Recent analyses of the firms as repositories of knowledge (e
Where Do Markets Come From?
Dr Michael Jacobides
Centre for the Network Economy
CNE WP06/2002
Where do Markets Come From?
Michael G. Jacobides
Assistant Professor of Strategic and International Management
London Business School
Sussex Place, Regent’s Park
London NW1 4SA, United Kingdom
Tel (011) 44 20 7706 6725; Fax (011) 44 20 7724 7875
mjacobides@london.edu
Where do Markets Come From?
Abstract
Where do markets come from? Taking a historical perspective, this question seems much more natural to ask, than the “traditional” question of when firms displace markets – as, in the real world, markets often just “come second”. Without analytical prejudice, this paper explores the evolution of the mortgage banking value chain over the last twenty years, and explores what drove the high degree of vertical specialization and intermediate market creation. From our evidence, it appears that several factors, largely unrelated to (and occasionally inconsistent with) the Williamsonian transaction costs – such as gradual learning, standardization of communication patterns, and information conditions – lead to the creation of intermediate markets and the vertical fragmentation of the value chain. Building upon this evidence, the paper explores what makes markets less “appropriate” than firms, and also what enables markets to appear and occasionally displace integrated organization of production. In so doing, we provide a fresh take on an old problem – namely, the relative merits of firm and markets. Firms and hierarchical structures, we argue, are not just or not primarily noteworthy because of their transactional guarantees. Rather, they also come as a response to the need to organize rich information in a manageable manner, and find effective means of coordination, even in the absence of incentive or transaction problems. The paper develops and discusses these coordinative and informational advantages of hierarchy, synthesizing and expanding upon existing research. It also elaborates on the determinants of value chain structure and its evolution, suggesting a rationale for the observed pattern of dis-integration in many industries, and also positing that this is a part of a larger long cycle of integration => dis-integration => re-integration driven by information standardization and new technology development. Finally, the paper looks at how recent Information Technology evolutions are changing the relative merits of firms vs. markets, and provides suggestions for further research.
Keywords: Market Genesis; Vertical Scope; Transaction Costs; Information & Coordination; Information Technology; Theory of the Firm
One of the most important advances over the last twenty years in institutional economics has been the work on the theory of the firm – work that has profoundly influenced management studies as well, having created sworn defenders and fervent foes. Almost uniformly, this work “is informed by the perspective that ‘in the beginning, there were markets’” (Williamson, 1985). The central question in that line of research is, in principle, to explain where firms come from, and, in practice, explain decisions to integrate vertically, abandoning market procurement. The canonical answer provided by transaction cost economists has been that firms (and hierarchical governance in general) come as a response for the need to provide a safe haven from the market’s potential transactional hold-ups and risks from opportunistic behavior (Williamson, 2000; Hart, 1995). The critics of Williamson –mainly Conner and Prahalad (1996), Kogut and Zander (1996) and Ghoshal and Moran (1996), counter the argument about the sources of the organizational advantage – i.e., provide an alternative explanation of why firms exist. However, neither institutional economists nor their critics study where markets come from.
The question then becomes, why not do so? As Simon (1991) observed ten years ago, markets are not the most prominent form of organization. Economic activities are carried out in organizations much more than they are in firms. And the very creation of an intermediate market is a noteworthy event. Taking a substantively different approach from Harrison White’s namesake paper (1981), which looked at the evolution of horizontal market positioning,[1] we examine “where (intermediate) markets come from”, what enables their genesis, and supports their function. In so doing, we uncover some new dimensions in how markets differ from firms.
1. Studying Market Genesis: Mortgage Banking in the USA
Asking a novel question such as “where markets come from” requires a grounded empirical setting, lest it devolve in misguided theorizing. We thus chose a setting where new intermediate markets have emerged over the last few years. Mortgage Banking in the USA has been a prime example of vertical dis-integration and intermediate market creation. Integrated firms (Savings & Loans associations) that used to produce, hold and service loans have given way to a specialized set of mortgage bankers, that originate and service loans yet sell them to securitizers or GSE’s such as Fannie Mae and Freddie Mac (see Figure 1). The process of gradual market creation and dis-integration happened on several different levels within the mortgage banking industry. The first instance was the dissociation between producing a loan, holding that loan as an asset and servicing it. The securitization trend in the 1970’s onwards allowed for the claim to capital to be sold, and what made this possible was the standardization of information.[2]
----------------------------------------------
Insert Figures 1 and 2 around here
----------------------------------------------
So after the 1970’s, a new, vertically specialized eco-system, with mortgage bankers, secondary markets for loans, and securitizers, competed with integrated loan providers. What’s more interesting, is that within the vertically specialized segment of mortgage bankers, further market creation has taken place. In the 1970’s and up to the early 1980’s, mortgage banks were integrated; they would do everything themselves, and only occasionally sub-contract some activities. But gradually, new markets appeared: Market for brokered loans; market for servicing rights; and so on. So these new markets enabled (and were driven by) vertical specialization, and new species such as mortgage brokers, loan wholesalers, loan warehousers, specialized and sub-servicers (all interconnected through new intermediate markets) appeared.
Mortgage origination in particular – the process of finding borrowers, selecting and selling a mortgage, warehousing it, and preparing it for the securitization process on the one hand, and servicing the loans, on the other – has seen significant change in its structure, including substantial dis-integration and re-distribution of profits. Mortgage brokers, for instance, who hardly existed before 1980, reportedly increased their origination volumes to as much as 50% of total originations in only a few years. Some mortgage bankers shed their origination branch network, and have focused on warehousing loans (providing the liquidity from the time a loan is closed to the time it is sold to the secondary market). Markets for closed loans; for freshly funded loans; for warehousing loans; etc., emerged linking the steps of the production process, displacing the only pre-existing solution – i.e., hierarchies.
Furthermore, the link between origination and servicing was severed from the late 1980’s onwards. Mortgage banks, conventional logic held, originated loans (usually at a net loss), to make money from the valuable servicing of a loan. So you had to originate, and then service. But this stopped being so simple in the 1990’s onwards. As some mortgage banks focused on servicing, a new intermediate market for servicing rights appeared, fuelled by and prompting further vertical dis-integration.
All in all, a myriad of specialized markets now mediate activities that used to be internalized within firms’ boundaries. Division of labor was increasingly made between, rather than within firms, and each of the boxes in Figure 1 (or steps in Figure 2) can now be performed by a single type of firm. So is this increasing “marketization” yet another instance for Transaction Cost Economics to show its analytical might? Alas, prima facie evidence does not seem to support this. In our extensive empirical coverage of the industry we could find little evidence that transaction costs explain this market creation and ensuing dis-integration. Then what did enable the creation of intermediate markets and the breakup of the production process?
To answer this question, we will look at three vignettes from an extensive study of the mortgage banking industry, underwritten by the Mortgage Bankers Association of America and the Wharton Financial Institutions Center. This large-scale project was based on archival and field research that included reading several hundred industry articles and manuals, interviewing over 150 executives, presenting the findings in a number of industry conferences (where the analysis has been very well received), and studying the industry for well over two years, with several thousand hours logged between October of 1998 and today (see Jacobides, 2000 for an overview; Jacobides and Hitt, 2001, for the quantitative research part of this project, not reviewed here; and Jacobides, 2001, for an industry-friendly overview of the historical findings).
1.1. The Development of the Mortgage Broker Segment
Mortgage brokers – the most upstream part of the mortgage banking process – are a relatively new segment. It started appearing in the early 1980’s, and by 1999 accounted for about 46% of total originations. Mortgage brokers are the individuals who contact the customers; take their application; prepare the loan; and either sell the qualified lead (and the concomitant organized and duly packaged paperwork) or go all the way to pre-closing the loan, that is obtaining the approval from a bank that collaborates with them (LaMalfa, 1998). These activities used to be performed by mortgage banks, who were, however, all too happy to find the additional option of “buying” rather than making them themselves. The question, however, is how did this segment appear, and what allowed for the creation of its supporting intermediate market. An article from the Mortgage Banking magazine on the development of this segment, provides so detailed a description of the very processes, that warrants replicating “as is”:
“In 1981 the first rumblings of change began, the start of a new paradigm, the unraveling and unbundling of the housing finance delivery system. First, an inverted yield curve and historically high bond rates caused money to flow out of the nation's thrifts and into money market funds. With no money with which to make mortgages, the S&L’s started round after massive round of layoffs in their loan origination staffs. Even though many, if not most, originators had compensation schedules geared toward commissions, the nation's thrift industry could not afford even modest salaries.
But as loan officers cleaned out their desks and headed for the door, they were often told something like the following: "Listen, Bob, you know we can't afford to keep you, you know that there's almost no business and that we have almost no money to lend. But we've always thought you put together high quality loans and we'd like to stay in touch. You have good contacts with Realtors, so if you can go out on your own and put together some loans, bring them to us and we'll fund them. And we'll figure out some way to split the loan fee." […]
Now things were different. A new kind of mortgage borrower who had access to institutional money appeared. The fellow who had been working at First Federal Savings was now an independent contractor working on his own – but he could place a borrowers loan with First Federal Savings. Initially, those originators who went out on their own would only broker to the institution they had previously worked for. But rates plummeted in August of 1982, and institutional lenders were flush with cash and needed mortgage product.
Our broker who had been placing loans with First Federal Savings was now brokering to other institutions. In need of loans, Second Federal Savings approached him with a conversation that went something like this: "We know you worked for and placed loans with First Federal Savings. But if you bring us some loans, maybe we can offer better programs, and perhaps we can offer you a better split on the fees."
The seeds of a new world had been planted. The mortgage broker became even more useful with his increased number of sources; and then origination fees started to shrink for institutional lenders. Before long, our broker was approved to place loans with a great many lenders. Calling themselves wholesalers, lending institutions set up whole departments not to originate loans at the borrower level, but to take them in from originating brokers. The origination process had been eliminated. All that was left were the underwriting, funding and servicing functions”. (Garrett, 1989: 30-32)
This historical account suggests that once the learning process allows for people to identify a simple way of trading something, once a standard is set, what used to be part of an integrated procedure within a firm, organized by authority / firm-based governance can now be organized as a market-mediated process. This need not be the result of conscious planning; it can be brought about by happenstance, as it did in our case, whereby a cyclical downturn and necessity made the actors think about setting up the standardized infrastructure for market exchange.
This setting also allows us to do away with another potential explanation for the changing role of firms and markets in an industry. In his seminal article, Stigler (1951) suggested that the degree of integration in an industry (and hence the very existence of intermediate markets) is determined by the size of the market. So, the theory would have it, markets are born when the overall industry grows large enough to sustain specialization. But that is simply not the case here. The mortgage banking industry has been one of the largest segments in the financial sector, amounting to a significant proportion of the GDP, and it has not grown to any significant extent.[3] So scale cannot be the culprit. Rather, learning seems to matter: with the progression of time and cumulative volume, information tends to get standardized, production and processes get modularized (initially for reasons of operational efficiency) and then intermediate markets start appearing. So perhaps even the presumed relationship between scale and specialization might be spurious, and really caused by cumulative volume and learning.
But there was more than just learning at play, as it emerged from the industry interviews. The availability of cheap and efficient communication technology (first, the possibility of mass-faxing interest rate sheets every morning to the broker network, and second, the increasing capabilities to transmit loan or rate information in real time) greatly facilitated the brokers and allowed them to integrate with bankers. It appears that this new technology enabled brokers to use the market to coordinate more effectively. As Garrett (1993: 20) observes,
“Our business has changed and new systems of information distribution have diminished [the marketing reps] value. It wasn’t that many years ago that rate sheets were mailed and actually delivered by hand. The facsimile machine and other technologies have made this essentially obsolete. And various overnight delivery services make it unnecessary for a field rep to physically pick up loans at the broker’s office.”
An even more important change has been occurring over the last few years, with the (inadvertent?) support of major mortgage banking technology providers like FiServ, or AllTell. Most of the current integrated technology solutions, including far-reaching ERP’s, or more limited MIS structures, now offer the option of integrating the broker channel almost as tightly as the retail production network itself. As an executive put it,
“Nowadays you can buy the product as easily from any source… it used to be easier to rely to your in-house people… but nowadays the systems make managing the brokers as easy, so it comes down only to price and cost.”
