Valuation and Risk Models

Valuation and Risk Models

• Bond Valuation

• Option Valuation

• Market Risk Model

• Credit Risk Model

• Operational Risk Model

• Country Risk
Bond Valuation
Basics ★★ 性质、计算

A t-period spot rate, or zero rate, denoted as zt, is

Spot Rate the yield to maturity on a zero-coupon bond that
matures in t-years.
Interest rates corresponding to a future period
implied by Tthe spot curve
T T 
， ft1,t2. T
 1  R 1   1  F1,2 
1 2 1
 1  R2  2

Forward Rate
F1,2  T2  T1  R 2 T2  R 1T1
e R 1T1  e  e R 2 T2  F1,2 
T2  T 1

d(t), for a term of (t) years, gives the present value

of one unit ofd currency 1 to be received at the end
t  ($1)
2t
of that term.  z t 
 1 
2 
Discount Factor


Internal rate of return found by equating the

Yield to Maturity present value of the cash flows to the current price of
the security.
Spot & Forward & Yield Curve ★★★ 性质、计算

Spot Interest Rate Interest Rate

&Forward &Yield 9 9
Curve 8 Forward Curve 8
7 Spot Curve 7
Yield Curve
6 6
Yield Curve
5 5 Spot Curve
Forward Curve
4 4
3 3
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Maturity(Year) Maturity(Year)
Upward-Sloping Term Structure Downward-Sloping Term Structure
Reinvestment Risk and Interest Rate ★★ 性质

Risk
Reinvestment Risk Future interest rates can be less than the yield to maturity at the time
bond is purchased, known as reinvestment risk.
Interest Rate Risk If the bond is not held to maturity, the investor faces the risk that he may
have to sell for less than the purchase price, resulting a return that is less
than the yield to maturity, known as interest rate risk.
Valuation ★★★ 性质、计算

CF1 CFT
Spot Rate Pr ice   ... 
1  z1  T
T
1  z
Forward Rate
CF1 CFT
Coupon Price  ... 
Bond
Discount Factor 1f0,1  1f0,1  1f1,2 ... 1fT1,T 
Bond Replication: Absent  CF1  d  1 factors,
Pr iceconfounding T  d T 
...  CFidentical sets
of cash flows should sell for the same price.

Valuation ★★ 性质、计算

   
2T

c  1   c
Annuity A  1     Perpetuity Perpetuity  F
y  y  y
   1   
 2  
Price Approximation

The actual, exact price P  f(y 0  Δy)

The duration estimate
P  P0  D*P0 Δy
The duration and convexity 1
estimate P  P0  D P0 Δy  CP0 (Δy)2
*

2
1
P(y 0  y)  P(y 0 )  f '(y 0 )y  f ''(y 0 )( y)2  …
2

dollar duration dollar convexity

Option Valuation
Binomial Trees ★★★ 性质、计算

Risk-Neutral Valuation: All cash flows should be

discounted to present using the risk-free interest rate.
Principles
Assumption: 1) Stock price follows geometric Brownian
motion; 2) u = 1/d.
t
One-Step
u  eσ ;d  e σ t

rt  r  q t f  pfu   1  p  fd  e rt

Calculation e d e d
p or p 
Two-Step ud ud
Calculation

f e 2rt
Makesure
p 2
f  2p(1  p)f  (1  p)2
f 
American uu that the option
ud value ddat each node is no less
Options
than the intrinsic value.
BSM Model ★★★ 性质、计算

The price of the underlying asset follows a lognormal

distribution.
Key The risk-free rate is known and constant.
Assumptions The volatility of the underlying asset is known and
constant.
c  SN  d   KerT
N d  or c  Se
The option valued are European Options.
1 2
qT
N d1   KerT
N d2 

