FOSS Trading

Updated Tactical Asset Allocation Results

2010-02-06T22:35:00.002-06:00

In November, I used the strategy in Mebane Faber's Tactical Asset Allocation paper to provide an introduction to blotter. Faber has updated the strategy's results through the end of 2009. For those interested, he expands on the paper in his book, The Ivy Portfolio.

R/Finance 2010: Registration Open

2010-02-05T15:46:00.001-06:00

As posted by Dirk Eddelbuettel on R-SIG-Finance:

R / Finance 2010: Applied Finance with R
April 16 & 17, Chicago, IL, US

The second annual R / Finance conference for applied finance using R, the premier free software system for statistical computation and graphics, will be held this spring in Chicago, IL, USA on Friday April 16 and Saturday April 17.

Building on the success of the inaugural R / Finance 2009 event, this two-day conference will cover topics as diverse as portfolio theory, time-series analysis, as well as advanced risk tools, high-performance computing, and econometrics. All will be discussed within the context of using R as a primary tool for financial risk management and trading.

Invited keynote presentations by Bernhard Pfaff, Ralph Vince, Mark Wildi and Achim Zeileis are complemented by over twenty talks (both full-length and 'lightning') selected from the submissions. Four optional tutorials are also offered on Friday April 16.

R / Finance 2010 is organized by a local group of R package authors and community contributors, and hosted by the International Center for Futures and Derivatives [ICFD] at the University of Illinois at Chicago.

Conference registration is now open. Special advanced registration pricing is available, as well as discounted pricing for academic and student registrations.

More details and registration information can be found at the website at

                      http://www.RinFinance.com

For the program committee:

    Gib Bassett, Peter Carl, Dirk Eddelbuettel, John Miller,
    Brian Peterson, Dale Rosenthal, Jeffrey Ryan

I hope to meet some of you there!

LSPM with snow

2010-01-10T17:01:00.003-06:00

My last post provided examples of how to use the LSPM package. Those who experimented with the code have probably found that constrained optimizations with horizons > 6 have long run-times (when calc.max >= horizon).

This post will illustrate how the snow package can increase the speed of the probDrawdown and probRuin functions on computers with multiple cores. This yields nearly linear improvements in run-times relative to the number of cores. (Improvements are nearly linear because there is overhead in setting up the cluster and communication between the nodes.)

The first optimization takes 346 seconds on my 2.2Ghz Centrino, while the second optimization (with snow) takes 193 seconds... nearly a 45% improvement.

# Load the libraries
library(LSPM)
library(snow)

# Create a Leverage Space Portfolio object
trades <- cbind(
c(-150,-45.33,-45.33,rep(13,5),rep(79.67,3),136),
c(253,-1000,rep(-64.43,3),253,253,448,rep(-64.43,3),253),
c(533,220.14,220.14,-500,533,220.14,799,220.14,-325,220.14,533,220.14) )
probs <- c(rep(0.076923077,2),0.153846154,rep(0.076923077,9))
port <- lsp(trades,probs)

# Optimization using one CPU core
system.time({
res1 <- optimalf(port,probDrawdown,0.1,DD=0.2,horizon=5,control=list(NP=30,itermax=100))
})

# Create snow socket cluster for both cores
clust <- makeSOCKcluster(2)

# Optimization using both CPU cores
system.time({
res2 <- optimalf(port,probDrawdown,0.1,DD=0.2,horizon=5,snow=clust,control=list(NP=30,itermax=100))
})

# Stop snow cluster
stopCluster(clust)

LSPM Examples

2010-01-02T15:35:00.001-06:00

I have received several requests for additional LSPM documentation over the past couple days and a couple months ago I promised an introduction to LSPM.

In this long-overdue post, I will show how to optimize a Leverage Space Portfolio with the LSPM package. Please use the comments to let me know what you would like to see next.

Some copious notes before we get to the code:

These examples are based on revision 26 31 from r-forge and will not work under earlier revisions (and may not work with later revisions). LSPM is still in very alpha status. Expect things to change, perhaps significantly.

These examples were run using ~~DEoptim_1.3-3 (and LSPM revision 26 depends on that version)~~ code from DEoptim_1.3-3 that has been bundled inside LSPM. We are working with the DEoptim authors to address issues with more recent versions of DEoptim. LSPM will un-bundle and use the most recent version of DEoptim as soon as the issues are resolved.

The first two examples are taken from Vince, Ralph (2009). The Leverage Space Trading Model. New York: John Wiley & Sons, Inc. The results will not match the book because of differences between optimization via DEoptim and Ralph's genetic algorithm implementation. Ralph believes his genetic algorithm is getting hung up on a local maximum, whereas DEoptim is closer to the global solution.

