Guidance

Early financial cost estimates of infrastructure programmes and projects and the treatment of uncertainty and risk

Updated 26 March 2015

0.1 Objective and context

This guidance supports better appraisal of the costs of infrastructure projects and programmes and more realistic estimation of contingency budgets. Cost estimates publicised in the early stages of appraisal can have a significant influence on expectations and subsequent behaviour leading to excessive cost estimates, over provision of contingency and inefficient project cost control. The guidance is in line with HM Treasury’s Green Book and Managing Public Money guidance.

The findings of the Infrastructure Cost Review 2010 recommended improvements to the estimation and management of contingency budgets. This guidance focuses on a framework for the production of early stage estimates, referred to hereafter as the Anticipated Final Costs (AFCs). See section 2 for further detail.

A distinction is made between organisations running programmes and pipelines of projects which can collect and exploit data from past projects and relatively unique one-off projects where relevant empirical data may not be available. It is also recognised that many unique large scale programmes are susceptible to the treatment recommended here when considered at the level of the smaller component projects that make them up.

0.2 Applicability and scope

The guidance relates to early project cost estimates and allowances for contingency in infrastructure programmes concerning, transport, solid-waste management, energy, communications water, waste water, flood risk and coastal erosion. However, it could be applied in other sectors for example schools, housing, hospitals where there are programmes of work involving creation of many essentially similar assets.

Where programmes and projects have not attained cost estimating maturity at level 4 or 5 on the NAO financial management maturity model or equivalent they will need to continue using the standard allowances for Optimism Bias set out in the Green Book supplementary guidance on this subject*.

The focus of this guidance is appraisal and evaluation and it concerns the allowance for contingency in early project cost estimates, it does not deal with the estimation of benefits, risk management or project management.

*http://www.nao.org.uk/help_for_public_services/financial_management/fmmm.aspx

1. The key principles of the guidance

Change	Advantages	Practical Advice* / Comment
A: Cost & Risk Estimation
The Standard Green Book recommended Optimism bias allowances should be used where adequate reference data is not obtainable. Measures should be taken to collect and utilise appropriate cost and risk outturn reference data to reduce reliance on generic Optimism Bias provisions. Projects should look for early transition to the use of quantified risk assessments as the basis for determining AFCs.	Removes ‘fear of the unknown’ from project costs, either releasing £billions from budgets or delivering the same outcome for less. Avoids project costs expanding to fit over-sized early estimates.	Any early-stage contingency for risks will need to be rigorously justified. Individual organisations may want to use detailed early-stage risk registers for maximum rigour.
Early-Stage total project costs to be quoted as estimate ranges rather than single point estimate figures.	Avoids having to give single point total costs early in a project when such accuracy is impossible. Reduces incentive to inflate total project estimates (to make early forecasts achievable, and appear spuriously accurate).	Strong industry consensus against spurious early-stage accuracy.
Cost & risk estimation to be integrated.	Limits duplication and double-counting of risk monies within cost plans and risk registers. Will lead to leaner budgets being established, and improve overall project budget estimation.	Integrated three point estimating using Monte Carlo simulation techniques (involving base cost uncertainty and risk register items), may be of benefit.

*An Industry report on cost estimation and contingency is available. View the report

2. The nature and structure of early cost estimates

2.1 The importance of early estimates

The early estimate of a project’s Anticipated Final Cost (AFC) is important in determining both the actual and perceived success of the project. Early estimates are also inherently uncertain and can result in the project cost being misrepresented and misunderstood.

Early cost estimates are vulnerable to influencing behaviours when allied to the desires of key stakeholders to ensure a proposal secures funding and meets (sometimes conflicting) objectives. These factors can undermine reliable cost estimation and representation of the risk and uncertainties and so affect delivery of public value.*

*By for example proceeding with a project or programme which on completion its cost to benefit ratio ends up lower than expected and lower than an alternative option that was excluded at an earlier stage of the investment decision process.

2.2 The nature of infrastructure cost uncertainties

The contingency element of any cost estimate needs to account for the likelihood and cost impact of three groups of factors:

• specific risks that are measured uncertainties,
• defined but unmeasured uncertainties around the estimate
• unknown uncertainties that at a given time are not known or understood

For projects that are relatively self-contained, contingency is primarily concerned with the first two of these. In the early stages of large infrastructure projects, a significant proportion of the risk exposure comes from the last bullet above, the unknown unknowns. These may derive from complex interfaces with the physical environment into which the infrastructure is to be built* (i.e. the route of the transport link or the site of a flood defence) and the unpredictable responses and requirements of stakeholders affected by the siting or performance of the infrastructure assets.

*Though the work of the Inspire programme aims to provide greater depth of information of what exists in a given place.

