Lecture 4: Forecasting with

option implied information

Prof. Massimo Guidolin

Advanced Financial Econometrics III

Winter/Spring 2019
Overview
 A two-step approach

 Black-Scholes single-factor model

 Heston’s two-factor square root stochastic volatility model

 Model-free implied volatility

 Option-implied correlations
 Model-free option-implied skewness and kurtosis

 Model-implied, parametric forecasts of skewness and

kurtosis
 Model-free forecasts of densities

 From risk-neutral to physical forecasts

Lecture 4: Forecasting with option implied information – Prof. Guidolin 2
The key point
 Derivative prices contain information useful to forecast any
twice differentiable function of the future underlying price
 Focus on European-style options, especially equity index
 Derivative prices contain useful information on the conditional
density of future underlying asset returns
 A derivative contract is an asset whose future payoff depends on
the uncertain realization of the price of an underlying asset
o Futures and forward contracts, swaps (e.g., CDS and variance swaps),
collateralized debt obligations (CDOs) and basket options, European
style call and put options, American style and exotic options, etc.
o Several of these classes of derivatives exist for many different types of
underlying assets, such as commodities, equities, and equity indexes
o However, some derivative contracts such as forwards and futures are
linear in the return on the underlying security, and therefore their
payoffs are too simple to contain useful reliable information
o Other securities, such as exotic options, have path-dependent payoffs,
which may make information extraction cumbersome
Lecture 4: Forecasting with option implied information – Prof. Guidolin 3
A two-step approach
 Forecasting with option-implied information proceeds in two steps:
① Derivative prices are used to extract a relevant aspect of the
option-implied distribution of the underlying asset
o E.g., ATM implied volatility (IV), closest to 30-day to maturity
② An econometric model is used to relate this option-implied
information to the forecasting object of interest
o E.g., realized 30-day variance is regressed on IV inferred from
observed option prices 30 days before
 In this lecture, brief review of the methods used in the first step
 The two most commonly used models for option valuation are the
Black and Scholes (1973, JPE) and Heston (1993, RFS) models
 Black and Scholes  constant volatility geometric Brownian motion

where r is a constant risk-free,  is volatility, and dz a normal shock

 The future log price is normally distributed and option price for a
European call with maturity T and strike price X is
Lecture 4: Forecasting with option implied information – Prof. Guidolin 4
Black-Scholes, single-factor model

(put-call parity)
 The formula has just one unobserved parameter, namely volatility
that can be backed out for any given option with market price as:

 The resulting option-specific volatility is called BSIV

 Although the BS formula is clearly non-linear, for ATM options, the
relationship between volatility and option price is virtually linear
o In general the relationship btw
volatility and option prices is
increasing and monotone
o Solving for BSIV is quick even
if it must be done numerically
o Vega captures the sensitivity of
the price w.r.t. changes in  5
Lecture 4: Forecasting with option implied information – Prof. Guidolin
Black-Scholes, single-factor model
 For equity index options, BSIV as an adjusted R2 of 62% and
summarizes all other information
o The sensitivity of the price vs.
 is the highest for ATM options
o In the following table we
show a predictive regression
of realized volatility on BSIV
o RVM, RVW, RVD are monthly,
weekly, and daily realized vol
(moneyness)
o RV is from intraperiod returns
and C nets jumps out

Lecture 4: Forecasting with option implied information – Prof. Guidolin 6

Black-Scholes, single-factor model
 Nonconstant patterns in BSIV vs. moneyness  misspecification

 The simple BSIV forecast is able to compete with some of the most
sophisticated historical return-based forecasts
 Index-option BSIVs display a distinct downward sloping pattern
commonly known as the smirk or the skew
 This is evidence that the BS model which relies on the normal
distribution is misspecified 7
Lecture 4: Forecasting with option implied information – Prof. Guidolin
Black-Scholes, single-factor model
 Nonconstant patterns in BSIV vs. maturity  misspecification

 Deep out-of-the-money put options (X/S0 << 1) are more expensive

than the normal-based Black-Scholes model would suggest
 Only a distribution with a fatter left tail (that is negative skewness)
would be able to generate these much higher prices for OTM puts
 BSIV for ATM (X/S0 = 1) tends to be larger for long-maturity than
short-maturity options 8
Lecture 4: Forecasting with option implied information – Prof. Guidolin
Heston, two-factor square root process
 Heston’s model makes variance stochastic and square root
 For variances to change over time, we need a richer setup than the
Black-Scholes models
 Most famous model that provides this result is Heston (1993, RFS),
who assumes that the underlying follows a square-root process:

where the two innovations are correlated with parameter ρ

 At time zero, the variance forecast for horizon T can be obtained as

 The mean-reversion parameter  determines the extent to which

the difference between current spot volatility and long run
volatility, (V0 - ), affects the horizon T forecast
 Whereas the BS only has one parameter, Heston has 4 parameters
Lecture 4: Forecasting with option implied information – Prof. Guidolin 9
Heston, two-factor square root process

