to change the probability measure one uses Girsanov theorem (formula):

$$ \frac{N_a(0)}{N_a(T)} d \mathbb{P}_a=\frac{N_b(0)}{N_b(T)} d \mathbb{P}_b$$

$$N_a$$ is numeraire (price of any non dividend paying asset, usually bond or bank account) and it’s correspoding probability measure is $$P_a$$

$$N_b$$ is another numeraire with it’s probability measure $$P_b$$

## Applications

### expectation of short rate is forward rate under T-forward measure

– prove that $$E^T(r_T|F_t)=f_{T->T+dT}(t)$$

where $$E^T$$ is a expectation under T-forward measure (where bond $$P_{->T}$$ is numeraire

and $$f_{T->T+dT}(t)$$ is an instantaneous forward rate for time T as seen at time t

– write bond price in two forms:

$$P_{->T}(t)=E^{RN}(e^{-\int_t^Tr_sds}|F_t)$$

$$P_{->T}(t)=e^{-\int_t^Tf_s(t)ds}$$

so

$$E^{RN}(e^{-\int_t^Tr_sds}|F_t)=e^{-\int_t^Tf_s(t)ds}$$

1) differentiate both parts by T

2) change measure from Risk-neutral to T-forward measure from time t to time T (as usual rule apply girsanov formula between times when the process is stochastic ,in this case not between 0 and t for example)

### Black Scholes formula

to calculate the term $$E(S_T 1_{S_T>K})$$ one uses girsanov to pass from risk neutral measure to forward measure where $$S_t$$ is numeraire

### Libor In Arrears Convexity adjustment

Simple example libor in arrears convexity adjustment to change from T1 measure to T2 measure