These issues may not become clearly defined until well into the detailed design or even early implementation phases (though the latter should be avoided wherever possible as it could mean incurring unnecessary extra costs). The number of undefined uncertainties should be regularly challenged and ultimately reduced through the application of quantified risk assessments as design and project maturity develops.

Experience shows that the same categories of uncertainty tend to recur across projects of similar type. Therefore the effective use of appropriate reference data should allow the cost impacts to be sufficiently well understood to form part of a reasonably robust estimate - even at early stages in the project life cycle.

The list originally produced by Mott MacDonald is a good starting point to undergo a structured assessment of the areas under which reference data should be sought.

Procurement

• complexity of contract structure
• late contractor involvement in design
• contractor capabilities
• government guidelines
• dispute and claims occurred
• information management system
• other (specify)

Project specific

• design complexity
• degree of innovation
• environmental impact
• other (specify)

Client specific

• inadequacy of the business case
• large number of stakeholders
• funding availability
• project management team
• poor project intelligence
• other (specify)

Environment

• public relations
• site characteristics
• permits/consents/approvals
• other (specify)

External influences

• political
• economic
• legislation/regulations
• technology
• other (specify)

2.3 Cost maturity

The need to use generic Optimism Bias or any other provision for unknown uncertainty in an AFC, implies that the cost estimate is likely to be ‘immature’. That is, there must be significant elements of the project (and perhaps even of the requirement) that are not defined or understood.

In systematically developing the AFC from the bottom-up there will come a point where for the first time the cost estimate can be said to be ‘mature’. This is a critical point in the project life cycle which should coincide with key decisions relating to budget setting, business approval and related commercial strategies. In order to demonstrate readiness to make these critical decisions, the point of cost estimate “maturity” can be defined by the achievement of the following criteria:

1. The project design - including its interfaces with the physical environment – should be sufficiently understood that the Base Cost is broadly complete (although still subject to refinement, and potentially growth, as the design becomes better understood and more developed).
2. A risk analysis should have been undertaken at a level sufficient to recognise all significant risks and allow them to be assessed using QRA techniques.
3. Any remaining contingencies within the risk allowance should relate to project specific risk items. In other words there should be no unallocated risk exposure and therefore contingency (that is contingency accounting for indefinable/undefined uncertainty).

It may be argued that the third criterion above is a residual optimism bias. However irrespective of this, the cost estimate used to determine the value proposition in the economic case must not be automatically converted into the final cost estimate for the financial case (see 2.8 and section 4) and subsequent budget.

This should also be the point at which:

• the organisation should feel comfortable, in budgeting terms, committing to go forward with a preferred option
• sponsor and client stakeholders are willing to commit to the AFC as a sufficiently reliable benchmark to assess the subsequent cost performance of the project

Readiness to make these commitments acts as a test of the estimate maturity. If either the organisation or the project feels uncomfortable in making them, then it is unlikely that estimate maturity has been reached. The reasons for this discomfort should be identified and addressed before proceeding.

In most cases this point of cost maturity coincides with the production of the Outline Business Case.

Until this first comprehensive and reliable iteration of the AFC has taken place; any decisions or communications relating to the AFC should reflect its immaturity and uncertainty. This is discussed further below.

2.4 Anticipated final cost

The Anticipated Final Cost (AFC) for any project or programme is, at its simplest, made up of two components:

• a Base Cost (which is the value the estimators believes represents the most likely expenditure required to deliver the requirement)
• a Risk Allowance (referred to hereafter as a Financial Risk Exposure “FRE”) which – in relation to the Base Cost - needs to account for (to varying degrees over the course of the project’s life) the cost consequences of:
  • the development and refinement of the design.
  • the greater understanding of the solution’s interfaces with its physical environment
  • legitimate changes in requirement scope*
  • a reducing provision for other areas of uncertainty which are not addressed by the above 3 bullets (as provided for through Optimism Bias in earlier high level assessments)
  • specific risks e.g. changes in key personnel during the project, pending legislative changes which would impact on the project

*If this is within the functional output commitment of the project. See also ‘The Importance of Project Scope’ in Section 5.

Figure 1 illustrates a simplified model of the three key stages of AFC development over the project life-cycle. Throughout, it stresses an overall reduction of total estimated cost and risk, reflecting the need for intensive, active management of these items to deliver leaner projects. Also, cost and risk assessment should be approached in an integrated way to avoid duplication.

For the initial AFC, the Base Cost is relatively low, reflecting the low level of definition of the solution required to achieve the scope. Simultaneously the Financial Risk Exposure needs to account for the project risks and those aspects of the overall solution which have either yet to be considered or are not understood.

