Portfolio Construction Portfolio Construction Strategies Strategies Using Cointegration Using Cointegration M.Sc. Student: IONESCU GABRIEL Supervisor: Professor MOISA ALTAR BUCHAREST, JUNE 2002 ACADEMY OF ECONOMIC STUDIES DOCTORAL SCHOOL OF FINANCE AND BANKING
33
Embed
Portfolio Construction Strategies Using Cointegration
ACADEMY OF ECONOMIC STUDIES DOCTORAL SCHOOL OF FINANCE AND BANKING. Portfolio Construction Strategies Using Cointegration. M.Sc. Student: IONESCU GABRIEL Supervisor: Professor MOISA ALTAR BUCHAREST, JUNE 2002. 1.Introduction. - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Portfolio Construction StrategiesPortfolio Construction Strategies Using Cointegration Using Cointegration
M.Sc. Student: IONESCU GABRIEL
Supervisor: Professor MOISA ALTAR
BUCHAREST, JUNE 2002
ACADEMY OF ECONOMIC STUDIES
DOCTORAL SCHOOL OF FINANCE AND BANKING
2
1.Introduction
Traditional models seek portfolio weights so as to minimize the variance of the portfolio for a given level of return.
Portfolio variance is measured using a covariance matrix which is not only difficult to estimate, but also very unstable in time.
Additionally, the mean-variance criterion has nothing to ensure that portfolio deviations (errors) relative to a benchmark are stationary, in the majority of cases being a random walk.
As a consequence the portfolio will drift virtually anywhere away from the benchmark unless is frequently rebalanced.
=>transaction costs=>negative influence on performances !
3
The Problem Root of the problem: MV analysis is based on returns (I(0)) rather than prices
(I(1)).
The difference:
So when we move from prices to returns we actually lose valuable information! What can be done?...
the prices are highly autocorrelated,
sharing long-term trends relative to
spreads and relative market
direction;
return analysis is based on low
autocorrelated market information,
having less stable, short-lived
portfolios, and little long term
predictive value, limited trend
information
4
Cointegration enables us to avoid this drawback because it measures how the prices, and not the returns, are moving together in the long run, having in contradiction to the classical correlation concept the advantage of using the entire set of information from the price levels.
If the spreads are mean-reverting, asset prices are tied together in the long run by a common stochastic trend => the prices are cointegrated
Cointegration tells us that when found, stable co-relationships between groups of assets will remain stable for some period of time as a result of prevailing market factors.
COINTEGRATION
5
Cointegration in portfolio management
Lucas (1997)
Alexander (1999)
DiBartolomeo (1999)
Alexander and Weddington (2001)
Alexander and Dimitriu (2002)
6
2. Portfolio Construction Strategies Using CointegrationPortfolio Construction Strategies Using Cointegration
Cloning strategies: aim to construct a portfolio, that clones a given benchmark, in terms of return and volatility, and preferably with the use of a small number of assets. Cloning portfolio will be strong correlated with the market.
Cointegration method: Engle -Granger (1987). Reasons: 1.we know a priori that we have a single cointegrating relation (portfolio weights) 2. Its simplicity; 3. For portfolio management the criterion of minimizing the variance is far more important than Johansen’s criterion of maximizing stationarity.
Once we ensured that the candidate asset price series are non-stationary, we will estimate a cointegrating regression, having as dependent variable the price series of the benchmark, and as independent variables the candidate clone portfolio components. Estimation will be made using a prespecified window of data, called calibration period.
7
n
ittAiitbenchmark PcP
1,, )log()log(
More formally, we will estimate by OLS the following equation:
where: Pbenchmark is the time series of (daily) benchmark price; PAi is the time
series of asset “i”; i are the estimated coefficients from the above
regression, coefficients that after normalization will play the role of portfolio
weights; and ε is residual series, which is nothing but the tracking error.
A simple algorithm of optimization
To fully benefit of the common stochastic trend followed by the asset prices that will compose the clone portfolio, it is paramount to select from the candidate assets, the basket that is the most cointegrated with the benchmark.