# Load the LSPM package

library(LSPM)

# Multiple strategy example (data found on pp. 84-87)

trades <- cbind(

 c(-150,-45.33,-45.33,rep(13,5),rep(79.67,3),136),

 c(253,-1000,rep(-64.43,3),253,253,448,rep(-64.43,3),253),

 c(533,220.14,220.14,-500,533,220.14,799,220.14,-325,220.14,533,220.14) )

probs <- c(rep(0.076923077,2),0.153846154,rep(0.076923077,9))

# Create a Leverage Space Portfolio object

port <- lsp(trades,probs)

# DEoptim parameters (see ?DEoptim)

# NP=30        (10 * number of strategies)

# itermax=100  (maximum number of iterations)

DEctrl <- list(NP=30,itermax=100)

# Unconstrainted Optimal f (results on p. 87)

res <- optimalf(port,control=DEctrl)

# Drawdown-constrained Optimal f (results on p. 137)

# Since horizon=12, this optimization will take about an hour

res <- optimalf(port,probDrawdown,0.1,DD=0.2,horizon=12,calc.max=4,control=DEctrl)

# Ruin-constrained Optimal f

res <- optimalf(port,probRuin,0.1,DD=0.2,horizon=4,control=DEctrl)

# Drawdown-constrained Optimal f

res <- optimalf(port,probDrawdown,0.1,DD=0.2,horizon=4,control=DEctrl)

Tactical asset allocation using blotter

2009-11-18T21:18:00.014-06:00

blotter is an R package that tracks the P&L of your trading systems (or simulations), even if your portfolio spans many security types and/or currencies. This post uses blotter to track a simple two-ETF trading system.

The contents of this post borrow heavily from code and comments in the "longtrend" demo script in the blotter package. Many thanks to Peter Carl and Brian Peterson for their hard work.

The first chart shows the result of holding an equal-weight portfolio of SPY and IEF from 2002-07-31 to 2009-10-31. The 2008 bear market led to a 30% drawdown in this portfolio.

The second chart shows the result of following Mebane Faber's tactical asset allocation approach using the same ETFs and time period. Though it did not perform as well as buy-and-hold through 2007, the 2008 bear market only caused a 5% drawdown for this strategy. Both observations are consistent with the conclusion in Faber's article.

Without further ado, here's the code:

# This code implements the strategy found in:

# Faber, Mebane T., "A Quantitative Approach to Tactical Asset Allocation."

# Journal of Risk Management (Spring 2007).

# The article implements a simpler version of the 200-day SMA, opting for a

# 10-month SMA because monthly data are more easily available in earlier

# periods and lower granularity should translate to lower transaction costs.

# The rules of the system are relatively simple:

# - Buy when monthly price > 10-month SMA

# - Sell and move to cash when monthly price < 10-month SMA

# 1. All entry and exit prices are on the day of the signal at the close.

# 2. All data series are total return series including dividends, updated monthly.

#    NOTE: For the purposes of this demo, we only use price returns.

# 3. Cash returns are estimated with 90-day commercial paper.  Margin rates for

#    leveraged models are estimated with the broker call rate.

#    NOTE: For the purposes of this demo, we ignore interest and leverage.

# 4. Taxes, commissions, and slippage are excluded.

# Data:

# This demo uses monthly data downloaded from Yahoo Finance for two ETFs: SPY and

# IEF.  These were chosen to illustrate the classic stock/bond asset portfolio.

# Though longer serires would be preferred, data for IEF begin in mid-2002.

# Load required libraries

library(quantmod)

library(TTR)

library(blotter) # r-forge revision 193

library(PerformanceAnalytics)

# Set initial values

initDate='2002-07-31'

endDate='2009-10-31'

initEq=100000

# Set currency and instruments

currency("USD")

stock("IEF",currency="USD",multiplier=1)

stock("SPY",currency="USD",multiplier=1)

# Load data with quantmod

print("Loading data")

symbols = c("IEF", "SPY")

getSymbols(symbols, from=initDate, to=endDate, index.class=c("POSIXt","POSIXct"))

# Adjust prices for splits/dividends (thanks pg)

#IEF = adjustOHLC(IEF)

#SPY = adjustOHLC(SPY)

# Convert data to monthly frequency (to.weekly() needs drop.time=FALSE)

IEF = to.monthly(IEF, indexAt='endof')

SPY = to.monthly(SPY, indexAt='endof')

# Set up indicators with TTR

print("Setting up indicators")

IEF$SMA = SMA(Cl(IEF), 10)

SPY$SMA = SMA(Cl(SPY), 10)

# Set up a portfolio object and an account object in blotter

initPortf(name='default', symbols=symbols, initDate=initDate)

initAcct(name='default', portfolios='default', initDate=initDate, initEq=initEq)

verbose = TRUE

# Create trades

for( i in 10:NROW(SPY) ) {

  CurrentDate=time(SPY)[i]

  equity = getEndEq(Account='default', CurrentDate)

  for( symbol in symbols ) {

    sym = get(symbol)

    ClosePrice = as.numeric(Cl(sym[i,]))

    Posn = getPosQty(Portfolio='default', Symbol=symbol, Date=CurrentDate)

    UnitSize = as.numeric(trunc((equity/NROW(symbols))/ClosePrice))

    # Position Entry (assume fill at close)

    if( Posn == 0 ) {

    # No position, so test to initiate Long position

      if( Cl(sym[i,]) > sym[i,'SMA'] ) {

        # Store trade with blotter

        addTxn('default', Symbol=symbol, TxnDate=CurrentDate,

          TxnPrice=ClosePrice, TxnQty=UnitSize, TxnFees=0, verbose=verbose)