1.2. Making Sense of the Development of the Mortgage Broker Segment
What seem to be the principles behind the creation of the markets? The evidence points to learning, information, and modes of coordination. As individuals learn, and as information becomes standardized, markets become less disadvantaged in handling information. With the appearance of market exchange comes more practice, and as market “rules” are established, market “capabilities” gradually increase. To put it in Adam Smith’s words,
“this division of labour, from which so many advantages are derived, is not originally the effect of any human wisdom, which foresees and intends that general opulence to which it gives occasion. It is the necessary, though very slow and gradual, consequence of a certain propensity in human nature which has no such extensive utility; the propensity to trade, barter, and exchange one thing for another.” (1776: 9, emphasis added).
The process consists of ironing out ways to measure and identify what is done, in order to then define dimensions of exchange (Barzel, 1982). This creates the conditions that make the market “be feasible” to begin with. What the market needs, in order to work, is a standardized mode of information provision, and the appropriate infrastructure (North, 1986); and a social process of learning the rules of market exchange and the concomitant social institutionalization process (DiMaggio and Powell, 1993).[4]
In our example, the increasing role of credit scoring and standardized underwriting rules also helped the mortgage broker segment to come of its own. Inasmuch as the information that needed to be conveyed from brokers to bankers was standardized, and the dimension along which a loan was evaluated were easily assessed and articulated as well as measured ex ante, it was technically feasible to have a market that would tackle the requisite information. Put otherwise, earlier on it was not clear how you could “trade” a loan. You could not efficiently make a loan through a broker, because it was hard to communicate the risk profile you wanted; you could do that more efficiently in-house, and market-based relationships were hard to think of. You could not even communicate the types of customers, nor could you specify what you wanted. However, with the rise of standardized instruments such as (a) standardized loans, on the product side, and (b) standardized information measures on the process side, such as Fair Isaac’s Fico score (a score that measured the credit-worthiness of a loan applicant), things were different. There now was a way around the problem. The complex information was not necessary; it was standardized. The tight coordination of the different parts of the value chain was not needed, and information to be transferred was not so “thick”; thus a market could work.
To re-iterate, a customer / loan application could be reduced to a series of unambiguous scores; and such “score-based” packages could be traded through the market almost as easily as organized within firm boundaries. The dimensions of the loan production side of the business were fixed and standardized, and this greatly facilitated the creation of the broker segment and the new intermediate market for closed loans. This learning process, pushed by historical accident, also implies a life-cycle of the balance between firms and markets in a value chain. History does not always support the economic conjecture that “at the beginning, there were markets”. More often than not, in the beginning, there were economic institutions other than the market: in many sectors, in the beginning, there were (integrated) firms. As the processes become better understood, firms begin standardizing their own production process, often for their own operational benefit. Inasmuch as a lingua franca can be instituted, inasmuch as work becomes divided in independent, separable sub-components, the market starts to work; timidly at first, and more aggressively thereafter.
1.3. The Evolution of the Market for Servicing Rights
For our second example of dis-integration, let’s consider the increasing separation between servicing and origination that has intensified over the last few years. Following the 1980’s crisis in the Savings & Loans (S&L’s), the government created the Resolution Trust Corporation (RTC) that took over loans of insolvent S&L’s. The RTC had to do two things. First, find a way to sub-contract its loans. Second, ultimately sell all these servicing rights, as the government was not in the business of running a mortgage servicing operation. The RTC’s interest in getting rid of the loans it had on its books, led it both to sell whole loans, but also pushed for the development of a market for selling servicing rights. As Garrett (1989: 34) reports,
“Servicing was rarely sold until recent years… because a relatively small amount was bought and sold, valuation was not a critical issue… Servicing was often sold only under special circumstances. When a company was sold or liquidated, that might be one time they sold their servicing portfolio. In 1981, San Francisco- based Baldwin & Howell was closed up by its owners. In its nearly 100 year history, it had never sold servicing.”
Four years later (1993: 19-20), in a follow-up article, he complements his description:
“When the Office of Thrift Supervision seized a Thrift and turned it over to the RTC, there was typically a cessation of all lending. Mortgage origination branches will die quickly if ordered to stop lending. Producers will leave overnight and soon the branch is dead. In either case, the RTC found that it could not sell the whole thrift; instead, it would have to sell off bits and pieces. Bidders for the remnants of the failed thrifts asked why should they create servicing when it was cheaper to buy it?”
This separated the servicing from the underlying loan, or contracted out the servicing to a third party (which did not assume the residual risks of a servicer.) While some individual sales of servicing did take place before that time, they never did so to any significant extent. The RTC helped mortgage banks develop the infrastructure (and the habit) of trading servicing through the open market by standardizing information, and helped the increasing specialization in servicing as well as the appearance of servicing-mostly mortgage banks.
The next stage in the development of the market for servicing rights came with the institution of MERS – the electronic depository for mortgage servicing rights, which was initiated in 1995. The motivation was that any servicing right transfer involved the cumbersome transfer of physical files, thus creating significant costs and discouraging a “real” market for (Mortgage Servicing Rights (MSR) to develop. Therefore, firms which were interested either in buying or selling MSR, decided that they should develop some type of a central depository for mortgage servicing information (and, ultimately, some central depository facility) about servicing, which would also allow for more complex sub-servicing agreements. Thus, after a couple of false starts, a number of important players got together and backed MERS, to act as a non-for-profit entity which would eventually have information for as many mortgages as possible, thus facilitating the transfer of information as well as the pure frictions of transferring the servicing rights & titles themselves. This innovation, backed both by IT evolutions and by the organization of a “transaction support” scheme, further facilitated using the MSR market.
1.4. Making Sense of the Evolution of the Market for Servicing Rights
In the dis-integration of this part of the value chain, and the appearance of a new market, standardization of information & coordination were, once again, important. Here, a new element – the role of critical mass – was also added to the list. Critical mass appeared to be important for two reasons. First, because it allowed for different parties to converge to the same “descriptions”. As Schelling (1978) observed, the issue is not whether we all agree that orange light should be the right color for cars slowing down or starting their engines; what matters is that we all understand that orange has one meaning, so that we do not need to re-define the rules of communication in real time. The S&L crisis, and RTC in particular created a “template”, a lingua franca people could agree on. Second, because some markets require a certain level of depth to work; they exhibit network externalities / increasing returns in the demand side. That is, with the increase of market participants, the market may become attractive enough; and there may be a “tipping point” after which the market becomes viable (Varian and Shapiro, 2000).
The second lesson from this vignette, as well as from the previous one, is that many a time the development of a market is driven by the desire to take advantage of one’s capabilities. For instance, firms which might make money on warehousing loans, rather than the origination and brokerage of the loans, would push for the development of an intermediate market for brokered loans, in order to outsource that part of the value chain. Conversely, if firms do not see significant underlying “gains from trade” from vertical specialization, specialization will not occur; hence the balance of the capability distribution in different parts of the value chain becomes a major driver of an industry’s vertical structure. (see Jacobides, 2000 for a formal proof, and Jacobides and Hitt, 2000, for quantitative evidence). In the case of the market for MSR’s, the differentials of good vs. mediocre servicers, which, according to industry analysts, have increased over the last fifteen years, have led to the need for an MSR market. Smith was right; the propensity to trade and barter, in order to take advantage of one’s capabilities (also see Demsetz, 1988; Langlois and Foss, 1998) is a major driver of vertical specialization.
Relatedly, vertical specialization and the growth of intermediate markets may also be the result of different parties trying to capitalize on their capabilities; that is, the extent of specialization may be a result of differential limits to growth. For instance, the increasing specialization in mortgage brokerage is due to the fact that the banks which are strong in warehousing cannot grow their origination (in-house brokerage) segment as quickly and efficiently as they can grow their brokerage functions, and hence start requiring increasingly more brokered business across their firm boundaries. Similarly, as the efficient servicers managed to scale up their operations quickly, they soon sought to create and support a market for MSR. [5]
There’s one more common thread between this and the previous vignette. In addition to the gradual, evolutionary learning process and the standardization of information, historical accident also plays a significant role in the institution of markets.[6] For the development of both the mortgage broker market / segment, and the market for servicing rights, some serendipitous turn of events ignited and strengthened intermediate markets. In the case of the mortgage brokers, the recession of the early nineties, and, for MSR, the collapse of S&L’s led to the establishment of rules that would eventually support a market exchange.
1.5 How Changes in Organization lead to Markets: Loan Commitments and Separability
One of the important changes in business practices in mortgage banks in the 1960’s has been the institution of more flexible commitments for mortgage bankers, on the behalf of the GSE’s (which, as Figure 1 indicates, buy loans from mortgage banks). In the “before” state, mortgage bankers had to produce only the types of loans which could be sold, in real time, to the investors / GSE’s they worked with (as they were not capitalized well enough to keep any loans which they could produce but not sell to the secondary market or to any other investors). In the “after” state, significant flexibility was added as a result of the institution of several new types of arrangements. In optional delivery schemes, mortgage bankers would commit to an option of delivering loans, which they were not forced to fulfill. Additional mechanisms were the “forward commitments”, whereby an agreement was made for a particular date, as well as the auctions for loans produced, that could allow mortgage bankers to sell excess loans at an attractive price (see Lederman, 1995.)
These institutional / business innovations reduced the coupling between retail production and secondary marketing / warehousing (see Figure 1); they made the coordination task easier. If a mortgage banker produced more loans than anticipated, she did not have to fear that they would be a burden, because of their inability to sell them to secondary investors; hence she did not have to worry at the time of production as for the sale to the secondary market. This dissociation between the two linked steps of the value chain (production of loan and sale to the secondary market), i.e. this reduction in coupling allowed mortgage bankers to focus on each of the two steps separately, managing origination without having to worry about the ability of the investors to absorb their production, and managing secondary marketing separately from origination. The availability of larger warehouse lines of credit from commercial banks had similar effects, as it enabled mortgage bankers to buffer direct origination from secondary marketing (i.e. selling the loan to the secondary market). To put it in Thompson’s (1967) terms, this reduced the interdependence, and made it sequential rather than pooled.
The implication of these changes was that gradually, origination (retail) and secondary marketing would stop being as unitary as they once were. Externalities from one side stopped being exerted to the other, thanks to the new buffers and reduced coupling, so hierarchical governance to manage this process was not as crucial. The task had become coordinatively simpler. The existence of these new business arrangements meant that the interactions between these two steps of the production process could be reduced. This led not only to a marked increase of the role of mortgage bankers (based on the operational efficiencies of modularization) but also, gradually, to the development of independent business models for origination and secondary marketing, and ultimately to the dissociation of brokerage / retail production from warehousing / secondary marketing, evidenced from the growth of the mortgage broker / pre-closed loan market in the 1980’s. So this change in coupling, reduced the merit of integrated / firm based governance between the production and servicing of loans, and enabled the development of the specialized brokers – warehouser model, with two entities linked through the market.
1.6 The Role of Coupling: Generalizing from Loan Commitment Methods and Separability
This vignette points to another important factor that indirectly shapes the nature of the value chain: technology. More specifically, the types of interdependencies between different stages of production (i.e. the underlying technology) determines the separability of these stages. In turn, separability enables the creation of a new market.
In several parts of the mortgage value chain, a tendency towards a modularization of production has led to the reduction in the extent of coupling between different stages. The pattern that emerges is the following: Increasing autonomy and standardized communication and coordination initially allows firms to separate their divisions clearly, often creating separate divisions. As a next step, firms place some parts of the production process squarely out of their own boundaries, creating long term contracts that further give rise to the market. That shows that the reduction in requisite coordination is an important precondition for the market to emerge, hinting at coordination weaknesses on the market side. It also shows that technology of production, and in particular modularization, significantly affects the degree of integration in a value chain.[7]
1.7. Making Sense of the Progression of Markets.
So what can we learn from this process? The answer is that we can see patterns recur in the creation of the new markets. Culprits involve information standardization, coordination between different parts of the value chain, underlying gains from trade from unequally distributed capabilities along the value chain (or differential growth rates in the value chain) and, in order for the market to take hold, learning that is needed to support mutually complementary market-mediated roles. Additionally, historical accident and information technology seem to provide a significant boost to the generation of these new ways of organizing, as we can see in Figure 3.
