Lecture 1 Finance Pro ject Somesh Jha 1 Goals of the - - PDF document

lecture 1 finance pro ject somesh jha 1 goals of the
SMART_READER_LITE
LIVE PREVIEW

Lecture 1 Finance Pro ject Somesh Jha 1 Goals of the - - PDF document

Lecture 1 Finance Pro ject Somesh Jha 1 Goals of the course T eac h the studen ts to systematic al ly design algorithms and systems for problems in the nance domain. T aking nance ideas from the ac


slide-1
SLIDE 1 Lecture 1 Finance Pro ject Somesh Jha 1
slide-2
SLIDE 2 Goals
  • f
the course
  • T
eac h the studen ts to systematic al ly design algorithms and systems for problems in the nance domain.
  • T
aking nance ideas from the ac ademic domain and making them real.
  • Maxim:
Learn b y doing. 2
slide-3
SLIDE 3 Nature
  • f
the course
  • Largely
indep enden t in nature.
  • Je
and I are here to guide y
  • u,
but y
  • u
will w
  • rk
as indep enden t teams.
  • Think
  • f
the course as a structur e d indep endent study c
  • urse.
3
slide-4
SLIDE 4 What w e will do in class
  • Je
and I will go through a systematic design
  • f
a system for pricing mortgage b acke d se curities
  • r
MBSs.
  • Use
the lectures as a guide for y
  • ur
pro ject.
  • There
will b e ve phases to the pro ject. These phases will b e dened later. 4
slide-5
SLIDE 5 Logistics
  • Studen
ts will form a team
  • f
3-4.
  • Eac
h team will select a pap er from a set pro vided in class.
  • Eac
h team will design a system based
  • n
the pap er they select.
  • Eac
h team will go through the v e phases (to b e describ ed later). 5
slide-6
SLIDE 6 Some useful tips
  • Pic
k a b alanc e d te am.
  • Stic
k to the sc hedule for eac h phase. 6
slide-7
SLIDE 7 Grading
  • Grading
will b e done dep ending
  • n
the
  • utcome
  • f
eac h phase.
  • No
tests and homew
  • rks.
7
slide-8
SLIDE 8 Fiv e phases Description
  • f
these phases will b e pro vided later, but here they are.
  • Requiremen
ts phase (Phase 1).
  • High-lev
el design phase (Phase 2).
  • Lo
w-lev el design phase (Phase 3). 8
slide-9
SLIDE 9 Phases (Con td)
  • In
class presen tation (Phase 4).
  • Protot
yp e (Phase 5) 9
slide-10
SLIDE 10 Protot yp e
  • This
will b e a sc ale d b ack v ersion
  • f
the design.
  • Mak
e as man y limiting assumptions as p
  • ssible,
but state them carefully .
  • Y
  • u
can use C,C++,JAVA to implemen t the protot yp e. Mak e sure y
  • u
tell Je the en vironmen t y
  • u
are using. 10
slide-11
SLIDE 11 Presen tation
  • There
will b e
  • ne
presen tation p er team.
  • The
presen tation will b e a synopsis
  • f
phase 1,2, and 3.
  • Eac
h presen tation will b e 20 min utes long.
  • Last
three lectures will b e all presen tations. 11
slide-12
SLIDE 12 P ap er 1
  • J.
Hull and A. White, Ecien t Pro cedures for V aluing Europ ean and American P ath-Dep enden t Options, Journal
  • f
Derivative, No 1, Pages 21-31, 1993. 12
slide-13
SLIDE 13 P ap er 2
  • J.
Hull and A. White, V aluing Deriv ativ e Securities Using the Explicit Finite Dierence Metho d, Journal
  • f
Financial and Quantitative A nalysis, V
  • l
25. No 1, Pages 87-99, 1990. 13
slide-14
SLIDE 14 P ap er 3
  • J.
Hull and A. White, Pricing In terest-Rate-Deriv ativ e Securities, The R eview
  • f
Financial Studies, V
  • l
3, No 4, Pages 573-592, 1990. 14
slide-15
SLIDE 15 P ap er 4
  • A.
