Daily Speculations

The Web Site of Victor Niederhoffer and Laurel Kenner

Dedicated to the scientific method, free markets, ballyhoo deflation, value creation and laughter. A forum for us to use our meager abilities to make the world of specinvestments a better place..

Write to us at: (not clickable).
Please include your full name, and omit attachments.

Economics of Location, by Victor Niederhoffer

The clustering of stores of the same category apparently is a universal phenomenon. In the old days, in the Moslem lands, grand viziers and sultans granted licenses to open stores of the same kind, like lighting stores in the same district. But I notice the same clustering of lighting stores, for example, on the Bowery in New York. The usual explanation is based on the economics of information. It's less costly for consumers to search, and less costly for the producers to keep up with technological change and hiring of proficient workers. Externalities also play a part with the traffic from big department stores spilling over to others in the same district, including related services like banks.

The tendency to clustering and high competition would seem to be opposite from the tendency of animal species to carve out niches where they don't have to compete for the same resources, the classic example being the different kinds of growth versus stability strategies of the plants in a rain forest, and the territoriality of lions and other carnivores. (I note in passing that the rain forest in Hilo, Hawaii, is one of the most exciting botanical gardens I have seen.) All these factors enter into the competitiveness of clusters of department stores in urban areas versus the single-purpose stores such as Home Depot and Bed Bath & Beyond.

A standard reference in this field is Harold Hotelling's "Stability and Competition," Economic Journal (1929). Peter Pashigian's Price Theory and Applications is another very good reference on this and similar questions.

To what extent are new issues in similar fields benefited by the success or lack of the predecessor? Do companies with the same alphabetical first letters or similar beginnings like "gen-" tend to perform with predictable leads and lags? Are innovations which change the traffic flow and information patterns for traditional stores especially detrimental to same? These are some of the thoughts engendered by a recent trip to Istanbul. -- Vic  (Jan. 6, 2004)

George Zachar observes:

I think the market metaphor for location is time. Fad clusters occur along a time axis [techs, biotechs, firms with "x" at the end of their names, tesobonos, the "carry trade", etc.], and historically the money on the long side is made early by momentum types and on the short side by later sell-to-the-public types. (1/5/4)

David Hillman observes:

In location theory, there are 16 classic location factors: availability of raw materials, transportation, market, energy, labor, capital, technical knowledge (r&d), scale, agglomeration, demand, competition, government influence, tradition, random factors, environmental factors, and perception.

The extent to which any is important or prioritized varies by industry and/or individual company management. The chair speaks of agglomeration, or the clustering of like-types of economic activity. He also alludes to a number of the others and correctly references Hotelling's "ice cream vendor on the beach" problem as the seminal work in respect to agglomeration.

Basically, the problem is two vendors on a beach, selling identical product at an identical price, located equidistant from one another at quartile points along the beach. People are uniformly distributed along the beach, and being distance minimizers, will patronize the closest vendor. Thus, the midpoint represents the market boundary between the two. Vendor A wishes to increase revenue/profit, so moves to the midpoint causing a shift in the market boundary between the two and garnering larger market share. Vendor B understands what Vendor A has done, has the same knowledge and motivation, so moves toward Vendor A, reestablishing the original midpoint and recapturing his market share. The conclusion is that in a spatial equilibrium on an isotropic plane, the optimal location for competing industry is adjacent to one's competition. The result, however, is that each vendor's revenues/profit are unaffected and consumers are less well served because they have to travel farther for product/service.

Assuming people will travel to given locations at which businesses have agglomerated, the problem then becomes one of differentiation. Take the example of fast food establishments. They tend to agglomerate. The idea is to draw maximum traffic and force the consumer to make a choice on the spot. Burgers are essentially burgers and people will pay only so much for them. Therefore, like burgers are priced within cents of one another. So, how to differentiate? Fish and chicken sandwiches. Salads and desserts. So-called breakfasts, etc. However, this leads to the classic innovator/imitator problem in economics. Differentiation works for the innovator only until the innovator appears. Differentiation then requires marketing, costs rise, prices rise, the consumer gets the same old stuff regardless of where they buy, but they pay more.

