Aurix Exchange

Free End-of-Day data for US markets ?

xwong

New member
Hi

Does anyone knows where to get free End-of-Day stock's data for US markets. I would like to download it some analysis.

Thanks in advance
Xwong
 

xwong

New member
Thanks JafSlow for the link. I've downloaded the software and its pretty cool. However would like to be able to have the EOD data downloaded into Excel or csv format for my study and analysis. Chartnexus will charge a fee for using their ExpertTrader module :(
Do you know of any other sites where US EOD data can be downloaded to csv or text format ?

Best rgds
Xwong
 

nlenz

Member
Thanks JafSlow for the link. I've downloaded the software and its pretty cool. However would like to be able to have the EOD data downloaded into Excel or csv format for my study and analysis. Chartnexus will charge a fee for using their ExpertTrader module :(
Do you know of any other sites where US EOD data can be downloaded to csv or text format ?

Best rgds
Xwong

How about Yahoo! Finance website? On a stock's page there is on the left "Historical Prices" where you can choose the period over which you want the data, then click on the Get Prices button. That will bring up the quotes on the web page, after which you should scroll to the bottom of that page and click on Download to Spreadsheet - which downloads a csv file of the quotes.
 

Naddmr

Member
You also might want to install "R" from hiere:

After installing this software you are able to query EOD-Data from yahoo, google etc. and to implement some more or less sophisticated forecasts based on that data.

For starters here's an AR() forecast of the SPY based on EOD-Data made in R:

Code:
require(quantmod)
# symbolName <- "DJIA"
symbolName <- "SPY"
z <- get( getSymbols( symbolName, from="1900-01-01" ) )
names(z) <- c("Open", "High", "Low", "Close", "Volume", "Adjusted")
#
ahead <- 60
ord_start <- 200
ord_end <- 300
ord_step <- 50
histrows <- 180
ylow_gap <- 0.1
yhigh_gap <- 0.1
#####################
xmin <- nrow(z)-histrows
xmax <- nrow(z)
#
ords <- seq(ord_start, ord_end, ord_step)
colors <- colors()[10:(10+length(ords))]
#
plothist <- c(z[,"Close"][xmin:xmax])
plotmin <- min(plothist)-ylow_gap
plotmax <- max(plothist)+yhigh_gap
# 
plot.zoo(plothist, xlim=c(time(first(plothist)), time(last(plothist)) + ahead), ylim=c(plotmin,plotmax) )
#
grid(nx=10, ny=0.05)
abline(v=last(time(z)))
abline(h=last(z[,"Close"]))
lines(z[,"Close"], col="black")
legend("topleft", legend=ords, lty=c(1,1), lwd=c(2.5,2.5), col=colors)
title(symbolName)
#
# One line for each AR-Order
#
j <- 1
for(i in ords ){
	model <- ar(z[,"Close"], method='ols', aic=FALSE, order.max=i)
	pred <- predict(model, n.ahead=ahead)$pred
	lines(data.frame(seq(time(last(plothist)), time(last(plothist))+ahead-1, 1), pred), col=colors[j])
	j <- j + 1
}

The data fetched by getSymbols() and stored in the variable "z" looks like this and is easily exported:

Code:
> tail(z)
             Open   High    Low  Close    Volume Adjusted
2013-04-09 156.50 157.32 155.98 156.75 100805700   156.75
2013-04-10 157.17 158.87 157.13 158.67 135157000   158.67
2013-04-11 158.70 159.71 158.54 159.19 110046600   159.19
2013-04-12 158.68 159.04 157.92 158.80 116342000   158.80
2013-04-15 158.00 158.13 155.10 155.12 216691600   155.12
2013-04-16 156.29 157.49 155.91 157.41 147290700   157.41
>

Edit: As I assume that you're interested in backtesting and strategies as well you also might want to install the "TTR" package. ( install.packages("TTR") from the R console )
 

Attachments

  • SPY.jpg
    SPY.jpg
    25.7 KB · Views: 0

nlenz

Member
You also might want to install "R" from hiere:

After installing this software you are able to query EOD-Data from yahoo, google etc. and to implement some more or less sophisticated forecasts based on that data.

For starters here's an AR() forecast of the SPY based on EOD-Data made in R:
...
Edit: As I assume that you're interested in backtesting and strategies as well you also might want to install the "TTR" package. ( install.packages("TTR") from the R console )

That's interesting. I wonder if applying statistics to markets will give good predictions.
 

Naddmr

Member
That's interesting. I wonder if applying statistics to markets will give good predictions.

I read about the integration of R into MT4 by prof7bit(*) and I wondered too. :)

And when I finally started to do some things in R then I saw, that most of all things which I conceived as "new" were already thought of, experimented on and nicely packaged in R. :)

Here's for example a guy who replicates an analysis of the frequency of the log-returns to build an indicator.



And here's the original analysis:



(*)
 

nlenz

Member
I read about the integration of R into MT4 by prof7bit(*) and I wondered too. :)

And when I finally started to do some things in R then I saw, that most of all things which I conceived as "new" were already thought of, experimented on and nicely packaged in R. :)

Here's for example a guy who replicates an analysis of the frequency of the log-returns to build an indicator.



And here's the original analysis:



(*)

Thank you! I'm trying the Auto Regression indicator from your last link in FF post #8 and it is very nice.
 

alwaterpolo

New member
Extremely interesting. I'm a quantitative methods grad student and use r frequently. I didn't think of using it for prediction.

Thanks!
 

Aurix Exchange
Top