Call p  KerTN d2   SN d1  or p  KerTN d2   SeqTN d1 

Calculation Put
Where: d1,2 

ln S Kert   1σ T or d1,2 

ln SeqT Ke rT  σ T
σ T 2 σ T 2

BSM Model for Early Exercise of American Options ★★ 性质、计算

For American call


Dn  X 1 e
r  T  tn 
 For American put

Dn  X 1 e
r Ttn 

The Greek Letters ★★★ 性质、计算

Ratio of the price change of the

option to the price change of the
underlying assets.
Call: 0 ~ 1; Put: -1~0
Long call: >0; Short call: <0; Long
put: <0; Short put: >0.
When t→T, delta is unstable.
Delta For at the money call, delta = 0.5
Portfolio Delta: Summation of the
product of each option position and its
delta. c c
Δ
 N(d1 ) Δ  eqTN(d1 )
S S
Delta Hedge: A position with a deltapof zero is calledpa delta neutral
Δ  N(d1 )  1 Δ   eqT  N(d1 )  1
position. To maintain a delta neutral
S position, the trader
S must re-
balance the portfolio.
The Greek Letters ★★★ 性质

Rate of change of the option’s delta

with respect to the price of the underlying
asset.
Providing added protection against
large movements in the underlying asset's
price.
Gamma is largest when an option is at-
Gamma
the-money.
Long call: >0; Short call: <0; Long put:
>0, Short put: <0.

Application in Delta Hedge: If gamma is higher, rebalancing more

frequently is required. On the other hand, when gamma is lower,
rebalancing is required less frequently.
The Greek Letters ★★ 性质

Rate of change of the value of the option

with respect to the volatility of the underlying
Vega asset.
Most sensitive to changes in volatility
when they are at the money.

Rate of change of the value of the option with respect to the

Theta passage of time with all else remaining the same.
Short-term at the money option has a greatest negative theta.

Sensitivity to the interest rates.

Rho In the money calls and puts are more sensitive to changes in
rates than out-of-the-money options.
Market Risk Model
Risk Measures Properties ★★ 性质

 Monotonicity : A random cash flow or future value R1 that is always greater

than R2 should have a lower risk.
R1  R2,then   R1     R2 

 Subadditivity: The portfolio’s risk should not be greater than the sum of
its parts.
Coherent  R1  R2     R1     R2 
Properties
 Positive Homogeneity : The risk of a position is proportional to its scale
or size.
  0,   R     R 

  R  c     RInvariance
 Translation  -c : Like adding cash for constant c,

Expected Shortfall ★★★ 性质、计算

 Conditional VaR ; Sub-additive ; Average of the worst 100×(1-α)% of losses.

Value at Risk ★★★ 性质、计算

VaR is the maximum loss over a target horizon and for a given confidence level.

Definition

 Time horizon: should depend on how quickly portfolio can be unwound.

Regulators in effect use 1-day for bank market risk and 1-year for
Parameters credit/operational risk.
 Confidence level: depends on objectives. Regulators use 99% for market risk
and 99.9% for credit/operational risk.
 Data Window

Properties  Not contain worst conditions/tail loss.

 Not sub-additive.

Calculation

VaR X%  E R  zX%  σ

VaR  X% Jdays  VaR  X% 1days  J

Value at Risk for Derivatives ★★★ 性质、计算

 Linear Approximation
VaR dP  D*P  VaR  dy 
Delta-Normal
VaR df   Δ  VaR dS

 Nonlinear Approximation 1
VaR  dP   D P  VaR  dy    C  P   VaR  dy 
* 2

2
Delta-Gamma 1
VaR  df   Δ  VaR  dS   Γ  VaR  dS 
2

 Full re-pricing of the portfolio under the assumption

that the underlying risk factor(s) are shocked to
experience a loss.
Full Revaluation
 Accurate but computationally burdensome.
- Historical/Bootstrap Simulation
- Monte Carlo Simulation
Volatility Estimation ★★★ 性质

 Cause: Volatility is time-varying.

