Download Functions of the Financial System

Asset Allocation and Monetary Policy:
Evidence from the Eurozone
Harald Hau
University of Geneva and SFI
http://www.haraldhau.com
Sandy Lai
Hong Kong University
http://www.sandylai-research.com
1
Motivation

Did low real interest rates contribute to the build-up of
risk leading up to the financial crisis?


Risk-Taking Channel of Monetary Policy




Rajan (2006, 2010); Borio and Zhu (2008); Adrian and Shin
(2010)
Banks: lowering credit standards at low real short rates (e.g.,
Maddaloni and Peydro, 2011)
Household risk taking in real estate (e.g., Taylor, 2008)
Investor risk taking in the stock market (our focus)
Asset price inflation as a side effect of an excessively
accommodating monetary policy
© Harald Hau, University of Geneva and Swiss Finance Institute
2
Identification Challenges





Hypothesis: A decrease in the real rate (nominal rate
minus inflation) creates more risk seeking and inflates
asset prices
Endogeneity of monetary policy: Use deviations of the
real rate from eurozone average to identify local
monetary policy conditions
Endogeneity of inflation rate: Use instruments, use
controls and segment local fund flows by investment
destination of the fund
Risk seeking measurement: Interpret fund flow as
revealed change about firm cash flow expectations or
risk tolerance
Quantification of the asset price effect: Compare flow
sensitive stocks to non-investable stock
© Harald Hau, University of Geneva and Swiss Finance Institute
3
Eliminate Endogeneity of Nominal Rate

Both monetary policy and portfolio investment react to changing
economic conditions

Examine real rate
variations orthogonal to
monetary policy:
Cross-sectional real rate
variations within a
currency union are (by
construction) exogenous
to monetary policy

Follows Maddaloni and
Peydro (2011)
© Harald Hau, University of Geneva and Swiss Finance Institute
4
Fund Flows as Revealed Preference Change


How to measuring local investor risk choices?
Home bias in fund investment: German investors mostly invest
through German funds, etc.


Inflow into equity fund and outflow from money market fund represent increase in
investor risk taking
Use Lipper fund data:


Eliminate Belgium,
Ireland, and
Luxemburg (large
share of non-local
investors)
Aggregate Analysis:
8 countries with fund
in- outfows using data
from 4,939 equity
funds and 1,441
money market funds
© Harald Hau, University of Geneva and Swiss Finance Institute
5
Investable and Non-Investable Stocks


How to measure effect on equity prices?
Identify equity fund inflows and use flow-return relationship.

Study return effect
relative to a local
benchmark of the 20%
most flow-insensitive
stocks (LFHI =Low
Fund Holding Index)

LFHI stocks should
still be subject to
discount rate changes,
but not to local risk
shifting into stocks.
© Harald Hau, University of Geneva and Swiss Finance Institute
6
Econometric Digression

Dynamic Panel Problem:




LSDV estimator: Fitting fixed effects does not give a consistent
estimator for small T.
“Difference” GMM:


Use lag 2 or higher for y, lag 1 or higher for x.
“System” GMM: add eq.



Fund flows have serial correlation;
Country fixed effects for flows
Use differenced instruments orthogonal to the fixed effect, Δy, Δx.
More efficient, but Δy, Δx ┴ θ.
Source: David Roodman: “How to do xtabond2”
© Harald Hau, University of Geneva and Swiss Finance Institute
7
Equity Fund Flows (Table 3)




LSDV: Biased in
small T samples
GMM: Use
lagged values as
instruments.
Results for
DGMM and
SGMM similar
Lowering the real short rate SR by 10 bps implies a 1% higher
contemporaneous local equity fund inflows (and 1.4% permanent effect).
© Harald Hau, University of Geneva and Swiss Finance Institute
8
Money Market Fund Flows (Table 4)


© Harald Hau, University of Geneva and Swiss Finance Institute
Same aggregate
flow persistence
as for equity
funds
Weaker evidence
that looser
monetary policy
conditions
(ΔSR<0)
correlates with
money market
fund outflows
9
Causality Issues: Is ΔSR the Flow Trigger?


Local short rate decrease ΔSR
(= local inflation ΔP>0) is a
result of positive AD and
negative AS shocks
Price
AD
AS
GDP
Equity flows could directly
react to the positive AD shocks
that also increase GDP and
firm profitability.
© Harald Hau, University of Geneva and Swiss Finance Institute
10
Causality Issues: Is ΔSR the Flow Trigger?
Base
Model
With Add.
Controls
Foreign
Focus


© Harald Hau, University of Geneva and Swiss Finance Institute
TEST 1: Control for GDP
growth and ΔROA
directly.
 Both variables are
better proxies for
macroeconomic
shocks than ΔSR
under nominal
rigidities
TEST 2: Use only funds
that invest mostly in
foreign equity.
 Cash flow shocks of
foreign stocks are less
likely to correlate with
the inflation rate (and
real short rate) of the
11
fund’s domicile.
Similar Flow Sensitivity to Real Rate Changes
© Harald Hau

Cash flow channel:
Fund flows into fund
with different
investment
destination should
show different real
rate sensitivity

Risk seeking
channel: Similar real
rate sensitivities are
predicted for flows
across investment
destinations
12
Robustness to Expected Short Rate Proxy
© Harald Hau
13
More Robustness
© Harald Hau
14
Equity Return Impact of Flows

Low real rates impact stock returns through



A lower discount rate (discount rate effect)
Additional equity inflows (flow effect)
Estimated flow effect in a simultaneous equation model:
 Eq.1: Identify equity flows that are triggered by ΔSR.


We can further relate
to lagged changes of short
term real interest rates. Small terms in
with k>1 are
ignored.
© Harald Hau, University of Geneva and Swiss Finance Institute
15
Equity Return Impact of Flows

Eq.2: Estimate flow effect on the return difference
between flow sensitive and flow insensitive stocks:

(Assume constant flow impact γ)

Substitution gives second equation:
with constraints:
© Harald Hau, University of Geneva and Swiss Finance Institute
16
Equity Return Impact of Flows
© Harald Hau, University of Geneva and Swiss Finance Institute

Results for the flow
equation are not
much changed
simultaneous
equation approach

Differential return
effect (flow effect) is
economically large:
10bp lower SR
implies a 2% return
increase
17
Equity Return Impact of Flows


© Harald Hau, University of Geneva and Swiss Finance Institute
Home bias reenforces link
between local
monetary
policy
conditions
and equity
return impact.
The equity
flow effect
more than
doubles: 10bp
lower SR
implies a 4%
return
increase
18
Summary




Lower real short rates are associated with investor
reallocation from money market funds to equity funds.
Fund flows reflect either changing firm cash flow
expectations (cash flow channel) or change in risk
aversion (risk seeking channel); identical flow sensitivity
across investment destinations is evidence for the latter
Equity flows from loose monetary policy conditions have
a large return effect, particularly in countries with greater
home equity bias
Risk seeking channel is Archilles’ heel of the currency
union
© Harald Hau, University of Geneva and Swiss Finance Institute
19