Using Decision Trees In Financial Management

Decision trees are diagrams that show the sequence of interconnected determinations and the expected consequences of taking one option over the other. Typically, more than one pick or option is available when you ‘re faced with a determination or, in this instance, possible results from a hazard event. The available picks are depicted in tree signifier get downing at the left with the hazard determination ramifying out to the right with possible results. Decision trees are normally used for hazard events associated with clip or cost.

Stairss in determination tree analysis

Main stairss in determination tree analysis are as follows:

1. Identifying the job and options

To understand the job and develop options, it is necessary to get information from different beginnings like marketing research, economic prediction, fiscal analysis, etc. As the determination state of affairs unfolds, assorted options may originate which are to be identified. There would besides be sorts of uncertainnesss in footings of market size, market portion, monetary values, cost construction, handiness of natural stuff and power, governmental ordinance. Technological alteration, competition, etc. Recognizing that hazard and uncertainness are built-in features of investing undertakings, individuals involved in analysing the state of affairs must be encouraged to show freely their uncertainties, uncertainnesss, and reserve and motivated to propose eventuality programs and place promising chances in the emerging environment.

2. Defining the determination tree

The determination tree represents the anatomy of determination state of affairs. It illustrates

determination points along with the option options available for experimentation and action at these determination points

opportunity points where results are dependent on a opportunity procedure and the likely results at these points

This determination tree graphically reflects the nature of determination state of affairs in footings of alternate classs of action and opportunity results which have been identified in the first measure of the analysis.

If myriad possible future events and determinations are considered, it can go really complex and cumbersome. As a consequence, it would non be a utile tool of analysis. If many luxuriant events are taken into history so it may obfuscate the critical issues. Hence it is necessary to simplify the determination tree so that focal point can be given on major future options.

3. Stipulating chances and pecuniary results

After defining the determination tree, chances matching with each of the possible results at assorted opportunity points and pecuniary value of each combination of determination option and opportunity result have to be gathered.

The chances of assorted results can be defined objectively. For case, based on nonsubjective historical informations the chance of good monsoon can be defined. On the other manus, chances for existent life results are slightly hard and can non be obtained. For illustration, one can non find the chances for success of a new car launch. These have to be defined subjectively and based on experience, judgement, apprehension of informed executives and their intuition. Besides, it is hard to measure hard currency flows matching to these results. So once more judgement of experts helps in specifying these hard currency flows.

4. Measuring assorted determination options

The concluding measure in determination tree analysis includes rating of assorted options. This can be done as follows:

get downing with the right- manus terminal of the tree and so we calculate the expected pecuniary value at assorted opportunity points that come foremost as we proceed leftward.

Given the expected pecuniary values of opportunity points in measure 1, evaluate the options at the concluding phase determination points in footings of their expected pecuniary values.

At each of the concluding phase determination points, select the option which has the highest expected pecuniary value and truncate the other options. Each determination point is assigned a value equal to the expected pecuniary value of the alternate selected at that determination point.

Continue backward ( leftward ) in the same mode, ciphering the expected pecuniary value at opportunity points, choosing the determination option which has the highest expected pecuniary value at assorted determination points, truncating inferior determination options, and delegating values to determination points, till the first determination point is reached.


A determination tree is a determination support tool that uses a tree-like graph or theoretical account of determinations and their possible effects, including opportunity event results, resource costs, and public-service corporation. It is one manner to expose an algorithm. Decision trees are normally used in operations research, specifically in determination analysis, to assist place a scheme most likely to make a end. Another usage of determination trees is as a descriptive agency for ciphering conditional chances. When the determinations or effects are modelled by computational verb, so we call the determination tree a computational verb determination tree [ 1 ] .

In determination analysis, a “ determination tree ” – and the closely-related influence diagram – is used as a ocular and analytical determination support tool, where the expected values ( or expected public-service corporation ) of viing options are calculated.

