πŸ“ˆ

FI-3 Lecture 3

Term Structure Models

Multiple Interest Rate Factors

Key Rates & Perturbations

We assume that
P=f(y1,y2,y3,y4)P=f(y_1,y_2,y_3,y_4)
where y1,y2,y3,y4y_1,y_2,y_3,y_4 are the 2-year, 5-year, 10-year and 30-year par coupon rates.
For some security with deterministic cash flows, the first-order approximation can be expressed as
Ξ”P=βˆ‚fβˆ‚y1Ξ”y1+βˆ‚fβˆ‚y2Ξ”y2+βˆ‚fβˆ‚y3Ξ”y3+βˆ‚fβˆ‚y4Ξ”y4\Delta P=\frac{\partial f}{\partial y_1}\Delta y_1 +\frac{\partial f}{\partial y_2}\Delta y_2 + \frac{\partial f}{\partial y_3}\Delta y_3 + \frac{\partial f}{\partial y_4}\Delta y_4
The quantities
βˆ’βˆ‚fβˆ‚yi10,000-\frac{\frac{\partial f}{\partial y_i}}{10,000}
can be thought of as the DV01s corresponding to the i-th key rate.
The quantities
βˆ’βˆ‚fβˆ‚yiP-\frac{\frac{\partial f}{\partial y_i}}{P}
can be thought of as the durations corresponding to the i-th key rate.
The second-order approximation takes the form
Ξ”P=βˆ‘i=14βˆ‚fβˆ‚yiΞ”yi+12βˆ‘i,j=14βˆ‚2fβˆ‚yiβˆ‚yjΞ”yiΞ”yj\Delta P=\sum_{i=1}^4 \frac{\partial f}{\partial y_i}\Delta y_i+\frac{1}{2}\sum_{i,j=1}^4\frac{\partial^2 f}{\partial y_i\partial y_j}\Delta y_i\Delta y_j
with multiple factors, duration is described by a vector and convexity is described by a matrix.
Key Rate Perturbations
Let’s assume that a change of one basis point in each individual key rate yiy_i leads to a change in the par-coupon yield curve of Yi(t)Y_i(t) basis points where the function YiY_i are as shown in the graphs below (Note that all the impacts in the graphs are our assumptions.)
notion image
notion image
notion image
notion image
notion image
 
Par Coupon yields and Discount Factors
Note that we can derive the discount factors from the par-coupon yields.
Recall that the par coupon yield y(n+12)y(\frac{n+1}{2}) corresponding to maturity n+12\frac{n+1}{2} is given by
y(n+12)=2[1βˆ’d(n+12)]βˆ‘i=1n+1d(i2)y\Big(\frac{n+1}{2}\Big)=\frac{2[1-d(\frac{n+1}{2})]}{\sum_{i=1}^{n+1}d(\frac{i}{2})}
starting with a par-coupon yield curve y(t)y(t), the discount factors can be computed recursively by
d(12)=11+y(12)2d(n+12)=2βˆ’y(n+12)βˆ‘i=1nd(i2)2+y(n+12)\begin{aligned} d\Big(\frac{1}{2}\Big) &=\frac{1}{1+\frac{y(\frac{1}{2})}{2}}\\ d\Big(\frac{n+1}{2}\Big) &=\frac{2-y(\frac{n+1}{2})\sum_{i=1}^nd(\frac{i}{2})}{2+y(\frac{n+1}{2})} \end{aligned}

Construct a Hedge using Key Rates

Generalization of the Method
Suppose we have sold a security SS whose price is given by
P(S)=f(x1,x2,...,xN)P^{(S)}=f(x_1,x_2,...,x_N)
and we wish to hedge the sale of this security using a portfolio of basic securities. (Typically we will want NN securities in the hedging portfolio). If the price of the hedging portfolio is given by
P(H)=g(x1,x2,...,xN)P^{(H)}=g(x_1,x_2,...,x_N)
then we choose the amounts of each of the securities in the hedging portfolio so that
βˆ‚βˆ‚xif(0,0,...,0)=βˆ‚βˆ‚xig(0,0,...,0),i=1,2,...,N\frac{\partial }{\partial x_i}f(0,0,...,0)=\frac{\partial}{\partial x_i}g(0,0,...,0),i=1,2,...,N
This is analogous to matching all of the key rate DV01s. (note that xi=0x_i=0 for i=1,2,...,Ni=1,2,...,N corresponding to the original yield curve, i.e., the perturbation is zero).