      }

    } else {

      # Have a position, so check exit

      if( Cl(sym[i,]) < sym[i,'SMA'] ) {

        # Store trade with blotter

        addTxn(Portfolio='default', Symbol=symbol, TxnDate=CurrentDate,

          TxnPrice=ClosePrice, TxnQty=-Posn, TxnFees=0, verbose=verbose)

      }

    }

  } # End symbols loop

  # Calculate P&L and resulting equity with blotter

  updatePortf(Portfolio='default', Dates=CurrentDate)

  updateAcct(name='default', Dates=CurrentDate)

  updateEndEq(Account='default', Dates=CurrentDate)

} # End dates loop

# Buy and Hold cumulative equity

buyhold = cumprod( ( 1 + 0.5*ROC(Cl(IEF)) + 0.5*ROC(Cl(SPY)) )[-1] )

# Final values

cat('Tactical Asset Allocation Return: ',(getEndEq(Account='default', Date=CurrentDate)-initEq)/initEq,'\n')

cat('Buy and Hold Return: ',tail(buyhold,1)-1,'\n')

# Plot Strategy Summary

png(filename="20091118_blotter_strategy.png", 720, 720)

charts.PerformanceSummary(ROC(getAccount('default')$TOTAL$End.Eq)[-1],main="Tactical Asset Allocation")

dev.off()

# Plot Buy and Hold Summary
png(filename="20091118_blotter_buyhold.png", 720, 720)
charts.PerformanceSummary(ROC(buyhold)[-1],main="Buy & Hold")
dev.off()

opentick alternatives

2009-11-05T18:54:00.005-06:00

I've been getting a bit of traffic from people searching for opentick (the defunct company), so I've started a list of similar (but non-free) data providers.

I'm not affiliated with any of these vendors, and the list is in no particular order. I'll update this post as more information becomes available.

IQFeed
- features, fees, API**
- starts at $60/month*
- OS: Windows
- Same API as DTN.IQ

eSignal
- features, fees, API**
- starts at $125/month*
- OS: Windows

eoddata
- features, fees, API
- some data are free, as low as 1-minute intraday data for $19.95/month
- OS: Windows

Notes:
* plus exchange fees
** API access may cost extra

Xasax closes shop

2009-10-18T11:21:00.003-05:00

Six months after shutting down opentick completely Xasax (opentick's parent company) has followed suit.

It looks like Xasax hit funding problems in August... Inside Market Data mentions the above in this story. Here is the full story (subscription required).

Update

2009-09-22T22:37:00.000-05:00

I can't believe it's been two months since I last posted... wow, time has a way of slipping through my fingers. Here's a short list of some upcoming posts:

An introduction to LSPM -- a new R package that implements Ralph Vince's leverage space portfolio model (co-authored with Soren Macbeth).
Updated charting with quantmod.
An introduction to blotter -- a transaction-based infrastructure for trading systems and simulation, providing support for multi-asset class and multi-currency portfolios.

David Varadi's RSI(2) alternative

2009-07-19T08:38:00.003-05:00

Here's a quick R implementation of David Varadi's alternative to the RSI(2). Michael Stokes over at the MarketSci blog has three great posts exploring this indicator:

Here's the R code:

DV <- function(HLC, n=2, bounded=FALSE) {
# "HLC" is an _xts_ object with "High", "Low", and "Close"
# columns, in that order.

# This is David Varadi's alternative to the RSI(2). Calculations
# taken from the marketsci blog -- http://marketsci.wordpress.com
# Author of this implementation: Joshua Ulrich

# Calculate each day's high/low mean
hlMean <- rowMeans( HLC[,-3] )

# Calculate the running Mean of the Close divided by the
# high/low mean, then subtract 1.
res <- runMean( HLC[,3] / hlMean, n ) - 1

# If we want the bounded DV...
if(bounded) {
    # Set the range to calculated the bounded DV
    rng <- 252:NROW(res)

    # Grab the index of the unbounded results, so we can convert
    # the bounded results back to an xts object.
    indx <- index(res)

    # A simple percent rank function hack
    pctRank <- function(x,i) match(x[i], sort(coredata(x[(i-251):i])))

    # Apply the percent rank function to the coredata of our results
    res <- sapply(rng, function(i) pctRank(res, i) / 252)

    # Convert the bounded results to xts
    res <- xts(c(rep(NA,251),res), indx)
}

# Return results
return(res)
}

RSI(2) Evaluation

2009-06-28T22:02:00.012-05:00

Despite my best efforts, it's been a month since the last post of this series. The first post replicated this simple RSI(2) strategy from the MarketSci Blog using R. The second post showed how to replicate the strategy that scales in/out of RSI(2).

This post will use the PerformanceAnalytics package to evaluate the rules that scale in/out of positions. I've also provided a simple function that provides some summary statistics. There is a lot of code, so I put it at the end of the post.

Table 1 contains output from my simple trade summary function (the wins and losses are in percentages, i.e. 0.69 is 69 basis points). The short side of the rule traded more often and had a lower win rate. The short side overcame its lower win rate via much higher mean and median win/loss ratios.

