# classify according to one variable and no classes

I have a problem, I think that is common but I don´t have idea about how to proceed. I am looking an elegant solution for it.

I have N groups and two variables for each group. After that I have ccreated a variable dividing the variable one by the variable two. Now I need to create groups according to this variable. In this case this variable represent the rate of sufficiency. It means how much times the amount of variable 1 is in variable 2. If it is greater than 1 we have problems, because it means that there is not enough money to cover expenses. Now I need to classify this in groups indicating the risk level, LOW, MEDIUM, HIGH.... I have generated the density of this variable and classifying vissually. BUt I dont know how to justify in an elegant manner.

FOr example I have these data set.

    V1   V2     Rate
Group 1 100  62     1.61
Group 2  80  32     2.5
Group 3  123 230    0.53
Group 4  88   145    0.60
Group 5  120   49    2.44
Group 6  102   77    1.32
Group 7  180   111    1.62


Now, since I know that if the rate is bigger than one are good I put those values with a tag LOW, then the values smaller than 1 are HIGH, but if are smaller than 0.8 are VERY HIGH because it means that the expenses are much bigger than the moeny that the company have.

But this is only logic, how to do something more statistical?