MANAGEMENT SCIENCEVol. 32, No. 10, OcKiber I9M

Piimed in t/SA.

STRATEGIC FACTOR MARKETS: EXPECTATIONS, LUCK,AND BUSINESS STRATEGY*

JAY B. BARNEYGraduate School of Management. University of California, Los Angeles. California 90024

Much of the current thinking about competitive strat^y focus^ on ways that firms cancreate imperfectly competitive product markets in order to obtain greater than normal economicperformance. However, tiie economic performance of firms does not depend simply on whether(H- not its strati^es create such markets, but also on the cost of imiriementing those strategies.Cteaily, if the cost of stiattgy imtdementation is greater than returns obtained from creatingan imperfectly competitive product market, then firms will not obtain above normal economicperformance from their strat^izing efforts. To help analyze the cost of implementing strat^es,we introduce the concept of a strat^c factor market, i.e., a market where the resourcesnecessary to imidement a strat^y are acquired. If strat^c &ctor markets are perfect, then thecost of acquirii^ strat^c resources will amnvximately equal the economic value of thoseresources once they are used to implement product market strat^es. Even if such strat^iescreate imperfectly competitive product markets, they will not generate above normal economicperformance for a firm, for their full value would have been anticipated when the resourcesnecessary for implementation were acquired. However, s t r a t a &ctor markets will be imperfectlycompetitive when di%rent firms have different expectations about the future value of a strategicresource. In these settings, firms may obtain above normal economic performance fromacquiring strat^ic resources and imidementing strategies. We show that other apparent strategic&aisx market imperfections, including when a firm already controls all the resources neededto implement a strategy, when a firm controls unique resources, when only a small number offirms attempt to implement a strat^y, and when some firms have access to lower cost capitalthan others, and so on, are all specMi cases of diflferences in expectations held by firms aboutthe future value of a strategy resource. Firms can attempt to (tevelop better expectations aboutthe future value of strategic resources by analyzing their competitive environments or byanalyzing skills and capabilities they already control. Environmental analysis cannot be expectedto improve the expectations of some firms betta than others, and thus cannot be a source otmore accurate expectations about the future value of a strat^c resource. However, analyzinga firm's skills and capaUIities can be a source of more accurate expectations. Thus, from thepmnt of view of firms seekii^ greater than normal economic performance, our analysis su^eststhat strategic chcnces shcnild flow mainly fix>m the analysis of its unique skills and capabilities,rad^r than bom the analysis of its competitive environment.(STRATEGY IMPLEMENTATION; EXPECTATIONS; LUCK)

Introdnction

Researeh on corporate growtii through acqtiisitions and mergers suggests the existenceof markets for buying and selling companies (Porter 1980, p. 350; Schall 1972; Mossin1973; O^wland and W^ton 1979). Mcst eminrical evidence seems to suggest thattiiese marked are reasonably competitive. That is, the price an acquiring firm will^nerally have to i»y to acqmie a firm in these nfiarkets is approximately equal to thediscounted |»esent value of the acquired firm (Mandelkar 1974; Halpem 1973; Ellert1976). Imleed, if above normal returns accrue to anyone in markets for companies,rraearch seems to atggest that they will most likely go to the stockholders of theacquired, rather than tiie acquiring firms (Porter 1980, p. 352; EUert 1976).

To » ^ e ^ t i ^ on averse, acquiring firms cannot expect above normal returnsfrom tiieir inve^meots in corpraate ^xiuiations is not the same as si^^sting that nofinns ever expaience such returns. In^«d, to the extent that imperfect competition in

* Accepted by AIM Y . Lewm; received Fetauary 10, 1984. This papa has bem wiA the author 6 monthsfor 5 leviaoBS.

12310025-1909/86/32 lOAXXJOSOl .25

Copfit^ ® I9M. The inttffle orMaBagnBcnl Scieaoa

1232 JAY B. BARNEY

these maiieets exists or can be created, acquiring firms may be able to obtain suchreturns. Porter (1980, i^. 353-354) has isolated several such competitive imperfections.

