 November 2018
2. Fully burdened with all benefits
Page 27
The Effect of a Changed Consumption Rate
Operational Balance
Stage 1 – The Operational Loop
Requests for ST Housing
Activity
Requirement
Consumption
Rate
Resource
Requirement
Labour
Consumption
Rate Searching for Housing
Demand Requirements
Strategy first drives ...
Change Capacity
+ 40 staff x 1,500 hours
= 60,000 hours
Shortage
Adjust Consumption
60 minutes / search
= 455,000 hours
0.65 searches/app
= 455,000 searches
700,000 applications
230 staff x 1,500 hours
= 345,000 hours
270
405,000
50
380,000
Adjust Capacity
Resource
Capacity
If we start with a
40% jump in
demand to 700K
apps, then holding
all else constant
we’ll need about
480,000 hours of
staff time (incl.
buffer) … but
1
Web search
technology or
process
improvement
allows us to reduce
search time to 50
minutes/search
2
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 28
Adding Buffer Capacity
Operational Balance
Stage 1 – The Operational Loop
Requests for ST Housing
Activity
Requirement
Consumption
Rate
Resource
Requirement
Labour
Consumption
Rate
Resource
Capacity
Searching for Housing
Demand Requirements
Strategy first drives ...
Change Capacity
+40 staff x 1,500 hrs
= 60,000 hours
Shortage
Adjust Consumption
Adjust Capacity
60 minutes / search
325,000 hours
0.65 searches / app
= 455,000 searches
700,000 applications
230 staff x 1,500 hours
= 345,000 hours
270
405,000
50
=380,000
Then we add 16-17
staff as a buffer 3
KEY POINT:
Buffers are added ‘at the end’
rather than at each stage. Imagine
the resource need if we added
10% to demand, searches per app,
minutes per search, and hours per
FTE ….. We’d need dozens more
FTE and then buffer that !!
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 29
Operational and Financial Balance
Stage 1 – The Operational Loop
Operational Balance
Activity
Requirement
Consumption
Rate
Resource
Requirement
#
Consumption
Rate
Resource
Capacity
Demand Requirements
Strategy first drives ...
Stage 2 – The Financial Loop
Financials
270 staff x $82,000
= $22,140,000Resource Cost
(incl. plan for increase next year)
$22,140,000 / 455,000 searches
= $48.66 / searchActivity Cost
$48.66 / search * 0.65 searches/app
= $31.63 / application Product Cost
Funding = $39.00 / application, total of $27,300,000
Activity Costs = $48.66 / search, total of $22,140,000
Oper. Surplus = $5,160,0000
Other Costs = $2,000,000
Total Surplus = $3,160,000 or 11.6% of Funding
325,000 hours
455,000 searches
700,000 applications
230 staff x 1,500 hours
= 345,000 hours
270
405,000
Change Capacity
+ 40 staff x 1,500 hours
= 60,000 hours
Shortage
Adjust Capacity
50 minutes / search
100,000 calls / campaign
Target
Results
Requests for ST Housing
Labour
Searching for Housing
380,000
Adjust Consumption
Building a Dynamic Scenario Based Forecast 21 November 2018
Financial Loop Update – The improved consumption rate allows
funding to be reduced to just about $35 / application !
Page 30
More Realistic: Two Activities & Services
Resource
Capacity
Labour
Initial Searches
Demand Requirements
Follow-up Searches
Requests for ST Housing Requests for LT Housing
Activity Requirements
700,000 applications 700,000 applications
Resource Consumption Rate:
minutes / search
Activity Consumption Rate:
ST searches / appl.
Resource Requirements
0.65
50 m/s
0.10
90 m/fus
554,000 hours
263,000 hours
= 817,000 hours
455,000 searches
210,000 searches
= 665,000 ST searches
0.15 0.30
105,000 searches
70,000 searches
= 175,000 LT searches
Resource Consumption Rate:
minutes / f/up search
Activity Consumption Rate:
LT searches / appl.
700,000 applications @ 0.65 ST searches/application plus 0.30 searches for longer term accommodation/resettlement
700,000 applications @ 0.15 FOLLOWUP ST searches/application plus 0.10 FOLLOWUP searches for longer term
accommodation/resettlement.
665,000 initial ST searches at 50 minutes/search ÷ 60 minutes per hour = 554,000 hours.
175,000 follow-up searches at 90 minutes per search ÷ 60 minutes per hour = 263,000 hours.
1
2
3
1 2
3
Building a Dynamic Scenario Based Forecast 21 November 2018
SIMPLIFYING ASSUMPTION: all initial searches are equal,
as are followups (i.e. no difference in effort between ST & LT
searches)
Page 31
Where Traditional Approaches Go Wrong
► As underlying or external demand changes, that
knowledge is not cascaded to lower-level units
► Consider the iimigration center:
► Does HR know that 30 or 40 staff need to be hired?
