Efficiency in the Financial Industry

"The green line shows "finance income," which can be understood as a measure of the value added by firms in the financial sector... it is calculated by taking the total revenue of a firm and subtracting the cost of all goods and services purchased from other firms.

The orange line measured on the right axis is "intermediated assets," which measures the size of the financial sector as the sum of all debt and equity issued by nonfinancial firms, together with the sum of all household debt, and some other smaller categories. Back in the late 19th century, the US financial sector was roughly equal in size to GDP. By just before the Great Depression, it had risen to almost three times GDP... [and by the mid-2000s it stood] at more than 4 times GDP.

An intriguing pattern emerges here: finance income tracks intermediated assets fairly closely. In other words, the amount paid to the financial sector is more-or-less a fixed proportion of total financial assets. It's not obvious why this should be so.... Does a bank need to incur twice the costs if it issues a mortgage for $500,000 as compared to when it issues a mortgage for $250,000? Does an investment adviser need to incur twice the costs when giving advice on a retirement account of $1 million as when giving advice on a retirement account of $500,000? Shouldn't there be some economies of scale in financial services?"

That was the Conversable Economist summarizing Thomas Philippon's article "Has the US Finance Industry Become Less Efficient?"

Am I mistaken or does this apparent lack of economies of scale not follow? Yes, total financial income tracks total intermediated assets as a percent of GDP, but this neglects composition effects. 

As financing reaches more (lower income) people, it might be that the average cost for the financial industry to intermediate a dollar increases. This may be due to smaller individual transactions or greater costs to vetting applicants because of an increasing selection bias problem.

There is also a vastly greater number of financial products today than just a few years ago. An evolving set of heterogeneous offerings may be masking realized economies of scale for any single product. This is especially true if newer offerings are more complex to deliver. 

Could there be negative network effects in finance?

The analysis also does not speak to scale advantages facing individual firms, which is what we should really care about.

Other factors?  

Though I am no champion of the modern financing culture, I doubt one could get away with such an argument in any other industry. But some trends that cannot be overlooked include increasing regulations and sector concentration, which also increases financial rents... "the same people doing the same job earn around 20% more when doing that job in the finance sector rather than the non-finance sector."