From a broader perspective, markets for comi)anies are just one example of strategicfactor markets^ Whenever the implementation of a strategy requires the acquisition ofresources, a strategic factor market develops. These markets are where firms buy andsell the resources necessary to implement their strategies (Hirshleifer 1980). In the caseof markets for companies, firms wishing to implement a strategy of product diversifi-cation may decide to do so by acquiring other firms. In this sense, because an acquiredfirm is the resource required to implement a firm's diversification strate^, the marketfor companies is a strat^c factor market

All strategies that require the acquisition of resources for implementation havestrategic factor markets associated with them. Thus, for the strat^y of being a lowcost producer, a resource necessary for implementation may include, among otherresources, large market share (Henderson 1979), and a relevant strat^c factor marketmay be the market for market share (Rumelt and Wensley 1981). For a strat^y oflow volume, h ^ margin sales (Porter 1980), a relevant resource may be a qualityreputation, and a relevant strategic &ctor market may be the market for corporatereputations (Klein, Crawford, and Alchian 1978). For a strategy of being a productinnovator, a relevant resource might be research and development skill (Thompsonand Strickland 1980), and relevant strategic factor markets may include the labormarket for research scientists (Hirshleifer 1980). For most strategies, man^ement skillwill be a resource required for successful implementation (Porter 1980). Thus, in thissense, managerial and other labor markets can also be strat^c factor markets.

The existence of strategic factor markets has important implications for returns toproduct market strat^es implemented by firms, for the size of the returns to productmarket strat^es wiU dqiend on the cost of the resources necessary to implementthem. And the cost of these resources will depend on the competitive characteristicsof the relevant strategic factor markets. If strategic factor markets are perfectlycompetitive, then the full value of product market strat^es vnll be anticipated whenthe resources necessary to implement these s t r a t ^^ are acquired, and firms will onlybe able to obtain normal returns from acquiring strategic resources and implementingstrat^es. Firms can only obtain greater than normal returns from implementing theirproduct market strategies when the cost of resources to implement those strategies issignificantly less than their economic value, Le., when firms create or exploit competitiveimperfections in strat^c factor markets.

From a normative point of view, the existence of strat^c factor maiicets suggeststbe importance of developing a conceptual framework that firms can use to anticipateand exploit competitive imperfections in strat^c &ctor markets. Such a frameworkwould ^sist firms in choosing h i ^ retum prodiK t maricet strategic to imidementThe primary objective of this paper is to b ^ n to develop such a con(%ptu^ framew(»4c.

We develop our discusaon of the competitive implications of stiat^c fitctor marlcetsin two parts. In the first part, we argue that firms tiiat wish to obtain expected abovenormal returns from implementing produiA market ^rat^ies must be consistentlybetter informed concerning ths future value of those strat^es tiian other firms actii%in the same strat^c fauctor maiieets. We also aigue that other apf^aent sources ofadvantj^ in strat^y imidementation are, in &cl, dther a manifestation of thesespecial in^hts into the future value of i^^at^ies, cnr a manif^tation of a firm's goodfortune and luck. In the second part of the paper, we outline some ways that firmsatD become better informed about the future value of sttategi^ being imi^mented,including throu^ the analy^ of a firm's competitive environment and th ro t^ theanaly^ of its unique skills ami a^abilities. We coiKslude thai, environmoital analysis,by itself, cannot create the required unique insets , while in ^nne drcunastances, the

STRATEGIC FACTOR MARKETS 1233

analysis of a firm's unique skills and capabilities can. In a final section, otir argumentsare summarized and some of their implications for the practice and theory of strat^yare discussed.

Competitive Imperfections in Strat^c Factor Markets

Perfect Strategic Factor Market CompetitionWhen firms seeking to acqtxire resources to implement a strategy (strat^izers) and

firms who currently own or control these resources (controllers) have exactly the same,and perfectly accurate, expectations about the future value of product market strat^esbefore Uiey are actually implemented, then the price of the resources needed toimplement these strategies will approximately equal their value once they are actuallyimplemented. This is a conclusion of normal returns consistent with all perfectinfonnation models of competition where no competitive uncertainty exists (Hirshleifer1980). Under these perfect expectation conditions, controllers will never sell theirr^ources if the full value of those resources is not reflected in their price, nor willstrategizers pay a price for a resource greater than its value in actually implementinga strat^y. In such markets, all pure profits that could have been had when the strat^yin question was implemented will be anticipated and competed away.

Expectations in Strategic Factor Markets

These perfect competition dynamics, and the normal returns from implementingstrategies they imply, depend, of course, on the very strong assumption that allstrat^izer and controller firms have the same, and perfectly accurate, expectationsconcerning the future vsdue of strategies. This is a condition that is not likely to existvery often in real strategic factor markets. More commonly, different firms in tliesemarkets will have different expectations about the future value of a strategy. Thesedifferences reflect uncertainty in the competitive environments facing firms. Becauseof these differences, some firm expectations will be more accurate than others, althoughfirms will typically not know, with certainty, ahead of time how accurate theirexpectations are. When different firms have different expectations concerning the futurevalue of a strate^, it will often be possible for some strate^izing firms to obtain abovenormal returns firom acquiring the resources necessary to implement a product marketstrat^y, and then implementing that strategy.