Definition 1. We will call cointegrating portfolio the linear combination
m
iAii PPC
1
)log( , m <= n,
with the property that log(Pbenchmark) is cointegrated with PC, n being the number of available assets.
D e f i n i t i o n 2 . W e s a y t h a t c o i n t e g r a t i n g p o r t f o l i o
1
11 )log(
m
iAii PPC i s m o r e c o i n t e g r a t e d ( i n E n g l e
G r a n g e r s e n s e ) t h a n p o r t f o l i o
2
12 )log(
m
iAii PPC , i f n o t i n g 2,1,)log( jPCP jbenchmarkj
t h e n t - s t a t (ε 1 ) < t - s t a t (ε 2 ) , w h e r e t - s t a t i s t a k e n f r o m t h e u n i t r o o t t e s t a p p l i e d t o r e s i d u a l ε j .
Definition 3 (optimality) . Let
!)!(
!|
kkn
njPCj be the set of all cointegrating portfolios that
can be formed with the n assets, using the same calibration period. Let
!)!(
!...2,1,)log(|
kkn
njPCP jbenchmarkjj be the set of residual series
corresponding to the above portfolios. We say that PCk is optimal cointegrating portfolio if and only if t-stat(ε k) < t-stat(ε h) for kh .
9
Data: MSCI equity indices for Eurozone countries
Series name Description
1 LAUSTRIA MSCI Austria Equity Index
2 LBELGIUM MSCI Belgium Equity index
3 LFINLAND MSCI Finland Equity index
4 LFRANCE MSCI France Equity index
5 LGERMANY MSCI Germany Equity index
6 LGREECE MSCI Greece Equity index
7 LIRELAND MSCI Ireland Equity index
8 LITALY MSCI Italy Equity index
9 LNETHERLANDS MSCI Netherlands Equity index
10 LPORTUGAL MSCI Portugal Equity index
11 LSPANIA MSCI Spain Equity index
12 LEURO MSCI EURO Equity index
13 LEUROPLUS MSCI EURO Equity index plus a spread of 2% p.a. uniformly distributed
14 LEUROMINUS MSCI EURO Equity index minus a spread of 2% p.a. uniformly distributed
10
We will estimate a cointegrating regression using all candidate assets as independent variables, and as dependent variable EURO MSCI index plus 2% p.a.
We will try to find the most cointegrated portfolio eliminating successively variables from the regression, and testing the stationarity of the resulting residuals.
11
Optimization rounds (eliminating FI, NE, SP, BE)
SERIES NAME ADF H0: I(1) vs I(0)RESIDAU -4.7471RESIDBE -4.7094RESIDFI -6.6327RESIDFR -6.1154RESIDGE -4.7469RESIDGR -5.2723RESIDIR -4.6777RESIDIT -4.9688RESIDNE -5.4707RESIDPO -3.9057RESIDSP -6.3582
SERIES NAME ADF H0: I(1) vs I(0)RESID01AU -7.1659RESID01BE -6.4796RESID01FR -6.5213RESID01GE -5.0125RESID01GR -6.318RESID01IR -6.4678RESID01IT -5.8177RESID01NE -7.3124RESID01PO -5.4401RESID01SP -6.8202
SERIES NAME ADF H0: I(1) vs I(0)RESID02AU -7.0592RESID02BE -7.1881RESID02FR -6.3224RESID02GE -5.2677RESID02GR -6.9923RESID02IR -7.0693RESID02IT -6.0594RESID02PO -6.2361RESID02SP -7.3568
SERIES NAME ADF H0: I(1) vs I(0)RESID03AU -7.1127RESID03BE -7.2266RESID03FR -5.5149RESID03GE -5.8952RESID03GR -7.0518RESID03IR -6.8181RESID03IT -6.893RESID03PO -6.9268
12
Suboptinal round
SERIES NAME ADF H0: I(1) vs I(0)RESID04AU -7.0568RESID04FR -5.525RESID04GE -4.8763RESID04GR -6.4989RESID04IR -6.7358RESID04IT -6.8774RESID04PO -6.0192
a further attempt to optimize the
portfolio composition will end up
in obtaining a suboptimal
portfolio, because eliminating the
Austrian equity index from
portfolio will lead to an error less
stationary (ADF t-stat of –7.0568)
comparing to the previous round
(ADF t-stat of –7.2266). We will
conclude that previous round
gives us the most cointegrated
portfolio.
