21 March 2003
Investment Appraisal, Output Costing &
1 To explain the differences between Investment Appraisal, Output Costing and Whole-Life
Costing and how they may be used by the Department to improve decision-making.
Investment decisions will increasingly be effected by the move to resource accounting
and whole life costing and, therefore they must be taken in the context of the new
2 The development and introduction of the three distinct types of costing techniques have
emerged as a response to different business management requirements. Government
Accounting states that “Investment Appraisal (or more generally option appraisal) is an
important part of good financial management, and it should always be applied to major
capital developments.” Project CAPITAL (the implementation of Resource Accounting
and Budgeting) was set against a vision of “a Department managing its resources at all
levels on an output basis”. To achieve the vision, CAPITAL aimed “ to integrate
commercial accounting, output costing and management planning into a single
management regime….” Resource Accounting also introduced the charge on capital and
depreciation. These are dealt with in the Resource Accounting Policy Manual – JSP472.
Soon after, the Strategic Defence Review called for the introduction of Whole-Life Costing
(WLC), which was aimed at the costs of procurement and logistics so that the cost of
equipment was examined through out its life, not just at its introduction.
Costing information for better decisions
3 Costing, in its own right, is not a value added activity in the MoD. It has value only insofar
as it helps managers to take better decisions. In the Department there are four main
categories of decision where better costing information should improve the quality of the
a. Strategic planning: eg what is the impact of increasing this defence capability or
what can we do to reduce expenditure on this equipment by say £400m over the next
b. Investment decisions: eg which of these options is the best value for money? Is
the preferred option affordable?
c. Output costing and output management: eg what is the cost of this Customer
Supplier Agreement (CSA)? What happens if we flex the volume of demand by 5% in
the first year? Where should we focus our improvement programme?
d. Cost management and forecasting: what do we need to charge for this service to
cover our costs? What would be a reasonable charge from contractors? What
are the forecast costs of this equipment maintenance programme?
21 March 2003
4 Although there can be overlaps between these types of decision, each will require
different information in terms of scale and range, timeframe, cost components, accounting
principles and degree of accuracy. The characteristics and linkages between the major
types of costing information are discussed below.
5 Costing at the strategic level is aimed at addressing questions such as how much of the
Department’s resource control total is consumed in supporting strategic capability. Such
as, for example, Deep Strike, and what the options are for reducing spending in this area
by £xmillion over the next y to z years? Answers to some questions can be found directly
through data held in the financial accounting systems. For example, the costs of
operations are usually recorded against specific codes, whereas the cost of some
functions (or intermediate outputs, such as training) can be readily related to the
organisational structure. However, management responsibility and hence costs of
defence equipment capability beyond the initial procurement phases (CADM) are shared
by DPA, DLO, Front Line Commands and Personnel and Training Commands.
6 The WLC Project has developed the prototype Options and Affordability Toolset (OATS)
that summarises the Cost Of Ownership statements (COOs) for defence equipment
projects and allows these to be summarised and reported by DEC, force element or
organisation. OATS has been designed as a database rather than as a traditional
accounting hierarchy and therefore has the flexibility to incorporate developments such as
“Defence lines of development” or “enabling components”, so long as the relationships
can be identified and expressed clearly.
7 Strategic Planning assumptions are high level and, as such, require only high-level
“roughly right” costings to support them. But as plans cascade down towards the delivery
end of the organisation i.e. BLB and/or Agency, and the time horizon switches from the
long to medium and short term, so the granularity of costing, and the depth of planning
assumptions required to support it should increase as reflected in the Through Life
Management Plan (TLMP). The objective of WLC is to support the more strategic
decisions about projects, where there are long lead times; it should prompt the right
questions in good time, when it is more useful to be “roughly right” now, rather than
“precisely wrong” later.
8 The introduction of Resource Accounting and Budgeting (RAB) has not changed the way
in which value for money is assessed. Value for money is measured in a cost-
effectiveness analysis, which in MoD’s case has been expressed as the Combined
Operational Effectiveness and Investment Appraisal (COEIA). The COEIA relies on a
forecast of resource flows and other benefits. These are discounted to give a net present
value or more appropriately, a net present cost. It should be noted that IA is constructed
to support a major decision and, as such, will seek to highlight differences between
potential options. Consequently, it will present information at a common price base,
excluding VAT and other transfer payments and ignoring costs that are common between
options irrespective of their materiality. As a result, the IA itself will not directly address
the question of affordability.
