Venture Capital Decision Making

Venture Capital Decision Making

Chapter 9 Venture Capital Decision Making Start-up companies often do not have access to sufficient capital, but, if they could obtain that capital,...

1MB Sizes 1 Downloads 59 Views

Chapter 9

Venture Capital Decision Making

Start-up companies often do not have access to sufficient capital, but, if they could obtain that capital, they may have the potential for good long-term growth. If a company is perceived as having such potential, investors can hope to obtain above-average returns by investing in such companies. Money provided by investors to start-up firms is called venture capital (VC). Wealthy investors, investment banks, and other financial institutions typically provide venture capital funding. According to the National Venture Capital Association, the total venture capital invested in 1990 was $3.4 billion distributed among 1317 companies, while by 2000 it was $103.8 billion distributed among 5458 companies. Venture capital investment can be very risky. A study in [Ruhnka et al., 1992] 343

344

C H A P T E R 9. V E N T U R E C A P I T A L DECISION M A K I N G

Figure 9.1: A simple influence diagram modeling the decision of whether to invest in a start-up firm.

indicates that 40% of backed ventures fail. Therefore, careful analysis of a new firm's prospects is warranted before deciding whether to back the firm. Venture capitalists are experts who analyze a firm's prospects. Kemmerer et al. [2006] performed an in-depth interview of an expert venture capitalist and elicited a causal (Bayesian) network from that expert. They then refined the network, and finally assessed the conditional probability distributions for the network with the help of the venture capitalist. As discussed in [Shepherd and Zacharakis, 2002], such models often outperform the venture capitalist, whose knowledge was used to create them. In Section 9.1 we present a simple influence diagram, based on their causal network, that models the decision of whether to invest in a given firm. This simple influence diagram is for the purpose of providing an accessible introduction. Section 9.2 shows a detailed influence diagram obtained from their causal network. In Section 9.3 we show the result of using the influence diagram to model a real decision.

345

9.1. A SIMPLE VC DECISION MODEL

A Simple VC Decision Model

9.1

A simple influence diagram modeling the decision of whether to invest in a start-up firm appears in Figure 9.1. The influence diagram was developed using Netica. The following report, produced by Netica, shows the conditional distributions and utilities in the diagram:

Dil

Risk

"

t.rue

fml

0 .. 9 5

0 _0 5

true

0 .. 0 5

0 ._9 5

false

Need

Add

Fund

t rue 0 _ 05 0 ._ 9 5

VC

Res

se

Need

false

VC

0 _ 95 0 _0 5

sufficien~ insufficient

Res

i.n s u f f i c i e n t 0 .75

R i.s k

.Ad j ::

hi gh

i ow

P.ay_Ho

0_95

0 ..0 5

long

0 .. 2,S

0 .7 5

sh o r t

s hQ r~

0.9

0._I

hi gh 0.. I 0 7

r

r"

I c, n g

Like

Fund

Fund"

0_ 25

Pay_Ho

Fund

'

s u f f i c i.e n t

VC

.A.dd

ly__Ret

i. o~4 0_893

Invest

Likely_Risk

VC_Ris.k_Adj_Ret_Exp

Dilution_Risk

8S 90

yes yes

high high

high high

t rue false

87 .5

yes y e s

high h i.g h

low 1 ow

true fal. s e

y e.s y e s y e ~

l,c~w low law

h ig h high i o~w

t ru e fa i se t rue

15 50

y e s no,

1 ew high

1 :o:w high

fa i s e true

50

no

high

high

false

50

n o

h ig h

1 ow

t.r u e

50 50

n e n :o

hi g h I c~w

l.ow h ig h

fa i se t ru e

50

n o

i ow

h ig h

f a i. s e

50 50

n C. n o

10 W low

i :D~ i ow

~ rU e false.

95 i i i0

._Adj_Ret

346

CHAPTER 9. VENTURE CAPITAL DECISION MAKING

Figure 9.2: When the root nodes are instantiated to their most favorable values, the utility of the investment is over 93.

