Invest in option or in its underlying stock ?

Yves Rakotondratsimba
17 septembre 2010

Résumé
Despite the suggestion made by investment banks and brokerage firms for private
investors to substitute one or all part of their stock investment by the associated option,
the real benefit obtained from dealing with an option rather than with its underlying
stock remains to be understood.
Using our general results ( in Theorems 1 and 4 ), market and synthetic data on option
prices, it appears that for many situations investing in an option may be less interesting
than investing directly in the related underlying stock.
However there is no definitive answer about the superiority or not of investing in
the option, since the situation depends on the level of asset change at the investment
maturity. Our contribution here to the above question is to provide a systematic analysis
of the return structures associated with the two alternative investments. Then, it is up to
the investor to decide the best investment support to choose depending on her views on
the future asset relative change and on the option characteristics which are available on
the markets. The fomulas we derive here may be helping in performing this task.
Keywords : investment decision, stock options, returns, Black-Scholes-Merton pricing.
Classification : G11, G13. 1

1 Introduction
Many investment banks and brokerage firms, involved in selling speculative investment
products, suggest to private investors to substitute one or all part of their stock investment
into the investment in the associated option ( see for instance [Os] and references in [Si] ). In the
leaflets related to the derivative products they propose, very few ( and most of time simplistic
) examples are provided, and these last do not reflect seriously the risk and complexity behind
the proposed instrument. For instance, the presentation is made in order to draw the client
attention on the option leverage effect rather than on the real return itself. Therefore for the
favorable case, the investment in option seems to be attractive due to the large amplification
of the underlying asset return obtained in comparison with a direct investment in the stock
itself. For the defavorable case, the sellers divert the client attention on the small size of the
premium associated to the considered option and recall that only one part of the investor’s
wealth would be suitable for the option speculative investment. The question of real benefit
obtained from investing in the option rather than in its associated underlying stock remains
to be understood in full generality and deserves some analyzes.
Consequently, in this paper, we will focuse on analyzing the returns corresponding to the
two alternative investments : in the option and its associated stock. Therefore any investor
1. ECE lab. ENRSF, 37 quai de Grenelle CS71520, 75 725 Paris 15, France ; e-mail : w_yrakoto@yahoo.com

1
Electronic copy available at: http://ssrn.com/abstract=1678468
may get from our result, by incorporating her view on the future asset relative change, a first
indicator on the possible return she could get related to her investement project. It is just a
first indicator since, as done in financial literature and some commercial leaflets, we do not
take into account the transaction costs. To get more realistic views on the consequence of any
investment choice, market frictions such as transaction costs should be incorporated into the
analyzes. Transaction costs are not taken into consideration here for clearness and also to avoid
tedious technical development. Nevertheless some precise hints about the approach under the
presence of costs are given in subsection 5.6 below. It should not be forgotten that an investment
with a theoretical positive return may be really less interesting when the transaction costs are
taken into consideration.
The speculative and hedging features of option, for the seller-standpoint, are largely studied
and developped over the financial literature. In this work, we will focuse on the speculative
aspect of option for the buyer-standpoint. Of course the option deals are not only intended for
an immediate return but are also helpful for an objective of portfolio diversification. However
for various investors ( as those who are involved in a portfolio with small size and possibly
made by one type of option ), the question is essentially reduced to materialize a maximum
profit and loss for a short or medium term horizon. The analyses we perform in this work may
be used to help these investors in taking their investment decisions. Next they can also viewed
as a departure for further research on general return structure related to a portfolio containing
derivatives. Indeed most of available results for such a portfolio are essentially simulation-based
or just lying on very particular market situations.
In contrast with the various claims presented inside commercial leafts and intended for (
private ) investors, by using our general results ( see Theorem 1, Proposition 3 and Theorem 4 ),
real market and synthetic data on option prices, it appears that for many situations investing
in option may be less interesting than investing directly in the associated underlying asset.
However, there is no definitive answer about the superiority or not of investing in option, since
the situation depends on the level of asset change at the investment maturity. Our contribution
here to the question is to provide a systematic analysis of the return structures associated with
the two alternative investments. Then it is up to the investor to decide the best investment
support to choose depending on her views of the future asset relative change and on the option
characteristics which are available on the markets. The return structures analyses we obtain
here may be seen as a starting point for implementing a tool for investment decision.
Our results are presented in Section 2. First we analyze in Theorem 1, the return structure
of the call-option. The main point here is about finding suitable intervals containing the reali-
zed asset relative change at the option maturity and for which the comparison between the two
investments ( in the option and in the underlying stock ) becomes clear. The fact that the in-
vestment in option may be less interesting in comparison with the investment in the underlying
stock leads us to ask about the returns associated with the well-known call-covered strategy.
Especially the question of finding sufficient conditions ensuring a positive return is raised here.
The answer is displayed in Proposition 2. Moreover in Proposition 3, we explicit the probability
for which the investment in option is superior compared with the investment in the underlying
stock when this last is assumed to follow the standard geometrical brownian motion process
(GBM). For the numerical examples presented in the Appendix part, these probabilities appear
to be of low level. As an application of our Theorem 1, we are able to clarify the comparison
between the returns of the considered call-option and its underlying stock under the GBM
process. Next in Theorem 4 we consider the return structures of the put-option and the asset
short-selling, which is naturally associated to taking a position in a bearish market. It may be
pointed here that the mechanism associated to the asset short-selling creates some difficulties
both for the theoretical analysis and the real practice. Similarly, to the case of call-option, we

