A Hidden Markov Chain Model to train Bank Nifty Futures Market Data and predict Long/Short Signals. This code takes into account the brokerage as well as transactions costs for trading future in NSE stock market.
The R code heavily relies on R libs - 'quantmod' and 'PerformanceAnalytics' as the base of financial modelling framework. At the same time, it relies on Hidden Markov Chain R lib - 'depmixS4' for modelling the HMM on actual data.
numPrevDaysForBackTest <- 100 - Number of previous days to consider for model
numCurrDaysForBackTest <- 5 - Number of previous days to consider for model
NDayLookforwardLowHigh <- 10 - Parameter used when classifing in sample data as in a trend or not
HmmLikelihoodTestLength <- 5 - How many days of data to calculate the likehood ratio on to compare models
txnCostinPerc <- 0.00057709 - Complete Transaction cost for each trade (considers both legs - long as well as short)
The code relies on fetching futures data from MySQL. Below are the Database parameters hostName <- "localhost" - Hostname of the database dbName <- "FuturesDB" - Database Name for the futures data tbName <- "FuturesIntradayTenMinTable" - Table Name for the futures data It expects the data to be laid in the column name of Date, Time, Open, High, Low, Close, Volume for OHLCV data.