We see that only three variables directly affect the utility of the investment. These variables are the Likely_Risk_Adjusted_Return, Dilution_Risk, and VC_Risk_Adj_Return_Exp. The first of these variables concerns the likely return of the investment, the second concerns the risk inherent in the firm's ability to repay the loan, while the third concerns the risk-adjusted expectation concerning how soon the firm will repay the loan. The utilities are not actual returns. The utility can be thought of as a measure of the potential of the firm, with 0 being lowest and 1 being highest. An investment with a value of 1 can be thought of as the perfect investment. If Likely_RiskAdj_Return is high, VC_Risk_Adj_Return_Exp is low, and Dilution_Risk is false, then the firm has the most possible potential the model can provide, and a utility of 95 is assigned. So even when the variables directly related to the value node have their most favorable values, the utility is still only 95. We compare the investment choice to a utility of 50. If the utility of the investment is less than 50, it would not be considered a good investment, and we would choose to not invest. Notice in Figure 9.1 that the utility of the investment is only 11.3 when nothing is known about the firm. This indicates that, in the absence of any

9.2. A DETAILED VC DECISION MODEL

347

Figure 9.3: When the root nodes are instantiated to their least favorable values, the utility of the investment is about 1.47.

information about a firm, the firm would not be considered a good investment. Figure 9.2 shows the influence diagram with the root nodes instantiated to their most favorable values. We see then that the utility of the investment is over 93. Figure 9.3 shows the influence diagram with the root nodes instantiated to their least favorable values. We see then that the utility of the investment is about 1.47.

9.2

A Detailed VC Decision Model

The unrealistic aspect of the model just discussed is that ordinarily we would not be able to estimate whether the likely risk adjusted return is high or low. Rather there are many other variables that impact this variable, and we could estimate the values of many of them. Figure 9.4 shows an influence diagram based on the complete causal model obtained from the expert venture capitalist. The conditional probability distributions for this influence diagram appear in the appendix of this chapter. Notice that when nothing is known about a particular firm, the expected value of the firm's potential is still only 11.3.

348

C H A P T E R 9. V E N T U R E C A P I T A L DECISION M A K I N G

Figure 9.4: The detailed model of the venture capital funding decision.

9.2. A DETAILED VC DECISION MODEL

349

Figure 9.5: The influence diagram in Figure 9.4 with nodes instantiated to assessed values for a firm in the chemical industry.

CHAPTER 9. VENTURE CAPITAL DECISION MAKING

350

9.3

M o d e l i n g Real Decisions

In order to test the system, Kemmerer et al. [2006] asked the venture capitalist to recall an example of an investment opportunity from the past which he chose and which ended up being successful. A firm in the chemical industry was chosen. Next, the venture capitalist was asked to assess values for as many variables as possible relative to this firm, but the assessed values should be the ones the venture capitalist would have assessed at the time the decision was made. Figure 9.5 shows the influence diagram in Figure 9.4 with nodes instantiated to these assessed values. Notice that the utility of deciding to invest is about 71.4, which is somewhat above the baseline value of 50. Next, the venture capitalist was asked to recall an investment decision which had a disastrous outcome and to enter assessed values for variables at the time the decision was made. The utility of deciding to invest turned out to be 57.4, which is still above the baseline. However, when correct hindsight information regarding the market potential was entered, the utility turned out to be only 25.4, which is in line with the disastrous outcome. So the relatively high utility seemed to be due to poor assessment of the values of the variables. Notice in Figure 9.5 that the instantiated variable O t h e r _ I n d u s t r y _ T r e n d s is not instantiated to high or low. Rather, the probability of it having value favorable has changed from .5 to .758. Instead of instantiating a variable to a precise value, Netica allows the user to enter the new probability of the variable based on the user's evidence E. In this case, the capitalist entered

P ( O t h e r _ I n d u s t r y _ T r e n d s - f avorablelE ) - . 7 5 . The probability then became .758 because the variable Others Assmt Market was set to high, and these variables have a dependency through the variable R i s k _ A d j _ M a r k e t _ P o t . Similarly, the probability of Dynamic_Competitive _Issues being favorable became .572 because the capitalist entered

P(Dynamic_Competitive_Issues - f avorablelE ) - . 5 .