Consider first the retum potential of a firm that has more accurate expectationsconcerning the future value of a particular strategy than other firms. Two likelypossibilities exist. On the one hand, Kveral other firms might overestimate a strategy'sretum potential. This overestimation will typicaUy lead to strat^c factor market entry,competition, and the setting of a price for the relevant strat^c resource greater thanthe actual value of that resource when it is used to implement a strat^y. In thissituation, firms with more accurate expectations concerning the retiun potential of astrat^y will usually not enter tiie strat^c factor market, for they will believe tiiat indoing so they wiD probably sustain an economic loss by paying more for a strat^cresource than that resource is worth in implementing a strategy. Thus, in the loi^ run,firms with m<ae accurate e5y}ectations will usually be able to avoid economic lossesassodated with Imying overpriced ^ ra t^c resources. Firms that do acquire theseowrpriced resources suffer from the "winna-'s curse," Le., the fact that they success-fully acquire tiw rraources in qu^tion ai^sests that tiiey overtnd (Bazerman andSamueteon 1983).

T1» second pc^lality feeing firms with more accurate expectations is that otherfinm, laflier than overe^imating the retum potential of a strat^y, mi^t under^timatethat strat^^'s tnw future value. Entry and competition in the strat^c &ctor market

1234 JAY B. BARNEY

would, in this case, typically lead to a strategic resource pii(% less than the actualfuture value of the strategy. In this situation, firms with more accurate expectationsabout the future value of the strategy in question will enter the strat^c factor marketand will pay the same for the relevant strate^c resource as firms with less accurate(i.e., pessimistic) expectations. Firms with more accurate expectations will not be ableto buy the relevant resource for less because of the inaccurate expectations held by ill-informed controllers and strategizers. And firms with more accurate expectations willcertainly not want to buy these resources for more. As strategies are implemented,equal above normal returns will accrue to all those firms that acquired the resourceand implemented the strategy, the well informed and ill-informed alike.

Thus, on the one hand, firms with more accurate expectations concerning the futurevalue of a strat^y can avoid economic losses due to the optimistic expectations ofother firms. On the other hand, these firms will also be able to anticipate and exploitany opportunities for above normal returns in strat^c factor markets when they exist.Thus, by avoiding losses and exploiting profit opportunities, these firms, over the longrun, can expect to perform better than firms with less accurate expectations about thefuture value of strategies.

Despite the advantages of havii^ a superior understanding of a strat^y's retumpotential when acquiring the resources necessary to implement that strategy, firmswithout this superior insight can still obtain above normal returns when acquiringresources to implement strategies. This can occur when several of these firms under-estimate the retum potential of a strategy. Because of this underestimation, the priceof the resources necesary to implement a strat^y will be less than the actual futurevalue of the strat^y. In this sense, these firms are able to buy a strat^y generatedca^ flow for less than the value of that cash flow. This is one definition of an abovenormal retum. However, this above normal retum must be a manifestation of ^ ^ efirms' good fortune and luck, for the price of Uie strate^c resource acquired was basedon expectations about the retum potential of that strat^y. Returns greater than whatwere exported are, by definition, unexpected. Unexpected superior economic returnsare just that, unexpected, a surprise, and a manifestation of a firm's good luck, not ofits ability to accurately anticipate the future value of a strat^y.

Even well-informed firms can be lucky in this manner. Whenever actual returns toa strategy are greater than expected returns, the resulting difference is a manifestationof a firm's unexpo^ted good fortune. The more accurate a firm's expectations about astrat^y's retum potential, the less of a role luck will {day in generating above normalreturns. In the extreme, thou^ probably very rare, case where a firm knows withcertainty the retum potential of a strat^y before that strategy is implemented, therecan be no unexpected retums to that firm from implementii^ strat^ies, ami thus nofinancial smpises. However, to the extent that a firm has 1 ^ than perfect expectations,luck can play a role in determining a firm's retums to implementii^ its

Other Apparent Competitive Imperfections

Firms with consistently more accurate expectations concerning tiie retum potentialof strat^ies diey are imi enKsnting can expect to enjoy h ^ ^ retums fixmi inq^n^atii^their strategies over the long mn. In this sense, differences in firm expectationsconstitute a strat^c &ctor market aimpetitive imperfection.