------------------------------------
Insert Figure 3 around here
------------------------------------
2. Tracing the origins of Markets: What can we learn about the relative merits of Firms / Hierarchies and Markets that we did not know already?
So far, we have looked at the drivers behind the creation of new markets, and the resulting value chain structure. What clearly emerged is that the “usual suspects”, i.e. the Williamsonian or broader transaction costs (TC), were nowhere near the importance of other factors; and often, the stories we heard were inconsistent with the TC perspective, as new markets would arise despite significant transactional difficulties. This may be because TCE has been designed to answer the question of whether market- based production should be preferred to in-house – presuming the make-vs-buy question can be asked. We could logically extend the TCE stories to explain why markets don’t exist, and why they appear; yet the evidence does not point to hold-ups or transactional difficulties. This fact, in and of itself, should be sufficient ground to look for new explanations. Looking for TC, even when the prima facie evidence points to different issues, would resemble the drunkard’s search for his keys under a lamppost, where “there is light”. This is a precarious practice, because if we look long enough, we might find some keys; but, most probably, not the right ones (Popper, 1974).
Instead, we should make an unbiased analysis of the factors that matter. Thus, the analysis of what allows for markets to be born, or, more broadly, the examination of a value chain / activity system as it evolves would be the natural thing to do. For a market to be born, some conditions must change. Hence we can see (as we did in the previous section) what factors made the market unattractive (or inconceivable) before, but plausible now.[8] That being said, given the prominence of TCE, it is still impingent upon this paper to show that TC considerations have not been driving these evolutions from the background.
2.1. What has been the Role of TC in the Genesis of Markets?
Let’s start with the traditional definition of Williamsonian TC – which focus on the possibility of post hoc opportunism. This, in our setting, does not seem to be particularly important, as no real asset-specificity is created, nor did such considerations seem to affect industry participants. We asked repeatedly about hold-ups and co-specialization, and little could be found to support such a role. Not only was it cognitively out of the manager’s agenda, but there did not seem to be any way in which we could convincingly apply the concept to this business. There may be some co-specialization in terms of the relations between brokers and bankers; but this has not stopped the market between them to take roots, nor does it seem to worry the executives. And, if anything, this co-specialization increased in the early to mid-nineties; so the market arose against the prediction of the standard TCE theory.
We then shifted our focus from the traditional TCE to some broader problems of “lemons”, i.e. unverifiable quality of loans, such as discussed in Akerlof (1970). That might have been a significant problem in the broker-banker relation. The potential issue is that information on loan applications could be mis-represented. Yet we were told that brokers were only marginally more likely to mis-represent loan information than the in-house agents. As an industry expert noted,
“I don’t think this is a major issue. Don’t get me wrong- you do get these lemons you talked about everywhere. But I don’t think you would chose how to procure your loans on the basis of default risk or fraud. You could have bad apples in your own retail branches- loan officers who make a killing and then go- as you could have greedy brokers or untrustworthy correspondents. I guess that if you trust fly-by-night brokers you can get in trouble, but you can monitor these things.”
So while in principle a TC (or, rather, a “lemons”) story could be constructed, in practice it does not hold – not because there are no risks in market transacting, but rather because the firm is not as safe a haven as presumed (as also suggested by Eccles and White, 1986).[9]
Finally, we did have some references to market frictions, broadly construed. In particular, executives referred to Information Technology, as it affected the costs of doing business through the market. As a senior consultant to the industry remarked,
“There is no doubt that the new technologies –EDI, integrated technology solutions, and especially the internet- are allowing firms to use their brokers and correspondents much more efficiently. There is less and less of a difference in how the loan is produced. It really has come down to a difference of cost to produce and price to purchase.”
But this is not an issue of transaction cost; it is an issue of information conditions, and the ability of markets to coordinate, as we will see in detail in the last section of this paper. Further (and strong) evidence that the “usual” Transaction Costs are not important in explaining this process of dis-integration was found by an analysis of a large-scale database of 180 mortgage banks over a ten-year period covering the vertical dis-integration phase (Jacobides and Hitt, 2000). So intermediate market creation in this industry has been largely unaffected by TC; and looking at the TC variables does not help us explain the growth of vertical specialization through market creation. Thus we can use this setting to explore the non-TC issues.
On the practical / empirical level, then, we can say with some confidence that there have been some “non-TCE” factors that have affected the creation of markets. Underlying differences in capabilities or differential up- vs. down-stream growth rates, as well as history and learning, enable markets to be created (see Jacobides, 2000). Yet there are also some real differences between markets and firms in their ability to coordinate; and once the information and coordination conditions change, then markets can appear. Still, could it be that these information & coordination differences are nothing but TC issues in disguise?
2.2. Sailing, Aligned Incentives and the Advantages of Firms over Markets
Our objective, then, is to establish that the “information- and coordination-” based advantages of firm / hierarchy based governance are really a separate category, worthy of further study, rather than an epi-phenomenon. One of the most potent defenses of TCE has been that most of the knowledge- (or information-) based arguments on the merits of the firm are, eventually, transactional problems; that is, that the real upstream driver of the ability of firms to handle information and ensure coordination is the inability to contract upon such dimensions in the market. On a “pure theory” level, what can we do to convincingly counter this critique, so that we can constructively propose new theory rather than engage in a bitter polemic? Quite simply, do away with the fundamental driver of the “traditional” TC, by creating an example where there are no incentives to renege, and where by construction TC cannot exist.[10] In so doing, we will complement our empirical findings with grounded theory development.
Before going into the analytics, let’s consider a sailing boat, where the actors have perfectly aligned incentives, and no interest in creating transactional problems ex post facto, contractual haggling or hold-ups. No transactional problems should exist. According to received theory, a "market" for the crew’s services could emerge. Yet anyone having sailing experience knows that absence of real authority and an “open”, “market-like” structure wherein anyone having timely information acts independently upon it, would lead to disaster. The reason is that actors cannot efficiently communicate if they set up a market interface, nor can they understand the externalities of all of their actions and hence coordinate effectively. The only way to get the sailing boat going, even with no transaction problems, is hierarchical governance. It’s not the Williamsonian TC that make a potential market inefficient; it is a different and ill-analyzed advantages of hierarchy.
This provides a new twist in comparing firms and markets. The issue here is whether a market is able, in real time, and not just in theory, to provide the template for efficient coordination of economic actors' efforts. Contrarily to Hayek's (1945) famous conjectures about market's informational superiority, it appears that often the market is not able to provide appropriate guides for action. And not for any of the reasons that TCE looks at. Rather, this failure is due to two factors. First, because the market's uni-dimensional signal (a price for a given quantity under given specs) is of lower dimensionality than the effort vector. That is, if "trimming the sails" of a sailboat is a multidimensional effort (you can take them in or leave them out, increase or decrease their twist) and the signal is one hypothetical price (how much you trim them) then no solution occurs. Prices are not able to capture the richness of information needed to yield efficient performance, and this leads to the need to coordinate via mandate, wherein information is richer and instructions more precise. In organizing by mandate, the skipper can constantly ask for specific actions, such as trimming tightly but having the twist lower down, etc.
To visualize this, consider a sailing boat with two actors- a skipper and a trimmer, who can be linked either through hierarchy (whereby the skipper dictates the action, and the trimmer undertakes it while minimizing her cost), or through a “market”, whereby (to be generous to the market feasibility) the skipper has an unlimited amount of “motivational units” to make the trimmer provide the optimum trim. These two different types of organization mimic, respectively, the firm – where compliance happens by authority, and the market – whereby compliance happens independently, through the price system. Consider, furthermore, that the speed of the boat is affected by two trimming-related features: the extent of twist on the sail; and the extent to which the sail is taken in or left out. For any given wind and course conditions, that is, for any given sailing strategy, there is an optimum combination of the sail trim and the sail tension, as shown in Figure 4a.
Let’s see how the boat will fare under the two alternative regimes, starting with authority. For the skipper, who knows the sailing strategy, it is simple to dictate the extent of trim and the amount of tension that will bring the boat to the optimal speed level for its course; and the trimmer will try to position as closely as possible to it, while economizing on her effort. (Strictly speaking, she will do this at the place where her lowest iso-effort surface meets that ascribed point.) Now let’s think about the market. Can it allow our sailing team to reach that optimum trim / twist combinations? It appears not. The reason is that the effort vector is two-dimensional: There is both twist, and tension. For all the “motivational points” the skipper might expend, the trimmer will not get the right combination, except for being lucky. Notice, however, that if only the tension mattered, the market would have been able to yield a solution, as Figure 4b demonstrates. That is, the firm is better when the effort vector has a higher dimensionality than the “price” vector – which is the case, most of the times!
----------------------------------------------
Insert Figures 4a and 4b around here
----------------------------------------------
The second reason for which the market cannot yield a solution is that economic agents are boundedly rational; they cannot process infinite information about the potential uses for their effort. In the sailing-boat context (and let’s now shift to a 10-strong crew off-shore), an excellent cook may spend his time cleaning the aft cabin, because he does not know that the rest of the crew are desperately hungry. If he is not able to monitor all the potential uses for his output (effort), then he will not be putting his effort to the most productive use. However, if there was some central administration (a skipper) who knew what the needs were, and spent his time doing just that, he could then help by assigning resources to their more value-adding uses. Such monitoring and assessment of what the needs are and how they can be met by limited resources is, purportedly, what managers (in any non-market situation) do, and why they add value.
To return to our previous visualization, consider now a situation where the trimmer has two possible outputs, and is facing two different ways in which effort can be allocated (trimming the main sheet and trimming the jib). Where should she go first? Even if the “price mechanism” as portrayed in 3b could exist, the trimmer might find it difficult to understand what the best allocation of her effort would be, especially if the two tasks are mutually exclusive. Even if prices are attached to each task, and a solution with “motivational units” could, in principle, be found, it would take time (and confusion) for the trimmer to select both the sequencing of the activities and the level of effort in each of them. The problem is that it takes time for the market mechanism to work, and that there is a cost in “market convergence” (cf. Weitzman, 1974). Limits in human cognition make hierarchical organization needed (Simon, 1947, 1991; March and Simon, 1958). A mythical Mensa Master, who can optimize perfectly and effortlessly, might be able to work out the implications of multiple prices and take the appropriate actions. Yet when the waves are hitting the sloop’s mask, and the boundedly rational trimmer does not know what is the best action to take, hierarchy makes more sense. In the same way, for an investment bank, it may be easier to “own” the account officers, as this will allow it to change the mix of clients, profiles, and types of deals much easier than what it would if it used the market.[11] If an individual actor cannot assess the best uses of his effort in a productive relationship, mandate serves as a valuable means to guide effort. It’s not an issue of identity, as Kogut and Zander (1992) suggest; it’s an outgrowth of bounded rationality (Simon, 1951; 1991).
Now let’s return to our question of markets and their appearance. “A market for the services of a trimmer? Rubbish!”, I hear you say. Rubbish indeed – but why? Because in sailing boats, as it is the case in a great part of the economic activity going on in the real world, we cannot find the simple conditions where a market may operate. The informational and coordination requirements are not met. If and once they are met – through the standardization of information, changing coordination structure, and given capability hererogeneity up- and down-stream, markets can emerge and specialization can happen. In the sailing boat example, this seems to be a low-probability event[12]; but not so in many other settings.
So let’s recap the basic theoretical claims we have made so far: First, even with full congruence on the behalf of economic actors, markets can fail. Markets fail, in these instances, not because there are TCE- or “lemon-” related problems, but rather because no “solution” can be found through the markets whenever the dimensionality in the output vector is higher than that of the price-setting vector, and whenever bounded rationality inhibits market signals from being sufficient guides to the allocation and direction of effort. But once the information becomes standardized, and if the coordination task becomes simple enough, then a market can be established, as we saw in the evolution of intermediate mortgage markets.
Now let’s push theory-building further, with yet another objective: Tie these new insights into existing research, and provide a structured analysis of the information- and coordination- based merits of hierarchical governance.
