Tải bản đầy đủ - 0 (trang)
5 Risk-Based Planning with Response, and Real Options

5 Risk-Based Planning with Response, and Real Options

Tải bản đầy đủ - 0trang


M. Rees

3 Benefits and Challenges

Some industries are already users of risk-based planning to a considerable extent;

these include insurance, financial markets, pharmaceutical, and oil and gas. However, even in these cases, the use of such processes is often confined to certain

functions or areas, and there is much unexploited potential. In this context, it is

therefore worth summarizing the type of benefits that are specifically achievable by

using such an approach, as well as mentioning some key challenges.




Information Alignment

As mentioned earlier, the consideration of uncertainty is a natural part of every-day

decision making. However, very often, the information provided to management is

of a static nature (e.g., a base case forecast) and it is left to the management to make

a decision based on incomplete information about the likelihood and impact of

other possible outcomes. The incorporation of uncertainty thinking and planning

can therefore provide management can with an information basis that more closely

matches their requirements, leaving them to focus on those areas where critical

judgment from senior decision-makers is still required.


Functional Alignment

Since uncertainty analysis requires an identification of factors that would lead

to a change in a base plan, such a process forces a consideration of issues of

a cross-functional nature. For example, to answer a question such as: “what

factors are likely to lead to a loss – or increase – in revenues?” many strategic,

operational and tactical issues must be considered (e.g., chance of competitor entry,

pricing policies, market growth and segmentation strategy, the targeting or new

customers, new product development and launch, potential operational problems,

the requirement to incorporate additional capital expenditure to allow the operational flexibility to expand, and so on).


Accuracy of Base Plans

Static model calculations (generally in an Excel® environment) contain a number

of potential biases or undefined cases.

For example, it is not generally clear what case an input variable to a static

model is showing, nor which case is shown by the output: is it the average, the most

Linking Strategy, Operations and Finance with Simulation-Based Planning Processes


likely, an optimistic or a pessimistic value? This issue is especially pertinent when

uncertainties are not symmetrically distributed around their assumed base values.

Typically the output value of a static model does not even shown the average

outcome, whereas this is the fundamental reference point according to the theory of

financial decision-making (Brealey and Myers 2003; Wilmott 2006).

The practice of decision-making also requires reference to the possible spread of

outcomes, in order to reflect risk tolerances and preferences. For example, one may

choose not to undertake a project, which is profitable on average, because some of

the outcomes are unfavorable. Equally, one may prioritize a potential project with a

lower average profit over a project with a higher average profit if the possible spread

of profits in the latter case were to be much larger.

Finally, an appropriate base case for decision-making would generally need

to reflect the inclusion of all those risk mitigation and response measures that are

considered worthwhile but exclude potential mitigation factors that are not

worthwhile (because they are judged too expensive to implement), and such action

may not have been contained in a static plan for which risks have not been

considered. There is effectively a new base case, which can be thought of as an

optimized one, with the resulting uncertainty in the outcome being a residual one

(Rees 2008).



There are also of course a number of challenges when implementing risk-based

planning processes.


Process First, Model Later

There can be a temptation to focus on quantitative aspects too early. A process

whose ultimate aim is to quantitatively support decision-making can be hindered by

a focus too soon on quantitative thinking. A good quantitative analysis is predicated

on the completion of a good qualitative analysis; such analysis requires identification, consideration and prioritisation of all relevant risk factors, uncertainties and

opportunities – whether in the area of strategy, market environment, operations or

financial aspects – and in fact requires and forces a cross-functional communication

and collaboration.

Risk models are often perceived to be of a complex or “black-box” nature, or

alternatively as box-ticking exercises. The slogan of “garbage in-garbage out” is

one that appeals to many critics of risk models. We believe that appropriate risk

modeling is generally an exploratory process, in which one attempts to understand a

situation better, whilst being explicitly aware of the limitations of the models.



M. Rees

Value-Added Insight Without Oversimplification

A consideration of risks and opportunities naturally leads to situation where models

may potentially contain many more variables. A desire to include “everything that

might happen” can lead to “perfection being the enemy of the good.” Participants

can find it challenging to keep the analysis “as simple as possible, but no simpler”

(to paraphrase Einstein), to focus on where the value-added is, and to make

appropriate trade-offs when setting priorities. However, in the nature of exploratory

thinking that is often required to generate value-added, “approximately right is

better than precisely wrong.”


Information Overload

Both process participants and decision-makers often find it a key challenge to

communicate the results of risk-based processes. By their nature, there is more

information that can be discussed and presented compared to static plans, there are

more issues that may arise, and it may be hard to “see the wood for the trees.”

