The Economist opposes the break up of large banks but is also critical of some of the regulatory limitations gradually being imposed on these financial institutions.  The two positions, in the messy real world, have an air of inconsistency.  

What are the arguments for leaving the size of large financial institutions untouched?  According to the Economist, there are three reasons why some favor a break up of the large banks:  there is something rotten about investment banking that infects commercial banking; there is a threat to financial stability because of the added complexity that comes with size and involvement in the securities markets; and universal banks "are a dreadful deal for investors."

Having defined the problem in a particularly inapt way, the Economist then demolishes each of its own characterization.   As it notes:  "The idea that all finance’s problems stem from the investment-banking 'casino' is a misdiagnosis."   But of course no one says that all problems in finance stem from involvement in the securities markets, only that involvement increases risk. 

What about financial stability?  The main argument is that involvement in securities markets permits commercial banks to diversify their investment portfolio, with a mix of loans and securities activities.  That is, of course, true, but it does not in any way assess the risks associated with particular types of investments.

As for the "dreadful deal" for investors, the article simply noted that this was "open to question."  Perhaps.  But isn't the issue a bit broader than that?  If we were only concerned about investors, there would be no problem with "too big to fail."  Let them fail and wipe out the equity holders.  Too big to fail is decidedly not a doctrine focused on investors but the impact of a failure on the financial system. 

Finally, the Economist did note that while "it is easy to call for banks to be carved up, it would be hellishly difficult in practice."  True.  But that goes to execution (no small matter).  It is not a commentary on the broader issue of whether a break up should occur.  Thus, the comment that they should not be broken up because "most are so large that simply slicing them in two would not solve the problem" begs the question of how deep to slice in any downsizing endeavor. 

Nonetheless, recognizing the problem of too big to fail, the Economist noted the proposal by the British Government to adopt a "ring fence" around retain and investment banking activities.  As the article noted: 

that would force the retail and investment-banking arms of universal banks to have their own capital buffers, shielding depositors from losses in the investment bank but retaining some diversification benefits. Add in new rules requiring all types of banks to hold more capital, and (crucially) efforts to impose losses on private creditors of a failing institution, and the case for radical surgery is blunted.

Put aside the logistical issues that would arise in connection with the implementation of such a scheme.  The solution of the Economist was do not downsize, but increase regulation.  

But having proposed additional regulation in place of downsizing, it did not take the Economist very long to criticize efforts to increase the regulatory framework for investment banks.  In a subsequent article, the Economist noted the inevitable consequences of increased regulation, a decline in profitability.  

regulations on capital and liquidity are starting to bite. These are reducing returns earned by banks as well as forcing them to shrink their balance-sheets and cut back on trading. Many banks are also starting to position themselves for proposed rules that are not yet in force, such as America’s Volcker rule, which aims to stop banks trading for their own account, and regulations that will shove over-the-counter derivatives, which command fat margins, onto clearing-houses and exchanges.

So regulation designed to more tightly regulate investment banking activities is starting to hurt, at least when measured by profitability.

It is a set of arguments that seek to have it both ways.  Start with the presumption that banks are in fact too big to fail.  Why?  The NYT Magazine said it nicely: 

The main lesson of Lehman’s collapse is that the response to a troubled financial system is, ultimately, determined not by technical regulation, but by politics. The F.D.I.C. can use its new powers only after receiving the consent of the Treasury secretary. And its new powers pertain only to those banks deemed systematically important, a designation determined by political appointees. So while the F.D.I.C. is working to formalize the rules governing its new powers, investment-bank lobbying has grown by nearly 60 percent since the crisis began. Bankers learned that they need to be closer than ever to politicians.

So there really are only two solutions.  One is to downsize the banks and make them small enough that they can fail without creating a political question.  The second is to increase the regulation of universal banks, not because it is good for investors, not because it makes the financial system more stable over all, but because it reduces the risk profile of universal banks and makes a failure less likely, reducing the instances of a possible failure. 

This is not an argument for either position.  Both solutions are difficult to implement and will cause plenty of problems.  It is mostly a commentary on those who seem to argue against both approaches.  There is a viable basis for doing so:  that too big to fail is a reality and should simply be accepted.  But to the extent there is concern with too big to fail, they can at least be ameliorated through adjustments in size and/or adjustments in risk.  For more thoughts on this subject, see The "Great Fall": The Consequences of Repealing the Glass-Steagall Act.