Bruno Dupire

Bloomberg LP

Lecture 9
Volatility
 Forward Equations (1)
• BWD Equation:
price of one option C (K 0 ,T0 ) for different (S, t )
• FWD Equation:
price of all options C (K , T ) for current (S 0 ,t0 )
• Advantage of FWD equation:
– If local volatilities known, fast computation of implied
volatility surface,
– If current implied volatility surface known, extraction of
local volatilities,
– Understanding of forward volatilities and how to lock
them.

Bruno Dupire 2
 Forward Equations (2)
• Several ways to obtain them:
– Fokker-Planck equation:
• Integrate twice Kolmogorov Forward Equation
– Tanaka formula:
• Expectation of local time
– Replication
• Replication portfolio gives a much more financial
insight

Bruno Dupire 3
 Fokker-Planck

• If dx = b( x, t )dW
∂ϕ 1 ∂ 2 (b 2ϕ )
• Fokker-Planck Equation: =
∂t 2 ∂x 2
∂ 2C
• Where ϕ is the Risk Neutral density. As ϕ =
∂K 2
2  ∂C 
 ∂ 2
C   ∂ 2
C
 ∂ 2  ∂  b 
2 2
∂  2 
 ∂t  =  ∂K  = 1  ∂K 
∂x 2 ∂t 2 ∂x 2

• Integrating twice w.r.t. x: ∂C b2 ∂2C

=
∂t 2 ∂K 2

Bruno Dupire 4
 FWD Equation: dS/S = σ(S,t) dW

δT C K ,T + δ T − C K ,T
Define CS K ,T ≡
δT
CK ,T +δT
CK ,T ST
K
CS Kδ T,T at T
dT
dT/2 σ 2 (K , T )
δT → 0 K 2δ K ,T
2

ST ST

K K

Equating prices at t0: ∂C

=
σ 2
( K , T ) K 2 ∂ 2
C
∂T 2 ∂K 2
Bruno Dupire 5
FWD Equation: dS/S = r dt + σ(S,t) dW

TV IV S T − Ke − rδ T
CS Kδ T,T at T = Time Value + Intrinsic Value
ST − K
(Strike Convexity) (Interest on Strike)
CK ,T
CK ,T +δT

Ke−rδT K ST

σ 2 (K , T )
K 2δ K ,T
2

IV rK δT → 0
TV
rKDigK ,T
K ST
Ke−rδT K ST

Equating prices at t0: ∂C σ 2 (K , T ) 2 ∂ 2C ∂C

= K − rK
∂T 2 ∂K 2
∂K
Bruno Dupire 6
 FWD Equation: dS/S = (r-d) dt + σ(S,t) dW

ST − Ke−rδT
CS KδT,T at T =
TVIV
TV + Interests on K
ST e − dδT − Ke − rδT – Dividends on S
CK,T+δT
CK,T
ST
(d−r)δT K
Ke
σ 2 (K , T )
K 2δ K ,T
δT → 0 2

(r − d )K (r − d ) K DigK ,T
Ke(d −r )δT K ST K ST
− d ⋅ CK ,T
− d ⋅ CK ,T

Equating prices at t0: ∂C σ 2 (K , T ) ∂ 2C ∂C

= K 2
− (r − d )K − d ⋅C
∂T 2 ∂K 2
∂K
Bruno Dupire 7
 Stripping Formula

∂C σ 2 (K , T )K 2 ∂ 2C ∂C
= − (r − d )K − d ⋅C
∂T 2 ∂K 2
∂K
– If σ (K , T ) known, quick computation of all CK ,T (S 0 ,t0 )
today,
– If all CK ,T (S0 ,t0 ) known:
∂C ∂C
2 + (r − d )K + dC
σ (K , T ) = ∂T ∂K
2 ∂ C
2
K
∂K 2
Local volatilities extracted from vanilla prices
and used to price exotics.