4 Examples

With the right tools, the quantitative aspects of risk-based planning are generally

easy to implement. In fact, simulation models can be built in Excel® (without the

use of add-ins), although doing so generally requires both the use of VBA code

(macros) and significant work to post-process the results. The use of commercially

available Excel® add-ins allows for more rapid, flexible and transparent model

building, as well as providing a wide set of distributions, the ability to correlate

them, and a range of post processing tools for analysis of the results; generally one is

better off using these tools than using Excel® with VBA macros. Examples of addins include The DecisionTools Suite® from Palisade Corporation (which contains

products such as @RISK®, RiskOptimizer® and PrecisionTree®), as well as

CrystalBall®, RiskSolver®, and ModelRisk®. The core elements of any of these

can be learnt in a few hours (from the perspective of software menus); each product

is fairly intuitive to understand, and does not require the modeler or the decisionmaker to have any advanced knowledge of mathematics, probability theory nor

statistics. Of course, the required issues in model framing, formulation and implementation are typically more challenging than the use of the software per se.

The following examples are all built using products from Palisade. The main aim

is to show simple yet potentially powerful models, and also to highlight how the

framing, formulation, and implementation of the modeling process would require a

close degree of functional co-operation.

Linking Strategy, Operations and Finance with Simulation-Based Planning Processes



Revenue Forecasting

Revenue forecasting is often a challenging task due to many sources of uncertainty

that could impact future revenues. A base static forecast may rely on many implicit

assumptions and most typically represent some imprecisely defined case within the

full possible range of outcomes. Consideration of the uncertainty in future revenues

is potentially important, not only for the impact on net income and cash flow, but

also because it forces a consideration of where additional external opportunities or

threats may lie within the overall market and competitive environment, as well as

the appropriate response to these. It may also be important for the setting of

appropriate targets for sales and other staff (such as “stretch targets”).

The following shows the output of a simple model, which reflects the range of

possible revenues over 5 years, taking into account the following assessment of


• The central most likely projection is for the market to grow at 5% p.a.

• It is estimated that there is a 10% chance that the market will decline (i.e. that the

growth rate will be less than 0% p.a.), and a similar chance that it will be above

8% p.a. in any given year.

• Market share in the central case is forecast to be constant, at 20%, but could

fluctuate by 1% in each period, and with the most likely value of the market

share in the next period derived from the ending market share in the current


• There is judged to be a 25% probability that a new application will be found for

the company’s products from year 3 onwards, and this will create an additional

market in which the company’s sales can be increased by 10% on top of the

original values.

• There is judged to be a 10% chance of a single new competitor entering at any

time, corresponding to the potential success of this competitor in making a

technology breakthrough. The impact of this is estimated as a reduction in

sales of 10–20%, with a most likely reduction of 15%.

In this case, the results (see Fig. 1) show that the sales in both years 1 and 5 are –

in approximately 2/3 of cases – likely to be below the base case forecast values, and

above these values in approximately one-third of cases (the delimiter lines are

placed at the base values for years 1 and 5). In other words, the base plan is perhaps

optimistic as a general reference case (because the bias of the actual outcomes is

below this base value). At the same time, the base plan may be too pessimistic from

the perspective of target setting (because the base will in any case be achieved in

one-third of cases, without further intervention).

Note that a key point here in terms of this text is that the making of these

assessments would, in practice, require a close cross-functional co-operation. In

addition, as the screenshot shows, only a very few lines of formulae are required to

generate a set of results that is potentially much more informative than a single

point forecast.


M. Rees

Fig. 1 Revenue forecast incorporating uncertainty


Optimal Capacity Planning

Faced with uncertainty in revenues, a company with possible future expansion plans

may wish to estimate how much new capacity should be put in place. A decision to

build too little capacity could result in a facility that does not fully exploit possible

opportunities if the market turns out to develop better than expected. A decision to

build too much capacity could mean that costs are incurred to invest in capacity that

would not be used. Once again, the consideration of this situation from an analytic

perspective would require a close cross-functional co-operation in order to establish

the drivers and range of uncertainty in areas such as demand, prices, investment and

operating costs. In the following example, we assume that:

• Volume and unit prices are uncertain (varying approximately Ỉ 20% and Ỉ 10%

around base case values).

• There is uncertainty around capital expenditure levels, and fixed and variable

operating costs (which each can vary approximately Ỉ 10% around base case


• A choice must be made as to how much capacity to build; in the model different

possible capacity levels are assumed and the model rerun to work out the profit

distribution for each case.