EXERCISES Section 9.1 E x e r c i s e 9.1 Using Netica, develop the influence diagram shown in Figure 9.1. Investigate the investment decision for various instantiations of the chance nodes.

EXERCISES

351

Section 9.2 E x e r c i s e 9.2 Using, Netica develop the influence diagram shown in Figure 9.5. 1. Investigate the investment decision for various instantiations of the nodes. 2. Try to obtain information on an actual firm that tried to obtain venture capital. Enter as much information as you know for the firm. Compare the recommended decision to that which was actually made concerning the firm. If the firm obtained funding, compare the recommended decision to the outcome.

CHAPTER 9. VENTURE CAPITAL DECISION MAKING

352

9.A

Appendix

The following report, produced by Netica, shows the conditional probability distributions and utilities in the influence diagram in Figure 9.4: Dil

Riskfal se 0.05 0 . 9S

true 0.95 0. O S

VC

Res

Need Add ~.rue false

Fund:

s:u f f i c i e n ~ 0.25

ins uf ficien~. 0 .75

M g m.t__Rep hi gh 0.8 0.2

Ac q

Mg mt

i ow 0 .2 0.8

Mg m.t_Q ual high low

f.al s e 0 _ 35 0.5

Mgmt_0 hi gh I.o w

-

~E~e 0.65 0_5

.St r o n g _ R e l _ N : e

ual

t"

9: . r u e 0_ ~S

0 _3 5

0. 3 5

0 _ 65

Mgmt_Qual high low

fal se 0 .35 0 _ 65

Mg m.t_Q ual high io w

fal ae 0_2 0.8

Mg:mt_0 ual hi. g h io w

i.o w 0 _25 0 _75

Risk hi. g h la w

fal se

S t rg_Adv

-

~.rue 0.65 0.35:

Mgmt_Exp

-

~.rue 0.8 0.2

Other

Mrkt

hi gh 0.75 0_7.5

Part

Mot

hi gh 0.25

I.o w 0.S

hi gh 0.5

Te.ch

"

Pot" I ow 0 ..7S

Mrkt

Fund

9.A. APPENDIX Tch

Know1 :OTg 0 .25 0 _75

h i gh 0_75 0.25

In d_Dv

unf 0 _0 0 _2 0 . 2 0 .9

0_75

0_05

Trends

unfavbl 0.5

0_5

e

Prog_B fast. f.a s t slow s i ow

ig_.Pl ay

Nd_Ind_I pr.e s e n ~ a b s en~ presen~ .ab s e n ~

e

Know-

hi gh 0_75 0.1

Mgm.t

1 ow 0 2 5 0.9 _

Qual

Prod

1 Q:w 0 .75

Olty:: I.o w G _ 07

hi gh

O. 9 3 0~83 0.55

0 _ 17

0 . 45 0 _ 9:5

0.05

Sc i_Te

ch_01

hi gh 0.25

Mr k t

ty " 1 ow 0._75

S t r gy -

mpprt 0.25

Po t_Pr

M g m t _ 0 u a i. hi. g h io w

"

hi gh 0_25

~rue 0.25

mvbl ~ 5 5 5

"

fa vb i e

~fkt

Mg m t_O u a 1 hi. g h 1o w

ipm.t "

fa vb le 0_94 0.75

Ind

353

ina pp r5 0 _75

od_Lblt

y

-

false 0_75

Sc i_Te high hi gh io w io w

ch_Ql

ty

T ch_Know h igh 1 ow

h ig h i ow

nnov

CHAPTER 9. VENTURE CAPITAL DECISION MAKING

354

Dy n_Comp

-

fa vb le 0_25

Nd

unfavbl.e 0_75

I n n o v :.