Projects should:

1. Strive for earlier definition of the risks and uncertainties facing the scheme, as early as possible within the project life cycle, ideally using ‘bottom-up’ techniques from the earliest point.
2. Leading organisations will make use of Quantitative Risk Analysis, Reference Class forecasting, and other relevant tools to achieve this before Strategic Outline Business Case stage.
3. Leading organisations will seek to justify all out-turn estimates as rigorously as possible. In all cases, use of Optimism Bias and similarly undefined cost uplifts should only be used if minimal risk definition can be undertaken, or very limited historical precedents exist for the project (e.g. global ‘one-offs’).
4. This early definition of specific risks and uncertainties facilitates active management; leading organisations will demonstrate intensive mitigation of these areas.

The mature AFC in the Figure shows that the Base Cost is larger than at the initial stage (reflecting improved definition of the requirement and preferred solution). At this point:

1. Major areas of cost uncertainty should have been clarified, with uncertainty only remaining around price variations and localised issues.
2. Ideally, only specific risks should remain in the Financial Risk Exposure.
3. Any residual provision for unallocated risks and uncertainties at programme or project level (referred to below as Management Contingency) should be identified and held at arms-length and not included in project level AFCs.

Correspondingly, the Financial Risk Exposure must be generated through a ‘bottom-up’ QRA, based on a comprehensive Project Risk Register and associated mitigation plans.

An organisation with a mature financial management and control capability would be expected to achieve an outturn cost below or close to the mature AFC or at least understand where the difference originated from. This should be fed back into the development of AFCs for future projects and programmes of work of a similar nature.

Once the AFC is fully mature (which is an expectation at the OBC stage) and the project budget has been fixed (at the end of the initial design phase):

• An element of the Financial Risk Exposure – normally addressing strategic and external risks beyond the control of the project (e.g. major political risks) – may be held by the business or externally as Management Contingency
• The remainder of the Financial Risk Exposure may in whole or in part be held/controlled by the project:
  • stand-alone projects may hold most or all of the Financial Risk Exposure under their control
  • projects within programmes are likely to see more of the Financial Risk Exposure held at programme level as both a means of control and also to allow hedging of common risks across a number of projects
• balance will need to be struck between financial rigour and the difficulty in accessing funds set aside for risk events in a timely fashion from corporate centres due to complex release processes
• duplication between organisational and project levels should be avoided, and managerial attention should consider how to achieve this

2.5 Mature and immature cost estimates

A Financial Risk Exposure will form part of the AFC throughout the development of the project but at the immature stage its derivation will be obviously less detailed than at the mature stage, where its derivation will have been subject to far greater detail and rigorous challenge.

The approaches recommended for the derivation of the Financial Risk Exposure for immature and mature estimates is set out below.

In setting out an approach to cost maturity that can be applied across a wide range of projects and programmes, it is important to recognise that:

1. initial estimates may be produced at the different points in the project life cycle
2. projects at the same stage in the life cycle can have widely differing (and inherent) profiles of uncertainty relating to their costs

Both these factors will affect whether an early project estimate is mature, and, if not, when and how quickly it will become so.

A clear example of an initial estimate that is immature would be where an AFC is required for a strategic analysis of river-crossing options where there are large uncertainties regarding tunnel/bridge design options, external stakeholder reaction, and the issues to be faced integrating the solution into its physical environment.

At the other extreme, in the case of a say a managed motorway project involving the construction of a near-commoditised, uncontroversial solution in very well surveyed terrain, the initial estimate should be fully mature and based on reliable reference cost and risk comparables.

2.6 The importance of project scope

An AFC must always relate to a particular project, a delivery timescale and a specific and unambiguous definition of the project’s scope (requirements, functionality and benefits).

At a certain level, the ‘scope’ of major infrastructure projects tends to remain uncertain and liable to change for a significant proportion of the project lifecycle. This reflects the fact that the interfaces with the physical environment are frequently complex. This means that a degree of legitimate* scope change will need to be managed and it follows that cost contingency will be required in order to allow such change to be implemented.

*c.f. non-legitimate change which would be where a significant and unplanned for scope change materially affects the validity of a previous appraisal.

It is therefore necessary to ensure that controls are in place on the scope to prevent undesirable trading-off between scope and cost, specifically:

• contingency being spent on scope change which is not necessary/legitimate
• scope being reduced to avoid over-spend.

It follows that the effective scaling and management of cost contingency requires the effective control of project scope through unambiguously:

• a statement of scope (setting out requirements, functionality and benefits)
• a definition of within what limits change to the scope is deemed ‘legitimate’ (that is, what is allowable within the parameters set by the project’s approval) and therefore costed for in contingency
• an effective control process, with access to additional funds if appropriate, to be followed where the scope needs to change beyond these agreed limits

The use of the 5 Case business model and good approvals processes can help ensure that ‘scope creep’ is controlled.