13
Once we found the composition, we determine the weights...Dependent Variable: LEUROPLUSMethod: Least SquaresSample: 4/30/1997 5/02/2001Included observations: 1046Variable Coefficient Std. Error t-Statistic Prob.
Figure 7. Historical and EWMA volatilities of the excess return for the two sub-strategies
0.94
0.95
0.96
0.97
0.98
0.99
1.00
5/01 7/10 9/18 11/27 2/05 4/16
CORR EWMA 30 PC rebalCORR HIST 30 PC rebal
0.94
0.95
0.96
0.97
0.98
0.99
1.00
5/01 7/10 9/18 11/27 2/05 4/16
CORR EWMA PClona EUROCORR hist PClona EURO
Figure 8. Historical and EWMA correlations between
a) rebalanced portfolio and market b) unmanaged portfolio and market
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
5/01 7/10 9/18 11/27 2/05 4/16
CORR HIST eroare1 euro CORR EWMA eroare1 euro
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
5/01 7/10 9/18 11/27 2/05 4/16
CORR EWMA res euro CORR hist res euro
Figure 9. Historical and EWMA correlations between
a) market and unmanaged residual b) market and rebalanced residual
6. Correlations
23
7. Distributional properties of cloning errors
Error Long P Long rbl P Short P Short rbl P Mean 0.010147 0.001207 0.006660 0.002074 Median 0.011981 0.002274 -0.006811 -0.005947 Maximum 0.032453 0.014586 0.080644 0.049457 Minimum -0.020394 -0.020299 -0.051729 -0.026488 Std. Dev. 0.011858 0.007568 0.034781 0.021043 Skewness -0.358671 -0.449998 0.601555 0.721445 Kurtosis 2.232754 2.301240 1.976044 2.103339
24
Arbitrage strategies
This type of strategies aims to construct a self-financing portfolio, which will generate positive returns irrespective of market direction, with a low volatility and in conditions of zero correlation with the market. To ensure the self-financing of the strategy, we construct two cointegrating portfolios: a long portfolio, which clone a benchmark plus a spread, and a short portfolio, which clone a benchmark minus a spread. The arbitrage portfolio will be given by the difference of the above portfolios, and will earn approximately the sum of the absolute values of the two spreads .
n
ittAiitplusbenchmark PcP
1,1,_ )log()log( (2)
n
ittAiitusbenchmark uPcP
1,2,min_ )log()log( (3)
25
We need to construct the short portfolio, which clones MSCI EURO minus 2%. Using the optimization algorithm we obtain:
VOL EWMA30 PArbitrage(PA)VOL EWMA 30 PA RBLVOL Hist 30 PAVOL Hist 30 PA RBL
Figure 12. Historical and EWMA volatilities of the arbitrage returns for the two sub-strategies
32
5. Correlation benchmark return - the arbitrage portfolio return
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
6/01 8/10 10/19 12/28 3/08
COR_PARBL_EURO_EWMACOR_PARBL_EURO_HIST
COR_PA_EURO_EWMACOR_PA_EURO_HIST
Figure 13. Correlation between MSCI EURO returns and arbitrage returns
33
3.Conclusions
we succeeded to find a cloning portfolio that systematically over-performed the benchmark in terms of returns, had a smaller volatility, and moreover was composed of a smaller number of assets than the original benchmark.
cloning strategy remained cointegrated with the benchmark during the entire testing period, even if the portfolio was left unmanaged
monthly rebalanced portfolio was more cointegrated than in the first case; also with a greater excess return and a reduced risk.
The performances of the model persisted even after accounting for brokerage fees.
The arbitrage strategy aimed to produce a positive return in all states of the nature. The enhanced stationarity of the tracking errors, gained by rebalancing, made it possible for the arbitrage portfolio to generate positive risk-free returns after deducting the corresponding transaction costs.