 Bakshi, Cao, and Chen (1997, JF) re-estimate the model daily
treating V0 as a fifth parameter to be estimated
 What if the model assumed to forecast volatility from option prices
turns out to be misspecified?
 The answer is tragic: nothing good can be expected of the forecasts
Lecture 4: Forecasting with option implied information – Prof. Guidolin 10
Model-free volatility estimation and forecasting
 Luckily a few methods to achieve model-free volatility estimation
are possible
 When investors can trade continuously, interest rates are constant,
and the underlying futures price is a continuous semi-martingale,
Carr and Madan (1998) and Britten-Jones and Neuberger (2000, JF)
show that the expected value of future realized variance is:

o Jiang and Tian (2005, RFS) generalize this result and show that it
holds even if the price process contains jumps

o In practice, a finite range, Xmax - Xmin, of discrete strikes are available

and Jiang and Tian consider using the trapezoidal integration rule

Lecture 4: Forecasting with option implied information – Prof. Guidolin 11

Empirical evidence
 Overall, the evidence indicates that option-implied volatility is a
biased predictor of the future volatility of the underlying asset
 Yet, most studies find that it contains useful information over
traditional predictors based on historical prices
o Option IV by itself often outperforms historical volatility
 BSIV is predictable and helps forecast volatility, but because arbi-
trage profits are impossible under transaction costs, predictability
is consistent with EMH (see Goncalves and Guidolin, 2006, JoB)
 There is recent, strong evidence that the variance risk premium
(VRP) can predict the equity risk premium
o VRP is the difference between implied variance and realized variance
 Bakshi, Panayotov, and Skoulakis (2011, JFE) compute forward
variance, the implied variance between two future dates, and find
that it forecasts stocks, T-bills, and changes in real activity
 Feunou, Fontaine, Taamouti, and Tedongap (2013, RoF) find that
the term structure of IVs can predict both equity risk and VRP
Lecture 4: Forecasting with option implied information – Prof. Guidolin 12
Option-implied correlations
 BSIV is useful in forecasting the volatility of individual stocks
 Implied volatility has also been used to predict future stock returns
 Information in options leads analyst recommendation changes
 The VIX is a priced risk factor with a negative price of risk, so that
stocks with higher sensitivities to the innovation in VIX exhibit on
average future lower returns
o VIX is a weighted average of BSIVs
o The CBOE computes VIX using OTM and ATM call and put options
o It calculates the volatility for the two available maturities that are the
nearest and second-nearest to 30 days.
 Certain derivatives contain very rich information on correlations
between financial time series
 E.g., in currency markets 
where R denotes a continuously compounded return
 Therefore
Lecture 4: Forecasting with option implied information – Prof. Guidolin 13
Option-implied correlations
 While implied correlations for currencies are derived from
the triangular equality, in the case of stocks only an implied
average correlation may be estimated
 The implied correlation is:

 Provided we have option-implied variance forecasts for 3

currencies, we can use this to get an implied correlation forecast
 Option-implied exchange rate correlations for the DM/GBP pair and
the DM/JPY, and USD/DM/JPY pairs predict significantly better
than historical correlations between the pairs
 There is a measure of average option-implied correlation between
the stocks in an index, I, Weight of
stock j

 Skintzi and Refenes (2005, JFM) use options on the DJIA index
Lecture 4: Forecasting with option implied information – Prof. Guidolin 14
Option-implied correlations
 Smirks (asymmetric smiles) in IVs indicate left-skewness in
the density of underlying returns, while symmetric smiles
point to excess kurtosis
 Implied correlation index is biased upward, but is a better predictor
of future correlation than historical correlation
 Implied correlations may be used to estimate betas and the litera-
ture finds that option-implied betas predict realized betas well
 However, using option-implied information in portfolio allocation
does not improve the Sharpe ratio or CER of the optimal portfolio
 We saw earlier that BS is unlikely to be correctly specified: the very
option prices (IVs) contain robust evidence of asymmetries and fat
tails in the predictive density of underlying asset returns
 Can we extract option-implied skewness and kurtosis?
 It is sensible to proceed with a model-free approach, called option
replication approach, see Bakshi and Madan (2000, JFE)
Lecture 4: Forecasting with option implied information – Prof. Guidolin 15
Model-free option-implied skewness and kurtosis
 For any twice differentiable function of the future underlying
price, there is a spanning portfolio made of bonds, stock, and
European call and put options
 Bakshi, Carr, and Madan show that any twice continuously differen-
tiable fnct, H(ST), of terminal price ST, can be replicated (spanned)
by a unique position in the risk-free, stocks and European options