Risk-Neutral Measure for Binomial Framework

Suppose that there are two trading times t=0,1t=0,1. Assume that there is a zero-coupon bond that will pay $1 at time 1 (i.e. the bond has face value F=1F=1 and maturity T=1T=1.) Denote the price of this bond at time 0 by B0,1B_{0,1}. Assume that there is another basic (traded) security that has initial price S0>0S_0 > 0 and price S1>0S_1>0 at time 1 taking one of two possible different values, each with prob > 0.
Ξ©={H,T}\Omega=\{H,T\}
Besides, assume S(T)<S(H)S(T)<S(H).
Consider a strategy described by (X0,Ξ”0)(X_0,\Delta_0), which means that the total initial capital is X0X_0 and that the portfolio will hold Ξ”0\Delta_0 shares of the risky asset.
Denote the capital of the strategy at time 1 by X1X_1, which is a random variable. The initial capital invested in the risky asset is Ξ”0S0\Delta_0 S_0 and the initial capital invested in the bond is X0βˆ’Ξ”0S0X_0-\Delta_0S_0. The terminal capital is given by
X1(w)=(X0βˆ’Ξ”0S0)1B0,1+Ξ”0S1(w)=(X0βˆ’Ξ”0S0)(1+R0)+Ξ”0S1(w)\begin{aligned} X_1(w) &=(X_0-\Delta_0 S_0)\frac{1}{B_{0,1}}+\Delta_0 S_1(w)\\ &=(X_0-\Delta_0S_0)(1+R_0)+\Delta_0 S_1(w) \end{aligned}
for all w∈{H,T}w\in\{H, T\}.
An arbitrage strategy means
  • X0=0X_0=0
  • P(X1β‰₯0)=1P(X_1\ge0)=1
  • P(X1>0)>0P(X_1>0)>0
That is, a strategy is an arbitrage iff
  • X0=0X_0=0
  • X1(T)β‰₯0,X1(H)β‰₯0X_1(T)\ge0,X_1(H)\ge0
  • X1(T)+X1(H)>0X_1(T)+X_1(H)>0
Absence of Arbitrage
S1(T)S0<1B0,1<S1(H)S0\frac{S_1(T)}{S_0}<\frac{1}{B_{0,1}}<\frac{S_1(H)}{S_0}
If we put
u=S1(H)S0,  d=S1(T)S0,  R0=1B0,1βˆ’1u=\frac{S_1(H)}{S_0},\ \ d=\frac{S_1(T)}{S_0},\ \ R_0=\frac{1}{B_{0,1}}-1
then
d<1+R0<ud<1+R_0<u
Replication and Pricing
Consider a derivative security that pays the amount V1(w)V_1(w) at time 1. Use it the replicate the payoff of the security by a portfolio with initial capital X0X_0 and Ξ”0\Delta_0 shares of the risky asset. Thus X0,Ξ”0X_0,\Delta_0 satisfies
(X0βˆ’Ξ”0S0)(1+R0)+Ξ”0S1(H)=V1(H),(X0βˆ’Ξ”0S0)(1+R0)+Ξ”0S1(T)=V1(T)\begin{aligned} (X_0-\Delta_0S_0)(1+R_0)+\Delta_0S_1(H)&=V_1(H),\\ (X_0-\Delta_0S_0)(1+R_0)+\Delta_0S_1(T)&=V_1(T) \end{aligned}
we can solve for Ξ”0\Delta_0
Ξ”0=V1(H)βˆ’V1(T)S1(H)βˆ’S1(T)\Delta_0=\frac{V_1(H)-V_1(T)}{S_1(H)-S_1(T)}
thus
X0(1+R0)=V1(H)+Ξ”0[S0(1+R0)βˆ’S1(H)]=p~V1(H)+q~V1(T)\begin{aligned} X_0(1+R_0) &=V_1(H)+\Delta_0[S_0(1+R_0)-S_1(H)]\\ &=\tilde{p}V_1(H)+\tilde{q}V_1(T) \end{aligned}
where
p~=S0(1+R0)βˆ’S1(T)S1(H)βˆ’S1(T)q~=S1(H)βˆ’S0(1+R0)S1(H)βˆ’S1(T)\begin{aligned} \tilde{p}&=\frac{S_0(1+R_0)-S_1(T)}{S_1(H)-S_1(T)}\\ \tilde{q}&=\frac{S_1(H)-S_0(1+R_0)}{S_1(H)-S_1(T)} \end{aligned}
and p~+q~=1\tilde{p}+\tilde{q}=1.
We can define another probability measure P~\tilde{P} on Ξ©\Omega by
P~(H)=p~,  P~(T)=q~\tilde{P}(H)=\tilde{p},\ \ \tilde{P}(T)=\tilde{q}
which is called the risk-neutral measure or pricing measure.
Since
X0(1+R0)=E~(V1)X_0(1+R_0)=\tilde{E}(V_1)
thus the arbitrage-free time-0 price of VV is given by
V0=E~[V11+R0]V_0=\tilde{E}\Big[\frac{V_1}{1+R_0}\Big]
Remark:
It is customary to define
u=S1(H)S0,  d=S1(T)S0u=\frac{S_1(H)}{S_0},\ \ d=\frac{S_1(T)}{S_0}
thus the no aribitrage condition is
d<1+R0<ud<1+R_0<u
and the risk-neutral probabilities are given by
p~=1+R0βˆ’duβˆ’d,  q~=uβˆ’(1+R0)uβˆ’d\tilde{p}=\frac{1+R_0-d}{u-d},\ \ \tilde{q}=\frac{u-(1+R_0)}{u-d}
For simplicity, questions in this course all assume that p~=q~=0.5\tilde{p}=\tilde{q}=0.5. But in more complex situations, we may calculate different p~,q~\tilde{p},\tilde{q}.

