A new BIS working paper finds that syndicated interconnectedness causes relatively mild shocks, if they affect the entire network, to be more problematic than the failure of a single active member

In 2007, before the financial crisis started unfolding, syndicated loans made up 40% of cross-border funding to US companies and two-thirds of cross-border funding to emerging markets, and banks rapid withdrawal from syndicated lending appears to be one of the main channels that allowed the stress to propagate through the financial system. But the net effect that syndicated lending has on financial stability right now still isn’t well understood.

“On the one hand, syndicated lending allows banks to diversify their credit risks, while also increasing lending in aggregate. On the other hand, complementarity in bank lending decisions due to dependency on syndicate partners can be a source of contagion,” write Makoto Nirei, Julian Cabellero, and Vladyslav Shushko in the Bank for International Settlements (BIS) working paper Bank capital shock propagation via syndicated interconnectedness.

They find that syndicated loan networks can survive the loss of an active member – even the most active member – surprisingly well, but that even a mild common shock can cause major problems.

Syndicated Lending loan network 0115

Syndicated lending turns small, common shocks into a big deal

The researchers start by creating a model that simulates some of the known qualities of the syndicated loan market: it responds rapidly to stress by decreasing the amount of new loans (as opposed to decreasing their size), and these lending contractions don’t start with the loan originator who typically tries to pick up the slack by increasing its own lending. The paper also assumes that banks will respond to a capital shock by decreasing lending. Technically issuing equity and selling liquid assets are also options, but during a crisis you’re unlikely to find anyone interested in a new offering and you’d rather hang on to liquid assets if at all possible, so the assumption seems reasonable.

The first thing that the researchers found was that syndicated lending allowed even a mild, common shock to create big disruptions, but independent banks could withstand a larger common shock without nearly as much systemic instability. That may not be surprising since the whole point of syndicated lending is to increase total lending (and therefore risk in the system), but it’s still important.

A single large shock produced moderate problems

What’s more surprising is that the loss of an active member of the network doesn’t bring everyone else down. When the researchers hit the largest member of their network with a ten-sigma shock, but left everyone else alone, they saw a moderate increase in withdrawals and loan dissolutions.

“This suggests that the failure of a very active bank may not necessarily generate a large systemic event in this market given the current empirical network structure,” they write. Big, idiosyncratic risks may get more headlines, but apparently it’s the broader, if milder, risks that we have to watch out for.