Some have st^^^ed that other differences b^ween firms, beacte di&renc^ in firmexpectations, can OKOte conujetitive imperfections in sttstegx fsxtaac markets. Hiesefirm differences, it is thought, can prevent oertsdn firms bom impleoKntiithat c^her firms can imidement Howevo-, <Aoss analyas of these othst

firms st^gests that, to the extent ttot they conj^tnte comp^itive impo&ctions

STRATEGIC FACTOR MARKETS 1235

in strate^c &ctor markets, they are actually a manifestation of different expectationsfirms hold about the future value of strategies being implemented. In this sense,differences in firm expectations are the central sotirce of above normal returns fromacquiring resources fi'om strategic factor markets to implement product marketstrategies. To see how other firm differences that can apparently give firms competitiveadvantages in strategic factor markets are actually manifestations of differences in firmexpectations, consider the following examples.

Lack of Separation. It has been suggested that a competitive imperfection in astrat^c factor market exists when a small number of firms seeking to implement astrategy already control all the resources necessary to implement it (Thompson andStrickland 1980). In this setting, these firms do not need to buy the resources necessaryto implement a strategy, and thus apparently stand in some competitive advant^e.An example of this lack of separation might include a uniquely well-managed firmseeking to implement a low cost manufacturing strategy. Such a firm already controlsmost, if not all, the resources necessary to implement such a strategy and thus is at anadvant^e compared to firms that would have to improve their efficiency in order toimplement such a strategy (Porter 1980). However, from another point of view,whether or not a lack of separation between strategizers and controllers of this type isa competitive imperfection depends on the expectational characteristics of earlierstrat^c factor markets.

Firms begin their history with a relatively small endowment of strategy relevantresources (Lippman and Rumelt 1982; Kimberly and Miles 1981). Most resources forimplementing strategies must be acquired fi-om a firm's environment at some point ina firm's history (Pfeffer and Salancik 1978; Hannan and Freeman 1977). Once acquired,they can be combined and recombined in a variety of ways to implement differentstrategies. It certainly may be the case that a firm, some time ^ o , acquired theresources necessary to implement some strategy that it would now like to implement.If the value of these resources in their current strategic use was anticipated bystrategizers and controllers at the time it was originally acquired, then these resourceswould have been priced at a competitive level. Thus, no competitive imperfectioncurrently exists, even though the firm controls all the resources necessary to implementa strat^y, because these resources were competitively priced in a previous strat^cfactor market.

If, as seems more likely, the resource was acquired for one purpose, and onlyrecently did its value in implementing another strategy become known, then its currentvalue was unanticipated when the resource was acquired. That is, this original strat^cfactor market was imperfectly competitive becatise of imperfect expectations held byfirms at that time. Thus, any current above normal returns to a firm because it controlsall the resources necessary to implement a strat^y are attributable to this priorimperfectly competitive strategic factor maricet. From our previous discussion ofexpectations in such markets, we can conclude that either the firm in question hadmore accurate expectations about the ultimate value of these strat^c resources whenthey were acquired, or that this firm did not expect these current advantages when theresources were acquired, in vtdiich case its current above normal returns fromimplementing a strat^y are a manifestation of its good luck.

Uniqueness. Others have argued that when only one firm can implement a strat^y,then a strat^c factor market competitive imperfection exists. Such a firm may havea unique history or constellation of other assets, and thus may uniquely be able topursue a strat^y. IBM, for example, has a very large installed bsse of users that allowsit to im{Aement strat^es that cannot be implemented by firms, like Honeywell andBurroi ^hs, without sudh a base (Peters and Watem^n 1982). In such settings.

1236 JAY B. BARNEY

competitive dynamics cannot unfold, and uniquely strat^izing firms could obtainabove normal performance from acquiring strat^c resources and implementingstrategies.

However, as before, a firm's uniqueness is actually a manifestation of the expectationalattributes of a previous strat^c factor market. The key issues become, how did thestrat^izing firm obtain the unique assets that allow it to develop the unique strategyit is implementing, what price did this firm have to pay for these assets, and whatprice must potential strat^izers pay in order to reproduce this set of oiganizationalassets so that they can enter and create a competitive strategic factor market? If thecurrent value of "unique" resources in implementing a strat^y was anticipated at thetime those resources were acquired, then tiiey would have been competitively priced,and any anticipated above normal returns would have been competed away. Thus,any current above normal returns enjoyed by a firm because of its ability to uniquelyimplement a strat^y must either be a reflection of that firm's more accurateexpectations of the value of that resource when it was acquired or, if the firm had nospecial expectations concerning the value of the resource when it was acquired, theseabove normal returns are a manifestation of a firm's good fortune and luck.