3. Developing Theory: Informational and Coordination Superiority of Hierarchical / Organization-based Production
The exhortation to look production rather than transactions is not new. Winter (1988) and Demsetz (1988, 1995) and later Langlois and Foss (1998), lamented the excessive emphasis on the transactional advantages of different organizing and governance mechanisms, and the focus on conditions of exchange, rather than production. Kogut and Zander (1996) suggested that firms are superior because they form identities and shape behavior. Conner and Prahalad (1996) went so far as to propose a model whereby being in the same organization allows for parties to share knowledge more effectively than across firm boundaries. Therein integration / firm based governance (of the dyad, in their hypothetical example) is due to the knowledge-based advantages of keeping “contracting” entities under the same institutional roof. Moving beyond the formal statement of these advantages, Monteverde (1995) suggested that the need for unstructured technical dialogue was the reason for semiconductor firms to opt to integrate. In the converse, a well-defined standardized technical dialog allowed for greater dis-integration. Argyres (1999), in studying the development of the Stealth bomber, found the existence of technical dialog and information processing to be particularly important in ensuring inter-firm coordination; it also seemed that one of the major enablers of working across firm boundaries was the existence of standardized information. So overall, some recent research has alluded to the existence of “non-TC” merits of firms, yet the analytics of why these differences exist have still not been developed (Foss, 1996). We still lack a consistent analysis of what really drives the presumed superiority of integrated production. Proposing one such classification, we suggest that these “superiority drivers” belong to two categories: Information and coordination conditions, reviewed seriatim.
3.1. Firms, Markets, and the Role of Information
A firm (rather than the market) adds value as an organizational alternative by allowing for richer, more contextual information to flow within its own boundaries – if such “thick” information is required for linking adjoining parts of the production process (or for completing a potential transaction.) Thick information may be needed, if we cannot fully articulate ex ante what we require (due to causal uncertainty, as discussed by Thompson, 1967), or because we can neither specify nor measure what we need (see Jacobides and Croson, 2001, for an extended discussion). Alternatively, continuous and smooth interaction between the “buying” and the “supplying” division or units may be required. In such cases, the market may not be able to cope with this information overload. Production through firms has three types of “information-thickness”- related advantages:
1. Information throughput advantages refer to the ability of a governance structure to cope with the information needed in order to make certain that the ‘appropriate’ good or service is procured, and to help the production process operate smoothly. While markets can only clear prices with qualities, given an agreed-upon specifications vector (Barzel, 1982), firms can process much more information – for a number of reasons. First, there may not be a ‘market’ for the services needed; the specs may be so opaque and generic, or so inarticulable and imperfectly assessed, that market-mediated transactions cannot be conceivable, even in a world populated by benevolent agents (cf. Langlois, 1992). Second, it is common that continuous and smooth interaction between the buying and the supplying parts may be needed. Designers of different parts of an airplane might have to be co-located, because of the need for continuous communication; members of a team designing apparel might need to literally work together, because they need tacit and inarticulable information – or plainly too much data – to flow in order to be efficient.[13]
2. The second type of informational advantage of hierarchy is superior information on the productive assets it employs; i.e., input information advantages (Alchian and Demsetz, 1972; Arrow, 1974). A firm knows what its own resources can produce, much better than it can ever know what to expect from outside contractors. More importantly, perhaps, a firm has the ability to constantly reevaluate the way an activity is being organized, and to use the feedback in order to adjust the configuration of the resources it employs, a point to which we will return in talking about coordination advantages. Thus firms are able to make better use of the stock of information on their productive resources and deploy them effectively.
3. Finally, communication advantages refer to the ability of a governance structure to provide a platform for efficient communication. As Pelikan (1969) and Arrow (1974) observed, organizations economize on their communications ‘programs’ - i.e., they create a communicative ‘shorthand’, they define shared meanings and hence can render communication and, thence, production efficient. The ability of firms to foster good communication through the creation of shared meanings is a significant reason for keeping activities in-house, and being integrated. The reason is that the stable patterns of interaction within a firm (‘routines’, per Nelson and Winter’s terminology) that characterize firms or stable networks, facilitate the development of these shared concepts.[14] To wit, firms often expand / hire rather than using contractors in order to work with people “who speak the same language” (and indoctrinates employees with such a language upon their joining). Thus, “common vocabulary” is a significant merit of firms, especially in the absence of universally acceptable and shared language and business semantics “shortcuts.”
3.2. Information Conditions in a Legal Context: Firm- vs. Market- Based Compliance
To help us better understand why firms are superior to markets with regard to the use and access to information we should consider the significant differences in legal and institutional structures (Masten, 1993). A major difference between integration / employment and market-based contracting is that in integration / employment the division / employee allows the HQ / employer to prescribe and monitor actions within a reasonable action space (Simon, 1951.) This allows firms to access and use information “at will”. In the case of authority-based structures, the principal (headquarters => divisions, or employer => employees) have the ability to reconfigure the agents without the need to re-contract (subject to their participation constraint), and to monitor their actions up to the point where their personal privacy is breached. (Marx, 1988; Batt, 1976). As Masten (1993: 200) notes,
"this [exception to the general contractual non-disclosure clause]… also has become codified in the law … as the employees' duties to disclose and account. [A] subordinate is obliged to communicate to [his employer] all facts which he ought to know."
By contrast, in across-firm boundary agency relationships, monitoring rights are usually limited, and specified in contractual terms (Kronman, 1978.) A buyer by default has limited legal rights to decide what information she will collect about the method of provision of goods and services intended for her. [15] So both legal infrastructure and business practice drive the superior ability of firm to handle and request information.[16] Furthermore, the fact that a firm can have its own MIS systems; that its employees are usually co-located; and, more important, that the organization can engineer the types of information flows that can be produced through organizational design (Galbraith, 1973) leads to a marked advantage that firms can have over the market, even if they do not always make the best use of it.
Having looked at the information conditions, we can now turn to the second type of factors that may render integration / firm-based governance a more desirable, or the only feasible, condition.
------------------------------------------
Insert Figure 5 around here
------------------------------------------
3.3. Differences in Coordination in Firms and Markets
Organization of economic activities can be obtained in one of two fundamentally different ways: either through authority, or through independent compliance with price as a signal. The market’s key distinguishing feature, and the source of its advantages (Hayek, 1945) as well as its shortcomings (Milgrom and Roberts, 1992), lies in its ability to achieve coordination through the decentralized information prices confer, and the resulting desirable properties of the price discovery mechanism (in free & well-behaved markets). Markets base their operation on ‘independent’ compliance, that is, the responses of economic agents to the market signals. On the other hand, firms achieve coordination through the use of administrative fiat and managerial authority, as we observed in our sailing boat example. This difference in compliance is behind hierarchy’s “coordinative” merits. Specifically, coordination advantages of hierarchical governance are based on the speed and extent of compliance.
1. As for speed of compliance, the market mechanism takes some time to respond to desired changes in supplied product attributes. If the company wants to make specific changes in quality, it is easier to mandate such changes by fiat rather than wait for the market to yield an efficient solution. As Langlois (1992) observed, even if a market could theoretically obtain, it will not materialize “in real time”. Time-sensitive activities, where quick coordination is needed, tend to be internalized within the firm; and casual observation suggests that the more important tasks, which need to be carried out fast, are kept within firm boundaries (consistent with contingency theory – cf. Thompson, 1967, Laurence and Lorsch,1973, Galbraith, 1973). Further, as firms have superior information on the nature of the productive resources they employ, they can expeditiously use mandated compliance to re-deploy them profitably. That is, firms are able both to identify the most promising returns for their resources (through information advantages), and to arrange the best possible deployment (through their ability to coordinate quickly by mandated compliance).
2. Fiat / mandate also ensures greater extent of compliance, as it can specify many compliance dimensions (e.g. quality attributes, design aspects, positioning etc.). To use the terminology of section 2, it enables a specification of a high-dimensionality output vector, especially when constant re-specification is needed. Also, managing by fiat takes (or can take) into consideration any operational or strategic externality. Consider a firm that wants to cut total costs or improve quality, and ponders about the merits of doing so under a market-based vs. an integrated regime. In the market-based regime, local responses of contractors to cost-cutting along different parts of the value chain may lead to a plethora of different, or even mutually incompatible solutions, for two reasons. First, the local response achieved by independent compliance may not be exactly what is desired, as the full detail of the desiderata cannot be specified or be handled by the market –to put it in the famous words of Benjamin Franklin, repeatedly used in IO textbooks, “if you want to do something right, do it yourself.” Second, each local choice (which can be the best cost-cutting or quality enhancing choice) may not be compatible with or integrate well to other parts of the evolving configuration along the value chain. Independent compliance often makes it hard for externalities and complementarities to be captured. [17] One of the reasons that strongly complementary activities are almost always within the same firm boundaries is that coordination obtained through the market will not take into account the complementarities and externalities (cf. Milgrom and Roberts, 1989; 1991). To use Thompson’s (1967) terminology, in the presence of complex interdependence and if there is a tight coupling in the production process, the market may prove a poor means to organize.[18]
In the sailing boat example of section 2.2, the skipper needed authority both because this would ensure higher speed of compliance than any type of a market, and because the actions between the different members of a crew are highly interdependent; hence no decentralized market system can solve such a coordination problem. Additionally, the difficulty in specifying simple metrics for operating a hypothetical market and the higher dimensionality of the output vector make hierarchical governance attractive. Inasmuch as a production process is characterized by intense and hard to assess interdependencies, it is not even conceivable to create intermediate markets, for the very same reasons that a complex, interdependent problem cannot be successfully decomposed (Simon, 1962; Baldwin and Clark, 2000).
4. From the Genesis of Markets to their Disappearance: The Long Cycles of Market Growth and Decline
So far, we have built a distinct argument as for why markets emerge on the first place, and subsequently systematized the analysis of the informational and coordinational advantages of hierarchies. Now let’s extend the argument, and see how these same factors relate to the old question of why markets disappear and become displaced by integration. While in that respect the TC analyses remain valid, the question is what more can we say about this problem. Having done that, we will be able to conclude our analysis, and come full circle to the examination of the patters of value chain evolution on the aggregate.
4.1. When do Markets Fail? The Coordination Story
To revisit the drivers of market apparition, consider the example of the Swiss watch making industry, discussed in Enright (1995). The watch making industry that had developed in the Jura mountains, close to Geneva, dominated the watch making business until the 1950’s. The industry had slowly evolved into a set of independent manufacturers, each occupying a part of the production process. Some watch-makers specialized in making hands for the watch, some finishing the cases, some working in the straps, etc. The limits of what each did were clear; and the intermediate markets had converged on a simple way to decompose the watch-making activity, with few interdependencies between the parts of the production process, and with well-defined dimensions on which the intermediate markets traded on. As long as improvements were needed within each of these modules / pre-defined parts of production, and as long as these related to the attributes that the intermediate market was based on, the industry prospered- and it did so for a long time.
However, when there was a new need to combine advertising and technology development with the watch making progress in the 1960’s, the vertically-disintegrated, market-based organization of the watchmakers in the Jura mountain was unable to respond effectively. Their ability to react was circumscribed by the limits implicitly imposed by the existing division of labor in the value chain. Independent watch-making firms were linked through the market, unable to coordinate effectively, like a skipper-less crew in rough and uncharted waters. There was no-one that could arrange for the tasks to be reconfigured appropriately, lest vertical integration were to take place – and that was exactly what happened: the need to coordinate and adapt to change led to vertical integration. Even in the high-end of the market (mechanical, hand-made watches), where the product itself has barely changed from the 1800’s to today, massive integration occurred, as individual producers could not coordinate their responses through the market.[19] It was not any newly found economies of scale; it was the inability to coordinate that mattered. (To wit, the resulting conglomerates such as the SHR kept the same portfolio of smaller production units, but simply changed the mode of compliance, using fiat in the place of market dynamics.)
What obtains, then, is that the market imposes some limits to the extent to which changes can happen along a value chain, as it prescribes an “architecture”, and defines how exactly production is to be orchestrated. It further hinders some types of systemic adaptation as the market cannot coordinate actions; there are too many externalities in such changes, and as we saw, markets may be ill-equipped to deal with such externalities. To use Henderson and Clark’s (1990) terms, the market can deal very well with “component” changes; yet it appears to suffer shortcomings in coping with more drastic changes and reconfiguration, such as re-shaping the way basic tasks are performed.[20] So, at least some type of change requires authority and fiat. This (rather than just hold-up and contractual problems) may be why integrated solutions are often in the forefront of radical innovation.[21]
To recapitulate, firms, because of their ability to process information and yield a stable solution to the coordination game, may be superior to the market. This can be particularly important under conditions of change, when adaptation is called for, as mandated authority can ensure a better solution to the pure coordination problems. The extent of the superiority of the firm, then, depends on the appropriateness of the market structure, and the relative changes of the economic environment (whether it calls for architectural reconfiguration or not). Thus, technological and institutional factors affect the potential coordinative superiority of the firm over the market.