The results (see Fig. 2) show the curve that is formed as chosen plant capacities

vary, and in which the average profit and the profit volatility are plotted as a

function of this variation. One can see that an average profit of about $150 m p.a.

is possible in the optimal case (i.e. where average profit is maximized), with a

corresponding volatility of about $25 m p.a. (volatility is the standard deviation of

profits and roughly equates to the Ỉ band around the average in which profits would

Linking Strategy, Operations and Finance with Simulation-Based Planning Processes


Fig. 2 Profit curves: Average profit and volatility of profit as chosen capacity varies

lie in two-thirds of outcomes). If one wished to reduce profit volatility, then this

could be achieved by sacrificing average profit; one could envisage that in some

circumstances this may be preferable. In this specific example, reading from the

curve, the volatility of the profits could be reduced by approximately 30% if one

were prepared to accept an approximately 5% reduction in average profit.


Enhanced Decision-Making, Incorporating Flexibility

In this example, we assume that one is faced with a situation in which a base static

calculation shows that the NPV (net present value) of a project is slightly negative,

but where our intuition is that the project is somehow worthwhile. It may be that the

situation contains flexibilities to react that are not reflected in the base static

calculations, such as expansion opportunities or risk reduction and mitigation


The screenshot (see Fig. 3) shows a model with a base static model in which total

discounted cash flow over the 5-year planning period would be negative, and hence

(typically speaking) the project would be rejected. However, the tree-based

approach allows for an enhanced model, which is able to capture that if the project

were to be conducted then – during the first year – more information about the

overall quality of the opportunity will be gained. After the first year, the project may

be expanded, continued as originally planned or abandoned. Such a model of course

requires additional assumptions about the additional expenditure required and the

relative return for additional investments (for example, that in a case in which the

quality of the project seems to be above the original anticipation, then the return on


M. Rees

Fig. 3 Incorporation of expansion and contraction flexibilities into project valuation

additional investment would typically be higher than the originally expected return

on investment). When such additional assumptions are reflected in the model, the

result is a situation in which the initial project evaluation is in fact positive from a

(revised) NPV perspective. Such an example forms the core principle behind real

options modeling.


Integrated Financial Planning

An integrated business and financial plan would ideally incorporate issues such as

those above i.e. it would reflect the risks, uncertainties, opportunities as well as the

optimal management response to these. Strategic and operational modeling can

then be integrated into overall financial planning. Such considerations can play an

important role in communications, target setting and the optimization of financing

structures (such as debt/equity mix, or the mix of different forms of debt- or equitylike obligations).

The following example shows a financial statement model that results from an

integrated business planning process, and in which some key risks are reflected in

the planning of many individual income statement and balance sheet items.

Uncertainties are assumed for the sales growth rate, production costs, unanticipated

exceptional losses, working capital, and capital expenditure levels. The results

show a distribution of possible outcomes for net income and for the amount of a

back-up financing facility that would be used. For example net income is in the

range $50–100 m about 50% of the time, and is approximately equally distributed

either side of these bounds. Similarly, the back-up facility is needed in about 15% of

Linking Strategy, Operations and Finance with Simulation-Based Planning Processes


Fig. 4 Uncertainty on earnings and on the financing facilities required

outcomes, with a level of borrowings in this facility of $30 m being exceeded in the

worst 5% of cases (Fig. 4).

In principle, such analysis can be applied to any relevant line item in the

financial statements, to calculate for example cash flow uncertainty, the effect of

unexpected costs, or of the outcome of product development processes, risks around

the realization of synergies or acquisitions, capital expenditure uncertainty, or risks

around taxes, unanticipated foreign exchange movements, goodwill or impairment

charges and so on.

5 Model Formulation and Implementation: Selected Topics

In practice there are a number of issues that need to be considered when deciding

the appropriate approach to use, and in formulating and building the appropriate

quantitative models. These range from pure qualitative issues to more quantitative



Risk Identification

The identification of sources of risk, uncertainties and opportunities and the

potential consequences of each is crucial to constructing an appropriate quantitative

analysis. The process of doing so is a key area of cross-functional co-operation, and

a key way to integrate strategic, operational and financial activities. Included in

such a process would also be – where relevant – the inclusion of management (or


M. Rees

other) flexibilities that are inherent in the situation, such as real options flexibilities

and so forth. The generation of this understanding is primarily of a qualitative

nature, and requires knowledge of the external market environment, and the internal

capabilities and activities of the business. It is nevertheless crucial to ensure

adequate work is done in this area before proceeding too far with detailed quantitative work that – once started – can generate a momentum of its own.



The sources and consequences of risk are expressed using distributions. For many

practical purposes a small set of input distributions is sufficient to build exploratory

models of adequate accuracy, and hence this topic need not be as complex as it may

perhaps appear at first glance, although some care of course still needs to be taken

(Bernardo and Smith 1994) and (Rees 2008).



