.

“WE WILL not kick you when you are down, at least not for a couple of days”: that is the gist of a putative deal struck by 18 global banks this week, which agreed not to pull abruptly out of contracts with each other if one of them hits the buffers. As modest as that may sound, regulators see it as the foundation of a firewall to halt the spread of future financial crises.

The agreement concerns derivatives, contracts whose value “derives” from the performance of an underlying asset such as a share, currency or bond. Banks use them to hedge themselves or speculate, to the chagrin of regulators who dislike how hard they are to value and how easily they can entangle financial institutions in a web of interdependency.

If a bank will benefit from invoking its right to demand early settlement of such contracts when a counterparty runs into trouble, it tends to do so, naturally enough. Lehman Brothers discovered this in 2008; bankruptcy lawyers are still untangling the mess. If everyone cuts and runs at once, however, the stricken party has to hand over cash when it can least afford it, deepening the crisis.

By contrast, a brief stay might give regulators time to right a listing bank, by forcing it to sell off still-healthy units, say. America and Europe have instituted such moratoriums since the crisis, but these do not apply to cross-border deals. Global rules which extend the principle to asset managers and others are in the pipeline.

Banks hope their voluntary fix will assuage regulators’ concerns that many of them are still “too big to fail”. Officials are increasingly anxious to find a way to close ailing financial firms without sparking a global panic. American regulators recently sent back for revision all 11 “living wills” they have received from banks, detailing how they might be wound down in a crisis. But at least one element of such a resolution is now clearer.