Bruno Dupire 8
 Smile dynamics: Local Vol Model (1)
• Consider, for one maturity, the smiles associated
to 3 initial spot values

Skew case Local vols

Smile S −
Smile S 0
Smile S +

S − S0 S + K

– ATM short term implied follows the local vols

– Similar skews

Bruno Dupire 9
 Smile dynamics: Local Vol Model (2)

• Pure Smile case

Local vols

Smile S +
Smile S −

Smile S 0

S− S0 S+ K

– ATM short term implied follows the local vols

– Skew can change sign

Bruno Dupire 10
 Summary of LVM Properties

Σ0 is the initial volatility surface

• σ (S,t ) compatible with Σ0 ⇔σ = local vol

•σ (ω ) compatible with Σ ⇔ E[σ

0 2 ST = K ] = (local vol)²

• σ̂ k,T deterministic function of (S,t)

⇔ future smile = FWD smile from local vol

Bruno Dupire 11
Volatility Replication
 Volatility Replication

dS
= σ t dW Apply Ito to f(S,t).
S
df = f S dS + f t dt + 12 f SSσ t2 S 2 dt
T
 T T

⇒∫ f SS ( St , t )σ S dt = 2  f ( ST , T ) − f ( S 0 ,0) − ∫ f t ( St , t )dt − ∫ f S ( St , t )dSt 
t
2 2

0  0 0 

European PF ∆-hedge

T
g
To replicate ∫ g ( S , t )σ dt ,find f : g ( S , t ) = f SS ( S , t ) S
2 2
: f = ∫∫
0
t
S2

Bruno Dupire 13
 Examples
Variance Swap S
g (S , t ) = 1 f ( S , t ) = − ln( )
S0
S
Corridor g ( S , t ) = 1[ a ,b ] ( St )
f ( S , t ) = − ln(
S0
) on [a,b]

Variance Swap + linear extrapolation

FWD Variance S
g ( S , t ) = 1[T1 ,T2 ] (t ) f (S, t) = −ln( )×1[T1,T2 ] (t)
S0
Swap
Absolute (S − S0 ) 2
g (S , t ) = S 2 f (S , t ) =
Variance Swap 2
Local Time at (S − K )+
g (S , t ) = δ K (S ) f (S , t ) =
level K K2

Bruno Dupire 14
 Conditional Instantaneous FWD
Variance
From local time:
T 2  C(K ,T )
E  ∫ σ t δ K ( S )dt  = 2 ×
0  K2

Differentiating wrt T:

[ ] [ ]
E σ T2δ K ( ST ) = E σ T2 ST = K ⋅ E [δ K ( ST )] =
K
2 ∂C
2
×
∂T
(K ,T )

∂ 2C
And, as: E [δ K ( ST )] = (K ,T )
∂K 2

∂C
(K ,T )
[ ]
K
2
E σ T2 ST = K = 2 × ∂2T
∂C
= σ loc
2
(K ,T )
(K ,T )
∂K 2
Bruno Dupire 15
 Deterministic future smiles
It is not possible to prescribe just any future
smile
If deterministic, one must have
C K ,T (S 0 , t 0 ) = ∫ ϕ (S 0 , t 0 , S , T1 ) C K ,T (S , T1 )dS
2 2

Not satisfied in general

K
S0

t0 T1 T2

Bruno Dupire 16
 Det. Fut. smiles & no jumps
=> = FWD smile
If ∃(S , t , K , T ) / VK ,T (S , t ) ≠ σ (K , T ) ≡ lim σ imp (K , T , K + δK , T + δT )
2 2
δK → 0
δT → 0
stripped from SmileS.t
K
Then, there exists a 2 step arbitrage:
Define ∂ 2C S0
( (K , T ) − V (S , t )) ∂K (S , t , K , T )
PL t ≡ σ
2
K ,T 2 S
(
At t0 : Sell PL t ⋅ Dig S − ε ,t − Dig S + ε ,t ) t0 t T
At t: if S ∈ [S − ε , S + ε ] 2
t buy 2
CS K, T , sell σ 2
(K , T )δ K ,T
K
gives a premium = PLt at t, no loss at T

