Option structures incorpo
delta & vol view
Choice of structure depends on view
For any given equity and volatility view a
referring to the below diagram
As 1x2 put spread has a theoretical prof
months) we consider it to be a bullish tra
Bearish Volatility View
Market View
CBOE
orate
w on equity and volatility markets
a suitable structure can be chosen by
file similar to short put (for maturities over c3
ade (often used as pseudo-protection)
Implied expensive
Bullish
Market
View
Implied
cheap
26
VIX and Volatility
Trading
What Does Implied Volatili
CBOE
ity Look Like?
Copyright © 2014 Chicago Board Options Exchange, Incorporated. All rights reserved. 28
What is VIX?
A measure of implied volatility:
Derived from S&P 500 index option
Calculated from nearby expirations
Uses the entire range of available s
– VIX is calculated from the va
the square of 30-day volatilit
CBOE
n prices
s for constant, 30-day volatility measure
strike prices
alue of a portfolio of options that replicates
ty
Copyright © 2014 Chicago Board Options Exchange, Incorporated. All rights reserved. 29
VegaThe Option Strip
A specially designed portfolio of options
Option weights inversely proportional to
Constant volatility exposure over sever
Delta-hedging P&L not path-dependent
Vega exposure of single option
50 65 80 95 110 125 140 155
UnderlyingPrice
CBOE
Vegas
o the strike price (K) squared
ral strike prices
t
Vega exposure of option “strip”
50 65 80 95 110 125 140 155
Underlying Price
Copyright © 2014 Chicago Board Options Exchange, Incorporated. All rights reserved. 30
VIX Futures/Options Tradi
Multiplier is 100 for options
Multiplier is 1000 for futures
Settlement is cash,
European-style options
Expiration is 30 days before the
following month’s SPX expiry.
Expires on opening print.
CBOE
ing & Settlement
Series Used in Settlement - August 20, 2014
173 Strikes used, 174 options prices
Date Expiration Strike P/C Trade Price Volume
20-Aug-14 19-Sep-14 1275 Put 0.05 673
20-Aug-14 19-Sep-14 1280 Put 0.05 505
20-Aug-14 19-Sep-14 1285 Put 0.05 470
20-Aug-14 19-Sep-14 1290 Put 0.05 467
20-Aug-14 19-Sep-14 1295 Put 0.05 463
20-Aug-14 19-Sep-14 1300 Put 0.05 489
20-Aug-14 19-Sep-14 1305 Put 0.05 486
20-Aug-14 19-Sep-14 1310 Put 0.05 483
20-Aug-14 19-Sep-14 1315 Put 0.05 478
20-Aug-14 19-Sep-14 1320 Put 0.05 475
………. ………. ………. ………. ………. ……….
………. ………. ………. ………. ………. ……….
………. ………. ………. ………. ………. ……….
20-Aug-14 19-Sep-14 1975 Put 21.9 387
20-Aug-14 19-Sep-14 1975 Call 22.6 64
………. ………. ………. ………. ………. ……….
………. ………. ………. ………. ………. ……….
………. ………. ………. ………. ………. ……….
20-Aug-14 19-Sep-14 2090 Call 0.35 721
20-Aug-14 19-Sep-14 2095 Call 0.2 366
20-Aug-14 19-Sep-14 2100 Call 0.2 364
20-Aug-14 19-Sep-14 2105 Call 0.15 333
20-Aug-14 19-Sep-14 2110 Call 0.15 333
20-Aug-14 19-Sep-14 2115 Call 0.25 719
20-Aug-14 19-Sep-14 2120 Call 0.15 330
20-Aug-14 19-Sep-14 2125 Call 0.1 807
20-Aug-14 19-Sep-14 2150 Call 0.1 1,045
20-Aug-14 19-Sep-14 2175 Call 0.05 1,132
95,301
Copyright © 2014 Chicago Board Options Exchange, Incorporated. All rights reserved. 31
VIX Futures and Options B
VIX is 30-day vol
Futures are 30-day forward vol
All options priced on that month’s futur
On Bloomberg: Futures Quotes - “V
Options Quotes - “V
Settlement Price - “
CBOE
Basics
res contract
VIX <index> CT”
VIX <index> OMON”
“VRO <index>”
Copyright © 2014 Chicago Board Options Exchange, Incorporated. All rights reserved. 32
VIX Futures Term Structur
CBOE
re
Copyright © 2014 Chicago Board Options Exchange, Incorporated. All rights reserved. 33
VIX and VVIXSM (VIX of VIX
CBOE
X)
Copyright © 2014 Chicago Board Options Exchange, Incorporated. All rights reserved. 34
SPX and VIX Indices
CBOE
Copyright © 2014 Chicago Board Options Exchange, Incorporated. All rights reserved. 35
Correlations
CBOE
Copyright © 2014 Chicago Board Options Exchange, Incorporated. All rights reserved. 36
Question
Is close to close volatility an unb
No (RED)
Yes (GREEN)
CBOE
biased measure of volatility?
