operations and finance...

Operations and Finance Interactions

John R. BirgeUniversity of ChicagoUniversity of Chicago

Booth School of Business

MSOM INSEAD, July 2013 1© JRBirge

Overview: What to Remember?Overview: What to Remember?• Assuming long-term arbitrage is dangerous

– Systematic risk is different from idiosyncraticSy y– Requires some adjustment to MSOM model

• Assuming no arbitrage can create• Assuming no arbitrage can create computational efficienciesI i f t (i l) k t ti• In imperfect (i.e., real) markets, operations-finance interactions matter

f h f d h l h– for the firm and the supply chain• Operational-financial considerations can

– guide policy and explain empirical observations© JRBirge MSOM INSEAD, July 2013 2

Assuming No Arbitrage• Arbitrage: the ability to start out withArbitrage: the ability to start out with

(possibly nothing) and never lose and to make a profit in some future statesmake a profit in some future states

• Easy to miss because:i d i i d l “ fi ” d bin a pure decision model, “profit” needs to be

examined – often not really present• In particular, financial instruments should

not able to provide (sustainable) arbitrage

© JRBirge MSOM INSEAD, July 2013 3

Risk Implications• With risky returns, there is a difference

between what can be hedged (idiosyncratic) and what cannot be hedged (systematic)

• To (a diversified investor):

Hedged Un-Hedged

© JRBirgeMSOM INSEAD, July 2013

4

Hedged

Risk TakeawayTh t f i k tt• The type of risk matters:– Idiosyncratic risk (without other market

i f ti ) h ld t b i dimperfections) should not be priced– Systematic risk should be priced by investors at

the market price of riskthe market price of risk– Lumping systematic and idiosyncratic risk is

inconsistent with no arbitrageinconsistent with no-arbitrage • Paying more for a market item than others

illi t i dare willing to pay is dangerous© JRBirge MSOM INSEAD, July 2013 5

Risk-Neutral Pricing ConclusionRisk Neutral Pricing Conclusion(Inter-temporal CAPM)

• The market and idiosyncratic components of cash flows can be separated p

• The market components should be priced by the market price of risk (with the remainderthe market price of risk (with the remainder diversifiable)

• Simple changes of the probability• Simple changes of the probability distribution then give consistent results


Implications for CapacityImplications for Capacity (JRB/MSOM 2000)

• USE CORRELATION TO THE MARKET?• USE CORRELATION TO THE MARKET?• Can measure for known markets (beta values)• If capacitated, depends on decisions

• Constrained resources capacity• Constrained resources - capacity• Correlations among demands

Revenue

A h t th d d di t ib ti t i k t l i l t b

DemandCapacity No revenue variation -

low discount

High revenue variation (risk) - high discount

• Approach: convert the demand distribution to a risk-neutral equivalent by removing the risk premium

– Capacity is like an option that can be priced using this approach

7/28/2013 MSOM 2013 7

Computing Capacity Value• Goal: Production value with capacity K

• Compute uncapacitated value based on CAPM:• S = e-r(T-t)c S dF(S )• St e ( )cTSTdF(ST)• where cT=margin,F is distribution (with risk aversion),• r is rate from CAPM (with risk aversion)

A S i kf l if i k• Assume St now grows at riskfree rate, rf ; evaluate as if risk neutral:

• Production value = St - Ct= e-rf(T-t)cTmin(ST,K)dFf(ST)• where Ff is distribution (with risk neutrality)

Capacity, K

V l T

Sales Potential, ST

Value at T

7/28/2013 J.R. Birge, Northwestern University 8

Computational Efficiencies• In capacity models (with log-normal demand):

– Demand distribution transformation is equivalent to a shift of capacity by the risk premium, e(rm-rf)(T-t)

• For the news vendor (JRB/Zhang, TEE1999), logic yields:

p− c ( )(T t) p− cFf (x

∗) =p c

p− s → F (e(rm−rf )(T−t)x∗) =p c

p− sor

x∗ e−(rm−rf )(T−t)F−1(p− c

)


x = e ( m f )( )F (p− s ).

Implications for Flexibility p yResources(Aytekin/JRB 2004)( y )

• Suppose firms can invest in capacity in multiple locations to take advantage of market p gfluctuations (exchange rates/demand)

• Implication:Implication: – Can identify limited set of possible configurations

Can develop closed form expressions of value with– Can develop closed-form expressions of value with computable bounds


Alternative Operations• Production Distribution

Domestic ???

Foreign

?

?ForeignAdvantage of foreign operations: When cost (exchange rate) is advantageous, can shift production.When cost (exchange rate) is advantageous, can shift production.

Excess capacity in D and F provide an option to shift production to the favorable location

© JRBirge UToronto 28 Oct , 2011

favorable location.

Valuing the Alternatives

• If sufficient flexible capacity, produce in the market with favorable exchange rateg

• Set thresholds for production shifts to overcome setup and changeover costsovercome setup and changeover costs.

