Message Boards Message Boards

0
|
6798 Views
|
3 Replies
|
2 Total Likes
View groups...
Share
Share this post:

How to code prior close in financial data?

Raymond Low
Posted 3 years ago

I would like to compute the range based on the high compared to the prior close. My data:

upstP = FinancialData["UPST", {2021}];
upstH = FinancialData["UPST", "High", {2021}];
upstL = FinancialData["UPST", "Low", {2021}];
upstC = FinancialData["UPST", "Close", {2021}];

I tried DatePlus[] without success

DatePlus[upstC, -1]
upstC[[1, 1]] // Normal // Short
DatePlus[upstC[[All, 1]], -1, "Day"]

I tried TimeSeriesShift[] without success 

TimeSeriesThread[{1, -1} . # &, {upstC, TimeSeriesShift[upstC, -1]}] //
   Normal // Short

and some other variations that I can no longer remember. How would I subtract yesterday's close from today's high -- continuous???
POSTED BY: Raymond Low
3 Replies
Sort By:

... high compared to the prior close....

I am not sure what you mean and I do not know anything about stock market (and do feel a very strong aversion to this stuff!), but for the sake of the syntax:

DateListPlot[upstH/TimeSeriesShift[upstC, Quantity[1, "Days"]], 
 ImageSize -> Large, Filling -> Bottom, PlotRange -> All]

enter image description here
POSTED BY: Henrik Schachner

Because these are irregular time series, meaning that the times of our data points can have gaps between them, TimeSeriesShift will produce output that won’t align with the following business day.

I think an easier approach is to get everything in a single time series and then to use MovingMap to take the difference.

data=FinancialData["UPST", "OHLC", {2021}];
diffs=MovingMap[#Values[[-1,2]]-#Values[[1,4]]&,data,Quantity[2,"Events"]];
DateListLogPlot[diffs]

A plot of rolling differences between the day’s high and yesterday’s close form UPST

Here I use #Values[[-1,2]] to get the high of the end of the sliding window and #Values[[1,4]] to get the close of the start of the sliding window, where the window size is 2 events in the time series. You can use a similar approach to compute rolling means, crossovers, etc. Also note that the date series is preserved so you don’t need to worry about weekends, holidays, or days where there may not be any data to market disruptions, quality issues, etc.
POSTED BY: Nicholas Engelking
Raymond Low
Raymond Low
Posted 3 years ago

Thank you Nicholas,

I tried working with your code and got closer to my target. I managed a 50 day moving average, which was interesting on upst, and questionable, seems like there is a mistake here?? Unless one or two very large jumps are moving through the average causing the distortion??

data = FinancialData["UPST", "OHLC", {2021}];

diff50HL = 
  MovingAverage[
   MovingMap[#Values[[-1, 2]] - #Values[[1, 3]] &, data, 
    Quantity[2, "Events"]], 50];
DateListLogPlot[diff50HL]

enter image description here

I got the differences between the high & low, the high & close, and the low & close -- 

diffHL = MovingMap[#Values[[-1, 2]] - #Values[[1, 3]] &, data, 
   Quantity[2, "Events"]];
diffHC = MovingMap[#Values[[-1, 2]] - #Values[[1, 4]] &, data, 
   Quantity[2, "Events"]];
diffLC = MovingMap[#Values[[-1, 3]] - #Values[[1, 4]] &, data, 
   Quantity[2, "Events"]];

Threaded the three differences together to determine the maximum difference -- 

diffT = Transpose[{Flatten@diffHL["Values"], 
    Flatten@Abs[diffHC["Values"]], Flatten@Abs[diffLC["Values"]]}] ;
diffMax = Max[#] & /@ diffT ;

And then I got stuck, try as I am, hours later, I still was not able to compute a moving percentage of the maximum difference and 50 day moving average of the high-low (range). The only result I got was diffidently not what I wanted

mm = MovingAverage[diffMax/diff50HL, 1];
DateListPlot[mm]

enter image description here

Everything else I tried crashed and burned. Thank you for any help that you can provide.

POSTED BY: Raymond Low
Reply to this discussion
Community posts can be styled and formatted using the Markdown syntax.
Tag limit exceeded
Note: Only the first five people you tag will receive an email notification; the other tagged names will appear as links to their profiles.
Reply Preview
Attachments
Remove
or Discard

Group Abstract Group Abstract