37
Markets do not random
walk in crisis
In normal markets equities random
• Market makers and stat arb funds remove
autocorrelation. In developed liquid markets with
short selling restrictions market makers and stat a
funds should remove any autocorrelation, hence m
normally random walk.
• In crisis equities no longer random walk. Beca
the risk, many market makers and stat arb funds
withdraw from stressed markets.
Reasons why not always random w
• Panic trading. Stop losses & panic trading can ca
market overreaction intraday, but this is normalise
a few days.
• Political “pin risk”. Politicians only do what is
necessary when markets plummet, and backtrack
they rise.
• Low volumes. Market moved by few big trades o
volume.
• Correction / misunderstanding of statements.
Markets moved by central bank comments, which
misread.
CBOE
m walk 115 Random walk / normal markets
110 Integrated vol = close to close vol = weekly vol
1234
h no 105 Random walk / normal markets
arb
Whipsaw / stressed markets
markets 100
ause of 95
90
walk 85 5
0
ause Days
ed after
k when
on low
h can be
38
Close to close overestima
volatility
Using intraday prices can improve
• When comparing volatility between regions, week
different time zones. This is only appropriate for la
advanced volatility measure is better.
• Close to close volatility needs c20 or more days o
close to close volatility is very noisy. An advanced
(C) is better for small samples.
Estimate Prices Handle Drift
Taken
Close to close C No
Parkinson HL No
Garman-Klass OHLC No
Rogers-Satchell OHLC Yes
OHLC No
G-K Yang-Zhang
ext
Yang-Zhang OHLC Yes
CBOE
ates
historical volatility measurement
kly volatility is better than daily to reduce effect of
arge data samples, if this is not available / practical an
of data to be accurate, for smaller periods e.g. 5 days
d measuring Open (O), High (H), Low (L) and Close
t? Handle Overnight Jumps? Efficiency
(max)
No 1
No 5.2
No 7.4
No 8
Yes 8
Yes 14
39
2 ways to profit from stres
markets
1) Delta hedging frequency
Hedge long gamma positions on the close. Wh
close to close volatility overestimates true volatility
crisis, you can extract this premium by delta hedg
close.
Hedge short gamma positions intraday. By del
hedging intraday (as frequently as is practical) you
extract the implied volatility premium to the true
(approximated by Yang-Zhang) volatility.
2) Long daily var and short weekly
Markets overreact when they are stressed. In a
markets often overreact (potentially due to stop lo
then correct after a few days. Investors can profit
this by going long daily variance and short weekly
variance.
No options or variance swaps need to be trade
variance swap is simply a delta hedged log contra
portfolio of long daily variance and short weekly v
can be replicated by delta hedging only (as log co
cancel). Alternatively a structured product can be
CBOE
ssed
hile Close to close – Yang-Zhang vol (30 day)
y in a
ging on 20%
lta 15%
u can
10%
5%
0%
var -5%
a crisis Performance long daily var short weekly
osses), 260 var
from 240 Long daily variance short weekly
y variance outperforms in a crisis
220
200
ed. As a 180
act, a 160
variance 140
ontracts 120
bought. 100
40
Question
Square root of time rule has vola
Near dated implieds move more than far da
Can adjust whole surface by adjusting impli
“one year implied vol move / T^p”.
