Transcript here - University of Nottingham

Comments on Session I
Paul Mizen
University of Nottingham
• I congratulate the authors on a very
detailed, well-focused set of studies.
• The key theme shows the importance of
capital flows and non-core liabilities in
the banking sector.
• I expect the papers to be highly
informative for regulatory and macroprudential policy.
Why this is important research
• A common feature of all these papers is the
view that it is unhelpful to consider EM and
DM economies separately. I agree!
• Many elements of DM crisis resembled EM
crisis experiences e.g. IR and ES housing
bubbles, withdrawal of capital etc
• Did we forget the self-fulfilling nature of
crises and become complacent about DM
debt and bubbly asset markets?
• Capital flows and growth of non-core
liabilities are critical to both EM and DM
economies.
Three Measures of Crisis
• First three papers explore the
usefulness of three types of financial
indicators:
– Market indicators of spreads, prices etc
– Ratios and gaps of balance sheet variables
– Liabilities, and their components
Let me discuss these papers first.
Three Measures of Crisis
Market indicators of spreads, prices etc tend to tell
us about current conjuncture, not future.
True, some indicators are better forward-looking
variables than others. VIX, CDS tend to be current,
but corporate bond spreads more forward looking.
Consider nature of contracts
• bonds are nominal, with fixed coupons – hence
investors explore future with care
• CDS written contracts that can be neutralised or
reversed with another contract
What about information in these spreads?
• CDS-fin v non-fin reveal exposure to sovr. Debt
• VIX may be a proxy for capital flows (direction and
reversal)
Three Measures of Crisis
Ratios of credit to GDP, core or non-core liabilities
to total liabilities offer indication of stage of the
cycle.
Agreed.
Perhaps worth looking at imbalances more generally.
• savings v. investment flows between EM and DM
• “uphill” flows of capital towards EMs
• Net foreign assets to GDP
• CA balance
• Core tier 1 capital ratio
• Private sector debt to income ratios, LTVs etc
Could these act as controls, corroboration, compl.
Info?
Three Measures of Crisis
Liabilities and the definition of core v. non-core
helpful.
Key contribution of these papers in looking at
• centralised lending decisions by global banks in
USD may lead to ‘supply push’ and rapid growth in
FCY liabilities of the banks.
• Non-financial firms acting as borrowers and the
rise of M2 deposit claims on banks in closed
economies.
Essentially these show the actions of global banks
or non-banks on the liabilities of the banking
system, creating systemic risk.
Might further consider the possibility that banks can
create large gross positions against each other.
John Moore calls these leverage stacks, and these
can also increase systemic risk.
Paper 1
Good to see evidence of
• Procyclicality of bank liability elasticities and corenon-core ratio with GDP
• very useful contributions to our understanding of
crises and may help us develop early warning
systems
• What about alternative sources of funds? US $ loans
v. dom LCY loans? What about bond and equity
funding? Swaps market?
Paper 2
Specifically on prediction of crises:
- Role for contagion, therefore consider neighbours,
close trading partners, countries with similar financial
characteristics
- Crises can jump between sectors e.g. currency crisis
to banking crisis, or vice versa. Can you find a causal
relationship between these different crisis measures?
Paper 3
Specifically on non-financial corporates:
- Changing relationship between USD loans and RMB
deposits in Hong Kong very significant
- Similarly, the development of the ‘dim sum’ bond
market is important for Chinese banks, FC Chinese
corporates e.g. property developers. Explore how
replaces expensive onshore funding for weaker
borrowers.
- Agree that non-financial corporates play a role in
transmission of funding. Two concerns: not through
banking system, firm to firm lending, and very short
term.
Papers 1,2 and 3
Some suggestions to consider:
- Financial development is likely to change the
relationships between these variables. For example,
greater use of RMB assets, and invoicing, will reduce
the significance of USD loans by global banks.
- Emergence of new (unregulated) areas of borrowing
and lending will create wider non-core definitions of
liabilities.
- Development of market finance might shift the focus
of concern away from banks and monetary
aggregates towards marketable debt instruments or
equity.
Papers 1,2 and 3
Some suggestions to consider:
- Lower bank dependence (disintermediation) over time
may reduce importance of banks.
- Greater emphasis on foreign investor participation in
onshore and offshore markets may alter financial
structure.
- Lengthening of maturity by firms may alter sensitivity
of firms to capital outflows/withdrawal of non-core
claims.
Paper 4
Questions whether there is a decline in equity home
bias for EM countries.
Evidence suggests that higher home bias in EM than in
DM countries.
But Japan is the exception, raising the possibility that
Asia has high home bias. Can you test this
hypothesis?
Clearly some countries are not financially open e.g.
China, should you drop these from your sample?
Perhaps you could refine your pooled probit by allowing
country fixed effects and an interaction term between
the home bias ratio and the country dummy.
Test equality of coefficients within groups (Asia, EM
Asia etc) using F tests.
Conclusions
Some very worthwhile work on the nature
and causes of financial instability.
The papers were impressive in their
scope and novelty.
Should provide a knowledge base on
which to base recommendations for
regulatory reform and macro-prudential
policy.