Conclusion: VK ,T (S , t ) independent of (S , t ) = VK ,T (S0 , t0 ) = σ 2 (K , T )

from initial smile

Bruno Dupire 17
 Consequence of det. future smiles

• Sticky Strike assumption: Each (K,T) has a fixed σ impl ( K , T )

independent of (S,t)
• Sticky Delta assumption: σ impl ( K , T ) depends only on
moneyness and residual maturity

• In the absence of jumps,

– Sticky Strike is arbitrageable
– Sticky ∆ is (even more) arbitrageable

Bruno Dupire 18
 Example of arbitrage with Sticky Strike
Each CK,T lives in its Black-Scholes (σ impl ( K , T ) )world
C1 ≡ C K 1 ,T1 C 2 ≡ C K 2 ,T2 assume σ 1 > σ 2
P&L of Delta hedge position over dt:

δ PL (C 1 ) = 1
2
((δ S ) − σ S δ t ) Γ
2
1
2
1

δ PL (C 2 ) = 1
2
((δ S ) − σ S δ t ) Γ
2
2
2
2
Γ1C 2
Γ2 C 1
Γ1Γ2 2 2
δ PL (Γ1C 2 − Γ2 C 1 ) =
2
(
S σ 1 − σ 22 δ t > 0 ) S t1 S t + δt
(no Γ , free Θ )
! If no jump

Bruno Dupire 19
 Arbitraging Skew Dynamics

• In the absence of jumps, Sticky-K is arbitrageable and Sticky-∆ even more so.
• However, it seems that quiet trending market (no jumps!) are Sticky-∆.
In trending markets, buy Calls, sell Puts and ∆-hedge.

Example:
K1 St
PF ≡ C K 2 − PK1
K2

σ 1 ,σ 2
S PF
∆-hedged PF gains
VegaK > Vega K
2 1 from S induced
σ 1 ,σ 2 volatility moves.
S PF
VegaK < Vega K
2 1

Bruno Dupire 20
Skew from Historical Prices
 Theoretical Skew from Prices

?
ð
Problem : How to compute option prices on an underlying without options?
For instance : compute 3 month 5% OTM Call from price history only.
1) Discounted average of the historical Intrinsic Values.
Bad : depends on bull/bear, no call/put parity.
2) Generate paths by sampling 1 day return recentered histogram.
Problem : CLT ðconverges quickly to same volatility for all strike/maturity;
breaks autocorrelation and vol/spot dependency.

Bruno Dupire 22
 Theoretical Skew from Prices (2)

3) Discounted average of the Intrinsic Value from recentered 3 month

histogram.
4) ∆-Hedging : compute the implied volatility which makes the ∆-hedging a
fair game.

Bruno Dupire 23
 Theoretical Skew
from historical prices
How to get a theoretical Skew just from spot price
history? S
Example: K
ST
3 month daily data
1

T1 t T2
1 strike K = k ST1
– a) price and delta hedge for a given σ within Black-Scholes
model
– b) compute the associated final Profit & Loss: PL(σ )
– ( ) ( ( ))
c) solve for σ k / PL σ k = 0
– d) repeat a) b) c) for general time period and average
– e) repeat a) b) c) and d) to get the “theorical Skew”

Bruno Dupire 24
IV. Volatility Expansion
 Introduction
• This talk aims at providing a better
understanding of:

– How local volatilities contribute to the value of

an option
– How P&L is impacted when volatility is
misspecified
– Link between implied and local volatility
– Smile dynamics
– Vega/gamma hedging relationship

Bruno Dupire 26
 Framework & definitions
• In the following, we specify the dynamics
of the spot in absolute convention (as
opposed to proportional in Black-Scholes)
and assume no rates:
dSt = σt dWt
• σ: local (instantaneous) volatility
(possibly stochastic)
• Implied volatility will be denoted by σ$

