# Banking-networks-and-contagion
Contains practice codes for Paper1
-links between nodes reflect credit or equity relationships and netwrok is directed
newnetwork: - test data to test the working of the code; change the first line to link to other codes.
entrpy :- entropy is the function of probabilty distribution and is a measure ofthe predeictibilty of exposures.
implemented maximum entropy pordecure using the algorithm provided by Mohammad Djafari(1991)
[http://arxiv.org/pdf/physics/0111126.pdf]
using the above paper; following codes were extracted-
fin1_x.m; fin3_x.m; me_dens1.m;
me_dens2.m; me_dens3.m; ME3.m; ME2.m; ME1.m
running ME1.m ME2.m and ME3.m calls other functions.
Connections b/w banks- To obtain the maximum entropy
distribution for the number of links that a bank has against other banks, using
the results of Bianconi (2009) that for uncorrelated networks, the maximum entropy
distribution for the number of links is a Poisson distribution. Uncorrelated
networks are those where the degrees of nodes are not correlated. for international banks Watts and Strogatz(1998) small world algorithm
is adopted
Some codes like fails.m and shockfails.m have codes for Prof.Agam's codes, they are yet to be modified.