Lack of Entry. Another source of an apparent competitive imperfection in astrategic factor market exists when firms that could enter such a market by becomingstrategizers do not do so. This lack of entry, however, like separation and uniqueness,is actually a special case of the expectations firms hold about the future value ofstrat^K. Lack of entry might occur for one of at least three reasons. First, firms that,in principle, could enter, might not because they are not attempting to act in a profitmaximizing manner. Second, potential strategizers may not have sufficient financialstrei^th to enter a strategic factor market and compete for strategic resource. Finally,fiiins that, in principle, could enter, may not know how to, for they may notunderstand the return generating characteristics of the strat^es that current strategizersare implementing. We will consider each of these possibilities in order.

Profit Maximizing. While certain examples of firms in strat^c factor markets notbehaving in (oofit maximizing ways can be cited (Porter 1980, p. 354), overall this isprobably a rare event. Usually, firms do not knowingly abandon profit maximizingbehavior (Hir^eifer 1980), although firms can be mistaken in their expectations aboutthe potential value of a strategy (Roll 1985). These incorrect expectations could leadthem to fail to enter a strat^c factor market when more correct expectations wouldsuggest that entry was appropriate. But this lack of entry is typically due to a firm'simperfect expectations about the true value of a strat^y, not the abandonment ofprofit maximizing behavior (RoU 1985).

Financial Strength. Another ap[»rent strat^c factor maiicet competitive imperfec-tion exists when only a few firms have enough financial baking to enter a strat^cfactor market and attempt to acquire the resources needed to im^dement a productmarket strat^y. Because only a few firms are competing for the relevantr^ources, perfisct competition dynamics are less likely to unfold, and it may beto obtain above nonnal economic returns firom Mwag tiie acquired resources toimplement a strat^y. IBM may, once again, be an examine of a firm with this type offinancial advanta^, for its v:^ financial resources allow it to e n ^ ^ in ^ratc^cbdtaviors not po^Ue for smaller firms. However, even such la i^ difference infinancial strength typically reflect expectational differences in strat^ic &ctor marketsratha* than differences between tiie financM strengtiB of firms, pear se. Two ways inwhich difference in finandal ^rei^th represent tt^se differences in firm exp«:tationsare conadered below.

First, in some drcum^ances, the actiial future valiK of a given strat^jr may be ^same for whatever firm implements i t In this case, if capital mariots are effident aad

STRATEGIC FACTOR MARKETS 1237

well informed concerning the actual future value of a strat^y, then funds will flow tofirms wishing to enter a strat^c factor market with anticipated above normal rettims.Sources of capital will rea^nize the possibility of above normal rettims and willprovide whatever funds are n«%ssary to ensure that potential strate^izers will enterand become actual strate^izers (Copeland and Weston, 1979). The same holds true forcontrollers. In this way, competition within a strat^c factor market will grow, andany anticipated pure profit will approach zero. This entry will only not occur if capitalsources are underinformed about the possibility that firms can obtain above normalreturns from acqtiiring resources to implement a strat^y. In this situation, potentialstrategizers and controllers would not be able to obtain adequate financial backingfrom underinformed sources of capital to enter into the strat^c factor market. Thislack of entry creates the possibility of ptire profits for firms that do enter.

However, when are capital sources likely to be underinformed concerning theanticipated returns fi-om implementing a strate^? If potential strat^izers and controllersare as well informed as acttial strat^izers and controllers, then it seems likely that therelevant information needed to generate return expectations fells into the generalcat^ory of "publically available information," and thus would be taken into consid-eration by capital sources in making funding decisions (Fama 1970; Copeland andWeston 1979). Thus, only when actual strategizers and controllers have expec:tationaladvantages over potential strategizers and controllers is it likely that sources of capitalwill be underinformed. Thus, in this case, the lack of entry into a strat^c factormarket due to insufficient financial backing is, once again, a reflection of theexpectational advant^es enjoyed by some firms in a strat^c factor market.

In an efficient capital market, when the actual future value of strategies does notdepend on which firm implements them, then the inability of firms to attract sufficientfinancial support to enter and compete for strat^c resources must reflect differencesin expecrtations among current and potentially competing firms. However, sometimesa strategy implemented by one firm will have a greater future value than that samestrategy implemented by other firms. In this situation, and under the assumption ofefficient and weD-informed capital markets, capital will flow to high return potentialfirms, while low return potential firms may not rec:eive such financial backing(Copeland and Weston 1979). This lack of financial backing may prevent entry, andthus constitute a competitive imperfection in a strat^c factor market.

