Adam Smith’s ideas about the competitive marketplace can work
inside companies as well. Internal electronic markets can convert the widely
dispersed knowledge, preferences and beliefs of people within an organization
into decisions that can improve resource allocation, predictive abilities—and
the bottom line. By Ajit Kambil
Outlook Journal, July 2002
To read offline:
Download this article (PDF, 443K)
PDF Help 
 dam Smith’s notion of the invisible hand
helped revolutionize our understanding of how markets work. He recognized that
markets provide real incentives to traders to reveal their beliefs—and by
extension to share information—and that the price mechanism is an efficient way
to assign value collectively to beliefs and information. Yet more than two
centuries later, it is the visible hand of management that still directs most
key corporate decisions.
Today, however, technological advances are enabling executives to bring
the benefits of a competitive marketplace described by Smith inside their
companies, resulting in more efficient resource allocation and improved
predictive abilities. In a corporate world where information is widely
dispersed within organizations, internal electronic markets can be effective
ways to reveal the knowledge, true preferences and beliefs of individuals and
to convert that information into decisions that have a significant bottom-line
impact.
Industry Leader Consider British Petroleum. For years it
has been a leader in practically every aspect of the energy business, including
the environment. In 1998 BP group chief executive Sir John Browne pledged that
by 2010 the company would reduce its greenhouse gas emissions by 10 percent
from 1990 levels.
He proposed to accomplish this goal in an innovative, highly efficient
way: BP would use a flexible, internal electronic market mechanism that would
allow the company’s diverse business units to trade emissions rights with one
another. This would help the company determine the most efficient means to
reduce overall emissions.
The result: Today BP is recognized as an industry leader in the reduction
of greenhouse gas emissions. It was no small task. BP is a large, decentralized
company with more than 90 business units. The costs for reducing emissions
could vary dramatically by unit: In one it could be extremely expensive to
reduce emissions by even 2 percent, while in another it could be relatively
inexpensive to reduce emissions by 20 percent.
Managers of the individual units were in the best position to know how to
reduce emissions efficiently and economically. Each business unit was given an
emissions reduction target. They were then free to pursue those targets by
investing in emissions reduction directly or, alternatively, by purchasing
reductions in emissions from other business units that were able not only to
reduce their volume of emissions but also to exceed their targets more cheaply.
For example, if a unit had an increased need for emissions rights due to
expanding production, it could use BP’s internal electronic market to purchase
those rights from another unit that was reducing production, or it could cut
its own emissions more cheaply.
The electronic market was set up on BP’s intranet. Business units could
sell emissions rights at a specific price or could post bids if they were
looking to buy. Sales occurred when bids and offers matched. In addition to
helping business units allocate their resources efficiently, the system
generated important information, such as the cost of abating emissions. For
example, emissions rights in March 2001 traded for close to $20 per metric ton.
An internal market like BP’s requires regulation and authentication, in
much the same way external markets do. To make it work, BP first had to
estimate and then verify its prior and current emissions. Next, divisions were
told they would incur a clear and tangible financial penalty for not meeting
their emissions targets. Finally, the emissions savings that were claimed
needed to be verified to build trust in the market.
Internal markets can be used in other ways as well—forecasting sales, for
example, which is a challenge for most organizations. Sales teams have the best
knowledge of their customers, of course, including their likely purchase
patterns. Yet salespeople often receive a bonus when they exceed their quota;
they therefore have an incentive to underestimate their sales forecast. This
underestimation can create problems with production.
To generate better forecasts, Charles Plott, an experimental economist at
the California Institute of Technology, and Kay-Yut Chen, a corporate
economist, created a special system of internal markets for predicting sales at
a major computer company. This system sought to tap the collective knowledge of
the entire sales staff rather than that of a select few salespeople who served
particular customers.
Each participant was given an initial allotment of approximately 20
shares in each of 10 different markets; each market was connected to one of 10
different sales scenarios for a specific product—zero to 1,500 units of the
product sold, for example, or 1,501 to 1,600, 1,601 to 1,700, and so on. Each
market was identified by the event to which it was connected.
The markets were open for several days, during which time participants
could buy and sell shares according to their beliefs about the likely level of
sales for a particular month. When the final sales figures were posted, those
who owned shares in the stock whose “event” or face value (for example, 1,501
to 1,600 units) included the actual sales (say, 1,585 units) were paid $1 per
share.
Because the share prices ranged from zero to 100 cents, they could be
interpreted as probabilities. Thus, if a stock that represented a sales range
of zero to 1,500 units had a price of 9 cents per share, the probability that
sales for a particular month would be in the range of zero to 1,500 would be
0.09, or 9 percent. The exercise was conducted 16 times inside the company. In
all but one case, the market mechanism generated a prediction that was
significantly closer to the actual sales than the traditional sales forecast.
Valuing Unknowns There are many different ways to use
internal markets. At the Massachusetts Institute of Technology’s Sloan School
of Management, researchers are testing the use of these markets to value
different product features. Caltech’s Plott notes the potential for
wide-ranging business applications, such as predicting the value of legal
judgments. In addition to sales forecasting, internal markets can be used to
evaluate product design, new technology developments, or investments in
research and development. They could even value ventures and other uncertain
opportunities.
 Once these markets are established,
management can enhance their value by including message boards on which traders
can post the rationale for a trade. Understanding the why behind a trade can
provide management with critical insights.
Using Adam Smith’s ideas to help create internal markets can produce real
benefits for a company. But as stock market bubbles teach us, the information
and signals generated by markets are not perfect—they reflect the best social
consensus of participants at any time.
Examples in both practice and laboratory experiments indicate that
markets can and do make mistakes. To improve the effectiveness of internal
markets and the value of the information they produce, companies must ensure
that there are enough traders in their markets (a critical mass); that the
people who trade represent the group that is knowledgeable about the issues the
market seeks to address; and that there are sufficient incentives to motivate
traders to trade and generate useful information.
In addition, using these new tools effectively will require executive
commitment, leadership and the alignment of market initiatives to the culture
of the company. Not all companies will be able to do this. Internal markets
require executives to give up some of their authority to the “invisible hand”
of many decision makers. Companies that are decentralized and allow significant
autonomous decision making are much better candidates for internal markets than
those with strong hierarchies or cultures in which information sharing is not
recognized and rewarded.
As multinationals get bigger, leaders will be challenged to manage
diverse operations while achieving consistent companywide goals. Instead of
imposing a one-size-fits-all solution from the center, executives can use
internal markets. Indeed, their use may be critical to making large companies
nimble and effective competitors in a more volatile world.
Ultimately, the leadership choice for CEOs will be to decide when the
certain authority and decision making of the visible hand is better than the
freedom of choice and uncertain outcomes of the invisible hand.
 Ajit Kambil is a Boston-based associate partner and
a senior research fellow at the Accenture
Institute for High Performance Business, where he leads diverse research
initiatives in supply chains, eCommerce, innovation and venturing. A widely
published author and frequent industry speaker, Dr. Kambil is also an executive
researcher in residence at Babson College in Wellesley, Massachusetts. His
latest book, Making
Markets: How Firms Can Design and Profit From Online Auctions and
Exchanges, was published by Harvard Business School Press in
June 2002.
Back to Contents
To Top |