Bruno Dupire 27
 P&L of a delta hedged option

Option Value P&L

Break-even
points
Delta
hedge

σ ∆t
Ct Ct + ∆t
− σ ∆t Θ S t + ∆t

St S St

Bruno Dupire 28
 P&L of a delta hedged option (2)
Correct Volatility higher than

1st 1st
1st 1st

St S t + ∆t St S t + ∆t

P&L P&L

Expected P&L = 0 Expected P&L > 0

Ito: When ∆t → 0 , spot dependency disappears

Bruno Dupire 29
 Black-Scholes PDE

P&L is a balance between gain from Γ and

σ2  From Black- σ 02

P& L( t ,t +dt ) =  Γ0 + Θ0  dt Scholes PDE: Θ0 = − Γ0
 2  2

=> discrepancy if σ different from

1 2
gain over dt = (σ − σ 02 )Γ0dt
2

• σ > σ 0: Profit 
 Magnified by Γ0
• σ < σ 0: Loss 
Bruno Dupire 30
 P&L over a path
Total P&L over a path
= Sum of P&L over all small time intervals

Gamma
P& L =
1 T (σ 2 − σ 2 ) Γ dt
2 ∫0 0 0

No assumption is made
on volatility so far Time

Spot

Bruno Dupire 31
 General case
• Terminal wealth on each path is:

wealth T = X ( Σ 0 ) + ∫ (σ 2 − σ 02 ) Γ0 dt
1 T

2 0
( X ( Σ ) is the initial price of the option)
0

• Taking the expectation, we get:

[ ]
T ∞
1
E wealthT = X (Σ0 ) + 2 ∫ ∫ E[Γ0 (σ 2 − σ02 )| S]ϕ dSdt
ϕ
0 0

• The probability density φ may correspond to the density of

a NON risk-neutral process (with some drift) with volatility σ.

Bruno Dupire 32
 Non Risk-Neutral world
• In a complete model (like Black-Scholes), the drift does
not affect option prices but alternative hedging strategies
lead to different expectations

Example: mean reverting process

towards L with high volatility around L
We then want to choose K (close to L)
T and σ0 (small) to take advantage of it.
L
Profile of a call (L,T) for
In summary: gamma is a volatility different vol assumptions

collector and it can be shaped by:

• a choice of strike and maturity,
• a choice of σ0 , our hedging volatility.
Bruno Dupire 33
 Average P&L

• From now on, φ will designate the risk neutral density

associated with dSt = σdWt .
•In this case, E[wealthT] is also X ( Σ) and we have:
T ∞
1
X ( Σ ) = X ( Σ 0 ) + 2 ∫ ∫ E [ Γ0 (σ 2 − σ 02 )| S ] ϕ dS dt
0 0

• Path dependent option & deterministic vol:

X ( Σ ) = X ( Σ 0 ) + 21 ∫ ∫ (σ 2 − σ 02 ) E [ Γ0 | S ] ϕ dS dt

• European option & stochastic vol:

C ( Σ ) = C ( Σ 0 ) + 21 ∫ ∫ ( E [σ 2 | S ] − σ 02 ) Γ0 ϕ dS dt

Bruno Dupire 34
 Quiz

• Buy a European option at 20% implied vol

• Realised historical vol is 25%
• Have you made money ?

Not necessarily!
High vol with low gamma, low vol with high gamma

Bruno Dupire 35
 Expansion in volatility
• An important case is a European option with
deterministic vol:

