Goldman Sachs’ report earlier this week recommending that JPMorgan be broken up has created a tremendous stir. Coming as it did after calls for a similar smack-down at BNY-Mellon and quieter, but potent demands on several other G-SIB behemoths, activist investors who see more value in big-bank parts than in a whole are firing up a break-up campaign just as the leviathans face a comparable and concerted assault from policy-makers and legislators around the world. Big banks that like their franchises just the way they are need to take this challenge very, very seriously and ready some persuasive arguments before the momentum shifts to the combined force of activist investors and populist policy-makers.

The basic idea behind Goldman’s paper is that JPMorgan would be more accretive to shareholders in three parts: a retail bank, a wholesale bank, and an asset manager. Each would duck the G-SIB surcharge – or at least its highest charge – and perhaps other costly regulatory requirements. From this, Goldman reasons that ROE must go up. So it might, at least for a week or two while activist shareholders cashed in. The real question for shareholders – and one that should also trouble policy-makers – is whether the value pop Goldman forecasts comes before long-term costs that result in far-reaching risk.

JPMorgan has long defended itself on grounds that its numerous activities create an array of synergies. These Goldman Sachs recognizes as a $6 to $7 billion income driver — an amount even JPM broken up would miss. Goldman also agrees that synergies not only provide this meaningful income boost, but also may well contribute to JPM’s most unusual position as a top performer in pretty much everything it does. Thus, if JPMorgan is broken up, Goldman thinks it quite possible the parts may well be less profitable after the break-up earnings pop fizzles not only due to lost synergies, but also less-attractive product offerings.

Whence cometh these synergies? Is it undue market clout that allows JPMorgan to force clients to take one service to get another? If so, the putative shareholder value of a break-up would come in tandem with a more transparent and fair financial market. If not, then JPMorgan’s customers would pay more for multiple services or give up some they want and, perhaps, stop using some products like hedging that make markets safer.

As a result, the first question that has to be asked about a big-bank breakup is whether lost synergy matters. If these are an unfair revenue driver in which JPMorgan revels, then let it go; if not, the fact that break-up results in artificial market design based on regulatory – not economic – factors needs carefully to be considered.

Big banks that don’t want to be broken up thus must address – carefully, not just rhetorically – how they come by these vaunted synergies. Which products go with what and why is it good for the customer or for the broader economy? Simple convenience is a reasonable part of the answer, but risks, conflicts of interest, and concentration need also to be considered if big banks are to raise a credible defense to their monolithic structures.

The second question big banks should not just answer, but also raise is what charter these new bank spin-offs assume. Goldman assumes that JPMorgan’s three severed parts – a retail and wholesale bank as well as an asset manager – would be new banking organizations. Why this must be so is not explained nor is it in any way inevitable. A wholesale bank without an insured depository is a very different entity and an asset manager similarly divorced from, say, a custody bank is again a very different company. Each of these new ventures might include an insured depository, but then why is it less of a G-SIB other than that it’s a bit smaller?

Some will surely argue that cutting off a G-SIB’s wholesale and asset-management operations from a traditional retail deposit system washes subsidies from non-traditional banking. This is the rationale for the Vickers ring-fencing in the U.K. and calls to follow suit in the EU. However, subsidy evisceration would be more likely to be true in the U.S. if an array of non-traditional insured depositories weren’t still available to wholesale banks – Goldman Sachs, anyone?

If all but the traditional bank remnant of JPMorgan is to be stripped of an insured depository, then why would the wholesale institution and asset manager choose to be bank holding companies? And, if they are not BHCs, then how deep grow the shadows? Who would regulate them how and how well does this address systemic risk? The G-SIB surcharge would disappear far faster than the Goldman Sachs’ analysis suggests since the two non-traditional entities aren’t banks, but would they then be declared SIFIs? Unless or until FSOC figures out how to handle non-bank broker-dealers and asset managers, the spun-off entities would go into the shadows. Better for shareholder, perhaps, but not so much for the rest of us.

Does this mean that JPMorgan or any of the other U.S. G-SIBs raise no risk by virtue of their heft, complexity, and concentrated market role? Far from it. It is, though, to argue that just because shareholders might get a higher return – at least for a while – from a broken-up big bank doesn’t mean policy-makers should view this as license to open up the hunting season.