A determination Tree consists of 3 types of nodes: –

1. Decision nodes – normally represented by squares

2. Chance nodes – represented by circles

3. End nodes – represented by trigons

Drawn from left to compensate, a determination tree has merely burst nodes ( dividing waies ) but no sink nodes ( meeting waies ) . Therefore, used manually, they can turn really large and are so frequently difficult to pull to the full by manus.

Analysis can take into history the determination shaper ‘s ( e.g. , the company ‘s ) penchant or public-service corporation map, for illustration:

The basic reading in this state of affairs is that the company prefers B ‘s hazard and final payments under realistic hazard penchant coefficients ( greater than $ 400K — in that scope of hazard antipathy, the company would necessitate to pattern a 3rd scheme, “ Neither A nor B ” ) .

Uses in learning

This subdivision requires enlargement.

Decision trees, influence diagrams, public-service corporation maps, and other determination analysis tools and methods are taught to undergraduate pupils in schools of concern, wellness economic sciences, and public wellness, and are illustrations of operations research or direction scientific discipline methods.

[ edit ] Advantages

Amongst determination support tools, determination trees ( and act upon diagrams ) have several advantages:

Decision trees:

Are simple to understand and construe. Peoples are able to understand determination tree theoretical accounts after a brief account.

Have value even with small difficult information. Important penetrations can be generated based on experts depicting a state of affairs ( its options, chances, and costs ) and their penchants for results.

Use a white box theoretical account. If a given consequence is provided by a theoretical account, the account for the consequence is easy replicated by simple math.

Can be combined with other determination techniques. The undermentioned illustration uses Net Present Value computations, PERT 3-point appraisals ( determination # 1 ) and a additive distribution of expected results ( determination # 2 ) :

[ edit ] Example

Decision trees can be used to optimise an investing portfolio. The undermentioned illustration shows a portfolio of 7 investing options ( undertakings ) . The organisation has $ 10,000,000 available for the entire investing. Bold lines mark the best choice 1, 3, 5, 6, and 7, which will be $ 9,750,000 and make a final payment of 16,175,000. All other combinations would either transcend the budget or give a lower final payment. [ 2 ]

Decision Making Tools: Decision Tree Analysis and EMV

Decision Makers ‘ Toolkit

“ Decision-making is the cognitive procedure of choosing a class of action from among multiple options. Every decision-making procedure produces a concluding pick. ” That ‘s what Wikipedia says anyhow. What it does n’t state is that some determinations must be made for results that will happen in the hereafter. However, there are a twosome of tools that can be put to utilize in assisting do complex determinations, viz. , Expected Monetary Value and Decision Tree Analysis.

Expected Monetary Value ( EMV )

EMV is a balance of chance and its impact over the scope of possible scenarios. If you have to do a determination between two scenarios, which one will supply the greater possible final payment?

Scenario 1

Best instance provides a 20 % chance of doing $ 180,000

BC = 20 %

X $ 180,000= $ 36,000

Worst instance provides a 15 % chance of fring [ – $ 20,000 ]

WC = 15 %

Ten ( – $ 20,000 ) = ( – $ 3,000 )

Most likely instance provides a 65 % chance of doing $ 75,000

MLC = 65 %

X $ 75,000 = $ 48,750

Entire Expected Monetary Value 100 % $ 81,750

Scenario 2

Best instance provides a 15 % chance of doing $ 200,000

BC=15 %

X $ 200,000 = $ 30,000

Worst instance provides a 25 % chance of doing $ 15,000

WC= 25 %

X $ 15,000 = $ 3,750

Most likely instance provides a 60 % chance of doing $ 45,000

MLC=60 %

X $ 45,000 = $ 27,000

Entire Expected Monetary Value 100 % $ 60,750

Which scenario do you take? Number one, because it has the highest EMV, or $ 81,750

Decision Tree Analysis

In determination tree analysis, a job is depicted as a diagram which displays all possible Acts of the Apostless, events, and final payments ( results ) needed to do picks at different points over a period of clip.

