Economic Simulation

Download Report

Transcript Economic Simulation

Theo Gutman-Solo
How to achieve this?
 Integrate current economic theories
 Elementary Macroeconomics
 Paul Krugman’s New Trade Theory
 Joseph Stiglitz’s theories on asymmetric information
 Agent-based modeling display on rectangular
cartograms
Economic Assumptions
 Assumptions made in model
 Sum game

Productivity and wealth is relative resulting in a sum game
and “wealth” cannot be created
 Bias towards supply & demand
 Efficient Market

Inaccurate assumption but required for complexity issues
 Minimal Transportation and procedural costs
Methodology
 Generate random world
 Evolve world
 Each Timestep


Business
 Choose stations of production and marketing
 Take money from marketing country and give money to
country of production
 Difference (profit) is given to country of incorporation.
Countries
 Reevaluate status (GDP, CPI)
Methodology - Modeling
 Agents
 Represent

Business


Contain countries of
 Production
 Incorporation
 Marketing
Countries

Contain
 CPI
 GDP
Languages
 Java
 C++
Heuristic
How countries choose locales of
production, incorporation, and
marketing
 Incorporation
 Random, assigned at creation
of world
 Production & marketing
 Each timestep weights
countries by potential profit
and chooses best
combination
Methodology - Display
 The issue
 The code generates
immense amount of
data.
 Resolution show real
time evolution of
country wealth using
cartograms
Cartograms
Cartograms are maps in which
country size is dependent on a
particular variable. They are used
as an efficient method to
communicate information on
relative country data
In this project a rectangular
cartogram scaled to GDP will be
used.
nD
2
Quarter Developments
 Problems
 Random Skews in nation wealth
Error in evaluation process
 Bias towards supply and
Demand
 Model lacks stable country
development
 Causes inordinate
fluctuations


Solutions
 Integrate new trade theory
 Dynamic business agents
 Scenario
 I began accumulation of
data to be able to make
model based on real
world statistics rather
than randomly
generated world
Possible Future Developments
 Economic
 Incorporation could be
skewed to wealthier nations
 Include liquidity bias scalar
 Movement of production of
marketing should not be
instant
 Include mergers, acquisitions
and conglomerates in agent
action.
 Programming
 Agent setup is quite
wasteful, changing
information storing
procedure could allow
for more accurate
simulation