I nd

pros O_S

en 5

P r og_B

absen O_S

ig_P

i ay =

fa st 0.5

s 1o w 0.5

Other

Prod

:

hi gh 0_75 0.2.5

Tech

i.o w 0 .25 0 _75

Unc

Adj

:

hi gh 0.95 0 ~5

Size

:St r o n g _ P

P.ay_Hor long short

1 ow 0 ~OS

nonQ 0_9

false 0 _ 18 0 . 75 1 0 _25 0 _ 95 1

Part high high h i gh low low low

i . n m p p rt. 0 _25 0 _75

.Mitt K n o w hi. g h io w

ar t :

t.r u e 0 _ 82 0_25 0 0_75 0.05 0

appr~ 0_75 0_25

1 OT,~ 0.0S 0.75

:

hi. g h 0.05

Prod

Prod

i [email protected] 0 _2.5

Risk

Part

Ris.k h i.g h low

rty:::

hi gh 0_75

VC

t

Mot

Strat=

Part high low no ne high l.ow none

Size

9.A. APPENDIX P a y_Ho

r "

Io n g 0.9

Need

sho rt 0 .i

Add

Fund

0.05 0.96

Risk

P rod

f a i se

VC

0 .95 0 ._05

su ff i c ient

10T~ 0_9 0.9 0._9 0._9 0..9 0..9 0..9 0_9 0_1 0_0 0_7 0.7 0_9 0_8 0.9 0_9

0~25 0..3 0~I 0..15 0..05 0.075

hi Me ly__Ret

hi gh 0_9 0_9 0..35 0..3 0_45 0..4 0.3 0.25 0_5 0.25 0.2 0..1 0..2 0..I 0..1 0.~01

Pe rs,Wgm

Res

Fund

in su ff i cien

t

"

hi gh 0.~ 0..05 0_05 0.05 0.05 0.05 0.05 0..05 0.9 0_95

apprt 0.55 0_45

355

5 5 5 5 5 5 5 5

Po t_Prod_Lbl t rue t rue t rue 9t r u e t9 r u e true. rue true fa Ise

5 5

fa fa fa fa

5 5 25

fa i s e f a l.s e fa i s e

ty

l.s e Is e i.~ e l.s e

Prod_Ql hi. g h h i gh hi gh hi gh io w low io w io w hi gh

ty

P rod_S trat mpprt m ppr t i nap pr ~ inmppr t a ppr 5 mppr t i nap pr ~ i nap prt a p p r t.

Teoh_Unc h i gh i ow high low high low high 1 ow h igh

high high high low

a p p r 4. i nappr ~ i nappr t mpprt

i ow high i ow high

low low low

mppz t inappr inmppr

1 ow high low

t t

" i o~g 0_i 0_i 0.6 5 0.7 0.5 S 0.6 0.7 0 . 7 .5 0.5 0.7 5 0.8 0.9 0.8 0_9 0_9 0.9 9

Mgmt_Qual hi. g h hi. g h hi gh hi. g h hi gh

Risk_Prod h i gh h i gh h ig h h ig h i c,w

Risk_Mrkt hi gh hi gh 1o w io w hi gh

Strong_Part ~ ru e f al s e ~ rue f m i se t rue

hi gh h i gh hi gh

Iow 1 ow I ow

hi gh 1o w io w

f a i se ~ ru e f al s e

io w Io w

h ig h h ig h

h i gh h i gh

~ ru e f a i ~e

Io w low Io w low 1o w 1o w

h h 1 i i I

1o w low hi gh hi gh io w io w

t f ~ f t f

i n m pp zt 0 _45 0 .55

~g m t _Q u a 1

t"

hi gh io w

ig h i gh ow ow ow ow

zu e al se rue ml se ru e a i se

rty

356 Ri s ~ _ M r ~

CHAPTER 9. VENTURE CAPITAL DECISION MAKING t =

h i gh 0_9S 0_95 0.95: 0_ 92 0.33 0.3 0_"8

low

o

0

~s

0.33 0.3 0.28 0.25 0.01 0.01 00l 0_01 0 95 0.92 0.87 0.78 0_28 0.26 0.24 0.22 0 28 0.26 0."4 0.22 0.01 0.01 0.01 0_01 0.69 0 6:3 0.59 0.53 0.19 0.17 0 16 0_15 0_19 0.17 0.16 0.15 0.01 0.01 0.01