Example of Decision Tree Analysis: A Manufacturing Proposal

Your corporation has been presented with a new merchandise development proposal. The cost of the development undertaking is $ 500,000. The chance of successful development is projected to be 70 % . If the development is unsuccessful, the undertaking will be terminated. If it is successful, the maker must so make up one’s mind whether to get down fabricating the merchandise on a new production line or a modified production line. If the demand for the new merchandise is high, the incremental gross for a new production line is $ 1,200,000, and the incremental gross for the modified production line is $ 850,000. If the demand is low, the incremental gross for the new production line is $ 700,000, and the incremental gross for the modified production line is $ 150,000. All of these incremental gross values are gross figures, i.e. , before deducting the $ 500,000 development cost, $ 300,000 for the new production line and $ 100,000 for the modified production line. The chance of high demand is estimated as 40 % , and of low demand as 60 % .

The development of a determination tree is a multi measure procedure. The first measure is to construction the job utilizing a method called decomposition, similar to the method used in the development of a work dislocation construction. This measure enables the decision-maker to interrupt a complex job down into a series of simpler, more separately manageable jobs, diagrammatically displayed in a type of flow diagram called a determination tree. These are the symbols normally used:

The 2nd measure requires the final payment values to be developed for each end-position on the determination tree. These values will be in footings of the net addition or loss for each alone subdivision of the diagram. The net gain/loss will be gross less outgo. If the determination to non develop is made, the final payment is $ 0. If the merchandise development is unsuccessful, the final payment is – $ 500,000. If the development is successful, the determination is to construct a new production line ( NPL ) or modify an bing production line ( MPL ) . The final payment for the NPL high demand is ( $ 1,200,000 – $ 500,000 development cost – $ 300,000 physique cost ) or $ 400,000. For a low demand, the final payment is ( $ 700,000 – $ 500,000 development cost – $ 300,000 physique cost ) or – $ 100,000. The final payment for the MPL high demand is ( $ 850,000 – $ 500,000 development cost – $ 100,000 physique cost ) or $ 250,000. For a low demand, the final payment is ( $ 720,000- $ 500,000 development cost – $ 100,000 physique cost ) or $ 120,000.

The 3rd measure is to measure the chance of happening for each result:

Development Successful = 70 % NPL High Demand = 40 % MPL High Demand = 40 %

Development Unsuccessful = 30 % NPL Low Demand = 60 % MPL Low Demand = 60 %

Probability Totals* 100 % 100 % 100 %

*Probabilities must ever be 100 % , of class.

The 4th measure is referred to as the roll-back and it involves ciphering expected pecuniary values ( EMV ) for each alternate class of action final payment. The computation is ( chance X final payment ) = EMV This is accomplished by working from the terminal points ( right manus side ) of the determination tree and turn uping it back towards the start ( left manus side ) choosing at each determination point the class of action with the highest expected pecuniary value ( EMV ) .

Decision D2:

New Production Line V. Modified Production Line

high demand + low demand = EMV high demand + low demand = EMV

( 4 0 % X $ 400,000 ) + ( 60 % X – $ 100,000 ) ( 40 % X $ 250,000 ) + ( 60 % X $ 120,000 )

$ 100,000 $ 172,000

Decision Point 2 Decision: Modified Production Line with an EMV of $ 172,000

Decision 1: Develop or Do Not Develop

Development Successful + Development Unsuccessful

( 70 % X $ 172,000 ) ( 30 % ten ( – $ 500,000 ) )

$ 120,400 + ( – $ 150,000 )

Decision Point 1 EMV= ( – $ 29,600 )

Decision: DO NOT Develop the merchandise because the expected value is a negative figure.

When making a determination tree analysis, any sum greater than zero signifies a positive determination. This tool is besides really utile when there are multiple instances that need to be compared. The 1 with the highest final payment should be picked.

