price_impact_function: price_impact_function

In Martin-Summer-1090/syslosseval: Provide data and functions to support the analysis in Systemic Loss Evaluation

Description

This function computes the price impact as a percentage loss in value of the different bond exposures if a certain volume is sold to the market. It needs to know the state variables, the data on bond volatilities and average daily volumes as well as the base_year for which the stress test is taken as well as the value of a constant of order unity. Finally it takes as argument a method. Currently only one method is implemented, which we call "square-root". This method implements equation (9) in the paper. For future versions we keep space to include other impact functions which have been used in the literature, in particular the functions of Cont and Schaanning 2017.

Usage

price_impact_function(
  del,
  mat,
  lb,
  data_idx,
  data_adv,
  base_year,
  constant,
  method = "squareroot"
)

Arguments

del	I times 1 vector of discount factors with each component in [0,1]
mat	A list containing the matrices which are the output of make_matrices.
lb	A real number specifiying the leverage threshold lambd_bar.
data_idx	A dataframe with price index data. It must contain the variables Country (chr), Date, and Value (num)
data_adv	A dataframe with average daily volumes. It must contain the variables Country (chr), Year (num), Volume (num), Unit (chr) and Currency (chr)
base_year	The base year for the stress test exercise
constant	The constant kappa in the square root formula for price impact
method	In our paper we use the square root law of price impact because we can derive it from evidence and the market-microstructure literature. But others, for example Schannning and Cont use different impact functions. We leave space for allowing for different methods by which impact is computed. For now we only implement the square root law.

Value

A list with An I times 1 vector of discount factors delta measuring the price impact of a quantity sold.

Examples

stress_data <- make_stress_data(eba_exposures_2016, eba_impairments_2016, 1, 2015)
matrices <- make_state_variables(stress_data)
delta <- rep(0, dim(matrices$S_0)[2])
price_impact_function(del = delta,
mat = matrices, lb = 33,
data_idx = sovereign_bond_indices,
data_adv = average_daily_volume_sovereign,
base_year = 2015,
constant = 1.5)

Martin-Summer-1090/syslosseval documentation built on Dec. 17, 2021, 3:14 a.m.