0.01 0.6 0.$4 0_51 0 46 0_17 0.1S 0 14 0_13 0.17 0.1S 0 14 0.13 0.01 0.01 0 01 0.01

0_05

0.05 0 05 0 .08 0 .67 0.7 0.7 ~

~

0 _67 0_7 0 _72 0 75 0 .99 0 .99 0 99 0 _99 0 05 0.08 0 .13 0 _22 0 .72 0 _74 0 76 0 .78 0 .72 0

74

0.76 0 .78 0 99 0 .99 0 _99 0 _99 0.31 0 _37 0.41 0.47 0 81 0 .83 0.84 0 _85 0 .81 0 _83 0 84 0 _85 0 .99 0 _99 0 99 0 .99 0.4 0.46 0 .49 0 54 0 _83 0 _85 0.86 0 87 0 _83 0 _85 0 _86 0.87 0 .99 0 .99 0 .99 0 .99

Dyn_Comp favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble favble un favble u n f a v b le u n f a v b le un favble un favble u n fa v b le u n f a v b le un favble u n f a v b le un favble unfavble un fmvble u n fa v b le un favble un favble un favb le u n fa-r le u n f a v b le u n f a v b le u n f a v b le un favble un favb le un favble u n f a v b le un favble un favble u n f ~ v b le u n f a v b le un favble u n f a v b le un favble un favble

Tech_Uncrty high high high high high hi gh high high high hi gh high high hi gh high high high low low low low low low low low low low low low low low low low high high high high high hi g h high high high high high high hi gh high high high low low low low low low low low low low low l~w low low low low

Tech_Pot high high high high high high

high high low 1 ow low low low low low low high high high high high high high high low low low low low low low low high high high high high h igh high high low low low low 1 ow lmw low low high high high high high high high high low low l~w low low low low low

Mrkt_Strgy apprt appr~ appz~ appr~ inappr~ inappr~ inapprt inappr~ appz~ app r t appz~ apprt inappz~ inappr~ inappr~ inapprt appr~ appr~ apprt apprt inappr~ inappr~ in appr~ inappr~ appr~ apprr apprt ~pprt inappr~ inapprD inappr~ inappr~ appr~ appr~ appz~ .mp p rD inappz~, i na p p r~ inapprr i n a p p r~ appr~ appr~ appr~ apprt i na pp r ~ i n a p p r~ inappr~ inappz~ .app r~ appr~ appr~ apprr inappr~ inappz~ i n a p p r~ inapprD appr~ appr~ .appr~ appr~ inappr~ in appr~ inappr~ in~ppr~

I rives t

Likely_Ret

VC_Risk_Adj

Dil.Risk

8S

yes

high

high

true

90

yes

high

high

fal.s e

87 .5 9S

y e s yes

high high

low low

true fa i s e

1

y e s

low

high

t rue

1 I0

yes yes

low low

high low

fa is e true

U

low

low

fa is e

50

no

high

high

true

SO SO

no no

high high

high law

fm is e true

50

no

high

low

fals

50

no

low

high

true

50 50

n o n o

low low

high low

fa is e t rue

5O

no

low

low

fals

IS

yes

e

e

Ind_Dvlpmt favble favble unfavble unfavble favble favble unfavble unfavble favble f avbl e unfavble unfavble favble favble unfavble unfavble favble favble unfavble unfavbl.e favble favble unfavble unfavble favble favble unfavble unfavble f~vble favble unfavble unfavble f avble favble unfavble u n f a%-ble favble f avbl e unfavble unf avble favble f avble unf avble unfavble f avb I e f mvble unf avble unfavble f avble favble unf avble unfavble favbl e favble unfavble unf avble favble f avble unfavble unfavble favble favble unfavble unf avble

Ind_Trends favble unfavble favble unfavble favble unfavble favble unfavble favble unf,avble favble unfavble favble unfavble favble unfavble favble unf avble favble unfavble favble unfavble favble unfavble fBvble unfavble favble unf avble favble unfavble favble unfavble f avble unfavble f avble unfavbl.e f avble unfavble f avble unfavble favble unfavble favble unf avble f av bl e unfavble f avble unfavble f avbl.e unf avble favble unf avble favble unf avble favble unf.mvble f avble unf mvble favble unf avble f avble unfavble f avble unf.svbl.e