krul presentation1

Information about krul presentation1

Published on April 22, 2008

Author: Pravez

Source: authorstream.com

Content

Calibration of Stochastic Convenience Yield Models For Crude Oil Using the Kalman Filter.:  Calibration of Stochastic Convenience Yield Models For Crude Oil Using the Kalman Filter. Delft – 22-02-08 www.ing.com Adriaan Krul Contents:  Contents Introduction Convenience yield follows Ornstein-Uhlenbeck process Analytical results Convenience yield follows Cox-Ingersoll-Ross process Analytical results Numerical results Conclusion Further research Introduction:  Introduction A future contract is an agreement between two parties to buy or sell an asset at a certain time in the future for a certain price. Convenience yield is the premium associated with holding an underlying product or physical good, rather than the contract of derivative product. Commodities – Gold, Silver, Copper, Oil We use futures of light crude oil ranging for a period from 01-02-2002 until 25-01-2008 on each friday to prevent weekend effects. Stochastic convenience yield; first approach:  Stochastic convenience yield; first approach We assume that the spot price of the commodity follows an geometrical brownian motion and that the convenience yield follows an Ornstein-Uhlenbeck process. I.e., we have the joint-stochastic process Slide5:  In combination with the transformation x = ln S we have Analytical results:  Analytical results Expectation of the convenience yield Variance of the convenience yield Analytical results :  Analytical results PDE of the future prices Closed form solution of the future prices Stochastic convenience yield; second approach:  Stochastic convenience yield; second approach We assume that the spot price of the commodity follows an geometrical brownian motion and that the convenience yield follows a Cox-Ingersoll-Ross process. I.e., we have the joint-stochastic process Slide9:  Together with the transformation x = ln S, we have Analytical results:  Analytical results Variance of the convenience yield Analytical results:  Analytical results PDE of the future prices Closed form solution of the future prices Kalman filter:  Kalman filter Since the spot price and convenience yield of commodities are non-observable state-variables, the Kalman Filter is the appropriate method to model these variables. The main idea of the Kalman Filter is to use observable variables to reconstitute the value of the non-observable variables. Since the future prices are widely observed and traded in the market, we consider these our observable variables. The aim of this thesis is to implement the Kalman Filter and test both the approaches and compare them with the market data. Kalman Filter for approach one.:  Kalman Filter for approach one. Recall that the closed form solution of the future price was given by From this the measurement equation immediately follows Slide14:  From we can write Kalman filter for approach one:  Kalman filter for approach one The difference between the closed form solution and the measurement equation is the error term epsilon. This error term is included to account for possible errors. To get a feeling of the size of the error, suppose that the OU process generates the yields perfectly and that the state variables can be observed form the market directly. The error term could then be thought of as market data, bid-ask spreads etc. Kalman filter for approach one:  Kalman filter for approach one Recall the join-stochastic process the transition equation follows immediately Slide17:  For simplicity we write Kalman filter for approach two:  Kalman filter for approach two From it follows Kalman filter for approach two:  Kalman filter for approach two Recall the join-stochastic process the transition equation follows immediately Slide20:  For simplicity we write How does the Kalman Filter work?:  How does the Kalman Filter work? We use weekly observations of the light crude oil market from 01-02-2002 until 25-01-2008. At each observation we consider 7 monthly contracts. The systems matrices consists of the unknown parameter set. Choosing an initial set we can calculate the transition and measurement equation and update them via the Kalman Filter. Then the log-likelihood function is maximized and the innovations (error between the market price and the numerical price) is minimized. How to choose the initial state.:  How to choose the initial state. For the initial parameter set we randomly choose the value of the parameters within a respectable bound. For the initial spot price at time zero we retained it as the future price with the first maturity and the convenience yield is initially calculated via Numerical results for approach one:  Numerical results for approach one Log future prices versus state variable x:  Log future prices versus state variable x Implied convenience yield versus state variable delta:  Implied convenience yield versus state variable delta Innovation for F1:  Innovation for F1 Kalman forecasting applied on the log future prices:  Kalman forecasting applied on the log future prices Kalman Forecasting applied on the state variables:  Kalman Forecasting applied on the state variables Conclusion:  Conclusion We implemented the Kalman Filter for the OU process. Both the convenience yield as well as the state variable x (log of the spot price) seems to follow the implied yield and the market price (resp.) quite good. Also, different initial values for the parameter set will eventually converge to the optimized set with the same value of the log-likelihood. This is a good result and tests the robustness of the method. The main difference between the systems matrices of both processes is the transition error covariance-variance matrix Vt. In the CIR model, this matrix forbids negativity of the CY. We simply replaced any negative element of the CY by zero, but since it is negative for a large number of observations, this will probably give rise to large standard errors in the optimized parameter set. Conclusion:  Conclusion The Kalman Forecasting seems to work only if there is no sudden drop in the data. To improve the Kalman Forecasting we could update it every 10 observations. Further research:  Further research Implement the Kalman Filter for the CIR model Inserting a jump constant in the convenience yield Compare both stochastic models Pricing of options on commodities, using the optimized parameter set