2.7 Interaction with the business case

As part of the decision to proceed from SOC to OBC, the strategic outline business case for a project should be tested against the full range of the ‘immature AFC’. Organisations should develop their own testing approach; chosen options should remain valid and robust even if the cited AFC rises by a significant amount. This stress testing will also help decision makers determine at which point a ‘stop’ decision point would be reached.

2.8 Application to large projects or programmes

The guidance on public sector business cases using the 5 case model must be used. This identifies Programme Business Cases and Project Business Cases, the latter typically involving three stages of development, the Strategic Outline Case (SOC), the Outline Business Case (OBC) and the Full or Final Business Case (FBC). There is some flexibility in this arrangement depending on the need for approval of procurement. For example where competitive dialogue is the chosen procurement route then there will be two approval stages for the FBC.

The programme business case may, where no procurement is required at the overall programme level involve only two stages, the SOC and the OBC, all procurement taking place at the sub programme level which is where delivery of the programme is planned and implemented in detail.

Programme Business Case (PBC) (previously known as Strategic Outline Programme Business Case (SOPBC)

For all major proposals requiring a number of sub projects and programmes a Programme Business Case (PBC) should be produced in support of the proposed scheme as set out in the Treasury Business Case Guidance. Among other things this should set out the need for change, how the proposal supports the organisation’s business strategy and the objectives that the scheme will support or deliver. The outcomes that are the objective of the scheme are defined in the SOC stage of the PBC which also examines a long list of alternative delivery options and narrows this to a short list, identifying indicative costs and benefits. At the OBC stage the PBC identifies the preferred way forward, it clarifies and plans the programmes, and sub-programmes, required for successful delivery and establishes an agreed spending envelope for the programme as a whole. In the absence of the need for a programme level procurement any further stages of the PBC are concerned with oversight of the progress of the programme and monitoring the associated spend within the agreed spending envelope and delivery of the planned outputs.

Strategic Outline Case (SOC)

The purpose of the Strategic Outline Case (SOC) is firstly to establish the case for change and the need for the spending or investment and secondly having examined a wide long list of options, to provide a short list of credible options and indicate a potential way forward for the scheme to obtain the early approval of management. Consequently, it provides the basis for the ‘initial agreement to proceed’ with further work on developing the scheme.

Outline Business Case (OBC)

The purpose of the Outline Business Case (OBC) is to identify the preferred option from the shortlist which optimises value for money (VFM). Then for the preferred option to conduct a more detailed risk analysis, creating both risk and benefit registers and to identify worst case scenarios and to explore switching values. The OBC prepares the scheme for procurement and puts in place the necessary funding and management arrangements for the successful delivery of the scheme.

Full (or Final) Business Case (OBC)

The purpose of the FBC is to identify the ‘market place opportunity’ which offers optimum VFM, set out the negotiated commercial and contractual arrangements for the deal, demonstrate that it is ‘unequivocally’ affordable and to put in place the detailed management arrangements including risk management for the successful delivery of the scheme.

The challenge with large projects or programmes is that at differing points along the timeline, elements of the scope and requirement will have varying degrees of clarity and surety and they may need to be revisited as one element changes and that change impacts on another elements. For example, a technological change in the communications network from a fixed line requirement to a mobile/radio requirement would mean the design might require less trenching and more masts and very different costs and delivery timescales.

Each sub-programme or project within a programme will, if it requires procurement, be subject to separate business case approvals if a spending authorisation is required or if it is regarded as novel, risky or contentious.

Achievement of a fully quantified risk assessment with mature AFC’s and the replacement of generic Optimism Bias allowances across all of the elements of a large Programme at the OBC stage would be virtually impossible and attempting to do so could seriously delay delivery. To prevent this it is expected that the generic OB allowances will be replaced in stages as the cost estimates are developed during development of the Programme and its constituent sub programme and sub project business cases, the requirements for this are set out in the following table.

Programme Business Case (PBC) - Previously called the strategic outline programme of PBC	Programme level generic OB may be applied to the financial case estimates but it must be clear which categories of risk are included in the OB. Broad brush uplifts should be challenged. Delegated funding envelopes for project management purposes must not include OB uplifts.
SOC for projects within a programme or elements of a large project	Greater disaggregation of OB into specific categories preparing the structure for a full QRA. Delegated funding envelopes should not include generic OB uplifts. Any residual unallocated risk allowance remaining should be centrally retained. Leading organisations will undertake bottom-up risk analysis and reference class forecasting based on empirical evidence base.
OBCs within a Programme or for elements of a large project	Any residual OB in financial estimates will not be included in delegated programme or project funding. Allowances for contingency should be centrally retained. Full QRA assessments with risk split between project, programme and management contingency are required.
FBCs for projects within a programme or elements of a large project	OB financial estimate completely replaced. Any residual risks fully quantified, tracked and managed. Provisions for contingency should be centrally retained.

