The New York Times presented a study by the Pew Forum that compares household wealth with education level for various religious affiliations within the United States.  Unfortunately, Sikhs are not included as a separate category. However, the results of the study are nonetheless interesting:

The most affluent of the major religions — including secularism — is Reform Judaism. Sixty-seven percent of Reform Jewish households made more than $75,000 a year at the time the Pew Forum on Religion and Public Life collected the data, compared with only 31 percent of the population as a whole. Hindus were second, at 65 percent, and Conservative Jews were third, at 57 percent…

The study found a very linear association between education level and household wealth:

Many factors are behind the discrepancies among religions, but one stands out. The relationship between education and income is so strong that you can almost draw a line through the points on this graph. Social science rarely produces results this clean.

Where would Sikhs fall on this chart?

As Sikhs, like Hindus, come from India, they would probably have the same cultural focus on education and thus might be located where Hindus do in this analysis.

The Pew Forum also puts forward a theory about the role that discrimination might play in why certain religious groups might not have as high of household wealth as their education level would predict:

And a few of the religions that make less than their education would suggest have largely nonwhite followings, including Buddhism and Hinduism. Pew also created a category of traditionally black Protestant congregations, and it was somewhat poorer than could be explained by education levels. These patterns don’t prove discrimination, but they raise questions.

One question about this study is whether the source of education of the respondents was accounted for.  Many professionally trained immigrants struggle when their education is not recognized in the United States, and are forced into occupations that may provide less income. 

Another factor that is not discussed in the article is the concept of joint families and how this relates to the definition of “household income”.  I believe that joint families are more common among Indian-based groups – such as Hindus and Sikhs – and perhaps this affects why the average household income of Hindus is near the top of the scale.