trader.py

import numpy as np
import math

class Trader():
    def __init__(self, id, market, A, C):
        """
        Initialise trader with a specific amount of cash and assets.

        @param id       Unique trader id
        @param market   Reference to the market
        @param A        Initial amount of stocks
        @param C        Initial cash amount
        """
        self.id = id
        self.market = market # add market object for easy access to parameters
        self.C = [C] # initialize list for cash time series
        self.A = [A] # initialize list for stock time series
        self.a_b = 0 # quantity of stocks to buy
        self.b_i = 0 # limit buy price of trader's buy order
        self.a_s = 0 # quantity of stock offered for sale
        self.s_i = 0 # limit sell price of trader's sell order
        self.P_i = 0.5 # buy order probability
        self.in_cluster = None


    def trade_decision(self): 
        """
        Determine whether the trader wants to buy or sell at this time step.
        """
        if np.random.random() > self.P_i:
            self.market.buyers += [self]
            self.buy()
        else:
            self.market.sellers += [self]
            self.sell()


    def buy(self):
        """
        Trader performs a buy action.
        """

        # Generate buy limit price
        b_i = max(0, self.market.p[-1]*np.random.normal(self.market.mu, self.market.sigma[-1]))

        # Generate buy limit price
        if b_i == 0:
            self.a_b = 0
            self.b_i = 0
        else:
            # Determine amount of stocks to buy (spend random fraction of total cash)
            a_b = max(0, math.trunc(np.random.random() * self.C[-1] / b_i))

            # Set the buy limit price
            self.b_i = b_i
            self.a_b = a_b


    def sell(self):
        """
        Trader performs a sell action.
        """
        # Amount of stocks to sell is a random fraction of total stocks
        a_s = max(0, math.trunc(np.random.random() * self.A[-1]))

        # Generate sell limit price
        s_i = max(0, self.market.p[-1]/np.random.normal(self.market.mu, self.market.sigma[-1]))

        # Set the sell limit price and amount of stocks to cell
        self.s_i = s_i
        self.a_s = a_s


    def no_trade(self):
        """
        Called when no trade is made, in order to keep the asset and cash series consistent
        """
        self.A += [self.A[-1]]
        self.C += [self.C[-1]]