N-Period Binomial Models

In the N-period binomial model, there will be N+1 trading dates 0, 1, 2, …, N. Denote Bn,mB_{n,m} for the price at time n of a ZCB that pays $1 at time mβ‰₯nm\ge n. Denote RnR_n for the spot rate that will prevail at time nn for loans to be settled at time n+1n+1 (one-period compounding).
Bn,n+1=11+RnB_{n,n+1}=\frac{1}{1+R_n}
Note that the interest rates RnR_n are known at time nn, but not earlier.
Short Rate Models v.s. Whole-Yield Models
Short rate models, also called equilibrium models, have no arbitrage (although they may give rise to negative interest rates). The spot rate curve is an output of the mode, rather than an input. Typically there are parameters in the model that are chosen to try to match the spot-rate curve as closely as possible.
Whole-Yield models, also called arbitrage-free models, take the current spot-rate as an input, and model the random evolution of the spot-rate curve. A key idea is to use forward rates. Prices are computed using a risk-neutrail measure.

Binomial Short Rate Models

Notations:
  • Dates: n=0,1,...,Nn=0,1,...,N
  • Sample Sapce: Ξ©={H,T}N\Omega=\{H,T\}^N
  • Interest Rate Process: (Rn)0≀n≀Nβˆ’1(R_n)_{0\le n\le N-1}
    • adapted process (values of the process at time n depend only on the information available at time n).
  • Pricing Measure: P~\tilde{P} (risk-neutral measure)
  • Discount Process: (Dn)0≀n≀N(D_n)_{0\le n\le N}
    • predictable (values of the process at time n are determined by the information available at time n-1)
Risk-Neutral Pricing Formula
For a security that makes a single payment of amount
Vm(w1,...,wm)V_m(w_1,...,w_m)
at a given date mm, the prices VnV_n of the security at the dates n=0,1,...,mn=0,1,...,m are given by
Vn=1DnE~n[DmVm]=E~n[DmDnVm]V_n=\frac{1}{D_n}\tilde{E}_n[D_mV_m]=\tilde{E}_n[\frac{D_m}{D_n}V_m]
where
Dm=1(1+R0)(1+R1)...(1+Rmβˆ’1)DmDn=1(1+Rn)(1+Rn+1)...(1+Rmβˆ’1)\begin{aligned} D_m&=\frac{1}{(1+R_0)(1+R_1)...(1+R_{m-1})}\\ \frac{D_m}{D_n} &=\frac{1}{(1+R_n)(1+R_{n+1})...(1+R_{m-1})} \end{aligned}
In particular, the time-0 price is given by
V0=E~[DmVm]V_0=\tilde{E}[D_mV_m]
Remark:
  • Note Dnβ‰ dnD_n \ne d_n, because dnd_n is known at time 0, but as for DnD_n, R1,...,Rnβˆ’1R_1,...,R_{n-1} are not known at time 0.
  • the discount factor for time nn is given by
    • d(n)=E~[Dn]d(n)=\tilde{E}[D_n]
      Given an interest rate process and a pricing measure, we can compute the zero-coupon yield curve (spot-rate curve).