21 March 2003
9 JSP 5071 sets out the Department’s policy on investment appraisal and evaluation and
specifically requires not only the value for money assessment but also the consideration
of affordability, including affordability in resource (RAB) terms. The construction of the
Cost of Ownership profile facilitates this analysis and, when combined with OATS data,
offers an overall context against which long-term affordability judgements can be made.
10 In the case of Defence equipment, the key planning assumptions underpinning both the
creation of the IA and the COO will be held within the Master Data and Assumptions List
(MDAL) of the TLMP. However, in order to facilitate through-life management and to
support the Equipment Planning process, the TLMP and COO should be updated as part
of the annual planning process.
11 The IAB Sec has published guidance2 that enables the Investment Approvals Board (IAB)
to approve projects on a RAB basis as well as the COEIA. This guidance covers both
defence equipment and other infrastructure (eg estates and IT); and also capital
expenditure by the Department directly and via PFI/PPP schemes. The RAB approval
adopts the same principles as Whole-Life Costing.
Output costing and output management
12 The CAPITAL vision was for a “Department managing its resources at all levels on an
output basis”. This included an expectation that outputs would be identified, costed and
used as part of a system of Customer Supplier Agreements, that would be a key part of
the planning round. To date, few CSAs contain clear expressions of outputs and costs;
fewer still articulate the relationships between costs, activities and outputs; and very few,
if any, form the basis of quasi-contractual negotiations. From STP03 onwards, however,
all STPs are to be supported by CSAs, except for overheads and centrally provided
services. There will be a structure of 4* CSAs between TLBs plus 2* CSAs between
HLBs or management groupings.
13 All costs in supplier budgets should be attributed to a CSA agreed with customers. So
far, TLBs have addressed the task in varying ways, reflecting their approaches to output
costing. The DLO and CINCFLEET have taken a whole-TLB approach, implementing an
activity based costing system known as Metify. Activity Based Costing (ABC) uses the
costs and relationships in the underlying processes and activities to relate changes in
volumes of activities to changes in output costs. This enables Metify, in common with
some other ABC tools, to be able not only to model from inputs to outputs but to run the
model in reverse, modelling changes in outputs and expressing these in terms of input
costs. This has the potential to strengthen the STP process.
14 Where TLBs have based the output costing on the STP, and where the output focus is on
defence equipment, it is possible to use the same source information for both output
costing and as a basis for the COO. However, there is a need for the clear distinction
between the two analyses to be recognised. The output cost model works at DFMS Chart
of Accounts (CoA) Level 2 (ie around 40 lines of data) for a single TLB (or sub-set)
detailed across the 4 year STP period or less. Whilst the COO covers six high-level
operating cost categories, extending from the present through to the out-of-service date,
MOD Guide to Investment Appraisal and Evaluation
Smart Approvals – Instructions and Guidance on IAB and Delegated Approvals, Edition 7, September 2002
21 March 2003
and covers all material stakeholders across all of MOD. The detail of other stakeholders
needed for WLC may be unnecessary in the output costing models. Forecasting costs
and activity volumes in detail beyond the STP period is a pointless exercise; and building
large unwieldy models is resource intensive without improving the quality of decision-
15 Output management implies not only planning, but also review of actuals. The current in-
year management (IYM) processes will normally require a consideration of outputs in
order to understand variances, as part of the monthly routine, but the use of output
costing opens up the possibility of reviewing planned output costs against actuals. Output
costing is not strictly required for in-year management, but it will open up scope for further
analysis. Experience elsewhere suggests that it will be better to review output costing
variances no more than twice a year, since there will be too much “noise” in the data.
16 In contrast, there is no plan at present to record WLC actuals, until the process has
settled down. A further difference between output costing and WLC is the treatment of
overheads. Within a TLB or HLB, output costing will typically try to identify the full cost of
the output, including overhead costs of the HQ. In contrast, the COO template omits
overhead costs, where there is no clear relationship between the overheads (eg Central
TLB costs) and specific defence equipment. It is reasoned that overheads should be
omitted from the COO on the grounds that:
a. decisions on specific equipment are not dependent upon the
apportionment of these overhead costs;
b. the increased accuracy of the full costs is spurious at this level and
c. overhead costs are better managed down through mechanisms other than
defence equipment planning.