Real options analysis: tools and techniques for valuing strategic…

A By Johnathan Mun

hypertext transfer protocol: // id=X47bm9Etd7IC & A ; pg=PA649 & A ; lpg=PA649 & A ; dq=decision+tree+applications+oil+and+gas & A ; source=bl & A ; ots=W47wkDY2Xt & A ; sig=YwtNvZ8KEDJ-60CEK87Xhodouis & A ; hl=en & A ; ei=Bnz3TP7kMI-srAec-63vDw & A ; sa=X & A ; oi=book_result & A ; ct=result & A ; resnum=2 & A ; ved=0CB0Q6AEwATge # v=onepage & A ; q=decision % 20tree % 20applications % 20oil % 20and % 20gas & A ; f=false


Remington: the scientific discipline and pattern of pharmaceutics

hypertext transfer protocol: // id=NFGSSSbaWjwC & A ; pg=PA743 & A ; lpg=PA743 & A ; dq=decision+tree+applications+pharmaceuticals & A ; source=bl & A ; ots=V64QMimyuo & A ; sig=cdQYBEgEJSf_lON-Alkhv4E6B4E & A ; hl=en & A ; ei=8oT3TLOOLcS3rAfw1azvDw & A ; sa=X & A ; oi=book_result & A ; ct=result & A ; resnum=6 & A ; ved=0CDYQ6AEwBTg8 # v=onepage & A ; q & A ; f=false


determination tree analysis

the undertaking director can utilize ‘ determination tree analysis ‘ when a determination involves a series of several interconnected determinations. The undertaking director computes the ‘ Expected Monetary value ‘ ( EMV ) of all schemes and chooses the scheme with the highest EMV.

Assume that the undertaking director has four option schemes, S1, S2, S3, S4. The resulatant values for each scheme at different chance degrees are R1, R2, and R3. Assume that the chance of happening of these consequences is 0.5, 0.2 and 0.3. the final payment matrix for this job is given in table 18.4.

Table 18.4. Payoff Matrix




















P=0.5The undertaking director can besides stand for this job as a ‘ determination tree ‘ . Figure 18.3. depicts the determination tree for the given job. The undertaking director eventually selects scheme S1 as it has the highest value.

EMV ( A ) = 0.5 ( 13 ) +0.2 ( 10 ) +0.3 ( 9 ) =

EMV ( B ) = 0.5 ( 11 ) +0.2 ( 10 ) + 0.3 ( 8 ) =

EMV ( C ) = 0.5 ( 10 ) +0.2 ( 12 ) + 0.3 ( 11 ) =

EMV ( D ) = 0.5 ( 8 ) +0.2 ( 11 ) + 0.3 ( 10 ) =

Reappraisal of literature

1. Introduction

R and D direction, by its really nature, is characterized by uncertainness since effectual R and D requires a complex interaction of variables. It is of import to equilibrate strategic direction ( allocate resources and do the right R and D ) with operational direction ( executing of undertakings ) and at the same clip take into history issues of people direction ( leading, motive, administration and teamwork ) ( Menke, 1994 ) . The strategic facet of R and D direction entirely requires the declaration of some really of import inquiries, viz.

Do we have the right sum R and D budget?

Are we apportioning it to the right concern and engineering countries?

Do we have the right balance of hazard and return ; of long- and short-run undertakings ; of research V development ; of incremental V invention?

Are we working on the right undertakings and programmes with the right attempt?

It is clear that for success in R and D it is critical to find what is ‘right ‘ for the peculiar company. The normal procedure for making this is through the development of a engineering scheme. In pattern, the attack used will be that which best fits the operating method of the company but, as Braunstein ( 1994 ) has pointed out, the attack is less of import than the end product, which has to associate the corporate ends and scheme to the company ‘s major functional units. Having defined what the concern aims should be for the R and D programme and the overall strategic model that will specify the engineering program, it is so possible to travel on to what is likely one of the most debatable parts of engineering direction, the choice of single R and D programmes. There is a comprehensive literature of possible methods which can be used ( Baker and Pound, 1964 ; Gear et al. , 1971 ; Souder, 1978 ) . Many of these compare undertakings with different distributions of possible results and hazard, frequently utilizing comparatively complex quantitative methods.