• Shift production when limits are exceeded.G i l b l• Gain: natural balance.

• Cost: additional capacity and transaction.


Value of Flexible CapacityValue of Flexible Capacity• Can evaluate the capacity as a perpetual option

using only correlation to the market for theusing only correlation to the market for the elimination of risk

• Value also increasing in the volatility of the• Value also increasing in the volatility of the exchange rate but reaches a limit:


Markets Imperfections andMarkets Imperfections and Operations-Finance

• Without market imperfections, operational and financial can be separated p(Miller/Modigliani)

• Market imperfections make it less possibleMarket imperfections make it less possible to create perfect hedges

• Result is that operations and finance cannot• Result is that operations and finance cannot be separated


Operations-Finance Links:pImplications for the Firm

• Imperfections:Imperfections: – Tax advantage of debt– Cost of financial distress (e g loss in default)Cost of financial distress (e.g., loss in default)

• Newsvendor extension (Xu/JRB 2004, NRL 2006, TEE 2008):TEE 2008):

*

max ( , ) [ ( )] ( )

[ ( )] ( )

xx

q

V x D px px cx rD f q dq

ps ps cx rD f q dq

τ

τ

∞= − − −

+ − − −

*

0( ) ( ) (1 )

. . (1 ) (1 )[1 ( )] ( )

b

b

b

q q

fq

qbf

pqf q dq pqf q dq cx r

s t D r D r F q pqf q dq

α

α

+ + − +

+ = + − +


0( ) ( )[ ( )] ( )

0f q pqf q q

D cx≤ ≤

Joint Decision Making in an Imperfect Market

12

Interactive Effects between Production d Fi i l D i iand Financial Decisions

Separate operational and financial decision making is sub-optimal Production decision is negatively related with the production cost Debt decision is positively related with the production cost Optimal production decision is a decreasing function of financial leverage

1000

1200

0.8

0.9

atioProduction Level

Leverage Ratio

Optimal production decision is a decreasing function of financial leverage

800

1000

Out

put L

evel

0.6

0.7

t Lev

erag

e RaLeverage Ratio

400

600

0.4 0.5 0.6 0.7 0.8 0.9

O

0.4

0.5

Mar

ket

Production Cost

13

The Effects of Decision MisspecificationpKey Observations: Firm’s value is a convex function of financial leverage The effect of over-leverage is more severe than under-leverage Low-margin companies especially exposed to mis-

specifying decisions

The effect of production decision misspecification is moresevere

Production Misspecification

1e

Leverage Ratio Misspecification

e

0.6

0.8

1

aliz

ed F

irm V

alue

c: 0.6

c: 0 7

0.9

0.95

1

aliz

ed F

irm V

alue

c: 0.6

c: 0.7

0.2

0.4

0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5

Production Level

Nor

ma c: 0.7

c: 0.80.8

0.85

0 0.2 0.4 0.6 0.8 1

Leverage Ratio

Nor

ma

c: 0.8

14

Production Margin and Capital StructureTradeoff-Theory

An lower profit margin results in higher leverage (Fama & French 2002)

Model results

Leverage ratio is convex and U-shaped in production cost

Empirical ResultsEmpirical Results

For very high margin, leverage ratio appear to increase

60%

80%

100%

rage

Rat

io

Book Leverage Ratio

Market Leverage Ratio

0.3

0.4

0.5

aver

age

Rat

io

0%

20%

40%

0 0.2 0.4 0.6 0.8

P d ti C t

Leve

r

0.1

0.2

75% 80% 85% 90% 95% 100%O

Mar

ket L

a

Production Cost

15

Pre-tax Operating Margin

Additional Empirical EvidenceAdditional Empirical Evidence(JRB 2011)

• With fixed costs in this model leverageWith fixed costs in this model, leverage should first increase and then decrease in operating marginoperating margin

• Leverage/Operating Margin:


Implications for the Supply ChainImplications for the Supply Chain (Yang/JRB 2011)

• Trade credit provides a risk-sharing mechanism p gbetween firms

• By borrowing from a supplier, a buyer inducesBy borrowing from a supplier, a buyer induces the supplier to share in the risk

• Chain coordination can then be enhanced• Chain coordination can then be enhanced• Policies should favor junior credit • Empirical evidence:

– Firms use trade credit then short-term when needed– Making suppliers senior seems to hurt trade credit


Summary ObservationsSummary Observations• Operational decisions should be consistent with

financial realities• Operational and financial decisions are inter-

dependent in practical marketsdependent in practical markets • Inter-dependencies extend throughout the

supply chainsupply chain• Operational analyses (particularly the role of

it t) l i b d b h icommitment) can explain observed behavior and guide policy

• Much more research is needed at this interface © JRBirge MSOM INSEAD, July 2013 22

Thank you to MSOM!Thank you to MSOM!


operations and finance...

Documents