slow and steady: drawdown behaviours in phased ...introduction methodology results summary slow and...
TRANSCRIPT
Introduction Methodology Results Summary
Slow and Steady:Drawdown Behaviours in Phased Withdrawal
Retirement Income Products
Igor Balnozan
University of New South Wales
May 2018
Supervisors:Associate Professor Anthony Asher
Scientia Professor Robert KohnProfessor Denzil Fiebig
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 1 / 23
Introduction Methodology Results Summary
Spoilers
1 Introduction
2 Methodology
3 Results
4 Summary
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 2 / 23
Introduction Methodology Results Summary
Spoilers
1 Introduction
2 Methodology
3 Results
4 Summary
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 2 / 23
Introduction Methodology Results Summary
Motivation
I Population Ageing
I Sustainability (DB to DC, Government Pensions)
I Retiree Risk Management and Wellbeing
I Phased withdrawal / account-based pension / allocated pension
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 3 / 23
Introduction Methodology Results Summary
Research Aim
Identify and explain drawdown behaviours in phased withdrawalproducts, to inform appropriate retirement income productdesign, policymaking and financial advice.
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 4 / 23
Introduction Methodology Results Summary
Data Description
I 44,000 pensioners
I Multiple super funds
I Up to 12 years of drawdowns observed
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 5 / 23
Introduction Methodology Results Summary
Spoilers
1 Introduction
2 Methodology
3 Results
4 Summary
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 5 / 23
Introduction Methodology Results Summary
Methodology
I Three ComponentsI Panel Data ModelsI Grouping and Cluster AnalysisI Categorical Regression
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 6 / 23
Introduction Methodology Results Summary
Panel Visualisation – Regular Drawdown Rate
0.00
0.05
0.10
0.15
0.20
0.25
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Reg
ular
Dra
wdo
wn
Rat
e
Account Type Legacy Modern
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 7 / 23
Introduction Methodology Results Summary
Panel Visualisation – Excess Drawdown Rate
0.00
0.05
0.10
0.15
0.20
0.25
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Exc
ess
Reg
ular
Dra
wdo
wn
Rat
e
Account Type Legacy Modern
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 8 / 23
Introduction Methodology Results Summary
Panel Visualisation – Regular Drawdown Amount
0
10000
20000
30000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Reg
ular
Dra
wdo
wn
Am
ount
Account Type Legacy Modern
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 9 / 23
Introduction Methodology Results Summary
Panel Visualisation – Account Balance (FY Start)
0
100000
200000
300000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Acc
ount
Bal
ance
at F
inan
cial
Yea
r S
tart
Account Type Legacy Modern
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 10 / 23
Introduction Methodology Results Summary
Manual Grouping – Follow Minima
0.000
0.025
0.050
0.075
0.100
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Exc
ess
Reg
ular
Dra
wdo
wn
Rat
e
Account Type Legacy Modern
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 11 / 23
Introduction Methodology Results Summary
Manual Grouping – Draw 10%
0.00
0.25
0.50
0.75
1.00
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Reg
ular
Dra
wdo
wn
Rat
e
Account Type Legacy Modern
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 12 / 23
Introduction Methodology Results Summary
Manual Grouping – Prefer Level Amount
0
10000
20000
30000
40000
50000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Reg
ular
Dra
wdo
wn
Am
ount
Account Type Legacy Modern
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 13 / 23
Introduction Methodology Results Summary
Cluster Analysis – Level Amount
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 14 / 23
Introduction Methodology Results Summary
Cluster Analysis – ‘Revise Down’
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
−10000
−5000
0
5000
10000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Cha
nge
in R
egul
ar D
raw
dow
n A
mou
nt
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 15 / 23
Introduction Methodology Results Summary
Cluster Analysis – Minimum Pre-GFC
0.000
0.025
0.050
0.075
0.100
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Exc
ess
Reg
ular
Dra
wdo
wn
Rat
e
Account Type Legacy
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 16 / 23
Introduction Methodology Results Summary
Cluster Analysis – Quickdraw
0.0
0.3
0.6
0.9
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Financial Year
Tota
l Dra
wdo
wn
Rat
e
Account Type Legacy Modern
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 17 / 23
Introduction Methodology Results Summary
Spoilers
1 Introduction
2 Methodology
3 Results
4 Summary
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 17 / 23
Introduction Methodology Results Summary
General Findings
Table: General Findings
Draw Regularly atMinimum Rates
Make AdhocDrawdowns
Draw Regularly atHigher Rates
Draw Ad-hoc atHigher Rates
Older More likely More likely Less likely More likely
Face Higher MinimumDrawdown Rates More likely — More likely Less likely
Female — — Less likely —
Larger AccountBalance — — Less likely More likely
Retired Later — — More likely —
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 18 / 23
Introduction Methodology Results Summary
Behavioural Findings
Two drawdown ‘rules’ explain > 75% of sample:I 48% used MDR as anchorI 28% liked level dollar amounts
I 35% ‘revise down’
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 19 / 23
Introduction Methodology Results Summary
Behavioural Findings
Table: Behavioural Findings
Follow MinimumDrawdown Rates
Draw Level$ Amounts
Female More likely —
Larger Account Balances More likely Less likely
Retired Later More likely —
Legacy Account More likely —
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 20 / 23
Introduction Methodology Results Summary
Implications
I PolicyI Gravity of minimum drawdown rates
I Product designI Stepped annuities OR better value
I Financial adviceI Modest annuities + liquid account
I Regulation (Prudential Standards)I Financial advice: whose best interests?I CIPRs: design and defaults
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 21 / 23
Introduction Methodology Results Summary
Spoilers
1 Introduction
2 Methodology
3 Results
4 Summary
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 21 / 23
Introduction Methodology Results Summary
Summary
I Analysed drawdown behaviour in account-based pensions
I Largely heuristic- and/or default-driven behaviour
I Opportunity to extend analysis
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 22 / 23
Introduction Methodology Results Summary
See More
I Working paper on APRA WebsiteI http://www.apra.gov.au/AboutAPRA/WorkingAtAPRA/
Documents/I%20Balnozan_Brian%20Gray_2017.pdfI Literature reviewI Technical detailsI More on panel data modelsI More pretty pictures
Igor Balnozan (UNSW) Drawdowns Working Paper May 2018 23 / 23