C ( Σ ) = C ( Σ 0 ) + 21 ∫ ∫ (σ 2 − σ 02 ) Γ0 ϕ dS dt

• The corrective term is a weighted average of the volatility

differences

• This double integral can be approximated numerically

Bruno Dupire 36
 P&L: Stop Loss Start Gain

• Extreme case: σ 0 = 0 ⇒ Γ0 = δ K

C ( Σ ) = ( S 0 − K ) + 2 ∫ σ ( K , t ) ϕ ( K , t )dt
T
+ 1 2
0

• This is known as Tanaka’s formula

S
Delta = 100%
K
Delta = 0%

Bruno Dupire 37
 Local / Implied volatility relationship
Differentiation

Local volatility
Implied volatility

31
23
27
21
23
19
19
17
15
15
11
13

strike spot time

maturity

Aggregation
Bruno Dupire 38
 Smile stripping: from implied to local

•Stripping local vols from implied vols is the

inverse operation:
∂C
∂T
σ (S ,T ) = 2 2
2
(Dupire 93)
∂ C
∂ K2

•Involves differentiations

Bruno Dupire 39
 From local to implied: a simple case

Let us assume that local volatility is a

deterministic function of time only:
dSt = σ ( t ) dWt
In this model, we know how to combine local
vols to compute implied vol:
T

σ$ ( T ) =
∫0 ( t )dt
σ 2

Question: can we get a formula with σ (S , t ) ?

Bruno Dupire 40
 From local to implied volatility
• When σ 0 = implied vol

1 ∫∫ σ 2
Γ0ϕ dSdt
∫∫ (σ 2
− σ 0 )Γ0ϕ dSdt = 0
2
⇒ σ0 =
2

2 ∫∫ Γ0ϕ dSdt

• Γ0 depends on σ 0 ⇒ solve by iterations

•Implied Vol is a weighted average of Local Vols

(as a swap rate is a weighted average of FRA)

Bruno Dupire 41
 Weighting scheme
•Weighting Scheme: proportional to Γ0 ϕ

Out of the
At the
money
money
case:
case: Γ0 ϕ
t
S

S0=100 S0=100
K=100 K=110

Bruno Dupire 42
 Weighting scheme (2)
• Weighting scheme is roughly proportional to
the brownian bridge density

Brownian bridge density:

[
BBϕ K ,T ( x, t ) = P St = x ST = K ]

Bruno Dupire 43
 Time homogeneous case
σ = ∫ α ( S )σ ( S )dS
$ 2 2
α( S) =
∫ Γ ϕdt
0

∫∫ Γ ϕ dSdt
0

ATM (K=S0) OTM (K>S0)

α(S) S0 α(S) S0 K
σ T
small
S S

α(S) S0 α(S) S0 K
σ T
large
S S

Bruno Dupire 44
 Link with smile

σ$ K 1
and σ$ K 2
are K2

averages of the same local S0

vols with different weighting K1

schemes t

=> New approach gives us a direct expression for

the smile from the knowledge of local volatilities

But can we say something about its dynamics?

Bruno Dupire 45
 Smile dynamics
Weighting scheme imposes
some dynamics of the smile for S1
a move of the spot: S0 K
For a given strike K,
S↑ ⇒ σ$ K ↓
(we average lower volatilities) t
Smile today (Spot St) 26

&
Smile tomorrow (Spot St+dt) 25.5

in sticky strike model 25

Smile tomorrow (Spot St+dt) 24.5

if σATM=constant 24

Smile tomorrow (Spot St+dt) 23.5

in the smile model

St+dt St

Bruno Dupire 46
 Sticky strike model
A sticky strike model ( σ$ K ( t ) = σ$ K ) is arbitrageable.

Let us consider two strikes K1 < K2

The model assumes constant vols σ1 > σ2 for example

2)
1C(

C(K
K 1)

Γ2 *
2)
K(

Γ1 /
1C(
1C

σ 2 dt
K 1)

σ 2 dt
σ1 dt σ1 dt

By combining K1 and K2 options, we build a position with no gamma and

positive theta (sell 1 K1 call, buy Γ1/Γ2 K2 calls)
Bruno Dupire 47
 Vega analysis
•If σ &σ 0 are constant
1 2
C( Σ ) = C( Σ 0 ) + (σ − σ 02 ) ∫∫ Γ0 ϕ dSdt
2