P is the power of the move (rule of thumb is
If 1 year implied volatility rise
month implied rise?
More than 2 pts, i.e. power more tha
2pts, i.e. power = 0.5 or square root
Less than 2pts, i.e. power less than
NB: 3 months = 0.25 years, and square root of
CBOE
atility surfaces move power 0.5
ated implieds.
ieds for maturity T by
s P=0.5, or square root of time rule).
es 1pt, how much should 3
an 0.5 (WHITE)
of time rule (RED)
0.5 (GREEN)
0.25 = 0.5 (& dividing by 0.5 = 2).
41
Surfaces move by
“square root of time”
Vol move weighted by square ro
• Near dated implieds move more than far da
adjusting implieds for maturity T by “one ye
move.
• Typically volatility move weighted by square
0.5).
• Surfaces also sometimes move in parallel (
• On average surfaces move power 0.44, hen
parallel. Volatility moving by s
23%
1 year implied moves ha
22% amount of 3 month impli
Implied vol 21% +2% +1%
20% -1%
19%
18%
17% 1 yea
3 months 6 months Fla
Rise in implied
CBOE
oot of time is roughly constant
ated implieds. Can adjust whole surface by
ear implied vol move / Tp”. P is the power of the
e root of time is approximately constant (power
(power 0).
nce usually square root of time but sometimes
square root of time
half
ied
%
% -0.5%
4 year implied moves half
amount of 1 year implied
ar 2 years 3 years 4 years
at term structure Fall in implied
42
Can compare different
term structure & skew
Term structures can be normalis
If assume term structure is a fixed vol for in
maturity and a square root of time bump, th
different term structures can be compared
Multiplying standard V2 – V1 term structure
√(T2T1)/ (√T2- √ T1) allows different term
structures to be compared
Normalised term structure puts term structu
same units as 1 year – 3 month term struct
Skew multiplied by square root
time is roughly constant
Skew is greater for near dated implieds tha
dated
Can compare different skews when multiply
square root of time
CBOE
sed Term structure (normalised)
nfinite Term structure x
hen √(T2T1)
e by (√T2-√T1)
ure in 5
ture
0
of
-5 Dec-06
an far Apr-07
y by A ug-07
Dec-07
Apr-08
A ug-08
Dec-08
Apr-09
A ug-09
Dec-09
Apr-10
A ug-10
Dec-10
Apr-11
A ug-11
Dec-11
-10
-15
-20
6 mths - 3 mths (normalised) 1 year - 6 mths (normalised)
Skew Skew (normalised)
(normalised √T) In 2010 Q2 skew spiked, particularly at the far end,
3.8 due to changes in US regulation
3.6
3.4
3.2
3
2.8
2.6
2.4
2.2
2
Dec-09 Mar-10 Jun-10 Sep-10 Dec-10 Mar-11 Jun-11 Sep-11 Dec-11
3 month skew (90-100%) 6 month skew (90-100%)
43
Question
How does the realised volatility o
compare to the realised volatility
Realised post earnings > reali
Realised post earnings = reali
Realised post earnings < reali
CBOE
of a stock after earnings
y pre-earnings?