There are a figure of mutualities that have to ‘come good ‘ before the undertaking eventually produces value for the company and it has been argued ( Morris et al. , 1991 ) that because many of the major determinations ( and many sub-decisions at intermediate mileposts ) can be taken singly, the overall procedure is less hazardous that might ab initio be thought. Not surprisingly, hence, Morris goes on to suggest that, when taking R and D undertakings, there is virtue in traveling for ‘long shootings ‘ since this is efficaciously the purchase of ‘options ‘ which can be dropped subsequently if the undertaking does non look like bearing fruit. Furthermore, the higher hazard undertakings ( about by definition ) tend to be the 1s that have the highest payback if they are successful ( see besides Kester, 1984 ) .

2. Decision doing under uncertainness

Uncertainty in a concern state of affairs is frequently expressed verbally in footings such as ‘it is likely ‘ , ‘it is likely ‘ , ‘the opportunities are ‘ , ‘possibly ‘ , etc. This is non ever really helpful because the words themselves are merely utile when they convey the same significance to all parties. It is clear that different people have different perceptual experiences of the mundane looks which are frequently used to depict uncertainness. Uncertainty exists if an action can take to several possible results and an indispensable, but, disputing facet of R and D direction is to place the likeliness or chance that these results or events will happen.

There are two chief readings of chance. The first is grounded in the appraisal of the chance of an event in footings of comparative frequence with which the event has occurred in the past and is normally referred to as ‘objective chance ‘ . The 2nd positions chance as being the extent of an person ‘s or group ‘s belief in the happening of an event and is normally termed ‘subjective chance ‘ . Subjective chance estimations are frequently included in the theoretical accounts suggested as utile for undertaking choice in R and D planning. Such chances might be derived from past experience with similar research undertakings plus any particular characteristics that make the current attempt unique or different and change the past up or down from this base line.

A figure of tools have been proposed to assist in the procedure of bring forthing chances, though they are by no agencies perfect. Schroder ( 1975 ) draws attending to some of the jobs that occur in deducing chances of proficient success and concludes that “ subjective chances are a instead undependable forecaster of the existent result of single success ” . He proposes a figure of grounds for this which he categorises as either knowing or unwilled ( witting biasing ) . To diminish the unwilled mistakes he suggests the undermentioned actions:

O guarantee that hazard assessors have sufficient expertness in their field and a comprehension of subjective chances.

O better the handiness of information and peculiarly certification.

O to the full exploit information systems and effort to use incentive systems which reward truth and dependability.

O analyse past public presentation in measuring chances to supply valuable penetration into possible betterments.

O utilise tested attacks to assist in the subjective chance appraisal.

It is apparent, nevertheless, that some assurance degrees need to be established and possibly the most obvious manner of accomplishing this is by the bite over a period of clip, of how anterior appraisals have compared with world. For this to hold echt value will necessitate a comparing of the premises that have been made at each appraisal.

3. The usage of fiscal methods for hazard analysis

Benefit/cost ratios have been popular for some clip, since they are simple and are an effort to understand the possible addition for the attempt required. In executing even a simple benefit/cost analysis, it is necessary for the decision-maker to supply quantitative information in order to impute a value to a undertaking. When this has been done, the undertaking can be viewed as a comparatively simple fiscal investing and hence capable to more standard fiscal investing tools. The danger of this is that it gives no consideration to the fact that proficient programmes are frequently aimed at a broad scope of strategic aims, a point made by Mitchell and Hamilton ( 1988 ) who made a separation into:

O exploratory/fundamental type work which is aimed chiefly towards the construct of cognition edifice.

For this type of work, the concern impact of which is frequently ailing defined and broad ranging and here R and D is frequently best considered as a necessary cost of concern.

O good understood proficient programmes normally associated with incremental betterments of bing merchandises which can be clearly defined. Here the R and D can be seen as an investing and treated consequently.

As usual with two extremes, the hard portion is the mid-ground where neither attack is peculiarly suited. Writers have attempted to utilize techniques borrowed from the fiscal community which frequently has to cover with uncertainness. Hazard analysis is a cardinal country in fiscal markets and several of the attacks used in fiscal analysis are besides found in the R and D direction country ; for illustration, determination trees and Monte Carlo analysis.