Relationship lending in a financial turmoil Stefania De Mitri Giorgio Gobbi Enrico Sette
Discussion of
Double Bank Runs and Liquidity Risk Management Ippolito Peydró Polo Sette by! Enrico Perotti! ! ! ! ! ! Barcelona GSE Summer Forum June 8-9, 2015!
A solid contribution to an established empirical literature! •
Empirical banking has progressed enormously !
•
Progress in proper identification of supply and demand effects, thanks to !
•
•
Clever use of more detailed loan data !
•
Use of proxies for important elements of the lending relationship (eg distance)!
As I am both jealous of this nice paper, and I am empirically incompetent, I will offer some cheap theoretical shots.!
Double runs! •
Banks offer liquidity insurance to firm via credit lines!
•
Evidence: at the peak of the crisis, corporate run to draw on credit lines (Ivashina Scharfstein)! •
Note: I thought the US corporate run was in 2008, not 2007!
•
Interpretation: preventive strike, anticipating crunch!
•
Contribution: in 2007, banks with greater chance of becoming illiquid were run most by firms with multiple credit lines!
•
But no overall evidence of excess drawdown for these banks!
Let us say the data is OK! •
I take the evidence at face value and discuss the theoretical logic!
•
Reasonable interpretation: firms understand that bank exposure to wholesale market is a sign of vulnerability!
•
They extract liquidity first from those who may run out of it.. those with more wholesale funding !
•
But in the end, these banks are well hedged (so they unfairly targeted ?)!
•
Explanation: the borrower run is concentrated among smaller and leveraged firms !
Narrative versus theory! •
I have a feeling that these days the empirical narrative easily takes flight from theory, so some logic may be lost !
•
The paper cites Hanson Shleifer Stein and Vishny (2014) as “financial institutions with more fragile sources of funding (uninsured wholesale finance) should hold assets with lower liquidity risk.... such as credit lines”!
•
But credit lines have HIGH liquidity risk !!
•
Held by uninsured commercial counterparts, they ensure immediate contingent outflows in distress!
What kind of firms have ! multiple bank relationships?! •
We know from this literature that firms with bank relationships sometimes break away and expand borrowing from more banks!
•
In general, they must have a more credit worthy status!
•
•
In Diamond’s (1994) work, opaque or financially weak firms use banks, until they graduate to capital markets!
•
Evidence: firms that expand bank relationships have share price gains and expand borrowing, but suffer in a crunch!
Are these randomly drawn firms ?!
What kind of firms have ! multiple bank relationships?! •
We know from this literature that firms with bank relationships sometimes break away and expand borrowing from more banks!
•
In general, they must have a more credit worthy status! •
In Diamond’s (1994) work, opaque or financially weak firms use banks, until they graduate to capital markets!
•
Evidence: firms that expand bank relationships have share price gains and expand borrowing, but suffer in a crunch!
•
Are these randomly drawn firms ?!
•
What if we apply the same logic to banks?!
What kind of banks have more wholesale funding ? ! •
Those perceived as more creditworthy, or the reckless ones ?!
•
Banks who outgrew their deposit base by lending to large borrowers ?!
•
Larger ones ?!
•
In the crisis, those banks proved to be the most systemic!
•
In this sense, I think the shock of 2008 may be more relevant for these!
Another narrative! •
Here is a provocative counter argument to the credit crunch !
•
Suppose there is excessive lending to some sector, fueled by lenders attracting wholesale funding !
•
Once conditions change, more exposed banks are run, thus they cut credit!
•
If (big if) the credit crunch involves only firms in the bloated sector (real estate, trade), this just corrects excess credit !
•
The crunched firms do cut investment, but this is efficient!
•
Thus: Add controls for sectors’ past credit growth (real estate)!
On banks as liquidity providers! •
The crisis proved we do not have a simple theory of who gets run, and who gets inflows. !
•
Kashyap Rajan Stein: in economic distress banks receive inflows, so they are partially hedged to provide liquidity insurance !
•
This depends on the root of the crisis ! When banks overplay liquidity provision, they became source of outflows in a shock!
•
Other example: in a crisis, investors flee to safe government debt. But what if the crisis is driven by sovereign risk ? Can you fly to government-insured banks ?!