When we are analysing data, we often need to compare data across individuals (sectors) and time periods. When we are analysing economic data, such as revenue growth, GDP growth, we can compare the data from all previous time periods, or from a certain part of the previous time period, the annual GDP growth rate is a comparison of the first type and the quarter-on-quarter growth rate is a comparison of the latter type.
Here raises a question that how we choose the correct type of comparison. This problem is generally solved by econometrics and statistics. Econometrics tell us if we want to check the influence across different periods (such as quarterly), we need to check the continuous time series first with a number of lags. In addition, when there are some important events that could make structural changes, we might see a break between our time series.
When we read reports, sometimes we only see the comparisons on quarter-on-quarter base, such report could mislead us, as we do not know whether such analysis is reasonable, especially for those key factors. Some factors are accumulated across time, a quarter-on-quarter comparison does not make much sense. Therefore, it is necessary for us to re-do data analysis when making important decisions.
No comments:
Post a Comment