jaymon0703’s gists

jaymon0703 / CSCV3.Rmd

Last active October 6, 2018 16:08

Marcos Lopez de Prado - CSCV_3 program from Financial Charlatanism paper- http://www.ams.org/notices/201405/rnoti-p458.pdf

	---
	title: A replication of the Practical Application section in 'The Probability of Backtest
	Overfitting' - Bailey et al.
	output:
	html_notebook:
	# html_document:
	# df_print: paged
	---

	In their paper "The Probability of Backtest

jaymon0703 / random_strategy_builder.R

Last active October 12, 2017 21:59

In preparation of my next blog post about txnsim, i would like to compare the txnsim output for some standard TTR backtests with the txnsim output for a purely randomised strategy. To do this, it makes sense to build a random strategy builder based on user specified stylized facts sampled from predefined distributions. Thanks Brian Peterson for …

	require(blotter)

	# Remove portfolio and account data if run previously
	try(rm("portfolio.txnsim_rnorm_port","account.txnsim_rnorm_acct",pos=.blotter), silent = TRUE)

	# load the example data
	currency("USD")
	stock("GSPC",currency="USD",multiplier=1)
	getSymbols('^GSPC', src='yahoo', index.class=c("POSIXt","POSIXct"),from='1998-01-01')

jaymon0703 / profit.hurdle.R

Last active August 14, 2017 18:58

This is the Profit Hurdle implementation in R from Harvey & Liu's 'Backtesting' paper, Exhibit 6. It works as expected, so putting the code here for safekeeping until i complete the roxygen documentation and create a branch for issue 66 in braverock/quantstrat

	###############################################################################
	# R (http://r-project.org/) Quantitative Strategy Model Framework
	#
	# Copyright (c) 2009-2017
	# Peter Carl, Dirk Eddelbuettel, Brian G. Peterson, Jeffrey Ryan, and Joshua Ulrich
	#
	# This library is distributed under the terms of the GNU Public License (GPL)
	# for full details see the file COPYING
	#
	# $Id$

jaymon0703 / perTradeStats_InitQty_EndQty.R

Created March 9, 2017 19:56

perTradeStats with added columns Init.Qty and End.Qty

	#' calculate trade statistics for round turn trades.
	#'
	#' One 'trade' is defined as a series of transactions which make up a 'round turn'.
	#' It may contain many transactions. This function reports statistics on these
	#' round turn trades which may be used on their own or which are also used
	#' by other functions, including \code{\link{tradeStats}} and \code{\link{tradeQuantiles}}
	#'
	#' @details
	#' Additional methods of determining 'round turns' are also supported.
	#'

jaymon0703 / perTradeStats_with_Increased.to.Reduced.R

Last active January 24, 2017 19:02

perTradeStats with "increased.to.reduced" assuming FIFO to line up start and end dates for round turn trades

	#' calculate flat to flat per-trade statistics
	#'
	#' One 'trade' is defined as the entire time the symbol is not flat.
	#' It may contain many transactions. From the initial transaction that
	#' moves the position away from zero to the last transaction that flattens the
	#' position is all one 'trade' for the purposes of this function.
	#'
	#' This is sometimes referred to as 'flat to flat' analysis.
	#'
	#' Note that a trade that is open at the end of the measured period will

jaymon0703 / mcsim_wCI.R

Last active January 4, 2017 18:07

	#' Monte Carlo simulate strategy results
	#'
	#' This function resamples the daily transaction, cash equity, or percent-return
	#' P&L from a portfolio of trading results. It may be applied to both real
	#' (post-trade) and backtested transactions.
	#'
	#' The general argument here is that you can compare the performance of real
	#' portfolio equity or a backtest equity curve to a sampled version of the same.
	#'
	#' We've chosen to use daily frequency for the resampling period. If your holding

jaymon0703 / Harvey & Liu - Profit Hurdle.R

Created November 4, 2016 20:12

R-view of Profit Hurdle in Harvey & Liu (2014)

	### Required returns due to testing multiplicity ------ Harvey and Liu
	### (2014): "Backtesting", Duke University

	Profit_Hurdle <- function (num_tests, num_obs, alpha_sig, vol_annual, RHO){


	###############################
	####### Parameter inputs ######

	### 'num_tests': No. of tests one allows for in multiple tests;

jaymon0703 / Haircut_SR with sample_random_multests.R

Last active November 17, 2016 19:42

Harvey and Liu - Backtesting

	require(MASS) # for mvrnorm()

	### Generate empirical p-value distributions based on Harvey, Liu and Zhu (2014) ------ Harvey and Liu
	### (2014): "Backtesting", Duke University
	sample_random_multests <- function(rho, m_tot, p_0, lambda, M_simu){

	###Parameter input from Harvey, Liu and Zhu (2014) ############
	###Default: para_vec = [0.2, 1377, 4.45890.1, 5.55080.001,M_simu]###########

	p_0 <- p_0 ; # probability for a random factor to have a zero mean

jaymon0703 / mcsim_multipleStats_trimmedHIST.R

Last active October 21, 2016 19:13

	#' Monte Carlo simulate strategy results
	#'
	#' This function resamples the daily transaction, cash equity, or percent-return
	#' P&L from a portfolio of trading results. It may be applied to both real
	#' (post-trade) and backtested transactions.
	#'
	#' The general argument here is that you can compare the performance of real
	#' portfolio equity or a backtest equity curve to a sampled version of the same.
	#'
	#' We've chosen to use daily frequency for the resampling period. If your holding

jaymon0703 / mcsim_wMultipleStats.R

Last active September 19, 2016 09:56

mcsim with Sharpe Ratio added

	#' Monte Carlo simulate strategy results
	#'
	#' This function resamples the daily transaction, cash equity, or percent-return
	#' P&L from a portfolio of trading results. It may be applied to both real
	#' (post-trade) and backtested transactions.
	#'
	#' The general argument here is that you can compare the performance of real
	#' portfolio equity or a backtest equity curve to a sampled version of the same.
	#'
	#' We've chosen to use daily frequency for the resampling period. If your holding

Jasen Mackie jaymon0703