• σ = σ 02 + ε
2

1
C(σ + ε ) = C(σ ) + ε ∫∫ Γ0ϕdSdt
2
0
2
0
2
{
∂ 2C
∂σ 2
∂C ∂C ∂σ 2 ∂C
Vega = = 2 ⋅ = 2 ⋅ 2σ
∂σ ∂σ ∂σ ∂σ
Bruno Dupire 48
 Gamma hedging vs Vega hedging
• Hedge in Γ insensitive to realised
historical vol
• If Γ=0 everywhere, no sensitivity to
historical vol => no need to Vega hedge
• Problem: impossible to cancel Γ now for
the future
• Need to roll option hedge
• How to lock this future cost?
• Answer: by vega hedging
Bruno Dupire 49
Superbuckets: local change in local vol

For any option, in the deterministic vol case:

X ( Σ) = X ( Σ 0 ) + 21 ∫ ∫ (σ 2 − σ 02 ) E[Γ0 | S ] ϕ dS dt

For a small shift ε in local variance around (S,t), we have:

X ( Σ ) = X ( Σ 0 ) + 21 ε E[ Γ0 | S ] ϕ
dX
⇒ = 21 E[ Γ0 ( S , t )| S ] ϕ ( S , t )
(
d σ (2S ,t ) )
dC
For a european option: = 21 Γ0 ( S, t )ϕ( S, t )
( )
d σ(2S ,t )
Bruno Dupire 50
Superbuckets: local change in implied vol

Local change of implied volatility is obtained by combining

local changes in local volatility according a certain weighting
dC dC d (σ 2 )
=∫
d (σ$ )
2
d (σ 2 ) d (σ$ 2 )

sensitivity in weighting obtain

local vol using stripping
formula
Thus:
cancel sensitivity to any move of implied vol
<=> cancel sensitivity to any move of local vol
<=> cancel all future gamma in expectation

Bruno Dupire 51
 Conclusion
• This analysis shows that option prices are
based on how they capture local volatility

• It reveals the link between local vol and

implied vol

• It sheds some light on the equivalence

between full Vega hedge (superbuckets)
and average future gamma hedge
Bruno Dupire 52
 Delta Hedging
• We assume no interest rates, no dividends, and absolute
(as opposed to proportional) definition of volatility
• Extend f(x) to f(x,v) as the Bachelier (normal BS) price of
f for start price x and variance v:

( y − x)2
with f(x,0) = f(x) f ( x, v) ≡ E x ,v [ f ( X )] ≡ 1 −

2πv ∫ f ( y )e 2v
dy
• Then,
• We explore 1
f v ( xvarious
, v) = f xxdelta
( x, v) hedging strategies
2

Bruno Dupire 53
 Calendar Time Delta Hedging
• Delta hedging with constant vol: P&L depends on the
path of the volatility and on the path of the spot price.
• Calendar time delta hedge: replication cost of
f ( X t , σ 2 .(T − t ))
1 t
σ + ∫0 0,xx replication
− σ
2 2
f ( X
• In particular, for
0 , .T )
sigma2 = f ( dQV0 ,u ducost
) of
f (Xt )

1 t
f ( X 0 ) + ∫ f xx dQV0,u
2 0

Bruno Dupire 54
 Business Time Delta Hedging

• Delta hedging according to the quadratic variation:

P&L that depends only on quadratic variation and
spot price 1
df ( X t , L − QV0,t ) = f x dX t − f v dQV0,t + f xx dQV0,t = f x dX t
2
QV0,T ≤ L,
• Hence, for t
f ( X t , L − QV0,t ) = f ( X 0 , L) + ∫ f x ( X u , L − QV0,u ) dX t
0

f ( X t , L − QV0,t ) f ( X 0 , L)
f (And
X 0 , Lthe
) replicating cost of is τ : QV0,τ = L
finances exactly the replication of f until

Bruno Dupire 55
 Daily P&L Variation

Bruno Dupire 56
 Tracking Error Comparison

Bruno Dupire 57
V. Stochastic Volatility Models
 Hull & White
•Stochastic volatility model Hull&White (87)
dS t
= rdt + σ t dW t
P