ised pre earnings (WHITE)
ised pre earnings (RED)
ised pre earnings (GREEN)
44
Option prices include implied
on reporting
Near dated expiries usually best
Backing out implied jump needs the “diffusi
volatility (volatility without jumps) to be estim
If an expiry exists before earnings date then
can be used, if not then forward volatility ne
be calculated
As a jump has a bigger relative effect for ne
dated expiries, using near dated expiries is
Stocks c25% less volatile after
reporting
Before reporting the uncertainty of the resu
lift volatility
In the 1-2 week period before reporting ana
typically issue research publications, which
move stock prices
After reporting stocks are usually ¾ as vola
they were before reporting
Most analysis for implied jumps assume co
diffusive volatility
CBOE
jump
t Using implied vol as vol assumption
ive” σJump
mated
n this Expiry before jump Expiry after jump
eeds to Time now
(T)
ear σDiffusive (σBefore jump)
best σExpiry after jump
ults can Using forward vol as vol assumption
σJump
alysts Time now (First) Expiry after jump Second Expiry
h can (T1) (T2)
atile as σExpiry after jump (σ1) σDiffusive (σ12)
onstant σ2
45
Need to adjust for index term s
Most analysis assumes flat term
structure
Assumption of constant diffusive volatility
implies flat term structure
Front end of volatility surfaces often have th
steepest term structure
Adjusting term structure improv
results
Adjusting equity term structure by index term
structure gives more accurate results
Using current index term structure, rather th
historic average equity term structure, ensu
the current market risks are priced in
Using average of peers term structure is les
accurate, due to wider bid offer spreads
CBOE
structure
m Equity and index term structure
he Implied volatility Implied jump is too big as expiry 3 implied is Implied
40% lifted by term structure and reporting date difference
ves 35% due to
30% reporting
m
han 25%
ures
ss 20%
15%
10%
5%
0% 23 Expiry
1 Reporting date
Equity Index Index term structure (diff to expiry 1)
Equity term structure adjusted by index
Implied volatility Adjusting for index term structure flattens equity
40% term structure to give more accurate implied jump
35% Implied
difference
30% due to
25% reporting
20%
15%
10%
5%
0% 23 Expiry
1 Reporting date
Equity - Index term structure Index term structure (diff to expiry 1)
46
Question
How much should ATM implied vola
(we assume skew is fairly priced)?
Amount equal to skew, i.e. stic
Between 1x and 2x skew (RED
Twice skew (GREEN)
CBOE
atility move by if markets rise/fall?
cky strike (WHITE)
D)
47
Local volatility is instantaneou
of underlying
Local volatility skew is double B
Black-Scholes is the average volatility of all
average of ATM local volatility and the local
ATM local volatility and ATM Black-Scholes v
Black-Scholes skew is half the local volatility
ATM volatility moves by twice the Black
If ATM is 20% and the 90% local volatility is 2
(average of 20% and 22%).
Local volatility 90-100% skew is 2% while Bl
If spot moves down 10% then the ATM Black
Underlying
price
S
Spot
Expiry
CBOE
us volatility
Black-Scholes skew
l paths from spot to strike, which is approx the
l volatility of the strike. This leads to 2 results:
volatility are identical
y skew
k-Scholes skew
22%, then the 90% Black-Scholes volatility is 21%
lack-Scholes 90-100% skew is 1%.
k Scholes (= ATM local volatility) is now 22%.
Low local volatility
Black-Scholes
implied volatility is
Strike
average volatility of
all paths between
spot and strike
High local volatility
y
48
Skew trading profitability dete
by volatility regime
There are 4 idealised regimes to
volatility surfaces
Sticky delta / moneyness. Constant volatility for
ATM constant)
Sticky strike. Constant volatility for options with s
Sticky local volatility. When markets fall Black-S
Jumpy volatility. Very high negative correlation b
Volatility surface with equities falling 10%
30%
28% Fixed strike implieds rise as markets fall
26%
24%
Implied vol 22%
20% Fixed strike implieds
18% decline as markets fall
16%
14% Markets fall 10%
12%
10% 40 45 50 55 60 65
35 Sticky strike (and before) Strike (€)
Sticky delta Sticky local vol Jumpy vol
CBOE
ermined
o describe movement of
options of same strike as percentage of spot (e.g.
same fixed currency strike (e.g. €50 strike constant)
Scholes implied volatility rises, and vice versa
between spot and implied volatility (panicked markets)
Volatility surface with equities rising 10%
30% Markets rise 10%
28%
26%
Implied vol 24% Fixed strike implieds increase
22% as markets rise
20%
18%
16%
14% Fixed strike implieds decline as markets rise
12%
10% 40 45 50 55 60 65
35 Sticky strike (and before) Strike (€)
Sticky delta Sticky local vol Jumpy vol
49
Skew trades breakeven with
sticky local volatility
Fixed strike implied
volatility change
Volatility Equity Equity
regime decline rise
Sticky delta Falls Rises
Sticky strike - -
Sticky local Rises Falls
volatility
Jumpy Rises Falls
volatility significantly significantly
CBOE
P&L breakdown for long skew
(e.g. long put, short call)
Remark Skew theta Total
+=
+=
+=
+=
50