2
Electronic copy available at: http://ssrn.com/abstract=1678468
present in Proposition 5 the return associated to the well-known strategy of covered-put. The
probability that the investment in put-option is superior, compared with the investment in
the underlying stock short-selling under the GBMP assumption, is performed in Proposition
6. The numerical examples, to shed light the statements of our results, are given in Section 3.
We use here both real market data and synthetic situations spanned by the use of practitioners
Black-Scholes-Merton option pricing. The conclusions are written in Section 4, where we also
describe the limits and further developments related to the present work. Proofs of our results
are displayed in Section 5. At the last part of Subsection 5.6, we outline the approach to use
when transaction costs are taken into consideration. Finally the tables corresponding to the
various examples presented in this paper are shown in the Appendix Part.
This work may be considered as the complete version of the author’s paper [Ra2], which is
essentially devoted to the result about call options. ∼

2 Main Results
2.1 Results related to call-options
Throughout this subsection we will consider european call-options with the exercise price
K, maturity T and, whose the underlying asset is some stock XX. It is assumed that 0 < K, T .
Let us recall that a call-option is a contract which gives to its holder the right to buy, at
maturity T from the call-seller, the underlying stocks XX at the unit price K. Let us denote by
S0 the stock value at the present time 0. The future time-T value of stock XX may be written
as
ST (·) = {1 + vT (·)}S0 (1)
such that
ST (·) − S0
vT (·) =
S0
represents the asset stock relative change during the time-period [0, T ]. Of course, from time
0, either vT (·) or ST (·) may be considered as a random variable, and we always assume that

−1 < vT (·) and 0 < ST (·).

These last inequalities mean that bankruptcy of the firm issuing the stock is excluded to
happen. For convenience we will make use of the quantity
K 
w0 ≡ −1 . (2)
S0
Using the stock relative change vT (·) between [0, T ] and (2), it is clear that
the call-option exercise, i.e. K < ST (·), is equivalent to the realization of

w0 < vT (·). (3)

This inequality means that the call-options may be exercised only when the stock relative
change vT (·) between [0, T ] is large enough. Our motivation in introducing (3) is that in practice
people deal very often in term of the asset relative change rather than in the asset value itself.
So we have immediatly here the call-exercise criterion without refering to both the strike and
asset value. Though vT (·) remains completely unknown at the present time 0 ( the truth will
be revealed only at time-T ), very often the investor has some feeling ( good or not ) about