4.2. Market Birth and Decline: Are Moving to a Dis-Integrated World?
So we now have a fuller overview of the drivers of market appearance and decline. Moving to the broader context, let’s consider the evolution of value chain structures. Is it true that industries tend toward greater specialization? Is the world becoming vertically more specialized and modular?
The evidence in the mortgage banking industry is, at least prima facie, consistent with such a view. This was driven by new markets, which appeared as soon as some basic dimensions became standardized, thus enabling for the production and servicing of a loan to be done by many different institutions, connected through the market. The first major market creation was the secondary mortgage loan market. In order for this market to be created, significant standardization was needed, and was offered mainly through the efforts of the GSE’s. Specifically, this standardization (and the subsequent vertical dis-integration) started with the government-secured loans, wherein some information on the customers was made redundant through the government-backed guarantee. (The securitization process in mortgages started with the government loans, which guaranteed the repayment of defaulting loans.) This made some of the information on the credit-worthiness of individual loans redundant, and hence enabled market mediation. Given the knowledge of the basic way to organize the mortgage process and the underlying loan so that it is produced in a dis-integrated manner, the learning was then applied to a much larger category, namely the “conventional” 1-4 home residential loans, where information could be more readily standardized – even without the government subsidy. The knowledge on how to organize this market-based loan production value chain, allowed for specialization in “trickier” loans, without guarantees, and so was born the bulk of the securitized “conventional” loans. After the production of such loans became a vertically specialized business, the integrated firms were working on non-traditional loans, such as jumbo, home equity, and commercial loans. Each of these more complicated, newer types of financings were also gradually affected by the dis-integration trend, albeit with a lag; the same learning had to progress with these newer loans, which became increasingly disintegrated – even at a lower absolute level, given greater difficulties in information standardization for unconventional loans. Thus the initial government guarantee itself was not that important in and of itself; what was more important is that it provided a template for loans to be sold. Once the “recipe” was known and once banks understood how to “break” the value chain and create a new market, the same process followed in non-government (conventional) loans, and later on more “exotic” species, giving way to the vertically dis-integrated value chain depicted in Figure 1.
So the first step was the break-up between making / servicing and holding a loan, and the development of the mortgage banker model. The next stage was vertical dis-integration and intermediate market creation within mortgage banks themselves (the “boxes” of Figure 1). With time, rather than having only unified firms doing all the mortgage banking activities reported in Figure 2, a market-mediated, dis-integrated picture started emerging. First, an intermediate market for “loans to be serviced” (or Mortgage Servicing Rights) appeared, and hence firms could now only produce or service loans, and the market could link the two. Then origination itself got broken up in two segments – upstream brokerage / retail production and downstream warehousing, mediated by the market for closed loans, a tendency that was documented in our three vignettes.
The reason was that the firm is superior to the market, only inasmuch as complexity and interdependencies create the need for fiat and authority. Inasmuch as a production process is separable and modular, however, the market can work just fine. Thus, the gradual evolution of production system into separable, interconnected yet not interdependent sub-modules allows for the creation of markets. In general, then, the creation of decomposable systems (as discussed by Simon, 1962) and modular designs (Baldwin and Clark, 2000) is an important forerunner and precondition for an efficient market. Modularity in production allows for the reduction of interdependence of each components, as well as the standardization of information. It thus creates the preconditions for a market to be feasible, and as efficient as (if not more than) integrated production.
But can we really generalize from this setting? Mortgage banking was, after all, selected because of the process of dis-integration it has gone through. The answer is that, on the one hand, these processes of vertical specialization do happen (and increasingly so) in several industries. Hence our setting is not unique; and we can learn a lot about this unbundling and market-creating process, by looking at its determinants. On the other hand, though, this may be part of a broader, cyclical trend, rather than a linear progression towards a world which is becoming “blown to bits” (Evans and Wurster, 1999).
So while learning, information standardization and the creation of discontinuities that leads to new markets being created, and to a dis-integrated value chain structure (as in mainstream mortgage banking), this is only half the story. This trend is partly countered by an “opposing force”, namely the development of new production technologies, or new products requiring subtly new technologies. Thereby, intermediate markets are not yet properly set up. Consider the example of the Swiss watch industry: The fine division of labor was not appropriate to cope with the new coordinational demands of new types of watches, and hence the intermediate markets gave way to more integrated solutions.
Even in the mortgage banking industry, new products such as “reverse mortgages” seem to be introduced in a vertically integrated form, as there is no way to easily partition them and have them manufactured or sold and serviced by linked entities; but there, too, with time intermediate markets creep in. So to some extent, the “market-ization” or the dis-integration of production is a function of the stage at which the underlying products or technologies are in. New technological possibilities lead some firms to integrate anew when tighter coupling is needed and re-organize the production process. In new products or new technologies, more intense coordination is needed (or is simply beneficial) and this may lead to the integrated, authority-based production being more efficient (or simply the only feasible) solution.
This happened in chemicals and food processing in the late 19th century, when some firms took advantage of the new technologies by integrating vertically and transforming the production layout (Chandler, 1977). Once the new technology or product structure becomes established, though, the centrifugal forces of dis-integration surface again, leading to the creation of new (or expanded) intermediate markets and to further specialization, until the shock from innovation and a more efficient yet “tightly coupled” production method comes along, thus making integration advantageous. This leads us to a cycle of dis-integration (driven by learning and improvements of the market), and re-integration (when new technologies or products require adaptation and coupling that cannot be effectively done through market-based compliance).
These dynamics are described in Figure 6. Casual empiricism suggests that in many industries, new products and technologies start off as being more integrated, and then the production process becomes more dis-integrated, as intermediate markets appear. A research project, currently under way, should be able to shed some light to the generalizability of this process.
------------------------------------------
Insert Figure 5 around here
------------------------------------------
5. Integrating Evidence, Theory, and Business Reality: The New Economy and the Changing Structure of the Value Chain
Let us now take stock. We started the paper by looking at the evolution of the mortgage banking industry, and the development of new markets. We then explored the drivers behind these new markets, and saw that information and coordination conditions seemed to matter. This prompted some theory development, whereby we distinguished between the TC described in theory as the factors that explain institutional choice, and the factors we observed in practice in looking at new markets, with the focus being on information & coordination, as well as learning, historical accident, and underlying capability or growth heterogeneity.
Having offered a new framework and linked it to existing literature, we can bring this all together and examine a last piece of empirical reality – namely, the implications of Information Technology (IT) on the development of intermediate markets, and on the resulting structure of the value chain.
1. IT and the Changing Value Chain Structure
IT, we are told, is re-shaping the vertical structure and the nature of markets on many an industry (Evans and Wirster, 1997, 1999; Rayport and Sviokla, 1995). But how so? Per our analysis, IT enables better information handling capabilities across firm boundaries, so much so that the creation of a vertical discontinuity, a market, is conceivable. By standardizing the requisite information, it renders the relative advantages of the firm less valuable or irrelevant. By standardizing the production process, it makes the market more appropriate.
Several information- and coordination-intensive transactions could, before the IT era, be economically performed only in a hierarchy, simply because the need for information transfer and communication would render any market-mediated procurement prohibitively expensive. Recently, and due to IT, in many instances where hierarchies were the only “practically feasible” solution, networks of suppliers or even the open market procurement have become economically viable. Electronic Data Interchange (EDI) has, for example, ended the prohibition on market procurement for many bandwidth-intensive coordination processes, or rendered the market procurement mechanism a much cheaper, faster and more reliable alternative than before. Designers need not sit in adjoining rooms or be under the same corporate roof; CAD/CAM now allows for efficient safe and cheap information transmission. IT has acted as a significant dis-integrator of the production function. As Johnston and Lawrence (1988) and Lawler (1988) suggest, by reducing the unit cost of information processing, networks (and ultimately, markets) substitute for the information processing formerly done by hierarchies.
Furthermore, the proprietary automated / information systems used by firms or even networks, which until very recently have not been interconnectable (Cohen and Apte, 1996; Kaminsky, 1982), have now started giving way to more modular and interchangeable, web-enabled systems. The open standards associated with the internet and, perhaps more important, the universally recognizable XML (eXtended Markup Languages) provide a universal “information Esperanto”. This allows for different proprietary systems to understand each other, and thus greatly facilitate the ability of parties linked through more occasional market transactions to coordinate effectively. XML’s are also important as they define and standardize not only the “meaning” of the digital bits of data exchanged between different systems, but also the business process itself. That is, many industries (including mortgage banking and the MISMO effort; see Tully and Williams, 2000) are trying to come to mutually agreed upon maps that define each step of the business process, in an effort that bears some similitude to the recently completed genome project. They are trying, through XML’s, to provide a language for the entire value chain, and the processes and intermediate steps therein. By so doing, they standardize, for the first time ever, the description of the way business is done. This, if fully implemented, means that market based-transactions will almost be at par, on an informational level, to hierarchical governance. Communication short-hands become industry-wide and non-proprietary, rather than firm-specific. Hence markets become significantly more attractive than they were before. Otherwise put, these efforts to standardize processes and information should lead (and have led) to a reduction of the extent to which activities are (or have to be) organized within one firm’s boundaries, as markets become better equipped. (See Evans and Wurster, 1997 and 1999).
5.2. A New Perspective of B2B’s: Market Creators to the Fore
Further evidence of the increasing impact of IT on the value chain structure is the stupendous growth of B2B exchanges in many sectors – including, of course, mortgage banking (Goldman Sachs, 2000; MSDW, 2000). These foremost “market creators” have been the outgrowth of, but have also contributed to the expansion of, information standardization and new technology of handling information. Regardless of current market capitalization, or their success as business models (Croson, Jacobides and Nguyen, 2001), these exchanges have changed the playing field.
The development of B2B firms in the mortgage banking sector such as OpenClose or IMX (which auction off loans from brokers to banks, or simply hook up brokers to banks – see Blutz, 2000) helped standardize the interface between mortgage bankers and mortgage brokers, thus boosting specialization in the loan origination process and hence reliance on markets. Likewise, firms like Pedestal and Ultraprise (cf. Quinn, 2000), that specialize in the trading of whole loans, have contributed to the creation of a more active market for closed, non-securitized loans, by their very institution of a set of descriptors that facilitates market transactions. B2B’s contribute to the dis-integration of the value chain by supporting the market mechanism with additional coordinative and information processing features, propelled by recent evolutions in IT. B2B’s also make the best of the market clearing technology, through the perfection of the auction-based mechanisms and the market clearing structures (Spubler, 1999), and through decision support algorithms, allowing for markets to cope with more difficult coordination tasks than those they could cope with before.
The fury of the stock-market to support such collaborative e-commerce in the last three years (MSDW, 2000; Goldman Sachs, 2000), and its willingness to continue some level of support even in such times of crisis in e-stocks, has also led substantial resources in devising new types and modes of intermediate markets. This has enhanced market efficiency and facilitated intermediate market creation, as these new markets have remarkable capabilities, and their fine-tuning has been funded as intensively as never before by investors. So while the tendency towards modularization is not a linear progression, as we observed in section 4, there is little doubt that the IT revolution has significantly increased the degree of “market-ization”.
5.3. New Perspectives and Old Deadlocks
This perspective may be particularly timely, as “traditional” TCE has fared quite poorly in explaining the impacts of the digital revolution. Despite the fact that these new information technologies create unforeseen opportunities to hold up and exploit strategically sensitive information (consider, e.g. outsourcing of IT systems and data) and although they don’t come with water-proof transactional guarantees, we still see a proliferation of market-based arrangements. This happens because now the market is able to coordinate economic activities in a more efficient way, and better process and handle information, aided by IT. In turn, this leads to a reduction of the advantages of firm-based governance. So the approach proposed in this paper could help us understand how IT matters.