► Is IT aware of the new telephony needs?
► Is the training group ready to handle the growth?
► Is there office space for 30 or 40 more people?
► Without this ‘cascading’ significant operational and
financial problems may arise
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 32
The 5 Levers of Budgeting
Stage 1 – The Operational Loop Stage 2 – The Financial Loop
Products & Services
Cost
Assignment
$
Adjust Resource Cost$
Adjust Funding
Financial
Balance
Adjust Consumption
Operational BalanceShortage or Excess
Activity
Requirement
Consumption
Rate
Resource
Requirement
#
#
... and also dictates
Consumption
Rate
Cost
Assignment
Resource
Capacity
Activities
Adjust Capacity
Target
ResultsFinancial ResultsAdjust Demand Demand Requirements
Strategy first drives ...
Resources
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 33
Case Overview & Instructions
NISP:
- New immigration program and the CFO has asked for your help to prepare a budget
- Case highlights
Now:
- Read case study document & review excel exhibits
- Think about two topics: building a budget/forecast and then factoring in options
Before & After lunch:
- Work in table groups
- Respond to each of the questions in the case
- Groups can present solutions then full room discussion
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 34
NISP Planning Request #1
METRICS: What are the top 3 metrics to project client & service level demands?
Exhibit 3 listed 8 metrics (top 3 bolded):
General Inflation
Consumer Purchasing
Nominal GDP Growth
Real GDP Growth
Housing Cost Growth
Unemployment Rate
Global Migration Growth
CDN Attractiveness Rank
Picking 3 requires that we know the intended use of each metric
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 35
NISP Planning Request #2
VOLUMES: Projecting client and service level volumes
See Solutions Exhibit #1
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 36
NISP Planning Request #3
COST TYPES: How does each admin cost type relate to service volume?
Exhibit 2 shows the 8 cost types (comments bolded):
Fin'l Acctg & Reporting: Fixed
Risk & Compliance: Fixed
FP&A - Short & Long Term Accomm'n: Step
FP&A - Lang/Cult: Step
FP&A - Employment: Step
FP&A – General: Step
HR – Office: Step
HR - Client Service: Step
Fixed – does not change with volume ‘over the relevant range’
Linear – changes with each discrete volume increment/decrement
Step – changes with volume but not by individual unit
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 37
NISP Planning Request #4
VOLUMES: Projecting work effort & head count
See Solutions Exhibit #2
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 38
NISP Options Analysis
1. CHANGES IN VOLUME: How can NISP adjust its plans if one type of application grows much faster than planned?
2. SHIFTS IN VOLUME: How can NISP adjust its plans if one type of application grows much faster than planned while others grow much more slowly?
3. GEO-POLITICS: How can NISP flex its plans to accommodate global trends?
4. ROLLOVER OF PLANS: How does NISP ensure that prior agency plans roll over to NISP?
SUGGESTION – Lets have each table discuss one of the 4 questions for 10 minutes, then pick one table per topic to present its ideas.
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 39
NISP Options Analysis, discussion #1
1. CHANGES IN VOLUME: How can NISP adjust its plans if one type of application grows much faster than planned?
Assessment of cost types and trends is key: which costs change on linear vs. step bases, or are fixed?
Is funding tied to volume or must NISP constrain its services or operate differently?
Can components of its budget be shifted to other areas?
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 40
NISP Options Analysis, discussion #2
2. SHIFTS IN VOLUME: How can NISP adjust its plans if one type of application grows much faster than planned while others grow much more slowly?
Usually a much simpler situation to face
Can components of its budget be shifted to other areas?
But need to assess resource skills to ensure service s are delivered by appropriate personnel.
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 41
NISP Options Analysis, discussion #3
3. GEO-POLITICS: How can NISP flex its plans to accommodate global trends?
Reflect back on Shell Oil – does NISP require ‘scenario planning’ resources to pro-actively ‘look outside’ and ‘look [far enough] ahead’ to see potential changes in trends
Is funding tied to volume or must NISP constrain its services or operate differently?
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 42
NISP Options Analysis, discussion #4
4. ROLLOVER OF PLANS: How does NISP ensure that prior agency plans roll over to NISP?
This is a topic for YOU, the experts in the room!
Building a Dynamic Scenario Based Forecast 21 November 2018
Page 43
Initial Questions for our CFO
1. As a CFO how do you assess the completeness &
reasonableness of a forecast/plan submission?
2. Can you describe 3 essential skills that a Financial
Advisor and a CFO must have to perform their role?
3. How do you see the role of a Financial Advisor?
4. In your opinion, how a Financial Advisor can contribute
to better manage the forecasting process and reduce the
financial risks?
5. As a CFO, What do you expect from financial advisors to
help you to manage the initial budget allocations and
initial financial pressures?
Building a Dynamic Scenario Based Forecast 21 November 2018
Top Related