Table 1: RSI(2) Trade Statistics - RSI steps = 5, Size steps = 0.25
Signal	# Trades	% Win	Mean Win	Mean Loss	Median Win	Median Loss	Mean W/L	Median W/L
-1.00	133	58	0.69	-0.44	0.53	-0.25	1.55	2.12
-0.75	173	49	0.62	-0.39	0.37	-0.25	1.60	1.48
-0.50	143	54	0.43	-0.36	0.28	-0.19	1.19	1.51
-0.25	158	56	0.21	-0.19	0.14	-0.13	1.15	1.08
0.00	1262	0	NaN	NaN	NA	NA	NaN	NA
0.25	117	53	0.26	-0.31	0.18	-0.21	0.83	0.86
0.50	137	58	0.51	-0.58	0.31	-0.35	0.87	0.89
0.75	143	62	0.88	-0.89	0.50	-0.71	0.99	0.70
1.00	119	63	1.34	-1.41	0.80	-1.11	0.95	0.71

Table 2 shows the output from the PerformanceAnalytics table.Drawdowns() function. The largest percentage drawdown occurred in late 2008, but only lasted a few weeks.

The table also shows the system is prone to drawdowns that trough quickly and take months to recover from. A week of bad trades can take months to recover from.

Table 2: RSI(2) Drawdowns - RSI steps = 5, Size steps = 0.25
From	Trough	To	Depth	Length	To Trough	Recovery
2008-10-06	2008-10-10	2008-10-28	-0.157	17	5	12
2001-08-30	2001-09-21	2002-01-23	-0.091	96	12	84
2002-07-19	2002-07-23	2002-08-20	-0.088	23	3	20
2000-03-22	2000-04-14	2000-07-05	-0.076	73	18	55
2009-02-17	2009-02-23	2009-04-27	-0.070	49	5	44
2003-03-14	2003-03-21	2003-05-09	-0.055	40	6	34
2000-10-09	2000-10-12	2000-12-06	-0.052	42	4	38
2002-08-29	2002-09-24	2002-10-10	-0.051	30	18	12
2008-01-02	2008-01-22	2008-03-11	-0.045	48	14	34
2001-04-18	2001-06-18	2001-08-10	-0.045	81	43	38

Table 3 shows the output from the PerformanceAnalytics table.DownsideRisk() function. The ratio of gain/loss deviation is encouraging. I have to defer to the PerformanceAnalytics documentation and vingettes to describe the rest of the table.

Table 3: RSI(2) Downside Risk - RSI steps = 5, Size steps = 0.25
Statistic	Return
Semi Deviation	0.0050
Gain Deviation	0.0094
Loss Deviation	0.0076
Downside Deviation (MAR=10%)	0.0099
Downside Deviation (rf=0%)	0.0092
Downside Deviation (0%)	0.0092
Maximum Drawdown	-0.1572
VaR (99%)	0.0160
Beyond VaR	0.0160
Modified VaR (99%)	0.0705

The chart below shows the output from the PerformanceAnalytics charts.PerformanceSummary() function. It shows the equity curves and drawdown from peak for the long and short sides of the strategy. The middle graph shows the *daily* returns for the combined strategy.

The code below has everything that created the results above. It also contains the same results for a modified RSI(2) strategy. The modified strategy uses RSI steps of 10 and sizing steps of 0.3 (i.e. RSI<10 -> size=1, 10<20 -> size=0.7, etc.).

# Attach packages. You can install packages via:
# install.packages(c("quantmod","TTR","PerformanceAnalytics"))
library(quantmod)
library(TTR)
library(PerformanceAnalytics)

# Pull S&P500 index data from Yahoo! Finance
getSymbols("^GSPC", from="2000-01-01")

# Calculate the RSI indicator
rsi <- RSI(Cl(GSPC), 2)

# Calculate Close-to-Close returns
ret <- ROC(Cl(GSPC))
ret[1] <- 0

# This function gives us some standard summary
# statistics for our trades.
tradeStats <- function(signals, returns) {
# Inputs:
# signals : trading signals
# returns : returns corresponding to signals

# Combine data and convert to data.frame
sysRet <- signals * returns * 100
posRet <- sysRet > 0 # Positive rule returns
negRet <- sysRet < 0 # Negative rule returns
dat <- cbind(signals,posRet*100,sysRet[posRet],sysRet[negRet],1)
dat <- as.data.frame(dat)

# Aggreate data for summary statistics
means <- aggregate(dat[,2:4], by=list(dat[,1]), mean, na.rm=TRUE)
medians <- aggregate(dat[,3:4], by=list(dat[,1]), median, na.rm=TRUE)
sums <- aggregate(dat[,5], by=list(dat[,1]), sum)

colnames(means) <- c("Signal","% Win","Mean Win","Mean Loss")
colnames(medians) <- c("Signal","Median Win","Median Loss")
colnames(sums) <- c("Signal","# Trades")

all <- merge(sums,means)
all <- merge(all,medians)

wl <- cbind( abs(all[,"Mean Win"]/all[,"Mean Loss"]),
abs(all[,"Median Win"]/all[,"Median Loss"]) )
colnames(wl) <- c("Mean W/L","Median W/L")

all <- cbind(all,wl)
return(all)
}

# This function determines position size and
# enables us to test several ideas with much
# greater speed and flexibility.
rsi2pos <- function(ind, indIncr=5, posIncr=0.25) {
# Inputs:
# ind : indicator vector
# indIncr : indicator value increments/breakpoints
# posIncr : position value increments/breakpoints