Concluding Discussion
In the last thirty years, much headway has been done in answering Coase’s (1937) question, “why a firm?”, but very little in answering Harrison White’s (1981) question, “Where do Markets Come from?” To tackle the latter question, this paper looked at an actual value chain as it evolved, and studied the genesis and evolution of markets in mortgage banking. The factors that we uncovered in so doing, differed significantly from those emphasized in extant research. Based on that evidence, we looked at what, other than TC, could help explain the evolution of markets. The desire of firms to take advantage of their capabilities; learning, and evolution of market roles; standardization and historical accident, were all catalysts (examined in depth in a related paper – cf. Jacobides, 2000). In addition, our analysis suggested that there are two significant factors that affect the merits of markets vs. hierarchies: coordination and information.
To provide a theoretical foundation to these issues, we juxtaposed them to the TCE analysis. So, on the theoretical level, we specified particular reasons under which, even assuming away any incentive incompatibilities (and thence the danger of reneging) firms and hierarchies present advantages over markets. The paper further developed these information- and coordination- based factors, integrating the analysis to the rather disparate discussion in the extant literature.
There are several take-aways from this analysis. On the programmatic level, it is clear that studying TC is not enough. In our setting, TC did not really matter; more to the point, when looking at an industry, we should look at what really matters rather than trying to prove or disprove TC. We need to follow value chains as they evolve, and study what drives their evolution. In that respect, both exploratory and confirmatory research are called for, and a research project currently under way should provide evidence from a number of different industries.
On the conceptual level, it is clear that the superiority of firms as an organizational interface is not an absolute theoretical given (Barber, 1977; North, 1986). Rather, it is an empirical question, whose answer depends on the nature of information that has to be processed between two adjoining steps in the value chain, the type of coordination needed – and, importantly, of the technology that permits to link the two.
On the pragmatic level, this approach also allowed us to recast the role of IT, and examine how it changes the relative merits of firms vs. markets. The ongoing information revolution is re-shaping the boundaries of the firm not because it affects “traditional” TC, but rather because it affects the relative capabilities of market-based governance to process complex information and to coordinate complicated activities. Learning and standardization of information also affect the possibilities of market in becoming a viable organizational alternative. As people try to take advantages of potential gains from trade, they create and support market exchange mechanisms and this renders the market informationally and coordinatively efficient – and hence able to displace firms. On the other hand, new technologies may require coupling, leading back to the need for firm-based governance.
In this time of changing firm boundaries, the need for a theory that can account for the dramatic transformations of the firm and industry structure is evident. We need to look at the facts, and to allow for factors that have been neglected in previous research, rather than try to use our existing analytical tools, hoping that the keys are to be found beneath our (few and well researched) existing theoretical lampposts.
References
Akerlof, G., 1970. The Market for Lemons: Qualitative Uncertainty and the Market Mechanism, Quarterly Journal of Economics, 84 (August): 488-500.
Alchian, A. A., and Demsetz, H., 1972. Production, Information Costs and Economic Organization, American Economic Review, 62 (5): 777-795.
Allen, F.H., and A. Santomero, 1997. The Theory of Financial Intermediation, Journal of Banking and Finance, Vol. 21 (11-12). p 1461-85. December
Aoki, M, Gustafsson, B. and Williamson, O.E. (Eds.), 1994. The Firm as a Nexus of Treaties, London: Sage, 133-161.
Argyres, N S. 1995 Technology strategy, governance structure and interdivisional coordination. Journal of Economic Behavior & Organization. 28(3): 337-358.
Argyres, N.S., 1999, The Impact of Information Technology on Coordination: Evidence from the B-2 ‘Stealth’ Bomber, Organization Science, 10(2): 162-180.
Argyres, N.S., and Liebeskind, J. 1998. Contractual Commitments, Bargaining Power and Governance Inseparability: Incorporating History into Transaction Cost Theory, Academy of Management Review
Armour, H. O. and Teece, D. J. 1980. Vertical integration and Technological Innovations. Review of Economics and Statistics. 62: 490-494.
Arrow, K., 1974. The Limits of Organization, The Norton Press, New York.
Asanuma, B., 1993. Interfirm Relationships in the Japanese Automobile Industry. Rivista Internazionale di Scienze Economiche e Commerciali. 40 (12): 1019-40.
Baldwin C. and Clark, K. 1999. The Modularization of the PC Industry, Working Paper, Harvard Business School
Baldwin C. and Clark, K. 2000. Modularization and Product Architecture, forthcoming, Harvard Business School Press
Barber, B., 1977. The Absolutization of the Market: Some Notes on How we got from There to Here in G. Dworkin, G. Bermant and P. Brown, Eds., Markets and Morals, Washington DC: Hemisphere: 15-31.
Barzel, Y. 1982. Measurement Cost and the Organization of Markets. Journal of Law and Economics. 25 (1): 27-48.
Batt, E.R. 1967. The Law of Master and Servant, London: Pittman and Sons
Becker, G. S., and Murphy, K.M. 1992. The division of Labor, coordination costs and knowledge. Quarterly Journal of Economics, 111: 165-194.
Blau, P.M. 1956. Bureaucracy in modern society New York, Random House
Blutz, W. 2000. Pool Party: Loan Exchanges Online. Mortgage Banking, March, 77-83
Bolton, P. and Farrell, J. 1990. Decentralization, Duplication, and Delay Journal of Political Economy 98 (4) : 803-26
Bradach, J.L., and Eccles, R.G. 1989. Markets versus Hierarchies: from Ideal Types to Plural Forms, Annual Review of Sociology, 15: 97-118
Camerer and Knez, 1997. Coordination in Organizations: A Game Theoretic Perspective, in Zur Shapira, Ed, Organizational Decision Making. Cambridge: Cambridge University Press
Chandler, A. D., 1962, Strategy and Structure, Cambridge, Mass: MIT Press
Chandler, A.D., 1977. The Visible Hand, Cambridge, Mass: Harvard University Press.
Clemons, E. K., S.P. Reddi, and M. C. Row. 1993. The impact of information technology on the organization of economic activity: The ‘move to the middle’ hypothesis. Journal of Management Information Systems, 1993, 10:2 Fall: 9-35.
Coase, R.H., 1937. The Nature of the Firm, Economica, 4:386-405
Coase, R.H., 1972. Industrial Organization: A Proposal for Research, in V.R. Fuchs (ed.), Policy Issues and Research opportunities in Industrial Organization, New York: National Bureau of Economic Research
Coase, R.H., 1988. The Nature of the Firm: Meaning , Journal of Law and Economics,4: 19-32
Cohen, M.A., and U. Apte, 1996 Manufacturing Automation
Conner, K.R., and Prahalad, C.K. 1996. A resource-based theory of the firm: Knowledge versus opportunism Organization Science. 7(5): 477-501.
Craft, G.R. 1998. The Changing Structure of Financial Services, Report of BancAmerica Robertson Stevens, March.
Croson, D.C., and Jacobides, M.G., 1999. Small Numbers Outsourcing: Efficient Procurement Mechanisms in a Repeated Agency Model Working paper, Department of Operations and Information Management, The Wharton School, Philadelphia, PA.
Croson, D.C., Jacobides, M.G. and Nguyen, A., 2001. “Explaining Entry Decisions and Crowdedness in the Business-to-Business Market”, Working Paper, Center for the Network Economy, London Business School.
Cyert, R. M., and March, J.G. 1963. Behavioral Theory of the Firm, Englewood Cliffs: Prentice Hall.
Demsetz, H. 1968. The Cost of Contracting. Quarterly Journal of Economics. 87: 33-53.
Demsetz, 1969. Information and Efficiency: Another Viewpoint. Journal of Law and Economics, 12 (April): 1-22
Demsetz, H., 1988. The Theory of the Firm Revisited, Journal of Law Economics and Organization, vol. 1: 141-161.
Demsetz, H., 1995. The Economics of the Business Firm: Seven Critical Commentaries, Cambridge: Cambridge University Press.
Demsetz, H., 1997. The firm in economic theory: A quiet revolution American Economic Review, 87: 426- 429.
Dennis, M. W. 1983. Mortgage Lending Fundamentals and Practices, Reston, Virginia: Reston Publishing Company, Inc.
DiMaggio, P.J. and Powell, W.W. 1983. The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields, American Sociological Review, 48: 147-160
Dosi, G. and D. Teece, 1998. Organizational Competencies and the Boundaries of the Firm. In R. Arena and C. Longhi, (eds.), Markets and Organization, New York: Springer-Verlag, 281-300
Eccles, R. G., and H. C. White 1986. Firm and Market Interfaces of Profit Center Control, in Approaches to Social Theory, S. Lindberg, J. S. Coleman and S. Novak, Eds., Russel Sage, New York: 203-220.
Enright, M. J., 1995. Organization and Coordination in Geographically Concentrated Industries, in N.R. Lamoreaux and D.M.G. Raff (Eds.), Coordination and Information: Historical Perspectives on the Organization of Enterprise, Chicago: University of Chicago Press and NBER: 102-142
Evans J. and T. Wurster, 1997. Strategy and the new Economics of Information Harvard Business Review
Evans J. and T. Wurster, 1999. Blown to Bits, Boston: Harvard Business School Press.
Farrell, J. and Saloner, G. 1988. Coordination through Committees and Markets, RAND Journal of Economics, 19 (2): 235-252
Follain J.R. and P.M. Zorn, 1990. The Unbundling of Residential Mortgage Finance, Journal of Housing Research (1) 1: 117-137.
Foss, N.J., 1996a. Knowledge-based approaches to the theory of the firm: Some critical comments Organization Science 7 (5): 470-476
Foss, N.J., 1996b. More critical comments on Knowledge-based theories of the firm Organization Science 7 (5): 519-522
Fransman, M., 1994, Information, Knowledge, Vision and Theories of the Firm. Industrial and Corporate Change 3 (3): 713-757.
Galbraith, J., 1973. Designing Complex Organizations, Addison-Wesley Publishing Company.
Garrett, J., 1989, A New Paradigm for Lenders, Mortgage Banking, May: 30-41
Garrett, J., 1993, The Unbundling of an Industry, Mortgage Banking, May: 12-21
Ghoshal, S., and P. Moran. 1996. ‘Transaction-Cost Economics: Bad for Practice’, Academy of Management Review, 21 (1): 1-47
Goldman Sachs, 2000, B2B: Collaborative Commerce, Report of the US Technology Group
Granovetter, M., 1985. ‘Economic Action and Social Structure: The Problem of Embededness’, American Journal of Sociology, 91: 481-510.
Grossman, S., and Hart, O. 1986. The Costs and Benefits of Ownership: A Theory of Lateral and Vertical Integration. Journal of Political Economy. 94: 691-719.
Hart, O., 1995. Firms, Contracts and Financial Structure, New York: Oxford University Press.
Hart, O. D., and J. Moore, 1990. Property Rights and the Nature of the Firm, Journal of Political Economy, 98, 4: 1119-1158.
von Hayek, F., 1945. The Use of Knowledge in Society, American Economic Review, 35, September: 519-530.
Helper, S., J-P. MacDuffie and C.F. Sabel, 1997. The Boundaries of the Firm as a Design Problem, Working Paper, Mimeo
Henderson, R., and K.B. Clark, 1990. Architectural Innovation: The Reconfiguration of Existing Systems and the Failure of Established Firms, Administrative Science Quarterly 35 (March) 9-30.
Hennart, J.-F., 1993, Explaining the Swollen Middle: Why Most Transactions are a Mix of ‘Market’ and ‘Hierarchy’, Organization Science, vol. 4 no.4 (November): 529-547.
Hodgson, G., 1998. Competence and Contract in the Theory of the Firm, Journal of Economic Behavior and Organization, 35: 179-201.
Hodgson, G.M., 1988. Economics And Institutions : A Manifesto For A Modern Institutional Economics, Philadelphia : University of Pennsylvania Press
Hodgson, G.M. 1993. The Coasean Tangle and the Problem of Historical Specificity, Working Paper, Judge Institute of Management Studies, University of Cambridge.
Holmstrom, B. 1996. The Firm as a Sub-economy. Mimeo, Massachusetts Institute of Technology.
Holmstrom, B., and J. Tirole, 1989. The Theory of the Firm, in R. Schmalensee and R.D. Willig (Eds.), Handbook of Industrial Organization, Amsterdam: North Holland.