Related presentations


Other presentations created by Pravez

BlueGene
19. 09. 2007
0 views

BlueGene

GRP
17. 04. 2008
0 views

GRP

Howe1 25
14. 04. 2008
0 views

Howe1 25

Panview 3en
10. 04. 2008
0 views

Panview 3en

Global Trends and Prospects
09. 04. 2008
0 views

Global Trends and Prospects

2002 12 22 Christingle
07. 04. 2008
0 views

2002 12 22 Christingle

303 151renewableenergy
30. 03. 2008
0 views

303 151renewableenergy

mechanics of materials
05. 01. 2008
0 views

mechanics of materials

genetic recombination
06. 08. 2007
0 views

genetic recombination

DRAMA
19. 10. 2007
0 views

DRAMA

jong
19. 09. 2007
0 views

jong

C102 Bergallo
03. 10. 2007
0 views

C102 Bergallo

text cat tutorial
04. 10. 2007
0 views

text cat tutorial

Tornado Safety
07. 10. 2007
0 views

Tornado Safety

newborn screening
12. 10. 2007
0 views

newborn screening

Switzerland Presentation
17. 10. 2007
0 views

Switzerland Presentation

CP15633
22. 10. 2007
0 views

CP15633

Hack lu cisco ddos
07. 10. 2007
0 views

Hack lu cisco ddos

DNA PowerPoint
15. 10. 2007
0 views

DNA PowerPoint

Membres du Bureau
23. 10. 2007
0 views

Membres du Bureau

23981
12. 10. 2007
0 views

23981

Joao LOBO ANTUNES
16. 10. 2007
0 views

Joao LOBO ANTUNES

Tomatoes
11. 12. 2007
0 views

Tomatoes

C avance
20. 11. 2007
0 views

C avance

Familias Unidas
22. 10. 2007
0 views

Familias Unidas

Module 20
23. 11. 2007
0 views

Module 20

media interview
08. 10. 2007
0 views

media interview

EO1 SysEngrColloq 03 02 04 Stu
05. 10. 2007
0 views

EO1 SysEngrColloq 03 02 04 Stu

hoffman arizona
30. 12. 2007
0 views

hoffman arizona

IA Emergency Preparedness
01. 01. 2008
0 views

IA Emergency Preparedness

Cohen NDIA
03. 01. 2008
0 views

Cohen NDIA

mathematicians
05. 01. 2008
0 views

mathematicians

hamm
23. 10. 2007
0 views

hamm

Prologue
15. 11. 2007
0 views

Prologue

geo inversion
19. 09. 2007
0 views

geo inversion

HIV AIDS Pandemic
06. 08. 2007
0 views

HIV AIDS Pandemic

Jeannie Armour Bearer
06. 08. 2007
0 views

Jeannie Armour Bearer

Hawaiian Cultural History
06. 08. 2007
0 views

Hawaiian Cultural History

jiscdigicon 07par1 1notay
06. 08. 2007
0 views

jiscdigicon 07par1 1notay

ESCpresentation
03. 01. 2008
0 views

ESCpresentation

Unit2 Lecture1
04. 01. 2008
0 views

Unit2 Lecture1

graficos ddet265
24. 10. 2007
0 views

graficos ddet265

Robin Room
07. 01. 2008
0 views

Robin Room

jageps
06. 08. 2007
0 views

jageps

Jewish Initiation
06. 08. 2007
0 views

Jewish Initiation

HIUG
06. 08. 2007
0 views

HIUG

singers
20. 02. 2008
0 views

singers

HGM2006 S7 01 Pedersen
06. 08. 2007
0 views

HGM2006 S7 01 Pedersen

ModuleA5Session3
04. 03. 2008
0 views

ModuleA5Session3

hlt 2004
06. 08. 2007
0 views

hlt 2004

Brussel Health Kuopio KPoutanen
20. 03. 2008
0 views

Brussel Health Kuopio KPoutanen

CIO Main Screens Sydney Final
26. 03. 2008
0 views

CIO Main Screens Sydney Final