Li, P . Ritc hk en, and L. Sank arasubramanian, Lattice Mo dels for Pricing American In terest Rate Claims, Journal
  • f
Financ e, V
  • l
L, No 2, Pages 719-736, 1995. 15
slide-16
SLIDE 16 P ap er 5
  • P
. Ritc hk en and L. Sank arasubramanian, V
  • latilit
y Structures
  • f
F
  • rw
ard Rates and the Dynamics
  • f
the T erm Structure,Ma th ematic al Financ e, V
  • l
5, No 1, Pages 55-72, 1995. 16
slide-17
SLIDE 17 P ap er 6
  • P
. Ritc hk en and L. Sank arasubramanian, Pricing the Qualit y Option in T reasury Bond F utures, ,Mathematic al Financ e, V
  • l
2, No 3, Pages 197-214, 1995. 17
slide-18
SLIDE 18 P ap er selection
  • Don't
c ho
  • se
pap ers 5
  • r
6 unless y
  • u
are comfortable with sto c hastic calculus.
  • F
  • cus
  • n
the tec hniques and algorithms in the pap er. It is OK if y
  • u
don't understand all the mathematical deriv ations. 18
slide-19
SLIDE 19 Goals
  • f
reading the pap er
  • Decide
what nancial instruments y
  • u
w an t to price after reading the pap er. Pic k 2-3 instrumen ts. Y
  • u
will b e required to understand these instrumen ts completely .
  • Y
  • u
should ha v e a clear idea ab
  • ut
the algorithm prop
  • sed
in the pap er.
  • Mak
e a note
  • f
adv an tages/disadv an tages
  • f
this tec hnique/algorithm. 19
slide-20
SLIDE 20 Requiremen ts do cumen t
  • Describ
e the nancial instrumen t in great detail.
  • Describ
e the assets the instrumen ts dep end up
  • n.
State the assumptions
  • n
the prices
  • f
these assets. Describ e the cash-o w c haracteristics.
  • Describ
e the nancial instrumen ts and there cash-o w c haracteristics. 20
slide-21
SLIDE 21 Requiremen ts do cumen t
  • In
an abstract sense w e are describing what is the seman tics
  • f
eac h
  • p
eration that the user can do.
  • In
this very sp e cic example this amoun ts to dening the precise seman tics
  • f
mortgage b acke d se curities (MBSs). 21
slide-22
SLIDE 22 Describing Mortgages
  • Fixe
d R ate: The ann ual in terest rate
  • f
the mortgage sta ys xed through
  • ut
the life
  • f
the mortgage.
  • A
djustable R ate Mortgages (ARMs): The ann ual in terest rate can b e adjusted b y the loaning agency . 22
slide-23
SLIDE 23 Fixed Rate Mortgages
  • Let
M B b e the
  • riginal
mortgage balance.
  • Let
c b e the simple mon thly in terest rate.
  • Let
M P b e the mon thly mortgage pa ymen t.
  • Let
n b e the n um b er
  • f
mon ths. 23
slide-24
SLIDE 24 Relationship b et w een M P and M B
  • The
follo wing equation should hold b et w een M P and M B : M B = M P n X i=1 (1 + c) i = M P 1
  • (1
+ c) n c
  • Hence
the mon thly mortgage pa ymen t M P is giv en in terms
  • f
the mortgage amoun t M B using the follo wi ng form ula: M P = M B c(1+c) n (1+c) n 1 24
slide-25
SLIDE 25 Principal at time t
  • Let
M B t the remaining mortgage balance at time t.
  • W
e ha v e the follo wi ng relationship b et w een M B t and M P . M B t = M P 1
  • (1
+ c) (nt) c
  • So
w e ha v e the follo wi ng equation b et w een M B and M B t : M B t = M B (1 + c) n
  • (1
+ c) t (1 + c) n
  • 1
25
slide-26
SLIDE 26 Breaking the mortgage pa ymen ts
  • A
t time t the mortgage balance is M B t1 (t
  • 1).