3. Establishing initial cost estimates

3.1 Introduction

The initial cost estimate of any project needs to be sufficiently reliable to ensure that:

• any options appraisal, of which it is a part, can identify the most economically and socially advantageous option for delivering the business objectives
• the sponsor and delivery organisations can determine whether adequate funding exists within their budgets, and the project is therefore affordable

In meeting these needs, and from an estimating and cost management standpoint, the best situation is one where the initial estimate of a project’s AFC can be produced bottom-up from activity and product components and is, in the sense described in Section 4, ‘mature’.

In order to achieve this, organisations should consider the extent to which:

1. the solution to the project in question is of a commoditised nature
2. historical cost and risk data from similar projects is readily available and reliable
3. it is possible to make use of cost and risk estimating models that are proven to produce reliable comprehensive outputs at this stage
4. the organisation has the skills, resources, capability and time to undertake the level of analysis required

Whilst it is recognised that all of these factors may not be present for all organisations/projects at an early stage of development, it is incumbent on organisations to seek to build towards incorporating them all into their approach to the assessment and development of their projects and programmes.

3.2 Key principles

Move towards a systematic and application-specific approach to building base cost and contingency estimates.

Ensure that the ‘Financial Risk Exposure’ is rigorously justified and developed with more detailed cost and risk analysis as soon as possible.

Cursory application of Optimism Bias provisions should not be used as a substitute for accurate estimating at any stage.

3.3 Approach

In order to produce a credible AFC that is properly representative of the expected out-turn a two step approach should be used:

1. generating the initial estimate
2. testing the estimate against appropriate project out-turn reference data

Generating the initial estimate

Within organisations that collect and make use of empirical project data, the areas of cost uncertainty that need to be incorporated into an immature AFC should be well understood (even if specific risks and uncertainties are not). It is then the responsibility of the estimating team to ensure that any provisions for cost, risk and uncertainty are mutually exclusive and collectively exhaustive. The objective should be to ensure that costs and risk provisions are not duplicated in the AFC.

For organisations with mature costing methodologies, the derivation of the Financial Risk Exposure should be clear from a combination of experience and an analysis of the cost performance of past projects.

In developing the Financial Risk Exposure, a clear understanding should be gained as to the degree to which the defined elements of cost (Base Cost and any initial risk analysis) are inadequate to cover the cost impact of the areas of uncertainty in the following table.

Uncertainty Area	Description
Design Definition	Legitimate changes and refinements to the design
Stakeholder Impact	Legitimate changes to the design/scope forced by public enquiries and other external stakeholder interventions
Physical Integration	Implementation/integration costs required by the specifics of the geography
Site Constraints	Environmental issues (including impact of wildlife habitats):
	Archaeological sites
	Changes in land use
	Ground Conditions
Residual Risk	Areas of risk that are particular to the project in question and that are not already included in the base costs. E.g. changes in Law or Government policy change

Financial Risk Exposure is not there to account for the cost consequences of:

• poor management of the project; or
• early under-estimating caused by psychological (e.g. genuine optimism) or political / institutional issues

These factors are better addressed through governance providing heightened scrutiny, transparency and accountability of the AFC estimating process.

Where these factors are to be addressed in the financial risk exposure it must be made explicit and justified and if necessary included in the management contingency.

Testing the estimate against appropriate reference outturn data

The most reliable way to crosscheck any early estimate (and/or scale of the overall Financial Risk Exposure) is using out-turn data for comparable projects (see below). Where whole project data are available it can be applied through straightforward benchmarking or by analysing the cost performance (under- and over-runs*) of comparable projects (see below: Reference Class Estimating).

*Referenced against the applicable point in the project life cycle when the initial estimate is being made.

If the normalised benchmarks are significantly higher than the estimate, this suggests that it is prudent to include adjustment to the bottom-up calculation to inform the overall Financial Risk Exposure.

When using benchmark data for these purposes, it is important to understand its origins and particularly the degree to which the key cost drivers of individual projects are directly comparable (e.g. risk profile, nature of contractor engagement, contract model etc).