Imai K., and H. Itami 1984. Interpenetration of Organization and Market, International Journal of Industrial Organization, 2: 285-310.
Jacobides, M.G., 1998. Information, Technology and Coordination in Firms and Markets, Working Paper 98-013, Emerging Technologies Program, The Wharton School of the University of Pennsylvania.
Jacobides, M.G., 1999. Value Chain Structure, Co-Specialization, Exportability of Competencies and Global Expansion in Mortgage Banking, Working Paper, Real Estate Department, The Wharton School of the University of Pennsylvania.
Jacobides, M.G., 2000, “Capabilities, Transaction Costs and Limits to Growth: A Formal Model of Scope and Profitability”. Working Paper #01-16, Wharton Financial Institutions Center, The Wharton School
Jacobides, M.G., 2001, “Mortgage Banking Unbundling”. Mortgage Banking (Annual Strategy Review: Cover Report), January: 21-40.
Jacobides, M.G., and L.M. Hitt, 2000, “Vertical Scope, Revisited: Capabilities vs. Transaction Costs in Mortgage Banking”. Working Paper #01-17, Wharton Financial Institutions Center, The Wharton School
Jacobides, M.G., and D.C. Croson, 2001, Information Policy: Shaping the Value of Agency Relations, (April), Academy of Management Review: 202-221
Jacobs, J.J., 1993, In Search of the Real Mortgage Bankers, Mortgage Banking, June: 21
Jaffee, D.M. and K.D. Rosen, 1990. Mortgage Securitization Trends, Journal of Housing Research 1 (1): 117-137
Johnston, R. and P. R. Lawrence, 1988. Beyond Vertical Integration - The Rise of the Value-Adding Partnership, Harvard Business Review, (July-Aug 1988): 94-101.
Klein, B., Crawford, R. G. and Alchian, A. A. 1978. Vertical Integration, Appropriable Rents, and the Competitive Contracting Process. Journal of Law and Economics. 21: 297-326.
Kogut, B. 1993. The Nature of the Firm: Origins, Evolution, and Development. Administrative Science Quarterly, 38(3): 503-507.
Kogut, B. and U. Zander, 1996. What Firms Do? Coordination, Identity, and Learning, Organization Science, 7 (5): 502-518
Kronman, A.T. 1978. Mistake, Disclosure, Information, and the law of Contracts. Journal of Legal Studies, 7: 1-34.
LaMalfa, T., 1998, “The Gians of Wholesale Lending”, Mortgage Banker, October
Langlois, R.N., 1992. Transaction Cost Economics in Real Time, Industrial and Corporate Change, 1: 99-127
Langlois, R.N., and P.L. Robertson, 1995. Firms, Markets, and Economic Change: A Dynamic Theory of Business Institutions, London: Routledge
Langlois, R.N., and N. Foss 1997. Capabilities and Governance: The Rebirth of Production in the Theory of Economic Organization, Working Paper, mimeo
Lederman, J. (ed.), 1995, Introduction to Mortgage Banking, Washington, DC: MBA Editions
Lewis, M., 1990. Liar’s Poker, Penguin Books.
Madhok, A, 1996. The Organization of Economic Activity: Transaction Costs, firm Capabilities, and the Nature of Governance, Organization Science, 7 (5): 577-590.
March, J. G., and H.A. Simon, H. A., 1958. Organizations, New York: Wiley Marshall,
Marschak, J. and R. Radner, 1972. Economic Theory of Teams, New Haven: Yale University Press.
Meckling, W. H., and Jensen, M. C. 1992. Knowledge, Control, and Organization Structure: Parts I and II. In L. Hijkander and H. Werin (Eds.), Contract Economics. 251-274. Cambridge, MA: Basil Blackwell.
Milgrom, P. and J. Roberts, 1992. Economics, Organization, and Management, Englewood Cliffs, NJ: Prentice Hall.
Monteverde, K. 1995. Technical Dialog as an Incentive for Vertical Integration in the Semiconductor Industry, Management Science, vol. 41: 1624-1638.
Montgomery, C.A. (ed.),1995. Resource-based and Evolutionary Theories of The Firm: Towards a Synthesis. Boston : Kluwer Academic Publishers
MSDW (Morgan Stanley Dean Witter), 2000. The Internet and the Financial Services Report.
Nelson, R. R., and Winter, S. G. 1982, An Evolutionary Theory of Economic Change. Cambridge, Mass.: The Bellknap Press of Harvard University Press.
Nisiguchi, T., 1994. Strategic Industrial Sourcing: The Japanese Advantage, New York: Oxford University Press.
North, D.C., 1986.The New Institutional Economics, Journal of Institutional and Theoretical Economics, 142: 230-327.
Pelikan, P., 1969.Language as a Limiting Factor for Centralization, American Economic Review 59 (4): 625-631.
Penrose, E., 1959. The Theory of Growth of the Firm, New York: John Wiley and Sons
Popper, K., 1974. Conjectures and Refutations: The Growth of Scientific Knowledge, London: Routledge and Kegan Paul.
Porac, J.F, H. Thomas, F. Wilson, D. Paton, and A. Kanfer, 1995. “Rivalry and the industry model of Scottish knitwear producers”, Administrative Science Quarterly, June, 40 (2): 203
Portner, F. 1995. (Re)Inventing Mortgage Banking. Mortgage Banking. October 1995: 82.
Posner, K., 1999 The Internet Mortgage Report, Morgan Stanley Dean Witter Research, New York: March
Rayport, J.F., and Sviokla, J.J. 1995. Exploiting the Virtual Value Chain, Harvard Business Review, (November-December): 75-85.
Richardson, G.B., 1972. The Organization of Industry, Economic Journal, 327: 883-96.
Schelling, T, 1960, The Strategy of Conflict, Cambridge, MA: Harvard University Press
Schelling, T., 1978, Micromotives and Macrobehavior, New York: Norton
Schilling, M. 2000, Towards a General Theory of Modularity, Academy of Management Review
Schneider, H. 1994. Great Strides in Industry Automation. Mortgage Banking. November 1994.
Simon, H.A. 1947. Administrative Behavior, Free Press
Simon, H.A. 1951. A Formal Theory of the Employment Relationship. Econometrica, 19: 295-305
Simon, H. A. 1962. Sciences of the Artificial. MIT Press.
Simon, H. A., 1991. Organizations and Markets. Journal of Economic Perspectives. 5 (2): 25-44.
Slesinger, P.K., 1994, Data Standardization: Streamlining Access to the Secondary Market, Mortgage Banking, February: 107
Smith, A., 1776, The Wealth of Nations. Glascow
Snowden, K.A., 1997, Mortgage Securitization in the United States, in Sylla (ed.), Evolutions in the Anglo-American System of Finance
Stigler, G., 1951. The Division of Labor is Limited by the Extent of the Market, Journal of Political Economy, 59 (June): 185-93.
Stinchcombe, A. 1990. Information and Organizations, Berkeley, CA: University of California Press
Szulanski, G. 1996. Exploring internal stickness: Impediments to the transfer of best practice within the firm. Strategic Management Journal. 17: 27-43.
Szulanski G., and Winter, S.G. 1999. Replication as Strategy. Forthcoming, Organization Science
Thompson, J.D.. 1967. Organizations in Action. New York: McGraw-Hill
Utterback, J.M., and Abernathy, W. 1975. A Dynamic Model of Process and Product Innovation. Omega 3: 639-656.
Van Zandt, T., 1998. Decentralized Information Processing in the Theory of Organizations, in M. Sertel (ed.), Contemporary Economic Development Reviewed, Volume 4: The Enterprise and its Environment, London: MacMillan Press Ltd
Varian, H., and G. Shapiro, 2000, Information Rules, Boston, MA: HBS Press.
Weber, R, 1999. Coordination Problems in Growing Firms: Insights from Experiments, Working Paper, mimeo.
Weitzman, M., 1974. Prices vs. Quantities, Review of Economic Studies, vol. 41: 4-16.
White, H.E., 1981, “Where do Markets Come From?” American Journal of Sociology, Vol. 87, No. 3. (Nov): 517-547
Williamson, O. E., 1975. Markets and Hierarchies: Analysis and Antitrust Implications, Free Press: New York
Williamson, O. E., 1985. The Economic Institutions of Capitalism, Free Press, New York.
Williamson, O. E., 1988. The Logic of Economic Organization, Journal of Law, Economics, and Organization, 4: 65-93
Williamson, O. E., 1991. Comparative Economic Organization: The Analysis of Discrete Structural Alternatives, Administrative Sciences Quarterly, 36: 269-296.
Williamson, O. E., 1996. The Mechanisms of Governance, Oxford University Press, New York.
Williamson, O.E., 1999. Strategy Research: Governance and Competence Perspectives, Strategic Management Journal,20 (12): 1087-1108.
Williamson, O. E., and S.G. Winter (Eds.), 1991. The Nature of the Firm, New York: Oxford University Press.
Winter, S.G., 1988. On Coase, Competence, and the Corporation, Journal of Law Economics and Organization, vol. 1
Winter, S.G., 1995. The Four R’s for Strategy: Routines, Resources, Rents and Replication, in C. Montgomery (Ed.), Resource-based and Evolutionary Theories of The Firm: Towards a Synthesis. Boston : Kluwer Academic Publishers
Zenger, T., and W. Hesterly, 1997. The Disaggregation of Corporations: Selective Intervention, High-Powered Incentives and Modular Units, Organization Science.
Figure 1: The Mortgage Banking Industry Structure – Dis-Integrated Model
[pic]
Figure 2: The Mortgage Banking Value Chain: Tasks and Layout
[pic]
Figure 3: Factors Driving the Appearance of Markets – and Evidence from Mortgage Banking
Figure 4a: Figure 4b:
Defining a Solution to the Market Problem: Defining a Solution to the Market Problem:
Two Dimensions of Effort One Dimension of Effort
Figure 5: Sources of Informational and Coordinative Superiority of Firm / Hierarchy Based Governance
Upstream Drivers Advantages of Hierarchies Moderators of these Advantages
Figure 6: The Value Chain Integration => Disintegration => Re-Integration Cycle
-----------------------
[1] White, as well as several contemporary researchers (see Porac et al, 1995), is interested in understanding the horizontal structure of the markets- that is, how markets are defined in terms of product attributes (e.g., the “sport-utility vehicle” part of the market in cars) as well as how the interaction between market players defines their profits and type of interaction. As White remarks, a market is an “act that can be ‘got together’ only by a set of producers compatibly arrayed on the qualities which consumers see in them” (1981: 519). On the other hand, we are interested in how intermediate markets get created – that is, how the car production can be permeated by new markets.
[2] While the creation of a secondary mortgage market is an interesting case of market genesis, it is also a very subtle one, as it is intricately tied to the creation of a purely financial asset. In this paper, we will not focus on securitization per se, but rather on the specialization of the mortgage market that has happened without any concomitant securitazible asset creation, to keep the setting simple. As it will become evident by the end of the paper, though, most of the principles that underlie intermediate market creation within the mortgage banking value chain also applies in the creation of the secondary market itself. Thus, an analysis of the secondary market for mortgage loans, graphically discussed in Lewis’s (1990) Liar’s Poker, could be an extension of the paper that would also provide a new twist to securitization research (cf. Allen and Santomero, 1997).
[3] To provide a sense of proportion, the secondary mortgage market last year was valued at $4.4 trillion– comparable to the US Government securities market. In terms of new mortgage loans originated, they were $1.3 trillion in 1999.
[4] Consider the similarities and differences of intermediate market creation (i.e. the dis-aggregation of the value chain) reviewed here, and horizontal market creation (i.e. the establishment of market “types” as perceived by the customers), analyzed by White (1981) or Porac et al (1995). In both cases, the social identification process plays a role – as White remarks, “markets are self-reproduced social structures… [Actors] evolve roles from observing each other’s behaviors” (1981: 518). Yet this happens in very different ways. For horizontal markets, the focus is on “producers [who] watch each other in the market” (White, 1981: 518). For intermediate markets, on the other hand, the focus is on working with the up- or down-stream side of the business, as either or both try to support a market. Identification is driven by “looking vertically”, working with the adjoining parts of the value chain, rather than by looking at peer participants. Also, while this social process is necessary, it is not sufficient to support a market; if communication cannot occur at the end of the day, a market cannot emerge, or collapses – irrespective of cognition.