Four Binomial Models

In all 4 of these models, we take the risk-neutral measure to be a binomial product measure (be H or T with prob of p & q), with probability of heads equal to 0.5. We put
Ξ”Rn=Rn+1βˆ’Rn\Delta R_n=R_{n+1}-R_n
Ho-Lee Model
Ξ”Rn=Ξ»n+1+ΟƒXn+1Rn=R0+βˆ‘i=1nΞ»i+ΟƒMn\begin{aligned} \Delta R_n &=\lambda_{n+1}+\sigma X_{n+1}\\ R_n &=R_0+\sum_{i=1}^n\lambda_i + \sigma M_n \end{aligned}
Remark:
  • Ξ»1,Ξ»2,...\lambda_1,\lambda_2,... are called drift parameters and Οƒ\sigma is called the volatility parameter.
  • The Ho-Lee Model can be written in the form
    • Rn(w1,...,wn)=an+bβ‹…#H(w1,...,wn)R_n(w_1,...,w_n)=a_n+bΒ·\#H(w_1,...,w_n)
      where
      an=R0βˆ’Οƒn+βˆ‘i=1nΞ»i,  b=2Οƒa_n=R_0-\sigma n +\sum_{i=1}^n \lambda_i,\ \ b=2\sigma
Ho-Lee with Constant Drift:
Ξ”Rn=Ξ»+ΟƒXn+1Rn=R0+Ξ»n+ΟƒMn=R0+Ξ»n+Οƒ[(nβˆ’#Hn)βˆ—(βˆ’1)+1βˆ—#Hn]=R0+(Ξ»βˆ’Οƒ)n+2Οƒ#Hn\begin{aligned} \Delta R_n &=\lambda+\sigma X_{n+1}\\ R_n &=R_0+\lambda n + \sigma M_n\\ &=R_0+\lambda n + \sigma [(n-\#H_n)*(-1)+1*\# H_n]\\ &=R_0+(\lambda-\sigma)n + 2\sigma\# H_n \end{aligned}
Vasicek Model
Binomial Verision of Vasicek Model: 0<k<10<k<1
Ξ”Rn=k(ΞΈβˆ’Rn)+ΟƒXn+1\Delta R_n=k(\theta-R_n)+\sigma X_{n+1}
The solution of this difference equation is given by
Rn=(1βˆ’k)n(R0βˆ’ΞΈ)+ΞΈ+Οƒβˆ‘j=1n(1βˆ’k)nβˆ’jXjR_n=(1-k)^n(R_0-\theta)+\theta+\sigma\sum_{j=1}^n(1-k)^{n-j}X_j
where XjX_j is a random walk.
Remarks:
  • This model exhibits mean reversion
  • ΞΈ\theta is called the long-term mean, and kk is called the speed of mean reversion
Hull-White Model
Binomial Version of Hull-White Model: 0<k<10<k<1
Ξ”Rn=k(ΞΈn+1βˆ’Rn)+Οƒn+1Xn+1\Delta R_n=k(\theta_{n+1}-R_n)+\sigma_{n+1}X_{n+1}
Black-Derman-Toy Model
Ξ”(ln⁑Rn)=βˆ’Ξ”ΟƒnΟƒn(ln⁑θn+1βˆ’ln⁑Rn)+Οƒn+1Xn+1\Delta (\ln R_n)=-\frac{\Delta \sigma_n}{\sigma_n}(\ln \theta_{n+1}-\ln R_n)+\sigma_{n+1}X_{n+1}
which can be rewritten as
Rn(w1,...,wn)=anbn#H(w1,...,wn)R_n(w_1,...,w_n)=a_nb_n^{\#H(w_1,...,w_n)}
 

Pricing

Zero Coupon Bonds
For each n,m∈{0,1,...,N}n,m\in \{0,1,...,N\} with 0≀n≀m0\le n\le m, let Bn,mB_{n,m} denote the price at time nn of a zero coupon bond with maturity mm and face value $1. Then
Bn,m=1DnE~n[Dm]B_{n,m}=\frac{1}{D_n}\tilde{E}_n[D_m]
where (Dn)0≀n≀N(D_n)_{0\le n\le N} is the discount process. In particular,
Bn,n=1B_{n,n}=1
Coupon Bonds
Given q>βˆ’1q>-1 and 0≀n≀m≀N0\le n\le m\le N, let Cn,mqC_{n,m}^q denote the price at time nn of a coupon bond with maturity mm, face value $1 and one-period coupon rate qq. If n>0n>0, we make the convention that Cn,mqC_{n,m}^q is the price after the coupon payment is received at time nn. Then
Cn,mq=Bn,m+qβˆ‘i=n+1mBn,iC_{n,m}^q=B_{n,m}+q\sum_{i=n+1}^m B_{n,i}
provided that mβ‰₯n+1m\ge n+1.
General Security with Adapted Cash Flows
Let (An)1≀n≀m(A_n)_{1\le n\le m} be an adapted process. The price at time 0 of a security that pays the amount AnA_n at each of the times n=1,2,...,Nn=1,2,...,N is given by
E~[βˆ‘n=1mDnAn]\tilde{E}\Big[\sum_{n=1}^mD_nA_n\Big]

