Consider the following stock price simulation using TensorFl…
Consider the following stock price simulation using TensorFlow. import tensorflow as tf tf.random.set_seed(0) # Stock parameters r = tf.constant(0.01) sigma = tf.constant(0.3) S0 = tf.constant(100.0) K = tf.constant(100.0) # Time discretization T = tf.constant(1.0) N = tf.constant(201.0) # Number of steps to split T dt = tf.constant(T / (N – 1)) N_int = tf.cast(N, dtype=tf.dtypes.int32) # Tranform the constant N into an int32 type. t = tf.linspace(0.0, T, N_int) # Simulating the normal shocks z = tf.random.normal(shape=[N_int – 1], mean=0., stddev=1., dtype=tf.dtypes.float32) z0 = tf.constant(0.0, shape=[1]) # We need to pass the shape here to avoid an error when concatenating z0 and z. z = tf.concat([z0, z], axis=0) # Construct the Brownian Motions Wt = tf.math.cumsum(z) * tf.sqrt(dt) lnSt = TO BE FILLED St = tf.math.exp(lnSt) StWhat should I replace TO BE FILLED with to get the desire result?
Read Details