Transcript (Banking), CIA - e-PBO!

Proxy Modeling & Fiscal Impact of Major
Projects
Ram Mathilakath, M.Com. (Banking), CIA
Global Network of Parliamentary Budget Offices
Community Meeting
June 10, 2014, Ottawa
Disclaimer: Any views or opinions included in this presentation are solely those of the author and do not necessarily represent those of the employer.
Need

Parliamentary committee debates are generally issue
based – cost of a public policy, project or initiative.
Decision support info requires:
–
–
–
–
Detailed bottom-up costs
Cradle to grave LCC vs. current FY appropriations – what
parliament sees
Disaggregated information on horizontal issues by
participating departments reported in a consistent manner
Timely costing info to support committee debates
Challenges

Parliamentary appropriations == Costs

Reporting to Parliament is not project specific

Data spread across departments

Coding

NPV of current & future obligations
Approach


Have an open mind & a healthy skepticism - In God we
trust, everyone else brings data
Data mining & construction of waterfalls
–


Data patterns…?
Apply business valuation principles – multiples &
benchmarks
Zero in on why costs are different when compared to
benchmarks
–
E.g.: Asset deployment rate in Afghanistan
Approach

Using ranges
–


High:low estimates
Understanding relationship of costs & drivers in
‘theatre of war’
Leverage power of collaboration & expert opinion
–
–
E.g. use of accelerated dep. on 4% & 8% of capital assets
deployed in theatre of war – CBO discussion
Expert opinions to support development of a consensus cost
estimate by benchmarking against the experiences of other
countries.
Principles

Independent peer review panel to eliminate bias
–

collaboration with CBO, CRS, U of New Mexico,
Queens, PSAB etc
Test of reasonableness
Modeling Options

Proxy modeling – use of surrogate in the absence of reliable data
or methodology
–
–
–
minimizes data collection burden
High reliability when used with caution & minimizes data collection
Provides useful results in much less time

Stochastic modeling such as Monte-Carlo simulation helps
estimate probability distributions of potential outcomes by
allowing for random variation in one or more inputs over time.

Multiples & WACC – similar thought process

Top-down approach/statistical analysis such as trend and
regression analysis to examine the correlation between cost
drivers and costs.
Modeling Options

Deterministic approach is the simplest method used to look at
best estimates. E.g.: Financial projections usually use the most
likely rate with regard to:
–
–
–
default on loan
investment return
Foreign exchange rate etc.

The result provides a point estimate - the best single estimate of
bank’s current solvency position or multiple points of estimate depending on the problem definition.

Challenges:
–
–
Selection and identification of parameter values are difficult except for
seasoned financial analysts.
Real world scenario is that a range of possible outcomes exists and
some are more probable than others.
When to use

When PBO requires life cycle costs

Bottom –up data simply not available

Data cannot be trusted

Time is of the essence

Cannot directly measure what needs to be measured
- an indirect measure provides sufficient insight
Examples of when Proxy Models could
be used

Situation 1: Impact on depreciation costs from troop rotation,
intensity of the tempo of war, usage rates, residual value of the
asset deployed
– e.g accelerated depreciation rate; range for caution
– International bench mark – e.g. accelerated depreciation rates
in the theatre of war; in the same area with similar tempo and
intensity of war
– Mimics thought process for business valuation of privately held
companies

Situation 2: Government makes an IT investment based on a
business case with a minimum ROI of 10% P.A. MC simulation
allows for volatility of investment returns in each future time period
or the chance that an extreme event (change in government,
drastic changes to SLA, huge cost over runs etc) in a particular
time period leads to an investment return less than the guarantee.
Public Policy Question
In-source, outsource, off-shore, P3?



New SSO has to show savings.
Unless mandated by Govt., revenues and expenditures
could vary significantly. They depend on rate of compliance
with SLAs, punitive payments for non-performance,
inflation, % of cost over runs, savings during the period,
and so on.
Savings calculation involves a set of projections, looking at
what is expected to happen, and thus coming up with the
best estimate for revenues and expenditures, and therefore
for level of savings.
Examples


Impact of changes in charge backs on revenues,
savings/profitability and demand
A serious question for government departments
operating on cost-recovery basis
Final Thoughts





Bottom-up costing would be the best approach, but
not always feasible
Cost proxies represents a possible approach,
requires careful judgment
Use a combination of publicly available data, proxies
when relevant, expert opinion and international
bench marks
Power of collaboration!
Don’t get hung up on the academic holiness of
models – focus on using models simply as a means
Notes

Examples of incremental costs attributable to the
Afghanistan
–
–
–
–
–
Accelerated capital asset depreciation and increased reset
costs
Fuel
O&M costs due to increased operational tempo in a theatre
of war
Increased death and disability (veterans) benefits payout
Reservists pay; imminent danger pay and all incremental
personnel costs related specifically to operations
Contact
Ram Mathilakath, Executive Director, Finance,
Canada Border Services Agency, Government of Canada
Tele: +1 613 806 5757
[email protected]