Anecdote: In the '70s, there was no hotter fast food than MCD, nor was there any greater single generator of traffic. Most fast food chains employed vast armies of location and real estate analysts. Not Pizza Hut. Their location policy was this: find the most suitable parcel of real estate as close as possible to a MCD, buy and build. Why? MCD genned huge amounts of traffic and Pizza Hut sold no burgers. They competed tangentially with MCD, etc. for fast food dollars, but were differentiated by selling pizza. Let MCD gen the traffic, then appeal to the consumer to make the decision as to whether a cheesy cardboard pizza wouldn't be a tasty alternative to a mystery meat sandwich.

Another problem arises when an intervening opportunity is presented, i.e., one can get the same product along the route a little closer to home. Take the example of hospitals. In most major cities, hospitals agglomerated in a mal-located inner-city pattern, primarily to take advantage of factors such as economies of scale, associations, and r&d, perhaps, but it's agglomeration nonetheless. In the early '80's, suburban outposts in the form of doc-in-the-boxes began to appear. These were the harbingers of things to come. The '90's has seen a migration of hospital plant to the suburbs, i.e., closer to the demand. They present an attractive intervening opportunity to those hospitals located in the city center, especially for outpatient and ancillary services. Of course, this leads to further imitation, and the pattern is now one of more suburban hospitals while inner city hospitals die off.

On open-list, Mr.Higgs cites excellent example of quick lubes moving from strip locations to residential neighborhoods. They will and they will be imitated insofar as they will be tolerated. E.g., none of us wants a hospital in our backyard until we have an emergency and we all agree there should be more prisons, just not in our neighborhoods. Accommodating commercial enterprise in residential neighborhoods is not a new thought, e.g. the corner store, but the nature of the enterprises invading another land use is changing.

Mr. Zachar mentions time in markets. The classic pattern in which economic activity develops in cities is in concentric circles. A central business district surrounded by an industrial sector and slums, then by housing, middle-class, wealthy and commuter, respectively. Over time, the cbd declines to a point, then push and pull factors create a sort of reversal and gentrification occurs. The wealthy move back to the inner city as they can afford the high economic rent, and both slums and industry migrate outward in search of more affordable space and settle on the fringe. This pattern becomes more pronounced as cities age. European cities, for example, exhibit a more mature pattern than most American cities, simply because they 're hundred of years older. But, there is no rhyme nor reason to how or when these patterns occurs, which brings us around to markets ..wait..the wheels are turning...ever-changing..what??

In markets, with some exceptions, mutual funds seem to provide a fairly clear example of agglomeration, not spatially, but theoretically. There's a bunch of them out there. They all manage money for a fee. The problem is how to differentiate? Style..capitalization, sector, value, growth, global, etc. Past performance (disclosure, disclosure), management, fees, services, etc. In the end, a lot of it is marketing the brand.

In the case of a new fund added to a family, the new fund has the benefit of proximity to existing funds and the marketing of the fund to existing fund-family clients. Do they really compete? In a 'friendly competitor' sense..not too unlike agglomerated businesses do frequently....what's good for one is often good for all. Then there are times when a fund manager leaves and new manager assumes the position. What might have been a well-performing fund may tank, or vice versa, not unlike a store "under new management." Same building, same name, same location, same product..but, for better or worse, it's just not the same. A poorly managed business in the best location will fail eventually.

In stocks, we might want to think of sectors coming into and falling out of favor, or the ebb and flow of cash between the bond market and stocks, much as we do of the rise and decline of quarters of a city or an agglomeration of retailers therein. Decent companies with decent performance may suffer simply because others in the sector are not doing well.

Online brokers present attractive intervening opportunities, i.e., alternatives, to traditional brokers.

The chair asks about new issues and success based upon predecessor performance. Internet and biotech IPO's of late '90's come to mind. But, it certainly wasn't the operating performance, it was the price performance of the IPO itself.

Perhaps these few thoughts on market parallels to location factors spawn a few more.

Those interested in the basics of central place theory can find same in the works of Alfred Weber and Walter Christaller. For a more modern look at the complexities of something called regional science, a hybrid of economics, geography, math and operations management, see Brian J.L. Berry, the most cited geographer of the last fifty years. Berry, and no one does contemporary location theory better, developed the concept of threshold (minimum population required to support) and range (maximum distance traveled) to define central places offering goods and services. And, importantly, he TESTED it SCIENTIFICALLY using EMPIRICAL data.

One last thing: Hotelling was a mathematician, not a geographer, so let's chalk one up for the countists. (1/6/4)