St
dσ t = α dt + β dZ t
P

•Incomplete model, depends on risk premium

•Does not fit market smile
Implied volatility

Implied volatility

Strike price Strike price

ρ Z ,W = 0 ρ Z ,W < 0
Bruno Dupire 59
 Role of parameters
• Correlation gives the short term skew
• Mean reversion level determines the long term
value of volatility
• Mean reversion strength
– Determine the term structure of volatility
– Dampens the skew for longer maturities
• Volvol gives convexity to implied vol
• Functional dependency on S has a similar effect
to correlation

Bruno Dupire 60
 Heston Model
 dS
 S = µ dt + v dW

dv = λ (v − v )dt + η v dZ dW , dZ = ρ dt

Solved by Fourier transform:

FWD
x ≡ ln τ =T −t
K
C K ,T ( x, v,τ ) = e x P1 ( x, v,τ ) − P0 ( x, v,τ )

Bruno Dupire 61
 Spot dependency

2 ways to generate skew in a stochastic vol model

1) σ t = xt f (S , t ), ρ (W , Z ) = 0
2) σρ (W , Z ) ≠ 0 σ

S0 ST ST

S0

-Mostly equivalent: similar (St,σt ) patterns, similar

future
evolutions
-1) more flexible (and arbitrary!) than 2)
-For short horizons: stoch vol model Ù local vol model
+ independent noise on vol.
Bruno Dupire 62
 Convexity Bias

 dS = σ t dW
 2
 d σ t = α dZ ⇒ [ ]
E σ t2 | S t = K = σ 02 ?
 ρ (W , Z ) = 0

[ ]
NO! only E σ t2 = σ 02

[
E σ t2 | St = K ]
σ 02

S0 K

σ t likely to be high if St >> S0 or St << S0

Bruno Dupire 63
 Impact on Models

• Risk Neutral drift for instantaneous forward

variance
• Markov Model:
dS
= f (S , t )σ t dW fits initial smile with local vols σ (S , t )
S

σ 2 (S , t )
⇔ f (S , t ) =
E [σ t2 | S t = S ]

Bruno Dupire 64
 Smile dynamics: Stoch Vol Model (1)
Local vols
Skew case (r<0)
σ Smile S −
Smile S 0
Smile S +

S − S0 S+ K

- ATM short term implied still follows the local vols

(E [σ 2
T ]
S T = K = σ 2 (K , T ) )
- Similar skews as local vol model for short horizons
- Common mistake when computing the smile for another
spot: just change S0 forgetting the conditioning on σ :
if S : S0 → S+ where is the new σ ?
Bruno Dupire 65
 Smile dynamics: Stoch Vol Model (2)
• Pure smile case (r=0)

σ
Local vols
+
Smile S − Smile S

Smile S 0

K
S− S0 S+

• ATM short term implied follows the local vols

• Future skews quite flat, different from local vol
model
• Again, do not forget conditioning of vol by S

Bruno Dupire 66
Forward Skew
 Forward Skews

In the absence of jump :

model fits market ⇔ ∀K , T E[σ T2 ST = K ] = σ loc
2
(K ,T )
This constrains
a) the sensitivity of the ATM short term volatility wrt S;
b) the average level of the volatility conditioned to ST=K.

a) tells that the sensitivity and the hedge ratio of vanillas depend on the
calibration to the vanilla, not on local volatility/ stochastic volatility.
To change them, jumps are needed.
But b) does not say anything on the conditional forward skews.

Bruno Dupire 68
 Sensitivity of ATM volatility / S

At t, short term ATM implied volatility ~ σt.

As σt is random, the sensitivity ∂σ 2 is defined only in average:
∂S
∂σ loc
2
(S , t )
Et [σ 2
t +δt − σ Sδt = St + δS ] = σ ( St + δS , t + δt ) − σ ( St − t ) ≈
t
2 2
loc
2
loc ⋅ dS
∂S

In average, σ ATM
2
follows σ loc
2
.
Optimal hedge of vanilla under calibrated stochastic volatility corresponds to
perfect hedge ratio under LVM.

Bruno Dupire 69