# Initialize result vector
size <- rep(0,NROW(ind))

# Long
size <- ifelse(ind < 4*indIncr, (1-posIncr*3), size)
size <- ifelse(ind < 3*indIncr, (1-posIncr*2), size)
size <- ifelse(ind < 2*indIncr, (1-posIncr*1), size)
size <- ifelse(ind < 1*indIncr, (1-posIncr*0), size)

# Short
size <- ifelse(ind > 100-4*indIncr, 3*posIncr-1, size)
size <- ifelse(ind > 100-3*indIncr, 2*posIncr-1, size)
size <- ifelse(ind > 100-2*indIncr, 1*posIncr-1, size)
size <- ifelse(ind > 100-1*indIncr, 0*posIncr-1, size)

# Today's position ('size') is based on today's
# indicator, but we need to apply today's position
# to the Close-to-Close return at tomorrow's close.
size <- lag(size)

# Replace missing signals with no position
# (generally just at beginning of series)
size[is.na(size)] <- 0

# Return results
return(size)
}

# Calculate signals with the 'rsi2pos()' function,
# using 5 as the RSI step: 5, 10, 15, 20, 80, 85, 90, 95
# and 0.25 as the size step: 0.25, 0.50, 0.75, 1.00
sig <- rsi2pos(rsi, 5, 0.25)

# Break out the long (up) and short (dn) signals
sigup <- ifelse(sig > 0, sig, 0)
sigdn <- ifelse(sig < 0, sig, 0)

# Calculate rule returns
ret_up <- ret * sigup
colnames(ret_up) <- 'Long System Return'
ret_dn <- ret * sigdn
colnames(ret_dn) <- 'Short System Return'
ret_all <- ret * sig
colnames(ret_all) <- 'Total System Return'

# Create performance graphs
png(filename="20090606_rsi2_performance.png", 720, 720)
charts.PerformanceSummary(cbind(ret_up,ret_dn),methods='none',
main='RSI(2) Performance - RSI steps = 5, Size steps = 0.25')
dev.off()

# Print trade statistics table
cat('\nRSI(2) Trade Statistics - RSI steps = 5, Size steps = 0.25\n')
print(tradeStats(sig,ret))

# Print drawdown table
cat('\nRSI(2) Drawdowns - RSI steps = 5, Size steps = 0.25\n')
print(table.Drawdowns(ret_all, top=10))

# Print downside risk table
cat('\nRSI(2) Downside Risk - RSI steps = 5, Size steps = 0.25\n')
print(table.DownsideRisk(ret_all))

# Calculate signals with the 'rsi2pos()' function
# using new RSI and size step values
sig <- rsi2pos(rsi, 10, 0.3)

# Break out the long (up) and short (dn) signals
sigup <- ifelse(sig > 0, sig, 0)
sigdn <- ifelse(sig < 0, sig, 0)

# Calculate rule returns
ret_up <- ret * sigup
colnames(ret_up) <- 'Long System Return'
ret_dn <- ret * sigdn
colnames(ret_dn) <- 'Short System Return'
ret_all <- ret * sig
colnames(ret_all) <- 'Total System Return'

# Calculate performance statistics
png(filename="20090606_rsi2_performance_updated.png", 720, 720)
charts.PerformanceSummary(cbind(ret_up,ret_dn),methods='none',
main='RSI(2) Performance - RSI steps = 10, Size steps = 0.30')
dev.off()

# Print trade statistics table
cat('\nRSI(2) Trade Statistics - RSI steps = 10, Size steps = 0.30\n')
print(tradeStats(sig,ret))

# Print drawdown table
cat('\nRSI(2) Drawdowns - RSI steps = 10, Size steps = 0.30\n')
print(table.Drawdowns(ret_all, top=10))

# Print downside risk table
cat('\nRSI(2) Downside Risk - RSI steps = 10, Size steps = 0.30\n')
print(table.DownsideRisk(ret_all))

Packages featured with Inference for R

2009-05-12T09:22:00.003-05:00

quantmod, TTR, and xts were (not so) recently featured on the Inference for R Blog. Inference for R is a Integrated Development Environment (IDE) designed specifically for R.

The post gives an example of how to easily perform advanced financial stock analysis using Inference in Excel.

I appreciate how they're making R more accessible to a general audience, even though I like a command line interface and my preferred development environment is vim. :-)

R/Finance 2009 Presentations Online

2009-05-04T10:21:00.002-05:00

Posted to the R-SIG-Finance mailing list today:

For those who missed it, the slides for the R/Finance 2009 tutorials
and presentations are now available on RinFinance.com

http://www.RinFinance.com/presentations

We want to thank everyone who traveled to Chicago to make this happen. With nearly 200 attendees, coming from 8 countries and 20+ states in the US, the conference exceeded all of our expectations. A very big thank you to our presenters for taking the time to join us, UIC's International Center for Futures and Derivatives for hosting the event, and our sponsors REvolution Computing and Microsoft for their support.

