Comments on Tang and Wu`s Trade Credit Bank Credit

Download Report

Transcript Comments on Tang and Wu`s Trade Credit Bank Credit

Comments on Tang and Wu’s
Trade Credit Bank Credit and
Financial Crises:
The Case of Taiwan
Andrew K. Rose
UC Berkeley, CEPR and NBER
Part of an Ongoing Debate
• Massive trade collapse in Great Recession
– Severe, sudden, synchronized (sectors, countries)
– Bigger collapse than GDP
• Bigger than Great Depression (Eichengreen and
O’Rourke)
– Why?
2
Potential Explanations
1. Demand for manufacturing goods collapsed
– Especially durables
• Alternatively inventories
2. Issues with supply chains
– Perhaps a new normal?
3. (Artificial) Trade Frictions rose
– Traditional Protectionism
• Evenett, Global Trade Alert
– Alternatively: trade credit
3
Work on Trade Credit
and Great Recession
• Much Evidence finds role for credit constraints
reducing trade
– Manova: credit constraints affect exports
– Amiti and Weinstein: Japanese exporters most
affected use banks most affected
– Feenstra et al: more evidence on exports
4
Still, Much Dispute
• Most find small role for trade credit in collapse
• Alternative explanations work better
– Alessandra et al
– Levchenko et al
– Eaton et al
5
Tang & Wu Contribute to Debate
• Innovations include:
1. Better data (firm-level panel)
2. Compare trade and bank credit
•
•
Substitutes or Complements?
A: Varies with recession
–
Complements in 2001; substitutes in 2008-09
3. Compare recessions
•
2001 dot-com vs. 2008-09 Great Recession
6
Lots of Strengths
• Great Motivation
• Terrific Literature Review
• Careful data work
7
Things I would do differently
1.
2.
3.
4.
Worry about Timing
Add Crisis
More Sensitivity Analysis
Miscellaneous Gripes
8
1: Timing Issues
• Choice of 2008Q4 (not 2008Q3) seems critical,
arbitrary, unwarranted
• When did the recession begin?
9
Taiwan Recession Began 2008Q3
10
Bumps in 2008Q4
(Little variation otherwise and …
big excess supply of trade credit throughout)
11
Need to Choose Dates Carefully?
• Do we Have to Choose Crisis Dates at all?
• Instead: add comprehensive set of time
effects, focus on those coefficients
12
2: Add Asian Financial Crisis
• 2001, 2008—09 recessions relevant for Taiwan
• But Asian financial crisis of 1997-98?
– (Taiwan mildly affected)
– Sample of firms here starts in 1995!
– Can add third (regional) crisis at cost of some
firms
13
3: Sensitivity Analysis
• If stick with current methodology, show
insensitivity with respect to exact dates
• Alternative: use event-study on crises
14
4: Miscellaneous Gripes
1. Simultaneity: aren’t bank credit and trade credit
potentially driven by same phenomena?
2. How important are these firms in Taiwanese
manufacturing? GDP?
3. Selection bias important? Only firms listed
throughout; no entry/exit
4. Interpolation makes me nervous; so does
removing extreme values (especially in crisis!)
5. Way too many digits/coefficients in tables
15
Generalizability
• Key Finding: 2001 and Great Recessions differ
– Completely Plausible
– But what about Future Recessions?
– Idiosyncrasies reduce policy implications
• Is Taiwan special/unusual?
16