One-Period Forward Rates

For 0≀n≀m≀Nβˆ’10\le n\le m\le N-1, denote Fn,mF_{n,m} as the interest rate agreed upon at time nn for a loan initiated at time mm and settled at time m+1m+1. If $1 is the amount borrowed at time mm then the amount to be repaid at time m+1m+1 is 1+Fn,m1+F_{n,m}. Then
Bn,m=(1+Fn,m)Bn,m+1B_{n,m}=(1+F_{n,m})B_{n,m+1}
note Fm,m=RmF_{m,m}=R_m.

Backward Induction

Sinlge Payment
Consider a security that makes a single payment of amount VmV_m at time mm. Start at time mm and work backward by using
Vn=11+RnE~n[Vn+1]V_n=\frac{1}{1+R_n}\tilde{E}_n[V_{n+1}]
for n=mβˆ’1,mβˆ’2,...,1,0n=m-1,m-2,...,1,0
Multiple Payments
Consider a security that makes payments An(w1,...,wn)A_n(w1,...,w_n) at each of the times n=1,2,...,mn=1,2,...,m. We can determine the price at time 0 by viewing this as a string of mm securties each making one payment and pasting together the results.
Let VnV_n be the price of the security at time nn after the payment AnA_n has been made. The backward induction is
{Vn=11+RnE~n[An+1+Vn+1]=An+11+Rn+11+RnE~[Vn+1]Vm=0\begin{cases} V_n=\frac{1}{1+R_n}\tilde{E}_n[A_{n+1}+V_{n+1}]=\frac{A_{n+1}}{1+R_n}+\frac{1}{1+R_n}\tilde{E}[V_{n+1}]\\ V_m=0 \end{cases}
for n=0,1,...,mβˆ’1n=0,1,...,m-1.
Notice that with multiple payments, the discounted price process will not be a martingale under P~\tilde{P}.
Number of Heads as a State Variable
Simplification is possible when the short spot rate at time nn depends only on ww only through the number of heads up to time nn, i.e.
Rn(w1,...,wn)=rn(#H(w1,...,wn))R_n(w_1,...,w_n)=r_n(\#H(w_1,...,w_n))
and the pricing measure is a binomial product measure (P~\tilde{P}).
For this scenario, for a security that makes a single payment VmV_m at time mm and the payment can be expressed as function of #H(w1,...,wm)\# H(w_1,...,w_m)
Vn(w1,...,wn)=vn(#H(w1,...,wn)),  βˆ€n=0,1,...,mβˆ’1V_n(w_1,...,w_n)=v_n(\# H(w_1,...,w_n)),\ \ \forall n=0,1,...,m-1
The functions vnv_n can be computed by backward induction:
vn(k)=11+rn(k)[p~vn+1(k+1)+q~vn+1(k)]v_n(k)=\frac{1}{1+r_n(k)}[\tilde{p}v_{n+1}(k+1)+\tilde{q}v_{n+1}(k)]
where p~,q~\tilde{p},\tilde{q} are the risk-neutral probabilities of heads and tails, kk is the number of heads up to time nn.
For a security that pays AnA_n at each of the times n=1,2,3,...,mn=1,2,3,...,m where each AnA_n can be expressed as a function of either #H(w1,...,wn)\#H(w_1,...,w_n) or #H(w1,...,wnβˆ’1)\# H(w_1,...,w_{n-1}), i.e.
An=an(#H(w1,...,wnβˆ’1)),  n=1,2,....,mA_n=a_n(\# H(w_1,...,w_{n-1})),\ \ n=1,2,....,m
Then the function vnv_n can be computed by backward induction:
vn(k)=an+1(k)1+rn(k)+11+rn(k)[p~vn+1(k+1)+q~vn+1(k)]v_n(k)=\frac{a_{n+1}(k)}{1+r_n(k)}+\frac{1}{1+r_n(k)}[\tilde{p}v_{n+1}(k+1)+\tilde{q}v_{n+1}(k)]