We look forward to doing it again in the future. Stay tuned...

On behalf of the committee and sponsors.

RSI(2) with Position Sizing

2009-05-01T20:58:00.010-05:00

Though it's more than two weeks later, here's the second post in the series that will demonstrate how to build, test, and implement a trading strategy with R. You can find the first post here.

The first post replicated this simple RSI(2) strategy from the MarketSci Blog. This second post will demonstrate how to replicate this strategy that scales in/out of RSI(2).

A couple notes before moving to the code:

The rsi2pos() function isn't necessary, but it provides an example of how to define a function. Plus, it enables us to test several ideas with much greater speed and flexibility.

The ifelse() function works on entire vectors at once, avoiding costly loops (loops are costly in R because it's an interpreted language). Since we can potentially modify the entire 'size' vector, we must be mindful of the order of the tests.

On to the code!

# Attach the quantmod and TTR packages.
# You can install packages via:
# install.packages(c("quantmod","TTR"))
library(quantmod)
library(TTR)

# Pull S&P500 index data from Yahoo! Finance
getSymbols("^GSPC", from="2000-01-01", to="2008-12-07")

# Calculate the RSI indicator
rsi <- RSI(Cl(GSPC),2)

# Calculate Close-to-Close returns
# (this assumes we open/close our positions
# at each day's close)
ret <- ROC(Cl(GSPC))
ret[1] <- 0

# Define our position-sizing function
rsi2pos <- function(ind, indIncr=5, posIncr=0.25) {
# Inputs:
# ind : indicator vector
# indIncr : indicator value increments/breakpoints
# posIncr : position value increments/breakpoints

# Initialize result vector
size <- rep(0,NROW(ind))

# Long
size <- ifelse(ind < 4*indIncr, (1-posIncr*3), size)
size <- ifelse(ind < 3*indIncr, (1-posIncr*2), size)
size <- ifelse(ind < 2*indIncr, (1-posIncr*1), size)
size <- ifelse(ind < 1*indIncr, (1-posIncr*0), size)

# Short
size <- ifelse(ind > 100-4*indIncr, 3*posIncr-1, size)
size <- ifelse(ind > 100-3*indIncr, 2*posIncr-1, size)
size <- ifelse(ind > 100-2*indIncr, 1*posIncr-1, size)
size <- ifelse(ind > 100-1*indIncr, 0*posIncr-1, size)

# Today's position ('size') is based on today's
# indicator, but we need to apply today's position
# to the Close-to-Close return at tomorrow's close.
size <- lag(size)

# Replace missing signals with no position
# (generally just at beginning of series)
size[is.na(size)] <- 0

# Return results
return(size)
}

# Calculate signals using the 'rsi2pos()' function
sig <- rsi2pos(rsi, 5, 0.25)

# Break out the long (up) and short (dn) signals
sigup <- ifelse(sig > 0, sig, 0)
sigdn <- ifelse(sig < 0, sig, 0)

# Calculate equity curves
eq_up <- cumprod(1+ret*sigup)
eq_dn <- cumprod(1+ret*sigdn)
eq_all <- cumprod(1+ret*sig)

# Replicate Michael's nice chart (again)
png(filename="20090430_rsi2_replication.png")
plot.zoo( cbind(eq_up, eq_dn), plot.type="single",
ylab=c("Long","Short"), col=c("green","red"),
main="RSI(2) Strategy, with Position Scaling:\n 2000-01-03 through 2008-12-07" )
dev.off()

# Calculate signals using the 'rsi2pos()' function
# with new values
sig <- rsi2pos(rsi, 10, 0.3)

# Break out the long (up) and short (dn) signals
sigup <- ifelse(sig > 0, sig, 0)
sigdn <- ifelse(sig < 0, sig, 0)

# Calculate equity curves
eq_up <- cumprod(1+ret*sigup)
eq_dn <- cumprod(1+ret*sigdn)
eq_all <- cumprod(1+ret*sig)

# Re-plot equity curves using updated values
png(filename="20090501_rsi2_updated.png")
plot.zoo( cbind(eq_up, eq_dn), plot.type="single",
ylab=c("Long","Short"), col=c("green","red"),
main="Updated RSI(2) Strategy, with Position Scaling:\n 2000-01-03 through 2008-12-07" )
dev.off()

Visual inspection of the charts seems to indicate the updated RSI(2) strategy has slightly higher returns, but more volatility and larger drawdowns. The next post will use the PerformanceAnalytics package to evaluate the volatility, drawdowns, and related metrics associated with these strategies. I will do my best to post it less than two weeks from now!

R/Finance 2009 Overview

2009-04-28T09:26:00.004-05:00

The first international R/Finance 2009 conference in Chicago, IL was a huge success! David Smith from REvolution Computing has written a great summary of the entire event. I'll take the lazy route and point you to his blog post. :)

The first international conference dedicated to the use of R in the finance industry, R/Finance 2009, was a great success. With over 150 attendees (my poor estimation skills notwithstanding), sold-out tutorials, and an outstanding lineup of invited and contributing speakers from around the world, this event really demonstrated the importance of R in the world of financial analysis.