However, when can one firm implementing a strat^iy obtain higher returns thanother firms implementing that same strategy? The answer must be that the higherreturn firm already controls other strat^cally relevant assets not controlled by firmswith a lower return potential (Chamberlin 1933; Copeland and Weston 1979). Thus,this firm's ability to attract financial backing is a reflection of its unique portfolio ofstratepcally valuable assets and resotirc^, r^ources not controlled by low returnpotential firms. In this sense, lack of entry is simply a spoaal case of a firmimplementing a unique strat^y, and our previous discussion of expectations in strat^cfactor markets applies here as weU. In short, firms with unique resources that givethem a higher return potential are either exfdoiting special insists they had into thefuture \^ue of those resources when those resources were acquired, or, if they enjoyedno such insights, they are amply enjoying their good fortune.

Lade of Understanding. The final reason entry might not occat is that entrantsmay not understand tide return generating processes underlying a strat^y. Firms formtheir return expectations about specific s t r a t^^ based on their understanding of theeconomic return generating pmoesses undeilying ttese strat^es, i.e., on their under-^^iding of tiie cause and effect relations between organizational actions and economicletums (Iii^man and Rumelt 1982). Some of this understanding may be of tte"teaming by doii%" variety (Williamson 197S), and tiius not available to potential

1238 JAY B. BARNEY

strategizers and controllers. When potential entrants do not understand the relationshipbetween organizational actions and returns as well as current actors in a strategic factormarket, potential entrants are likely to incorrectly estimate the true value of strategies.If they underestimate this value, then these firms will not enter the strategic factormarket, even when expectations set with a more complete understanding of a strategy'sreturn generating processes would suggest that entry was appropriate. Again, this lackof entry, and the competitive imperfection that it might create, reflects the differentexpectations firms have about the return potential of strategies to be implemented.

Developing Insights into Strategic Value

Thus far we have argued that, in perfectly competitive strategic factor markets, thecost of the resources necessary to implement a strategy will approximately equal thediscounted present value of that strat^y once it is implemented. We have also ai^edthat competitive imperfections in this market can give firms opportunities for obtainingabove normal returns when implementing strat^es, but that the existence of theseimperfections depends on diflferent firms having different expectations concerning thefuture value of a strategy. Other apparent competitive imperfections in strategic factormarkets, including lack of separation, uniqueness, and lack of entry, in fact, reflect theexpectational characteristics of either current or previous strategic factor markets.

In imperfectly competitive strategic factor markets, firms can obtain above normalreturns from acquiring the resources ne<»ssary to implement strategies in one or acombination of two ways. First, firms with consistently more accurate expectationsabout the future value of a strategy than other firms can use these insights to avoideconomic losses and obtain economic profits when acquiring resources to implementstrat^es. Second, firms can obtain above normal returns through luck when theyunderestimate the true future value of a strategy. Thus, because luck is, by definition,out of a firm's control, an important question for managers becomes, "How can firmsbecome consistently better informed about the value of strategies they are implementingthan any other firms?" Firms that are successful at doing this can, over time, expectto obtain higher returns from implementing strategies than less well-informed firms,although, as always, firms can be lucky.

There are fundamentally two possible sources of the informational advantagesnecessary to develop consistently more accurate insights into the value of strat^es:the analysis of a firm's competitive environment and the analysis of organizationalskills and capabilities already controlled by a firm (Barney 1985a,b; Porter 1980;Stevenson 1976; Lenz 1980). We briefly consider each of these possibilities below.

Environmental Analysis

Of these two sources of insights into the future value of strat^es, environmentalanalysis seems less likely to systematically generate the expectational advantages neededto obtain expected above normal returns. This is because both the methodolc^es forcollecting this information (Porter 1980; Thompson and Strickland 1979) and theconceptual models for analyzing it (e.g.. Porter 1980; Henderson 1979) are in thepublic domain. It will normally be the case that firms applying approximately thesame publicly available methodoli^ to the analysis of the same environment willcollect about the same information. And these same firms applying publicly avulabkconceptual fhimeworics to analyze this information will typicaUy come to dmilarcondtisions about the potential of strategies. Thus, ^lalyzing a firm's competitiveenvironment cannot, on avraage, be expected to generate the expectational advantagesthat can lead to expected above normal returns in strat^c Victor markete.

Some would su^e t that it is not tte availatnlity erf' tbex environmental methods

STRATEGIC FACTOR MARKETS 1239

of data collection and analysis that is important, but rather the skill with which thesemethods are applied. More skilled firms can thus generate the required expectationaladvant^es through an analysis of the competitive environment. However, the skills ofenvironmental analysis can be "rented" from various investment bankii^ and consultingfirms, and thus skill advant^es in analyzing competitive environments will typic^yonly be temporary.