  • The
in terest I t
  • n
this is mortgage balance is: cM B t1
  • The
mortgage pa ymen t M P at time t is brok en in to t w
  • parts:
inter est p ayment I t and p ayment applie d towar ds princip al P t . W e ha v e the follo wi ng equation: M P = I t + P t 26
slide-27
SLIDE 27 Scenarios
  • A
requiremen ts do cumen t for a large soft w are system has a h uge n um b er
  • f
scenarios.
  • Basically
, sc enarios describ e what should happ en in sp ecic cases.
  • F
  • r
example, in the requiremen ts do cumen t for an
  • nline
brok erage system a scenario migh t describ e what should happ en when a user logs
  • n
and buys a sto c k. 27
slide-28
SLIDE 28 Examples
  • In
this case, scenarios are simply examples
  • f
cash
  • ws.
  • Consider
a mortgage
  • f
100; 000, ann ual mortgage rate (12c)
  • f
9:5% and time p erio d
  • f
30 y ears (360 mon ths).
  • Chec
k that the mon thly mortgage rate M P is 840:85. 28
slide-29
SLIDE 29 Example con tin ued
  • Chec
k that I = 791:67 and P = 49:19.
  • Chec
k that I 215 = 574:95 and P 215 = 265:90.
  • I
t is a decreasing function
  • f
t and P t is an increasing function
  • f
t. (Why?). 29
slide-30
SLIDE 30 ARMs
  • ARMs
start
  • ut
with an initial in terest rate.
  • ARMs
in terest rate can b e adjusted b y a margin m at a frequency sp ecied in the con tract.
  • Lifetime
c ap c L : This is an upp er b
  • und
that the in terest rate cannot exceed.
  • Lifetime
  • r
c F : This is a lo w er b
  • und
  • n
the in terest rate. 30
slide-31
SLIDE 31 ARMs
  • Let
us the in terest rate is c(t
  • 1)
at time t
  • 1
and w e are adjusting at time t,
  • The
new in terest rate c ( t) is giv en b y the follo wi ng cases: { if x(t) + m > c(t
  • 1)
min [x(t) + m; c L ; c(t
  • 1)
+ c P ] { if x(t) + m
  • c(t
  • 1)
min [x(t) + m; c F ; c(t
  • 1)
  • c
P ] 31
slide-32
SLIDE 32 Explanation
  • f
terms
  • x(t)
: Underlying index sp ecied in the con tract. Tw
  • widely
used indices are cost
  • f
funds index (COFI) and a constan t maturit y (one y ear
  • r
v e y ear) T reasury index.
  • c
L and c F are the lifetime cap and
  • r
resp ectiv ely .
  • c
P denotes the ARMs p erio dic cap, i.e., cannot adjust b y more than this amoun t. 32
slide-33
SLIDE 33 New Mortgage P a ymen t
  • Assume
that curren t time is t, the in terest rate, mortgage balance at time t
  • 1
are c(t
  • 1)
and M B t1 .
  • The
new adjuste d in terest rate is c(t).
  • The
new mortgage pa ymen t M P (t) is giv en b y the follo wing expression: M B t1 1
  • (1
+ c(t)) (nt) c(t)
  • Ev
erything else sta ys the same. 33
slide-34
SLIDE 34 Action Items
  • Pic
k y
  • ur
teams (3-4 studen ts) and send e-mail to Je
  • r
me.
  • Pic
k a pap er (Norene Mears has a cop y
  • f
eac h pap er) and cop y it. R eturn the master c
  • py.
  • P
ap ers should b e in London and New Y
  • rk
in 2-3 da ys.
  • Read
the pap er and decide
  • n
the nancial 34
slide-35
SLIDE 35 instrumen ts y
  • u
are going to price.
  • Start
reading ab
  • ut
the nancial instrumen ts y
  • u
are going to price. 35