Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
MotivationData
Research DesignPreliminary Results
Appendix
Property Derivatives in the Strategic AssetAllocation
ERES 2009 - Doctoral Session
Bertram SteiningerIRE|BS, University of Regensburg
June 24, 2009
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Table of Contents
Motivation
DataData CollectionInterpolation
Research Design
Preliminary Results
AppendixMarket CapitalizationTransaction CostsTrading Volumes
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Motivation I
I real estate as a major asset classI real estate’s characteristics:
I stability of their valuesI opportunity to hedge against inflationI specific risk-return characteristicsI low co-movements with traditional stock and bond marketsI lot size transformationI transaction costsI no short possibility
However, insufficient asset class for individuals!?
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Motivation II
Benefits from property derivatives
I term transformation and liquidity
I transaction costs
I bridge finance and efficient leverage
I short possibility
I lot size transformation
I diversification
I alpha generating
I physical portfolio management
I no property knowledge
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Motivation III
Drawbacks from property derivatives
I price to pay
I mark to market risk
I counterparty risk
I liquidity drying up
I lack of traditional alpha
I underlying risk
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Motivation IV
UK IPD Certificate: Key Features
I underlying: IPD UK Annual Index under ”All Properties TR”
I ”100% exposure to physical UK commercial property”
I Issuer: Goldman Sachs Jersey (Limited)
I Guarantor: Goldman Sachs Europe and The Goldman SachsGroup, Inc. (A, A1, A+; outlook: -, -, )
I minimum investment: GBP 10.00
I issue date: 26 June 2006
I expiry date: 31 March 2011
I liquidity: continuously quoted on the LSE
I fixed leg: 2.80% p.a.
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Motivation V
Studies suggest a optimal allocation of real estate in a mixed assetportfolio of 5-25% (for an overview see e.g. Hoesli, Lekander andWitkiewicz (2004)).The difference between suggested and actual allocation to realestate is considered to be a puzzle in real estate research (Chun,Sa-Aadu and Shilling (2004)).
Can property derivatives solve this puzzle?
I mean-downside-risk analysis
I by using forward contracts with optimal hedge ratios
I 130/30-portfolio strategy
I comparison between ex ante and ex post adjustments
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Data CollectionInterpolation
Data Collection
I included asset classes: stocks, bonds, and real estatederivatives
I based on quarterly data from Q1 1996 to Q4 2008
I investment countries: the USA, the UK, France (FRA), andGermany (GER)
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Data CollectionInterpolation
Interpolation
Problem: For FRA and GER only annual real estate dataSolution: Interpolation?!
I Nearest neighbor interpolation
I Linear interpolation
I Cubic spline interpolation
I Modified cubic spline interpolation
I Monte-Carlo simulation
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Data CollectionInterpolation
Interpolation Comparison
Figure: Comparison of different Spline Interpolation Methods
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Data CollectionInterpolation
Estimation Errors for the USA
The nearest neighbor interpolation (NNI), the linear interpolation (PLI), thecubic spline interpolation (CSI), the modified cubic spline interpolation (MCSI),and the Monte-Carlo simulation (MCS), are compared with the real returns(RR) by dint of the mean (µ), the standard deviation (σ), the coefficient ofvariation (CV), the mean squared error (MSE), and the root mean squarederror (RMSE).
RR NNI PLI CSI MCSI MCS
µ 11.47 11.36 11.74 11.94 11.39 11.37σ 7.67 6.43 4.94 4.87 6.89 6.66CV 0.67 0.57 0.42 0.41 0.60 0.59
MSE 0 20.35 15.67 17.26 7.82 21.89RMSE 0 4.51 3.96 4.16 2.80 4.68
Table: Estimation Errors for the USA
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
AA in a Shortfall-Risk Framework I
Initial Position: International Asset Allocation
I no normality
I different currency areas
Research design:
I Mean-Downside-Risk Analysis
I by using forward contracts with optimal hedge ratios
I 130/30-portfolio strategy
I comparison between ex ante and ex post adjustments
I stochastic dominance analysis (EWP, MRP, TP)
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
AA in a Shortfall-Risk Framework IIModifying Harlow (1991), the optimization approach is defined as:
minxi
→ 1
T
TXt=1
max
rf ;t −
NXi=1
xiRi ;t ; 0
!2
, (1)
subject to the constraints:
NXi=1
xiR i = RP ,
NXi=1
xi = 1,
NXi=1
(xi |xi < 0) ≥ −0.3,
i = 1, 2, ..., N.
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
AA in a Shortfall-Risk Framework III
T is the number of observed periods,
R i is the mean return on the asset alternative i over all periods,
RP is the prescribed portfolio return,xi is the portfolio weight of asset alternative i , andrf ;t is the risk free rate at the beginning of period t.
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
AA in a Shortfall-Risk Framework IV
I the returns are realized in different currency areas
I the returns has to be converted for an euro-investor
Ri = (1 + Rai ) (1 + ea) − 1 = Ra
i + ea + Rai ea (2)
Rai is the the uncertain monthly return on asset alternative i
in country a, andea is the exchange rate return between the euro and the
foreign currency area.
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
AA in a Shortfall-Risk Framework VI
Rhi = Ri + hi (f
a − ea) (3)
The forward premium f a between currency a and the euro isdefined as:
f a =F a
t+1 − Sat
Sat
(4)
F at+1 is the forward price at the point in time t for the delivery
of the currency at the point in time t + 1from the perspective of the home country, and
hi is the hedge ratio.
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Preliminary Results
monthly data from 1998-2008/6; mean − lpm1
Figure: Efficient Portfolio Sets
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Preliminary Results
Figure: Optimal Weights for the MRP and TP Portfolios
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Market CapitalizationTransaction CostsTrading Volumes
Market Capitalization
Country Stocks Bonds Real Estates
United States 13,552.2 4,478.3 1,287.2People’s Republic of China 4.230.2 1,136.7 NAJapan 3.211.0 4,854.4 201.2United Kingdom 2.621.5 613.5 411.3France 1.882.8 954.6 203.3Germany 1.430.5 946.4 277.6
Source World Bank BIS IPD
Table: Market Size in 2007 (billion euro)
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Market CapitalizationTransaction CostsTrading Volumes
Transaction Costs
Figure: Transaction Costs
The values for Belgium, the United Kingdom, Australia, Hong Kong, Japan,
Singapore, and the United States are averaged in subject to the different legal
basis points in their countries.
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Market CapitalizationTransaction CostsTrading Volumes
Transaction Costs
Total annualised cost of buying a commercial property unit trustand selling it after 5 years
Country
United States 0.75%United Kingdom 0.25%France 0.50%Germany 0.35%
Source: GS (2008)
Table: Transaction Costs
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation
MotivationData
Research DesignPreliminary Results
Appendix
Market CapitalizationTransaction CostsTrading Volumes
Trading Volumes
Figure: Notional Amount of Derivatives
(in billion euro)
Bertram Steininger IRE|BS, University of Regensburg Property Derivatives in the Strategic Asset Allocation