It may be the case that, in the collection of information concerning the value of astrategy from a firm's competitive environment, a firm might "stumble" onto someinformation that gives it an expectational advantage over other firms. However, if suchinformation was obtained through the systematic application of environmental analysistechniques, then other firms besides the firm that has this information would haveobtained it, and it would no longer give an advantage. Thus, only if the informationwas obtained through nonsystematic means can it give a firm expectational advanti^es.However, such information, because it does not result from the systematic applicationof environmental analysis methodologies, must be stochastic in origin. Any informationaladvantages obtained in this manner must reflect a firm's good fortune and luck, nottheir skill in evaluating the return potential of strat^es.

Organizational Analysis

While firms cannot obtain systematic expectational advantages from an analysis ofthe competitive characteristics of their environment, it may be possible, under certainconditions, to obtain such advant^^ by turning inwardly and analyzing informationabout the assets a firm already controls. Firms will usually enjoy access to this type ofinformation that is not available to other firms. If these assets also have the potentialto be used to implement valuable product market strategic, and if similar assets arenot controUed by large numbers of competing firms, then they can be a source ofcompetitive advantage. Examples of the types of oi^anizational assets that mightgenerate such expectations include special manufacturing know-how (Williamson1975), unique combinations of business experience in a firm (Chamberlin 1933), andthe teamwork of managers in a firm (Alchian and £>emsetz 1972). Firms endowedwith such oiganizational ddlls and abilities can be consistently better informedconcerning the true future value of strat^es they implement than other firms byexploiting these assets when choosing strat^es to implement.

SunuiMiiy and IiiipUcati<His

In summary, firms seeking to obtain above normal returns from implementingproduct market strategies must have consistently more accurate expectations about thefuture value of those strat^es when acquiring the rraources necessary to implementthem, although firms can be lucky. Moreover, while it is usually not possible to obtainthese advantages through the analysis of a firm's competitive environment, firms cansometimes obtain them when choosing to imi^ement strategies that exploit resourcesalready under their control.

These conclusions have important implications for the practice and theory of^rategy. For example, firms tiiat do not look inwardly to exploit resources they alreadycontrol in choosing strategies can only expect to obtain normal returns from their

^ d efforts. For a strat^y of diver^oition throi%h acquiation, this i l ithat firms tiiat fail to discover unique synei*^ between themselves and potenti^acquisitions, but rather rely only on puUicly available information when pridi^ an

can only expect normal retunis from their acquiation strat^es, tfiou£)ifirms mi^^t be lud(y and ^xiuire a firm with an unantidpated synergy. For a

low cost manu&ctudng strat^;y, our aiguments sugg^ that firms without any speaal

1240 JAY B. BARNEY

skills at low a)st manufacturmg can only expect normal returns from imitating thelost cost manufacturii^ strat^es of other firms, while firms with cultural or otheradvantage in low cost manufacturing, if few other firms have these same advantages,can exploit them to obtain above normal returns from implementing a low coststrat^y (Ouchi 1981; Peters and Waterman 1982). Also, firms that currently enjoyabove normal returns may do so because of unique insights and abilities they controlledwhen the strategies generating high current returns were chosen. On the other hand,these firms mi^t also have been lucky. Thus, above normal economic performancemay not always be a sign of strat^izing and managerial excellence (Peters andWaterman 1982).

Our emphasis on competition for the resources needed to implement strat^esdiffers from much current work in the field of strate^. Much of this research is basedon the observation that firms which compete in imperfectly competitive productmaiicets enjoy above normal returns (Porter 1980). As a description of the correlationbetween imperfect product market competition and above normal returns, this researchhas ^gnificant theoretical and empirical suiqx>rt (Hirshleifer 1980). Its implications formanagers are less clear. Simply because firms that compete in imperfectiy competitiveproduct markets enjoy above normal returns does not necrasarily imply that firms thatadopt strat^es to create these product market imperfections will enjoy above normalreturns. As we have sug^sted, this will depend on the competitive characteristics ofthe maiicets through which the resource nec^sary to implement these strat^es areacquired, that is, on the competitive characteristics of strat^c fector markets.'

' This work was made possMe by a grant bom the Office of Naval Researdi. Additional support was{Bovicted by IBM, Westingtouse, the CJenwal Ele<aric Foundaticw, the Alcoa Foundation, the MellonFoundation, and Amp IIM;. Many Ol these kteas woe devdoped in discusaons with Dick Rumelt, RotanWenstey, Kll Ouchi, Barbara Lawimce, Connie Geisick, Bill McKelvey, and the Oiganizational EconomicsSeminar at UCLA.