Source:
R/Finance 2009 roundup
David Smith
Revolutions, 4/27/2009

opentick is no more

2009-04-17T14:46:00.003-05:00

After a year of "we plan to accept new subscribers shortly", opentick has shut its doors completely. As of March 20th, the opentick service is no longer available.

From opentick.com:

3/16/2009

To opentick subscribers, friends, supporters, contributors and the rest of the community...

It has been quite a journey for opentick, and for those of you who have been with us for the ride we cannot thank you enough for the support, contributions and guidance you have given us over the course of the last 5 years. We could not have come as far as we have without you.

However, we are sad to say that the time has come for us to close the doors for opentick. However, this isn't goodbye. In fact, it's a new beginning. We will be introducing a fresh service with all the bells and whistles we've been slaving on over the course of the last year under a new name, a new website and a new level of service. Check back here in the near future for more information about the forthcoming new company and service. Of course amongst all these changes, there are some things that will remain the same - we still aim to provide a reliable free market data service, with an open architecture for a wide range of software platform support.

If you are a current paying subscriber, this March billing cycle will be your final billing cycle; at your next billing date service will be terminated. If you are a delayed or historical data user not currently paying for service, your account will be deactivated as of Friday, March 20th.

Sincerely,
opentick Staff

One day, we shall come back. Yes, we shall come back. Until then, there must be no regrets, no tears, no anxieties. Just go forward in all your beliefs, and prove to us we are not mistaken in ours.

I'm not optimistic their new offering (most likely their sister company, Xasax) will be available anytime soon. The managers of the two companies sub-optimally managed expectations with opentick and potential Xasax customers are sure to keep that in mind.

At least I learned quite a bit creating the 'opentick' package, since it is now useless... perhaps Xasax will have a similar API, but I won't get my hopes up.

Related Posts:
Xasax closes shop
opentick alternatives

Testing RSI(2) with R, First Steps

2009-04-13T22:35:00.001-05:00

This is the first of a series of posts that will demonstrate how to build, test, and implement a trading strategy using my favorite FOSS, R. I chose the RSI(2) strategy because it has gotten considerable attention on trading blogs over the past 6 months.

In particular, I will be replicating and extending some of the results from Michael Stokes' excellent MarketSci Blog. This post will focus on replicating this simple RSI(2) strategy.

Without further ado, let's get to some code...

# We will need the quantmod package for charting and pulling
# data and the TTR package to calculate RSI(2).
# You can install packages via: install.packages("packageName")
# install.packages(c("quantmod","TTR"))
library(quantmod)
library(TTR)

# Pull S&P500 index data from Yahoo! Finance
getSymbols("^GSPC", from="2000-01-01", to="2008-12-07")

# Calculate the RSI indicator
rsi <- RSI(Cl(GSPC),2)

# Create the long (up) and short (dn) signals
sigup <- ifelse(rsi < 10, 1, 0)
sigdn <- ifelse(rsi > 90, -1, 0)

# Lag signals to align with days in market,
# not days signals were generated
#sigup <- Lag(sigup,1) # Use lag() to avoid Toby's error
#sigdn <- Lag(sigdn,1) # Use lag() to avoid Toby's error
sigup <- lag(sigup,1) # Note k=1 implies a move *forward*
sigdn <- lag(sigdn,1) # Note k=1 implies a move *forward*

# Replace missing signals with no position
# (generally just at beginning of series)
sigup[is.na(sigup)] <- 0
sigdn[is.na(sigdn)] <- 0

# Combine both signals into one vector
sig <- sigup + sigdn

# Calculate Close-to-Close returns
ret <- ROC(Cl(GSPC))
ret[1] <- 0

# Calculate equity curves
eq_up <- cumprod(1+ret*sigup)
eq_dn <- cumprod(1+ret*sigdn*-1)
eq_all <- cumprod(1+ret*sig)

# Replicate Michael's nice chart
plot.zoo( cbind(eq_up, eq_dn),
ylab=c("Long","Short"), col=c("green","red"),
main="Simple RSI(2) Strategy: 2000-01-02 through 2008-12-07" )

# Wait a few seconds before making next chart...
#Sys.sleep(5)

# Create a chart showing the S&P500
#chartSeries(GSPC, type="line")

# Add the total equity line
#addTA(eq_all)

My version of Michael's chart is below. Up next, scaling in/out of positions with RSI(2).

Registration for R/Finance 2009 is Open!

2009-02-20T12:20:00.000-06:00

The conference website has details on:

the agenda and speakers,
travel accommodations,
registration, and
sponsors, who made the conference possible.

Hope to see you there!

TTR_0.2 on CRAN

2009-02-15T18:51:00.000-06:00

I am happy to announce a long-overdue update to the TTR package (version 0.2) is now on CRAN.

This update represents a major milestone, as TTR useRs are no longer restricted to using matrix objects. TTR 0.2 uses xts internally, so all major time series classes are now supported.