Fat Tails
 Solution: Regime-switching volatility model

 Parametric Model
Assumes asset returns are normally or lognormally
distributed with time-varying volatility

Historical-Based Approach  Nonparametric Model

No underlying assumptions

 Hybrid Approach
Both parametric and nonparametric method

Uses derivative pricing models such as the Black-

Implied Scholes-Merton option pricing model to estimate an
Volatility-Based Approach
implied volatility based on current market data rather
Credit Risk Model
EL and UL ★★★ 性质、计算

 Expect to lose on average AE  OS    COMU

over a predetermined EL  PD  EA  LR
EL period of time.
n
 Portfolio EL is linear and ELP   ELi
i 1
additive.
 Risk arises from the UL  EA  PD  LR
2
 LR 2PD
2

variation in loss levels - 2  PD  1  PD

PD  
which for credit risk is due
UL to unexpected losses. ULP
 Portfolio UL need consider  2UL2  2UL2  2  UL UL
1 1 2 2 1 2 1 2
the effects of
UL21  1,2UL1UL2 n
diversification.
ULC1  ULP   ULCi  ULC1  ULC2
ULP i 1
UL Contribution
Operational Risk Model
Operational Risk ★★★ 性质、计算

 Inadequate or failed internal processes, people, and systems

Classification
or from external events.
 Loss Frequency Distribution: The distribution of the
number of losses observed during the time horizon. A
common probability distribution is the Poisson Distribution.
 Loss Severity Distribution: Is the distribution of the size of
Loss Distribution a loss, given that a loss occurs. It is typically assumed that
loss severity and loss frequency are independent. For the
loss severity distribution, a lognormal probability
distribution is often uses.
 Loss Distribution: Monte Carlo Simulation
Data Issues ★★★ 性质

Frequency: Severity:
Internal Internal & External, encourage using scenario analysis.
Operational Risk Regulatory Capital ★★★ 性质、计算

Basic Indicator 
i  last three years
 GIi  α 
Approach
K operational,BIA 
3
Corporate Finance (18%) Commercial Banking (15%)
Payment and settlement (18%) Asset Management (12%)
Trading and Sales (18%) Retail Banking (12%)
Standardized Agency Services (15%) Retail Brokerage (12%)
Approach

 
  max   line18
 GI  βline1 8 ,0 

 
i  last three years
K operational,SA
3
AMA  The capital charge for AMA is calculated as the bank’s operational value at
Approach risk with a one-year horizon and a 99.9% confidence level.
Features of Stress Testing ★★ 性质

Advantage Disadvantage

 Complements value at risk.  Highly subjective.

 Simple and intuitive.  Implausible scenarios.
 Directly examines the tails.  Difficult to interpret.

Types of Stress Testing ★★ 性质

 Unidimensional scenarios focus ”stressing” on key one variable. The key

Unidimensional weakness of a unidimensional analysis is that scenarios cannot account for

And correlation.
Multidimensional  Multidimensional attempts to stress multiple variables and their
relationships.
Country Risk
Sources of Country Risk ★★ 性质

Sources Effect
Economic
 More mature markets/companies within those markets are less
Growth Life
Cycle risky than those firms/countries in the early stages of growth.

 Continuous versus Discontinuous Risk (Risks in democracies vs

Risks in dictatorships)
Political Risk  Corruption
 Physical Violence
 Nationalization and Expropriation Risk.

 The protection of property rights and the speed with which

Legal Risk
disputes are settled affect risk.

Economic  A disproportionate reliance on a single commodity or service in

Structure
an economy increases a country’s risk exposure.
Sovereign Default ★★ 性质

 Foreign Currency Default

- Countries are often without the foreign currency to meet the debt
obligation and are unable to print money to repay the debt. This
makes up a large proportion of sovereign defaults.

 Local Currency Default:

- The use of the gold standard prior to 1971 made it more difficult
Causes
for some countries to print money;
- Shared currencies make it impossible for countries to control their
own monetary policy;
- Some countries must conclude that the costs of currency
debasement and potentially higher inflation are greater than the
costs of default.
Sovereign Default ★★ 性质

 GDP growth  Trade retaliation

Consequences  Sovereign ratings and  Fragile banking systems
borrowing costs  Political change
 The country’s level of indebtedness
 Pension funds and social services
 Tax receipts
Influence Factor
 Stability of tax receipts
 Political risk
 Backing from other countries/entities
 Changes occur in real time
Advantage  Granularity
Sovereign
 Adjust quickly to new information
Default Spread
 Need for a risk-free security
Disadvantage  Cannot compare local currency bonds
 Greater volatility