Bibliography

ALOIIAN, A. A . AND H. C kfSETZ, "Production, Information Costs, and Economic Oiganization," Amer.Ecommic Re*.. 62 (1972), 777-795.

BARNEY, J. B., "Ratiooal Expectations Markets iac Strategy bnidemoitation: Asymmctrk; ^cpectatioos,Luck, and the Theory of Strategy," Unpublished, Grtduate Sdiod of Management, UCLA, 1985a., "Strategizing Processes and Returns to Strategizing," UnpuUidied, Graduate Sdiool of Mai»gement,UCLA, 1985b.

BAzmtMAN, M. AND W. SAMUELSON, "The Winner's Curse: An Eminrical Investigation," Leaure Notes inEcaumia mdMiah. Systems, 213 (1983), 186-^)0.

CHAMBQUJN, E H. . 7*e T^my qf Monopolistic Ccmpetition, Harvard Umversity Press, Camlxid^ 1933.Q»«LAHD, T. AND J. F. WESTON, Financial Theory and Corpwate Policy. A<fclison-Wesley, Reading, Mass.,

1979.ELLERT, J. C , "MMgw, Antitnis* Law Enforcement, and Stoddiolder Returm," / Finance. (May 1976),

715-732.FAMA, E. F., •^WbdetA Oqritd Marias: A Rev^w <rf Theosy and &n|Hrkal Work," / Finance. (May 1970),

383-417.HALPERN, P. J., "Empokad Estimates of the Amount and IXstdbution of Gaim to Conqnnies in Metgers,"

/ Business. (1973), 554-575.HANNAN, M. T. AND J. FRYMAN, "The P( >ulati(Hi Eccdc^y of Organizatrans," Amer. J. Sociology. 82

(1977), 929-964.HEhO^RSON, B. D., Herukrson on Cwpame Straegy. Meii»», l * w Ywt; 1979.HmsmJBBem, J., Pritx Thmsry md Aj^kmioiu. 2BA Ed, ftentfce-IiM, Ea^ewood Oife, NJ., 1980.

Y, J., R. MUJES AM) ASSOOATES (H>.), OrgmutOtmat Ufe Cydes. Jossey-Bass, San F1981.

N, B., R. CRAWPMID AMI A. A . A U S B A N , "Vorticte laK^ation, ^>p««>ris«e Rente, aad fteQMnpctitive Cfflmwaiiig PTOCMB," / . Law md Ecmonucs, 21 ,2 (1978), 297-326.

STRATEGIC FACTOR MARKETS 1241

LENZ, R. T., "Strategic Capability: A ConcqH and Framework for Analysis," Acad. Management Rev., S, 2(1980).

LiPBiiAN, S. AND R. RUMELT, "Uncertain Imitatrility: An Analysis of Interfirm Differences in EfficiencyUnder Competition," BdlJ. Economics. 13 (1982), 418-453.

MANr«LKER, G., "Risk and Return: The Case of Merging Firms," / Financial Economics, (1974), 303-335.

MossiN, J., Theory of Financial Markets, Prentice-Hall, Ei^ewood Oiflfe, N.J., 1973.OUCHI, W. G., Theory Z, AtMison-Westey, Reading, Mass., 1981.PETERS, T. AND R. H . WATERMAN, In Search of Excellence, Harper and Row, New York, 1982.PFEFTER, J. AND G. R. SALANOK, The External Control cf Organizations: A Resource Dependence Perspective,

Haiper and Row, New York, 1978.PORTER, M . E., Competitive Strategy. Techniques for Analyzing Industries and Con^etitors, Free Press, New

York, 1980.ROLL, R., "The Hubris Hypothesis," Unpunished, Graduate Sdbool of Management, UCLA, 1985.RuMELT, R. AND J. R. C. WENSLEY, "Market Share and the Rate of Return: Testing the Stodiastk

Hypothesis," Unpublished, Graduate School of Management, UCLA, 1981.SCHALL, L. D., "Asset Valuation, Firm Investment, and Firm Dlverafication," / Business, (1972), 11-28.STEVENSON, H., "Defining Corporate Strengttis and Weaknesses," Sloan Management Rev.. 17, 3 (1976).THOMPSON, A. A. AND A. J. STRICKLAND, Strategy Formulation and Implementation. Business PuUications,

Dallas, 1980.WauAMSON, O. E., Markets and Hierarchies: Analysis and Antitrust Implications, Free Press, New York,

1975.