NEW FEATURES:

- Added the zig zag indicator: ZigZag()

- Added volatility estimators/indicators: volatility(), with the following calculations
- Close-to-Close
- Garman Klass
- Parkinson
- Rogers Satchell

- Added Money Flow Index: MFI()

- Added Donchian channel: DonchianChannel()

CHANGES:

- All functions now use xts internally, adding support for all major time series classes. If try.xts() fails on the input object(s), they will be converted to a matrix and a matrix object will be returned.

- Added 'bounded' arg to stoch() and SMI(), which includes the current period in the calculation.

- Added naCheck() and implemented it in the moving average functions.

- Moved maType argument default values from function formals to function body for the following functions:
ADX, ATR, CCI, DPO, EMV, KST, MACD, RSI, TRIX, BBands, chaikinVolatility, stoch, SMI

- momentum() in CMO() no longer sets na=100

- Replaced 'na' argument in momentum() and ROC() with 'na.pad'

- Added 'multiple' argument to TDI(), allowing more user control

- getYahooData() now returns an xts object

- Added colnames to output for ADX, EMV, and CLV (for xts)

- Added unit tests using the RUnit package
- Used checkEquals on object attributes as well as values

- Removed .First.lib function and added .onLoad with package version.

BUG FIXES:

- Corrected NaN replacement in CLV()

- Corrected williamsAD(): AD=0 if C(t)=C(t-1)

- Corrected runMedian() and runMAD(). The argument controlling which type of median to calculate for even-numbered samples wasn't being passed to the Fortran routine.

- aroon() calculation starts at period n+1, instead of n

- Added NA to first element of closeLag of ATR()

- Corrected BBands() and CCI() for rowMeans use on xts objects

- Made changes to Rd files to pass R CMD check on R-devel (2.9.0)

Please do contact me with any questions, concerns, bug reports, etc.

IBrokers Featured on Quantitative Trading

2009-01-17T09:01:00.000-06:00

Jeff Ryan's IBrokers package was mentioned on Ernie Chan's blog, Quantitative Trading. Though the package is still in pre-alpha stage, it is generating quite a bit of interest.

Source:
Ernie Chan
Friday, January 16, 2009
Quantitative Trading: Algorithmic Trading Technology Update

R/Finance 2009: Applied Finance with R

2008-12-30T23:54:00.000-06:00

Call for Papers

The Finance Department of the University of Illinois at Chicago (UIC),
the International Center for Futures and Derivatives at UIC, and
members of the R finance community are pleased to announce

R/Finance 2009: Applied Finance with R

on April 24 and 25, 2009, in Chicago, IL, USA

Confirmed keynote speakers include:

Patrick Burns (Burns Statistics)
David Kane (Kane Capital)
Roger Koenker (U of Illinois at Urbana/Champaign)
David Ruppert (Cornell)
Diethelm Wuertz (ETH Zürich)
Eric Zivot (U of Washington)

We invite all users of R in Finance to submit one-page abstracts or
complete papers (in txt/pdf/doc format). We encourage papers both on
academic research topics and related to use of R by Finance practitioners.

Presenters are strongly encouraged to provide working R code to accompany
the presentation/paper. Datasets need not be made public.

Please send submissions to committee@RinFinance.com.
The submission deadline is January 31st, 2009.
Submissions will be evaluated and submitters notified via email
on a rolling basis.

Additional details about the conference will be announced as available.

For the program committee:
Gib Bassett, Peter Carl, Dirk Eddelbuettel, John Miller,
Brian Peterson, Dale Rosenthal, Jeffrey Ryan

Computational Finance with R

2008-12-05T20:39:00.000-06:00

Krishna Kumar, Jan Vecer and the great folks at REvolution computing put on a great event at the beautiful Columbia University campus on the Upper West Side of New York.

Presentations by Whit Armstrong, Anthony Brockwell, Bryan Lewis, Scott Payseur, Peter Carl, Brian Peterson, and our own Jeff Ryan made for an impressive display of the power of R in quant finance. Jeff's presentation slides can be found here.

Some great things are happening in this community, and we are certainly glad to be part of them.

Welcome to FOSS Trading

2008-09-28T15:51:00.001-05:00

This blog will highlight the development and use of free open-source software to research, test, and trade financial markets.

Meet the authors:

Joshua Ulrich is currently the author and maintainer of four R packages:

TTR - Technical Trading Rules - a suite of technical analysis functions
opentick - an R API to the opentick databases
xts - eXtensible Time Series - a time-based data class that integrates all of the current time-series classes (co-authored with Jeff Ryan)
pack - convert binary to and from formats other programs and machines can understand

Jeff Ryan is currently the author and maintainer of four R packages:

quantmod - specify, build, trade, and analyse quantitative financial trading strategies
IBrokers - provides native R access to Interactive Brokers Trader Workstation API
xts - eXtensible Time Series - a time-based data class that integrates all of the current time-series classes (co-authored with Joshua Ulrich)
Defaults - create global function defaults

As you may have guessed, most of the software we develop is in the R language/environment, but we also write code in C and Fortran for efficiency. There are several reasons we use R:

R allows for rapid prototyping, since it is a scripted language.
The list of finance-oriented R packages is large and growing.
The R-Finance community continues to grow rapidly.
A multitude of statistical routines are available in contributed packages.

We hope, with your help, this blog advances the use of free open source software in the finance / trading community!