Relationships between sources of risk most frequently take the form of one or other

(sometimes both) of correlated sampling or parametric dependence (e.g., where the

probability of occurrence of an event increases if another uncertain event has

happened). Correlated sampling is implemented as a technique in most commercial

add-ins to Excel® (such as those mentioned above), whereas relationships of

parameter dependency are generally directly implementable through formulae.



Part of the aim of risk modeling is to understand the range, likelihood and important

drivers in a situation. Very often, an appropriate risk model (i.e. one which is

focused on and built around the key sources of uncertainty, their impact, and

potential management responses etc.) is less detailed and built at a more aggregated

level than a static model.

6 The Role of the CFO in Leading Risk-Based Planning


The CFO is a natural champion of the implementation of risk-based planning

processes, and must lead, coach and support these processes if they are to generate

value-added insights. There are a number of areas where leadership and active

management is likely to be required.

Linking Strategy, Operations and Finance with Simulation-Based Planning Processes



Project Selection and Framing

To maximize the success of any initial applications on these ideas, it is of course

important to choose projects appropriately. An obvious starting point is to ensure

that the selected area of application is one where the consideration of uncertainty is

intuitively an important factor when making the final decision. Once this area has

been selected, it is important to provide some frame and scope for the work, so that

right questions are addressed. It is important to make clear that the project is not a

box-ticking exercise on risk controlling, but a creative and holistic one, in which

new thought processes and techniques may be required.


Retaining Focus

One danger when initiating risk-based thinking is that “perfection becoming the

enemy of the good.” The enhanced scope of possible models compared to traditional static planning approaches can result in attempts to include every possible

variable, uncertainty and outcome within the scope of a quantitative model. The

challenge is to keep the process creative, yet focused on those aspects that provide

the most insight and value-added; these areas are typically themselves those where

the consideration of risks and opportunities means that base plans may need to

be changed. The CFO must encourage an environment where “approximately

right is better than precisely wrong” and where the analysis “as simple as possible,

but no simpler” to paraphrase Einstein. This may apparently contradict the traditional perception of financial analysis as a set of precise calculations, and so the

support of the CFO in the creation of such an environment is therefore of the

essence. This applies as much to those staff who are involved in the process but

are not from finance functions as it does to those within the finance function, who

very often have a strong accounting background but less of a modeling and

forecasting one.



The CFO needs to take a key role in integrating a risk-based approach into overall

planning processes in a way that is appropriate for the company. Ideally, risk-based

planning would form a fully integrated part of the overall planning process. In some

cases, risk-based planning can be introduced as an immediate part of the regular

planning process. However, more often it may be more appropriate to have some

separation between the processes, at least initially. A pilot project that is positioned

outside of the main planning processes can also ensure that participants are open to

more creative thinking and do not become focused too early on trying to produce


M. Rees

detailed numerical deliverables. The results of a pilot project (where deemed

successful) can then be used to inform modifications that need to be made to an

overall base case plan (and where less successful, appropriate lessons need to be

drawn). Over time, with more experienced gained in this area, risk-based planning

can become fully integrated as a natural part of the regular planning process.


Timing and Resources

The more complete and also exploratory nature of introducing risk-based thinking

into planning processes may mean that – certainly initially – some additional work

is required. At a later stage, when risk-based thinking is more naturally embedded

in organisational processes, this additional work will be minimal, and conceivably

could even lead to a reduction in the total work involved in producing plans (a more

complete plan from early on can lead to a reduction in rework, as well as agreement

that there is a valid range of outcomes, rather than trying to agree on a somewhat

artificial static case). This additional work affects not only finance staff, but also a

wide range of staff from other functions. It is therefore crucial to ensure that there is

sufficient cross-functional involvement and resources, and indeed that the process

does not become dominated by financial staff.



Senior finance support is typically required in order that results of risk-based

planning are appropriately communicated, both internally and externally. Internal

communications revolve around general support to the process, and the linkage with

issues covered above, such as supporting processes, which are often of an exploratory – rather than calculatory – nature. Communications with external players (such

as public financial market participants) may not always be appropriate, but such

communication is potentially much more challenging than internal communications;

whilst there is a requirement for appropriate disclosure of risks and opportunities,

ways must be found to achieve this, which do not jeopardize market or competitive

position or otherwise damage the business.

7 Concluding Remarks

It may well be that one day essentially every planning or analytic exercise will

explicitly incorporate uncertainty into it. Such a belief would be justified in that

sense that – since uncertainty is present in any situation – decision-makers should

be provided with a solid basis of information in this area before making any

decisions or plans. This Chapter has discussed some key approaches, benefits and

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

5 Risk-Based Planning with Response, and Real Options

Tải bản đầy đủ ngay(0 tr)