Reference Class Estimating

Reference Class Estimating is a forecasting tool, used to predict the outcome of a planned action, based on actual outcomes of similar actions taken in the past. It involves three steps:

1. Identification of a relevant reference class of past, similar projects. The class must be broad enough to be statistically meaningful but narrow enough to be truly comparable with the specific project and adjusted for known changes such as new safety rules or improvements in technology.
2. Establishing a probability distribution for the selected reference class. This requires access to credible, empirical data for a sufficient number of projects within the reference class to make statistically meaningful conclusions.
3. Comparing the specific project with the reference class distribution, in order to establish the most likely outcome for the specific project.
4. Establishing a probability distribution for the selected reference class. This requires access to credible, empirical data for a sufficient number of projects within the reference class to make statistically meaningful conclusions.
5. Comparing the specific project with the reference class distribution, in order to establish the most likely outcome for the specific project.

3.4 Controls

It is good practice generally, and especially where estimates have been required to include a Financial Risk Exposure, to ensure that the estimating process is as transparent as possible through:

• the use of standard cost and risk structures and templates
• the documentation of all agreed assumptions
• the opportunity for an outside team to test and challenge the estimate, especially the rationale for any Financial Risk Exposure
• the establishment of clear audit trails
• the establishment of a cost and risk management process to be implemented across the life of the project

4. Building a mature cost estimate

4.1 Introduction

Having produced the initial estimate(s) through the process outlined in Section 3, the project team’s focus should move to the disciplined bottom-up production of an AFC which can be described as ‘mature’ according to the criteria set out in Section 2.4.

4.2 Key principles

1. Require project teams to focus on building a mature estimate of AFC as soon as possible to support the OBC.

2. A mature AFC is built bottom up and can all be clearly explained and accounted for in terms of estimate and risk. The Financial Risk Exposure has now been fully challenged and assigned to specific items.

3. Ensure that the overall Financial Risk Exposure (including any Management Contingency provisions) is only used to test the financial risk exposure to the organisation and not used to establish the project specific budgetary or funding envelope, which should set more challenging targets.

4.3 Approach

The focus of the cost estimation and risk analysis must be on producing a comprehensive and mature bottom-up estimate of AFC. This requires that they build on, but are not unduly influenced by the less rigorous assumptions which may have been required to create the early estimate(s).

The broad spectrum Financial Risk Exposure should be scrutinised, challenged and analysed such that each and every underpinning assumption is either assigned or incorporated into the Management Contingency or project specific risk categories.

During the development towards the Mature Estimate, elements of base cost or risk/uncertainty allowance should only be added to the estimate when they are sufficiently well defined and understood to allow the costs to be established.

An important discipline in ensuring that the development of the cost and risk estimates are mutually consistent and properly aligned with the development of the project’s solution is to set, share and challenge an agreed set of design/implementation assumptions.

The use of standard breakdown structures for cost estimating and risk will help ensure complete coverage and also facilitate the reuse of data and effective benchmarking and will help inform the commercial case.

As with initial estimates, it is important to maintain checks that all elements of the Base Cost and Financial Risk Exposure are mutually exclusive and collectively exhaustive. In particular, where specific contingencies are included, the estimate should be structured to avoid the layering of one contingency on another.

In order to ensure that the Financial Risk Exposure in the mature AFC estimate is lean, it is important that:

• active and iterative risk mitigation is established as soon as risk identification and quantification commences and well before the design is finalised
• where there is high confidence (90%-95% confidence level) in the value of the risk, this should be removed from the FRE and included in the base cost
• there are suitable incentives to achieve on-going cost effective risk reduction through different methods e.g. Risk Mitigation Targets, a Ring-Fenced Risk Model or the application of the methodologies in RAMP
• senior management scrutiny and control processes are established to review the effectiveness of risk mitigation
• external challenge should be used for the mitigation of major risks (defined as the product of likelihood and impact, and including very large impact risks irrespective of likelihood)
• contingency is held by the party best able to manage them and to control their use; across programmes and businesses, some contingency is often held centrally, especially where for example, it may be possible to hedge risks across a number of projects and reduce the overall level of contingency held
• clear, practical rules are established for the assessment of risk (e.g. does Post-Mitigation assume all mitigations are successful, how sensitive is the business case’s viability to a range of different post mitigation scenarios on key parameters)
• duplicate risks are avoided (either within the project’s risk register or between the project and the contingency held by other project stakeholders or other public projects occurring in the same place at the same time)
• rules are established, and checks run, for the inclusion of risks with low probability and high/very high impacts* taking account of the possibility that the true probability may be much higher than has been estimated

*Such risks may be better managed at a programme or business level or may be insurable.

4.4 Controls

As suggested above, the objective of achieving a lean Financial Risk Exposure as part of the first mature estimate of AFC can be facilitated by subjecting the AFC and risk assessment to challenge by external parties, focussing on the key points in the previous section. This will also encourage transparency in the way the estimate is developed.