[5] The “history-friendly” model in Jacobides (2000), based on computational general equilibrium theory, provides an illustration of this point. The quantitative evidence in Jacobides and Hitt (2000) supports the finding.
[6] Consider the evolution of the vertically dis-integrated production pattern in post WWII Japanese industries such as automobiles or electronics, chronicled by Nishiguchi (1994). The basic reason for which Japanese firms relied on outside contractors was not choice, but necessity. The shortage of capital, which forced industrial corporations to remain as dis-integrated as possible, also led them to create the channels, communication modes which allowed for both small-numbers outsourcing, and, later, the development of fully-fledged market.
[7] Note that this is not unique to the mortgage industry. The evolution of the PC industry, meticulously chronicled by Baldwin and Clark (2000), exhibits these very properties; the initial separation of the integrated PC into clearly separable sub-components with a standardized interface and the architectural limitation of the connections between each of these components allowed for a separation of the production and innovation process for each of the components, and this also led to a vertically dis-integrated and non-hierarchical industry structure. Therefore, the recent tendency towards modularization (Langlois, 1999; Schilling, 2000) may reduce the advantage of integrated firms, and enable intermediate markets and vertical disintegration.
[8] Asking why are not even conceivable to begin with, and looking at what enables the value chain to be broken in few and easily identifiable parts through newly emerging intermediate markets also changes our empirical focus. Looking only at the “make-vs.-buy” decision would miss the point, whereas explaining “technological discontinuities” is right on target.
[9] It should be noted that the standardization of information had some impact on the ability to mis-represent information, and thus this has helped alleviate the lemons problem. Earlier in the paper, however, we argued that standardization of information was more important as an enabler of simple communication and coordination. The question then becomes, is there a way to tell which of these two impacts of standardization mattered most? While a direct test is hard to make, there is some indirect evidence. In particular, we can look at the data on the loan types that brokers specialize in. (For details, see Jacobides and Hitt, 2000). Despite the fact that the same information type is used across loan types for banks to evaluate broker production, brokers focus mostly on the “straightforward”, plain-vanilla 15- and 30- year Fixed Rate Mortgages. (The percentage of loans generated by brokers as opposed to in-house retail production is much higher in these loan types). As information asymmetry cannot explain this, the only remaining reason is that this simplicity in standard loans facilitates coordination & communication. So we know that simplicity in communication does support (or is related to) market-based organization. For additional empirical analysis, see Jacobides and Hitt (2000).
[10] Recall that TCE problems, at least in their Williamsonian incarnation, are always created by the incentive to hold up, or otherwise take advantage of a situation – by construction and by definition.
[11] To return to our first vignette, what are these information- or coordination- related advantages of firms when compared to markets? The merit of retail production (as opposed to using the market, i.e. outside brokers) in terms of coordination & information, is (a) that information on what loans need to be produced can be updated in real time; (b) that production officers can be responsive to the needs of the HQ in the type of loans which they generate, and when they produce them; and (c) that they can adapt to the portfolio composition desires on HQ in terms of non-standard information. Brokers, on the other hand, will not respond directly to the mortgage banks daily whims; they have to be “convinced”, through attractive pricing, to provide the mortgage bank with the right types of loans (and will thus have a lower speed of compliance, as well as a disregard of any non-contractible dimension of the loans which may matter to mortgage bankers.) Additionally, retail branches have the ability to relay information more economically. However, in line with the analysis in section 1, this balance is changing and the markets are “catching up” – as a consequence of information standardization and reduction in requisite coordination.
[12] Note, however, that if a crew always travels in the same waters, with the same conditions, then perhaps some type of “market” for its services may arise if the “script” for each agent is standardized and does not change (hence requiring a skipper to reconfigure resources.) However, whenever there is a need for immediate response to change, and whenever there are externalities to the crew’s actions that cannot be managed (each of them trying to gauge, imperfectly, the best possible allocation of their effort), authority becomes indispensable.
[13] That also implies that standardization of information in goods & services, then, reduces the extent of a firm’s information throughput advantage, as the task or good at hand is simplified, and the market can cope with this more limited set of information; thus markets become more attractive as a result.
[14] However, the role of communication advantages should not be pushed too far in explaining firm boundaries and governance structures, as market-mediated transactions, especially in a ‘socialized’ conception of the market (see Granovetter, 1985), can result to such shared concepts and solid routines almost as efficiently. Interestingly, looking at communication advantages moves us away from the comparison of ideal types (firms vs. markets) to the structural examination of organizational alternatives. In particular, the distinction between anonymous markets without any communicative interaction beyond prices and quantities for given specs, vs. eponymous (often localized) markets with behavioral codes and shared meanings is a very significant one, and has not received the attention it merits, neither from economists, nor from sociologists (see Granovetter, 1985). In this sense, market-mediated communities such as the Silicon Valley (Saxenian, 1994) allow for richer information to flow across firm boundaries and hence substitute, to some extent, the need for firm-like governance structures. Likewise, firms differ with regard to their ability to promote smooth and efficient inter-firm communication.
[15] The non-disclosure clause generally protects agents without fiduciary duties to the principal (such as outside suppliers), unless expressly specified in contractual terms (Restatement [2nd] of Torts, §551; Restatement of Agency, §381; Kronman, 1978).
[16] The very noteworthiness of the Japanese supplier system, whereby suppliers accept being monitored openly by the buyer, shows how extraordinary such an informational structure is for across- firm-boundary agency relations.
[17] Argyres (1995) provides one example where a dis-integrated response to a procurement problem proved inferior. He compared GM and Ford’s procurement strategy for adopting systemic innovations. GM, by using the incentive-intensive market system, got some remarkable but incompatible technological solutions for its shop-floor automation. Ford, on the other hand, which opted for development based on in-house products, managed to arrive to a more mediocre but smoothly functioning solution for its shop-floor automation. The use of administrative oversight in order to synchronize the different parts of the automation project allowed it to reap the value from operational complementarities and externalities. The system was up and operable, and proved successful. As for GM, despite its having some more efficient components, its automation system proved a costly failure, because of its sourcing strategies, and its inability to coordinate the automation project.
[18] Additionally, as we know from experimental economics (e.g. Camerer and Knez, 1997; Weber 1999), pure coordination games are quite hard to resolve, absent authority. As Shelling (1962, 1978) has convincingly argued, authority often serves as a device to coordinate economic agents.
[19] Ironically, Hayek’s (1945) famous elegy of the market’s informational superiority is founded on the ability of the market to cope with change; yet Hayek considers the situations whereby each agent knows exactly what they can produce, so that the price is a sufficient signal to canalize their energy to the most effective aim. What we have argued so far is that in many situations, there cannot exist any such simple “price” that can guide action, and this leaves the agent powerless, and unable to be part of the “marvel of the market” – as our trimmer cannot know whether the sail is best taken in or whether the twist should be put lower down.
[20] Even Williamson seems to concede that to cope with change and adaptation there may be some benefit to hierarchical governance. As he remarks, “Some disturbances… require coordinated responses, lest the individual parts operate at cross-purposes or otherwise sub-optimize. Failures of coordination may arise because autonomous parties read and react to signals differently, even though their purpose is to achieve a timely and compatible combined response.” (Williamson, 1991: 278) Authority in firms may thus matter when firms have to change or make choices; adapt to changing situations; and coordinate (or re-configure) their activities. The more stable and repetitive the environment is, the greater the possibility of a decentralized, authority-void, market-like system to succeed. This may explain why, as Milgrom and Roberts (1992: Ch 4) observed, markets are more “brittle” than firms. Markets cannot accommodate adaptation or initiation of drastic changes, as each of the components links in a very autonomous and limited way with the others.
[21] This problem was identified, from a very different vantage point, in the economic development literature by Rosenstein-Rodan (1943), who suggested that market-based mechanisms cannot ensure the necessary coordination, thus paving the way for authority (here, the government) in coordinating activities between different sectors. Similarly, problems of coordination in markets were indirectly brought up by the literature on coordination through committees and markets (Farrell and Saloner, 1990), where economists observed that coming to the emergence of, e.g., a standard, cannot be done by relying to the market alone as markets are inefficient coordination devices.
-----------------------
Performance / effort level
Trim - Intensity
Extent of twist
Trim - Intensity
Performance / effort level
Motivational “price”
Performance f (effort)
Optimum (reached)
Performance f (effort)
Integration /
Firm Governance
Conditions
• Tight Coupling
• Need for Coordination
• Externalities between parts of the Production Process
• Lack of Standardized Information
• Too many dimensions on the output side
Increasing Dis-Integration / Intermediate Market Growth
Conditions
• Looser Coupling
• More autonomous components
• Few Externalities
• More Standardized Information
• Specific, universally understood dimensions on the output side
Conditions
• Tight Coupling, again
• …etc (due to new technology)
•
Re - Integration /
Firm Governance
New Technology / Architecture
Market unable to coordinate
Market for new tech components not developed
Learning-
willingness to trade
Standardization of Information
Modularization – coupling reduction
Creation of Market
“Language”
Information Technology
Technology
B2B Support…
Optimum- not
Reachable by the 1-dimension “market”
usually
occasionally
Coordination Advantages:
Extent of Compliance
• Externalities across parts of value chain not captured by market price
• Tight coupling leads to problems in coordinating through the market
• Specificity in desiderata – no “market” exists
Speed of Compliance
• Need for quick reaction hard through market
• Market “in real time” may not obtain
• Given bounded rationality, fiat helps to comply
Information Advantages:
Information throughput
• Tacit info hard to transmit across firm boundary
• If what is needed cannot be articulated, transaction cannot obtain due to no specs
Input Information
• Firm can know better the value of resources
• Ability to know better resources, and ability to have someone scan their best possible use=> => better allocation than through market
Communication- Shared Meaning
• “Codes”, shared meanings and short-cuts => =>Better coordination (all speak ‘one language’)
=> Economize on communication costs
Dimensionality of specs vector higher than output
i.e., market cannot convey and / or transmit all the relevant information
Bounded rationality and
Info. processing limit
i.e., individuals cannot see the best allocation of their effort or assess needs
Need for change:
=>Makes coordination needs greater
=>Authority most important there
Market Structure Dictates Architecture:
=>Division of labor prescribes limits to adaptation and coordination
Information Standardization
And Information Technology
=>Allow the market to be “as good”
=>Open networks & open standards reduce hierarchy’s advantage
Information Advantages driven by:
Legal structure
• Awards firm information rights
Common practice
• & co-location / common MIS / means to exchange information
Ability/ right to configure info flow
• … through organization design
Up- vs. Down-stream Differences:
• Different capabilities up- vs. downstream
• Different growth rates up- vs. downstream
⇨ desire to support market exchange
Case Evidence
Servicing Market: Loan production growth limit and high-end servicer success=> need for new market
Brokerage Market: Origination advantages of brokers & desire to profit from warehousing=> need for broker mkt
Historical Accident
• Enables creation of language
& market support
Case Evidence
Servicing Market: RTC rules for selling loans
Brokerage Market: 1990’s Recession leading to market based transactions
Underlying Catalyst
Enabler
Learning and Standardization
• Allows for market roles to solidify
• Provides & shapes templates
Case Evidence
Servicing Market: Loan sale templates
Brokerage Market: Growing standardization of loan sale structure
Facilitator
New Intermediate Market Creation
Information Technology & Standardization
Case Evidence
Servicing Market: MERS
Brokerage Market: Bank-broker software; XML; MISMO
Supporter
Underlying Principles:
• Reduction of Firm-based advantages in coordination
• Reduction of Firm-based advantages in information
• Changes in technology reducing “coupling”
Case Evidence
Servicing Market: New commitment methods => ability to de-couple production with servicing; independent servicers
Brokerage Market: technology-enabled information transfer to brokers, and new ways of managing brokers => ability to communicate & coordinate
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related searches
- title ix of the education amendments act of 1972
- recent arguments in the news
- recent products on the market
- the importance of knowledge quotes
- recent massacres in the us
- recent experiments in the news
- recent studies in the news
- the study of knowledge philosophy
- a recent survey of consumer search and brand choice yielded the following result
- vice president as president of the senate
- to what extent did the decade of the 1950s deserve its reputation as an age of p
- the problem of knowledge philosophy