Information Technology Reference
In-Depth Information
It is not only spatially embedded markets which may limit the ubiquity of market
information. Traders in a financial market have ready access to all trading information.
However, in this case the shear quantity of information may segregate the market. The
traders incur very little cost in gathering information, instead the main cost is that of
analysis. Analysing information takes time, meaning that it may be impossible for a sin-
gle trader to study and accurately respond to all of the information within the market.
Traders are therefore likely to ignore some of the information available and fail to take it
into account when making decisions. In effect the trader will not be hearing some of the
information even though it is available in principle. One possible consequence of this
is to focus the attention of traders on a small subset of market products, leading to spe-
cialisation. There is, however, an important difference between these cases. Although a
market may be segregated in terms of information flow, trade is not as restricted as it is
in the spatially extended case.
In either of these cases, however, assumptions about centralisation of market pro-
cesses no longer hold. Different traders within the market have access to different his-
tories of bids and shouts and, potentially, a propensity to deal with particular partners
rather than others. These problems aren't necessarily limited to human traders. It is pos-
sible to conceive of markets that are sufficiently large and complex that even computer
programs would find it inefficient to analyse all information present, or consider trad-
ing with every agent in the market. Recently models have started to appear that examine
these types of problems. For instance [8] and [9] have both examined trading scenarios
that take place across networks.
This paper aims to investigate the valuation of information within distributed markets.
As has previously been described, traders in these markets will have access to different
information sources and therefore different pictures of the market state. This will be
particularly apparent if some traders are more connected than others, i.e., they have more
information sources and trading partners. These better connected traders are, on average,
likely to have a better understanding of the market than those traders who are less well
connected. This paper will first examine the advantage this inequality provides to the
better connected traders along with the effect this has on trading within the market.
The effect of this imbalance is important because to some extent the degree to which
a trader is connected can be altered by the trader itself. It is well known that resources
must be expended to gather information and that properly analysing information takes
time. In many situations it is possible for a trader to change the proportion of its re-
sources dedicated to gathering and analysing information, however, it is important to
know under which circumstances to do this. This paper will begin to investigate this
question. It will consider a market where both trade and information flow are restricted
in a manner represented by an explicit, fixed network of possible agent-agent interac-
tions. The network will govern which agents are able to communicate with each other
and, therefore, which agents are able to trade with each other. Importantly, this network
will not be complete (fully connected), i.e., some traders within the market will not
be able to communicate directly with others. In this initial work we wish to gain an
understanding of the value of information in a simple segmented market so the mar-
ket network is fixed. Trader are not permitted to change their connections during the
simulation. In future we hope to develop this system so as to better understand the cir-