5. Communicating cost estimates

5.1 Introduction

Whenever the AFC of a project is quoted, either within the delivery organisation or publicly, it will become something against which the project is measured. This is especially so with early AFCs that become a benchmark against which future AFCs are compared and the cost performance of the project (the extent to which it is deemed to have over or under-run) is measured. This in turn may encourage the project/programme owners to be over-cautious in declaring a value, and encouraging the AFC to become unnecessarily inflated.

In the context of seeking to achieve better value out-turns, it is therefore important that any such publications and announcements are sensitive to the context and nature of the number being quoted.

5.2 Key principle

As far as possible ensure that AFCs which have yet to reach a mature level are expressed / communicated appropriately.

5.3 Approach

There are two important characteristics of an AFC during the early phases of a project which need to be taken into consideration.

1. In the very early stages, the AFC will legitimately vary – sometimes significantly –either upwards or downwards as project definition identifies/resolves important issues of design and scope are considered, analysed and negotiated.
2. Throughout the project life cycle, the AFC is an estimate and therefore a variable*. Hence, to express the AFC as a single value requires a clear understating and articulation of how the number should be qualified.

*That is it is by nature a variable number and not a single point value.

It follows that if an AFC is to be quoted during this period, care needs to be taken that:

• it is clearly understood what the number represents
• that the uncertainty of the number’s context is understood
• that the consequences of publicising the number are understood and addressed

Cost Maturity Phase	Nature of AFC	Appropriate form of communication
Leading up to first Quantitative Risk Assessment and ‘Mature Estimate’	Immature. Indicative forecast / range of likely cost	Limited, cautious and caveated*
Leading up to the fixing of costs (end of the Design phase)	Increasingly mature	Range or agreed Pxx value (e.g. P50, P80)
Leading up to completion of implementation	Mature single figure with increasingly small variably element	Single figure

*See Section 6 on communicating the cost estimate.

It is recognised that this simple model will not be readily applicable to all situations; nevertheless the principle of recognising both the nature of the estimate at any given stage and the consequences of quoting/misquoting are universally valid.

Communicating immature cost estimates

In communicating any ‘immature’ cost estimate as part of an appraisal, it is essential that the full context of AFC is provided alongside the overall number. Specifically the following should be included:

• The total value of the AFC (including a description of variability) comprising:
  • Base Cost
    • How it was produced (including checks and controls applied)
    • Main assumptions
    • Key inclusions/exclusions
  • The Risk Allocation (if available)
    • Key inclusions/exclusions
  • Financial Risk Exposure
    • Breakdown of uncertainty areas covered (see Section 4.3.1)
    • Justification (the reasons for requiring each item)
    • Estimating method
  • The Scope
    • The output-based requirement and functionality
    • A design summary statement
    • The delivery timescale

The immature AFC should not be used as the basis of setting a budget or funding envelope as it is only an indicative forecast of likely out-turn cost.

6. Glossary

The following list has been developed by the industry group RiskUK. Some of the terms are used throughout this document, though not all. Users of this guidance should be aware that there is as yet no universal acceptance of a standard set of terms and that differences and ambiguities need to be recognised and addressed if terms are taken and used outside the context of this guidance.