Movimento Pentecostal 2 ok
01. 10. 2007
0 views

Movimento Pentecostal 2 ok

guyana csw 2000
06. 08. 2007
0 views

guyana csw 2000

05f 1800 lys a3
19. 02. 2008
0 views

05f 1800 lys a3

Guze 07 Sarah Pedersen
06. 08. 2007
0 views

Guze 07 Sarah Pedersen

Normativa Vittoria marzo06
18. 06. 2007
0 views

Normativa Vittoria marzo06

new Media Activism
18. 06. 2007
0 views

new Media Activism

LKD ACPC2005autoshow
18. 06. 2007
0 views

LKD ACPC2005autoshow

LAS VEGAS PRESENTATION
18. 06. 2007
0 views

LAS VEGAS PRESENTATION

lab storia marcia mille
18. 06. 2007
0 views

lab storia marcia mille

Kreutz
18. 06. 2007
0 views

Kreutz

kinder verzamelingen 05
18. 06. 2007
0 views

kinder verzamelingen 05

kereso halo
18. 06. 2007
0 views

kereso halo

japan overview
18. 06. 2007
0 views

japan overview

jaden
18. 06. 2007
0 views

jaden

2007 dossier pere
18. 06. 2007
0 views

2007 dossier pere

2005 NAIDC Promo
18. 06. 2007
0 views

2005 NAIDC Promo

1212 SLIDES CSR Briefing final1
18. 06. 2007
0 views

1212 SLIDES CSR Briefing final1

07PSPTribute
18. 06. 2007
0 views

07PSPTribute

bioprez1
17. 11. 2007
0 views

bioprez1

Chapter18
19. 09. 2007
0 views

Chapter18

LUXva
18. 06. 2007
0 views

LUXva

f38 math meets maps
15. 06. 2007
0 views

f38 math meets maps

Extraction of Beam into TT40
15. 06. 2007
0 views

Extraction of Beam into TT40

Experiencing HaShem
15. 06. 2007
0 views

Experiencing HaShem

Euthyphro
15. 06. 2007
0 views

Euthyphro

ethnic humor
15. 06. 2007
0 views

ethnic humor

Essentials of UP 1
15. 06. 2007
0 views

Essentials of UP 1

Catv2
15. 06. 2007
0 views

Catv2

Cartoons pig
15. 06. 2007
0 views

Cartoons pig

cartoons stick men 1
15. 06. 2007
0 views

cartoons stick men 1

attr 159897 115
12. 10. 2007
0 views

attr 159897 115

kskee
19. 09. 2007
0 views

kskee

akrishna
28. 11. 2007
0 views

akrishna

18cm
18. 06. 2007
0 views

18cm

BrazilSeminar
22. 10. 2007
0 views

BrazilSeminar

PRSarahL
13. 03. 2008
0 views

PRSarahL

prezent1
15. 11. 2007
0 views

prezent1

power law tut
19. 09. 2007
0 views

power law tut

EthicalEgoism
15. 06. 2007
0 views

EthicalEgoism

petersuter ppt
17. 10. 2007
0 views

petersuter ppt

Internet Equivalence
06. 08. 2007
0 views

Internet Equivalence

slogany
21. 11. 2007
0 views

slogany

Blue Chalkboard
19. 09. 2007
0 views

Blue Chalkboard

Gogol at the Edge of Being
27. 09. 2007
0 views

Gogol at the Edge of Being

ScottKsander10 10
02. 11. 2007
0 views

ScottKsander10 10

labrolazegbe
18. 06. 2007
0 views

labrolazegbe

chap12sp06 groups day 3
14. 12. 2007
0 views

chap12sp06 groups day 3

Resources Potpourri Final2006
26. 10. 2007
0 views

Resources Potpourri Final2006

powers tami portia topics
19. 09. 2007
0 views

powers tami portia topics

APGridPMA EU 200605
10. 10. 2007
0 views

APGridPMA EU 200605

Higgins Webinar Slides 070425
19. 09. 2007
0 views

Higgins Webinar Slides 070425