Is there a rising trend in Number Series I am trying to build a stock screening utility. 
What  I am trying is to find if there is a rising trend in a time series of profit margins of a company. I know there can be dips in some years but I want to find if mathematically that overall there is a rising trend as well as towards the recent years there is higher margins as compared to previous years. 
Is there any way we can find it mathematically?
        Profit Return
Year,Company A,Company B,Company C
2005,38.97,20.44,22.01
2006,22.4,20.59,20.63
2007,22.4,20.59,20.63
2008,21.56,22.01,23.87
2009,25,20,23.52
2010,25,22.38,18
2011,30,24.13,24.85
2012,32,25.43,27.47
2013,30,20,22
2014,35,27,20

 A: Sounds like a job for correlation coefficients, or more precisely significance tests based on them, as you want to find out, whether there is a significantly positive correlation between the profit margins and time.
To briefly address the possible choices of correlation coefficients:


*

*Pearson’s r regards your time series’ absolute values but makes some assumptions regardang the nature of the correlation, namely it should be linear. Therefore you need to have some a priori knowledge that the correlation is linear or at least approximately linear (or if you have some other knowledge regarding the general functional form of your time series, you can use this to transform your data).

*Spearman’s ρ and Kendall’s τ are rank-based, i.e., they only regard the order of values and thus outliers do not have a huge impact and also no assumptions are made as to the correlation being linear. Whether this is what fits you hypothesis is something you have to decide.


As your time series seem to be comparably short, make sure that you use an implementation of these coefficients that does not use too strong approximations for small datasets.
