1

I have several variables, some calculated from 1 July to 30 June and the remainder calculated from 1 January to 31 December.

Now how I can change one of them into financial or calendar year?

Nick Cox
  • 48,377
  • 8
  • 110
  • 156
SAAN
  • 531
  • 5
  • 16
  • I removed "data-visualization" tag. The simplest method is (e.g.) to calculate 2011 as average of 2010-11 and 2011-12. Much fancier methods can be imagined based on differential weighting or smoothing or modelling the data. But this can't be a new problem. The question is whether there are enough people around here to comment on what people do. You may be better off in a finance forum. – Nick Cox Jul 03 '13 at 08:10
  • I assume there must be statistical method to do this. – SAAN Jul 03 '13 at 08:21
  • Indeed: there are lots of ways to do it, depending on your data and what assumptions you are willing to make. In practice, however, it is best to find out what people do first. – Nick Cox Jul 03 '13 at 08:28
  • For variables that represent total amounts or average amounts it's essentially the same problem as [this one](http://stats.stackexchange.com/questions/59418/interpolation-of-influenza-data-that-conserves-weekly-mean), because any solution amounts to predicting the values for semiannual periods. Any of the approaches mentioned there (in comments as well as answers) would apply. – whuber Jul 03 '13 at 14:34

1 Answers1

1

The problem seems similar to non-contemporaneous trading.

See here: https://quant.stackexchange.com/questions/7650/data-synchronization

From my personal studies I think the best solution is estimating data between tick (GARCH process).

Community
  • 1
lcrmorin
  • 1,380
  • 16
  • 31