Units of underlying
Units of risk-free bond

o H’’(X)dX are units of OTM call and put options with strike price X
o From a forecasting perspective, for any H(•), there is a portfolio of
risk-free bonds, stocks, and options whose current aggregate market
value provides an option-implied forecast of H(ST)

Lecture 4: Forecasting with option implied information – Prof. Guidolin 16

Model-free option-implied skewness and kurtosis
 Under mild assumptions, the prices of OTM puts and calls can
be used to infer risk-neutral volatility, skewness, and kurtosis
 Consider now higher moments of simple returns:

 We can use OTM European call and put prices to derive the
quadratic, cubic, and quartic contracts as

o High option prices imply high volatility

o High OTM put and low OTM call prices  negative skewness
o High OTM call and put prices at extreme moneyness  high kurtosis
Lecture 4: Forecasting with option implied information – Prof. Guidolin 17
Model-free option-implied skewness and kurtosis
 Now compute option-implied volatility, skewness, and kurtosis:

 Using S&P 500 index options over January 1996 - September 2009
we plot higher moments of log returns for the one-month horizon
o The volatility series is very highly correlated with the VIX index, with
a correlation of 0.997
 The estimate of skewness is negative for every day in the sample
 The estimate of kurtosis is always higher than 3
 Both skewness and kurtosis do not show significant or persistent
alterations during the 2008-2009 Great Financial Crisis
Lecture 4: Forecasting with option implied information – Prof. Guidolin 18
Model-free option-implied skewness and kurtosis

 In February 2011, the CBOE began publishing the CBOE S&P 500
Skew Index computed according to this methodology
Lecture 4: Forecasting with option implied information – Prof. Guidolin 19
Model-based option-implied skewness and kurtosis
 Jarrow and Rudd have proposed an option price appro-
ximation based of Edgeworth expansions of BS,

 There are also parametric, model-based methods to extract

skewness and kurtosis from option prices
 Some are based on functional «expansions»/approximations of BS
 Most famous approach is due to Jarrow and Rudd (1982, JFE) who
propose a method where the density of the security price at option
maturity, T, is approximated using an Edgeworth series expansion
Lecture 4: Forecasting with option implied information – Prof. Guidolin 20
Model-based option-implied skewness and kurtosis
o Jarrow and Rudd’s pricing formula is:

o Kj is the jth cumulant of the actual density, Kj() is the cumulant of the
lognormal density and other quantities are reported in Appendix A
o If one approximates around a log-normal density, then

o The model now has 3 parameters to estimate, VAR, SKEW and KURT
 As an alternative to the Edgeworth expansion, Corrado and Su
(1996, JFR) consider a Gram-Charlier series expansion:

Lecture 4: Forecasting with option implied information – Prof. Guidolin 21

Model-based option-implied skewness and kurtosis
 Also parametric jump-diffusion models may be used to
forecast skewness and kurtosis from option prices

 Additional parametric models have become popular in the

literature to infer skewness and kurtosis from option prices
 E.g., in Bates (2000, JoE), the futures price F is assumed to follow a
jump-diffusion:
Correlated with
coefficient ρ

 q is a Poisson counter with instantaneous intensity , and k is a log-

normal return jump,
 Higher-order moments can now be computed as a function of the
unknown parameters, to be estimated
 Options can also be used to forecast the density of underlying asset
returns
Lecture 4: Forecasting with option implied information – Prof. Guidolin 22
Model-free risk-neutral density forecasts
 The option-implied conditional density for the underlying at
maturity T is the forward second derivative of an ATM call
 Breeden and Litzenberger (1978, JBus) and Banz and Miller (1978,
JBus) show that the option-implied density can be extracted from a
set of European option prices with a continuum of strike prices
o This result is a special case of Carr and Madan’s result reviewed above
 The value of a European call, C0, is the discounted expected value of
payoff on expiry date T, i.e., under the implied measure, f0(ST):

 Take the partial derivative of C0 with respect to the strike price X:


 The conditional density function (PDF) denoted by f0(X) is obtained
as:

Lecture 4: Forecasting with option implied information – Prof. Guidolin 23

Model-free risk-neutral density forecasts
 A discrete strike approximation of the conditional PDF in
terms of calls is:

 Because of put-call parity, can use puts instead:

 In practice, we can obtain an approximation to the CDF using finite

differences of call or put prices observed at discrete strike prices:

o In terms of the log return, the CDF and PDF are

Lecture 4: Forecasting with option implied information – Prof. Guidolin 24

Parametric and approximated RN density forecasts
 The key issue in implementing this method is that typically only a
limited number of options are traded, with a handful of strikes
 Tricks exist: e.g., the simple but flexible ad-hoc BS (AHBS) model
constructs the density forecast off a BS implied volatility curve
 In a first step, estimate a second-order polynomial or other well-
fitting function for implied BS volatility as a function of strike and
maturity, to obtain fitted BSIV values:

 Second, using this estimated polynomial, we generate a set of fixed

maturity IVs across a grid of strikes
 Call prices can then be obtained using the BS formula:

o Option-implied density can be obtained using the second derivative

o The figure shows the CDF and PDF obtained when applying a
smoothing cubic spline using BSIV data on 30-day OTM calls and puts
on the S&P 500 index on October 22, 2009 vs. the lognormal
Lecture 4: Forecasting with option implied information – Prof. Guidolin 25
Parametric and approximated RN density forecasts

o The implied distribution is clearly more negatively skewed than the

lognormal distribution
 The limited empirical evidence on alternative methods fails to
reach clear conclusions as to which method is to be preferred
 Because the resulting densities are often not markedly different
from each other using different estimation methods, it makes sense
to use methods that are computationally easy 26
Lecture 4: Forecasting with option implied information – Prof. Guidolin
Parametric and approximated RN density forecasts

 Because of computational ease and the stability of the resulting

parameter estimates, the smoothed implied volatility function
method is a good choice for many purposes
 So far we have constructed forecasting objects using the so-called
risk-neutral measure implied from options
 When forecasting properties of the underlying asset we ideally
want to use the physical measure
Lecture 4: Forecasting with option implied information – Prof. Guidolin 27
From risk-neutral to physical forecasts
 Because forecasting occurs in the physical measure space, it
is often useful to know the mapping between Q and P
 Knowing the mapping btw. the two measures is therefore required
o Use superscript Q to describe the option-implied density used above
and we use superscript P to denote the physical density
 Black and Scholes (1973) assume the physical stock price process:

where  is the equity risk premium

o In the complete markets, BS world the option is a redundant asset
perfectly replicated by trading the stock and a risk-free bond
 The option price is independent of the degree of risk-aversion of
investors because they can replicate the option using a dynamic
trading strategy in the underlying asset
 Principle of risk-neutral valuation, all derivatives can be valued
using the risk-neutral expected pay-off discounted at the risk free
rate:
Lecture 4: Forecasting with option implied information – Prof. Guidolin 28
From risk-neutral to physical forecasts
 Using Ito’s lemma implies that log returns are normally distributed

which shows that the the only difference in drift is represented by

the equity risk premium
o In a BS world, the option-implied RN density forecast will therefore
have the correct volatility and functional form but a mean biased
downward because of the equity premium
 Because the RN mean of the asset return is the risk-free rate, the
option price has no predictive content for the mean return
 In the incomplete markets case we can still assume a pricing
relationship of the form

 But the link between the Q and P distributions is not unique and a
pricing kernel MT must be assumed to link the two distributions
Lecture 4: Forecasting with option implied information – Prof. Guidolin 29
From risk-neutral to physical forecasts


 The pricing kernel (or stochastic discount factor) describes how in

equilibrium investors trade off the current (known) option price
versus the future (stochastic) pay-off
 For instance, Heston’s model allows for stochastic volatility
implying that the option, which depends on volatility, cannot be
perfectly replicated by the stock and bond
 Heston (1993) assumes that the price of an asset follows

where the two innovations are correlated with parameter ρ

 The mapping between the P and Q-parameters is given by

Lecture 4: Forecasting with option implied information – Prof. Guidolin 30

From risk-neutral to physical forecasts
 The P and Q processes imply a pricing kernel of the form

where  is a variance preference parameter

 The risk premia  and  are related to the preference parameters by

 In order to appreciate the differences btw. P- and Q-forecasts of

variance, let’s examine the role played by :

 Under P-measure the expected variance in Heston’s model differs

from the RN forecast

Lecture 4: Forecasting with option implied information – Prof. Guidolin 31

From risk-neutral to physical forecasts

 For short horizons and when the current volatility is low then the
effect of the volatility risk premium is relatively small
 However for long-horizons the effect is much larger.
Lecture 4: Forecasting with option implied information – Prof. Guidolin 32
Conclusion
 The literature contains a large body of evidence supporting the use
of option-implied information to predict physical objects of interest
 It is certainly not mandatory that the option-implied information is
mapped into the physical measure to generate forecasts
 However, some empirical studies have found that transforming
option-implied to physical information improves forecasting
performance in certain situations
 We would expect the option-implied distribution or moments to be
biased predictors of their physical counterpart
 Yet this bias may be small, and attempting to remove it can create
problems of its own, for instance because based on imposing
restrictions on investor preferences
 More generally, the existence of a bias does not prevent the option-
implied information from being a useful predictor of the future
object of interest
Lecture 4: Forecasting with option implied information – Prof. Guidolin 33
Appendix A: Meaning of coefficients in Jarrow-Rudd’s formula

Lecture 4: Forecasting with option implied information – Prof. Guidolin 34