3-point quantification	a 3-point quantification of an uncertainty where the probability is set equal to 1
4-point quantification	the quantification of a risk (2) with a probability (of the risk event occurring) and maximum, most likely and maximum values of the consequence (if it does) – a common practice. There is also a 2-point variant in which all three consequences set equal leaving just a probability and an impact.
AFC	anticipated final cost
anticipated final cost	an estimate or forecast of the final cost made before the project is complete which takes on board the risk exposure at the time the estimate is made using risk analysis
base cost	an assessment of the cost of the project without cost risk and uncertainty. The ground rules on which the base cost estimate has been prepared need to be recorded and understood. Care is needed here: base cost may mean the cost derived from reference data which therefore includes the occurrence of risk.
budget	set of authorised costs for financial control purposes which may contain a contingency. This may be broken down by projects and programmes. It is not necessarily the same as the prospective costs used for appraisal or business case purposes.
contingency	that part of a budget not allocated to specific activities but retained to deal with uncertainties crystallising and risks materialising. It may be allocated at project or programme level but this does not necessarily imply that expenditure of contingency is delegated to the relevant project or programme manager.
cost risk and uncertainty	the concept that we do not know what the final cost of a project is going to be until the project is complete. The term cost uncertainty would suffice (see risk) but we have added ‘risk’ to be clear that this includes the impact of events which may or may not materialise, as advocated by some practitioners - see also uncertainty.
draw down	allocation of a contingency to a specific activity for spending. This will be subject to appropriate governance procedures.
early stage risk analysis	the risk analysis that is carried out during the initial stages of the project, especially appraisal or feasibility, which is likely to be less detailed than during implementation, for example, reflecting the materiality of different risk issues
EFC	estimated final cost or expected final cost, both equivalent as far as this report is concerned to anticipated final cost
expected cost	mathematical term for the average cost predicted by a risk model taking the probability element into account
expert assessment	the quantification of risk models using the experience and knowledge of suitable people
final cost	the eventual cost of the project which is subject to risk exposure until the project is completed
FRE	financial risk exposure
Financial Risk Exposure	a term used by the Green Book supplement for the risk allowance during a project and related to the base cost and AFC through the relationship AFC=BC+FRE; where there is a probability distribution this would be the expected value according to the Green Book
Green Book supplement	the companion document ‘Determining risk and uncertainty in the early cost estimates of (infrastructure) projects and programmes’ that is being produced to update guidance on appraisal in the presence of risk
initial cost estimate	the cost estimate that is created at project inception in the model used in the Green Book supplement, built from initial risk estimates and reference class forecasting, and associated with the strategic outline business case (SOBC)
mature cost estimate	the cost estimate that is created prior to commitment in the model used in the Green Book supplement, built from bottom-up risk analysis, and associated with the outline business case (OBC)
mean	same as expected value
Monte Carlo	see risk model
optimism bias (1)	belief that things can be built more quickly and cheaply than is the case
optimism bias (2)	an adjustment (or uplift) made to the base cost of projects at the appraisal stage, originally to compensate for optimism bias (1)
out-turn cost	same as the final cost of a project
P50, P80, etc	see percentile
percentiles	a measure of confidence constructed using probability. For example the 80th percentile cost (also known as the P80) is such that the probability of the final cost being less than P80 is 80%. (P50 is also known as the median.)
portfolio effect	the concept that the relative variability of a portfolio (or programme) of projects is less than that of the individual projects. This is a mathematical concept (valid only where there are not strong dependencies between the projects) that justifies, for efficiency reasons, holding contingency at programme level rather than allocating it to projects. In other words the programme P80 will be significantly less than the sum of the project P80s.
probability	mathematical construct used to quantify the likelihood of an event occurring
reference class forecasting	the use of data from other projects to estimate the final cost of new projects
quantitative schedule risk analysis	the construction of a risk model aimed at estimating the risk exposure of project milestones, including the delivery date; generally implemented by Monte Carlo analysis of a logically linked project programme
QSRA	quantitative schedule risk analysis
reference class forecasting	see reference data
reference data	information about the final cost of previous projects which can be used to characterise the risk exposure of future projects using reference class forecasting
RBS	risk breakdown structure
risk (1) (the concept of risk)	(ISO 31000) the effect of uncertainty on objectives
risk (2) (a specific risk)	a description of a specific event which may or may not occur, together with its causes and consequences
risk allowance	an amount added to base cost to recognise risk exposure for application in management processes. It is context-specific depending on the process in question: appraisal, budgeting, implementation, etc. It might be set at the expected value, P50, P80, best/worst case, etc.
risk analysis	the process of estimating risk (1). This may include uncertainty, sensitivity to scoping options and so on.
risk breakdown structure	a hierarchical expression of the possible risks (2), or types of risk, in a risk analysis and therefore likely to be a key element of a risk model
risk exposure	the output of risk analysis, a representation of the range of final costs (in this context) which credible, essentially the same as the risk profile. May be expressed as best and worst case, a confidence interval (e.g. P10-P90) or a complete probability distribution.
risk management	(ISO 31000) coordinated activities to direct and control an organisation with regard to risk (1)
risk mitigation	the part of risk management which is focussed on identifying and implementing actions to reduce risk
risk model	quantified model of cost risk and uncertainty constructed using probability. Risk models are often calculated using Monte Carlo methods, a technique involving random sampling.
risk profile	essentially the same as risk exposure, though generally has an implication of the many types of impact of risk (safety, reputation, etc) and is therefore less used in this report which focuses on cost risk
risk reduction	a quantification of the extent to which risk has been reduced, for example the reduction in an AFC
risk register	a list of risks (as per risk (2)) together with other information such as the likelihood of the risk materialising, planned risk mitigation, etc
risk release	release of a contingency from a budget so that it can be allocated elsewhere, a different project or a higher level contingency. This is necessary for financial efficiency.
scope	defined in the Green Book supplement as a statement of the requirements, functionality and benefits of the project with a view to emphasising that the scope may legitimately vary until quite a late stage in the lifecycle - in which case it needs to be controlled - or may be fixed early on
uncertainty	see cost risk and uncertainty. Uncertainty is often used to describe situations where the outcome is not known, but there is no identified event which may or may not occur, the impact of future inflation for example. Often uncertainty has an upside whilst risk is generally a downside.
uplifts	a generic term for estimating risk allowances which does not benefit from specific project risk analysis, for example optimism bias (2)