17 In order optimise the synergies between Output Costing and WLC there would be, ideally,
a Departmental framework. This would allow TLBs to develop the detail as they needed
for local decisions, but would structure the output costing models so that information
could be shared across TLBs and with other processes. This does not mean that there
should be a single system. On balance, it would be more practical to co-ordinate more
closely the development of output costing and WLC, rather than to try to create a single
all-encompassing model. In the DLO the proposed development of output costing using
“Tier 1 (Platform) ” and “Tier 2 (System or Equipment)” IPTs is a helpful further
convergence of output costing and WLC.
Cost management and cost forecasting
18 In most businesses there is a constant need to examine and manage costs at the local
level and to forecast costs of projects. These activities can typically be distinguished from
the output costing discussed above by the level of detail required. Examples of the
costing in this category would be the detailed activity based costing centred on individual
processes, services or manufacturing plants. Various management accounting
techniques are available to assess product costs and to underpin pricing regimes.
Detailed cost forecasts will also be required for project management and in contract
21 March 2003
19 The Department, understandably, does not have a single one-size-fits-all approach to
these costing activities, although in the case of cost forecasting a range of models,
employed by DPA/Pricing Forecasting Group (formerly SPS), is being validated by D
Scrutiny & Analysis for use to support investment decision making and project
management. Given the wide variety of the models used there is no standard way in
which they can be linked to other costing and planning processes.
20 For some areas it may be possible to use the General Ledger (in the financial accounts)
or to use the output costing models, particularly to shape the questions and to identify
areas to focus on. However, one-off decisions, cost reduction exercises and more
detailed decision support will need more information about the controllable cost items and
will typically need to make decision-specific assumptions about which costs are fixed and
which are variable.
21 The following table summarises some of the key differences between the four types of
decision taking discussed above.
21 March 2003
Strategy Investment Output Cost management
decisions management and forecasting
Scale and Departmental scale Departmental; in Business unit eg Process, element
focus expressed in terms some cases wider Agency or IPT to of business unit or
that decisions are economic issues are understand the full part of a platform.
being based eg also relevant. The costs of the unit’s Local focus
platforms, force focal point is the outputs set out in a typically.
elements or defence project CSA.
Timeframe From STP period (4 From today’s date to Short to medium term Various depending
years), through EP out of service eg next budget year on decision eg
period (10 years), to (including disposal to STP period. planning period,
long term up to 30 period). contract duration.
Cost Related to the The IA should include All costs of the May omit common
components resource control all material costs, but business unit plus the costs and
framework eg distinguishing costs that are overheads on
Resource DEL and between options communicated into marginal cost
Capital DEL. Costs needs may ignore that unit eg the costs decisions. Likely to
could exclude common costs. The of accommodation or focus on areas on
overheads when affordability overheads should be to the decision is
decisions on marginal assessment should passed through to the most sensitive.
changes are being be based on the business unit to give Specific rules apply
explored eg defence RAB-based WLC a full cost. This to PSC
equipment costs cost of the should be related to calculations.
could exclude Central investment, related to the resource
TLB or other generic the control accounts
support functions framework.
Accounting RAB basis for Cash for the IA; Using the same data Calculating costs
principles elements related to excludes tax and as the STP as the from first principles,
the control transfer costs. reference point for concentrating on
framework. For Affordability needs to the planning process. the areas that are
periods beyond the be assessed in the In year management going to be most
STP, constant prices context of the control will require actuals as decisive. Primarily
are easier to framework ie RAB. well as budget. cash based, but
understand care must be taken
to reflect RAB
Degree of “Roughly right” within The IA accuracy Detailed enough to Detailed enough to
accuracy the timetable for the needs to focus on create sensible take the related
decision and the those areas to which relationships between decision on a
materiality levels. the decision is most activities, costs and sound footing.
sensitive. Careful outputs. Accurate Sensitivity analysis
assessment of enough to prompt the may help the
affordability required correct management decision making
Other Consistency is Needs to follow Typically built using May be adjusted for
important along with Treasury guidance activity based costing risks (eg PSC) and
clarity in the where IA is software & has strong probabilities (eg
assumptions concerned. requirement for cost modelling)
17 March 2003
Page 7 of 7