3
a rough interval which contains it. It should not be forgotten that having some views on the
future market change is a necessary prerequisite before taking any investment position. At the
point where the investor is assumed to have a rough idea about the possible values of vT (·), it
may be useful to focuse on the value of w0 . To this end we observe that
– for a in-the-money (ITM) call ( i.e. K < S0 ), then w0 < 0 ;
– for a at-the-money (ATM) call ( i.e. K = S0 ) then w0 = 0 ;
– for a out-of-the-money (OTM) call ( i.e. S0 < K ) then 0 < w0 .
Therefore intuitively, when thinking that vT (·) will be certainly strictly positif, then we expect
that an ITM or ATM-call will be easily exercised. For the OTM-call the status seems not clear
since, by (3), it depends on the magnitude of w0 . Numerical values of w0 may be seen over all
the Tables displayed in the Appendix Part.
Since we do not take into account the transaction cost, the investment return when taking
a long position on the stock XX during the time-period [0, T ] is given by
ST (·) − S0
ret_asset(·) = = vT (·). (4)
S0
For shortness we use the notation ret_asset(·) instead of the more complete one ret_asset0,T (·).
Note that in real investment we have to deal with an integer number of stocks XX but not just
with one-security as used over this paper. Similarly taking a long position on the associated
call-option, during the time-period [0, T ], leads to the return
{ST (·) − K}I{K<ST (·)} − C0  S0  
ret_call(·) = = vT (·) − w0 I{w0 <vT (·)} − 1. (5)
C0 C0
Here C0 denotes the call premium which corresponds to the price of the right to exercise at
the maturity T . To avoid arbitrage, we have always 0 ≤ C0 < S0 , and will focuse on the
significant case 0 < C0 . The notation I{K<ST (·)} is used to indicate the characteritic function,
that is I{K<ST (·)} = 1 for K < ST (·), else I{K<ST (·)} = 0.
S 
0
For an ATM-call, i.e. w0 = 0, and when vT (·) is large enough in comparison with 1 then
C0
one has S  S 
0 0
ret_call(·) ≈ vT (·) = ret_asset(·).
C0 C0
Since 1 < CS00 then the leverage effect linked to option, as emphasized and pointed by investment
banks and brokerage firms appears here. The advantage with investing in the call-option for
the other cases ( as for no ATM-call, or CS00 vT (·) with small or moderate size, . . . ) deserve some
analyses, we are performing from now.
In order to state a general result, not depending on particular situation or data, we introduce
the following two quantities
C0 1
w0∗ = w0 + = (K + C0 ) − 1 (6)
S0 S0
and
w0∗
w0∗∗ = (7)
1 − CS00
such that
w0 < w0∗ < w0∗∗ . (8)
Remind that the call can be exercised whenever the future asset change vT (·) goes above w0 .
Here w0∗ may be seen as the call-break-even point, that is the call return becomes nonnegative

4
only when vT (·) is above w0∗ . As will seen over the proof of Theorem 1, the call return is less
important than the underlying stock return as long as vT (·) remains below w0∗∗ .
Theorem 1 Recall that
ret_asset(·) = vT (·).
1. Case : −1 < vT (·) ≤ w0 . Here we have

ret_call(·) = −1 = −100% (9)

such that
ret_call(·) < ret_asset(·) (10)
and
0 < ret_asset(·) − ret_call(·) ≤ 1 + w0 . (11)
2. Case : w0 < vT (·) ≤ w0∗ . Here we have
 S  
0
ret_call(·) = vT (·) − w0∗ (12)
C0
and  S n  C o 
0 0
ret_asset(·) − ret_call(·) = 1− −vT (·) + w0∗∗ (13)
C0 S0
such that
−1 < ret_call(·) ≤ 0 (14)
(10) is satisfied (15)
and
 S n  C o  n  C o
0 0 ∗∗ ∗ 0
0< 1− w0 − w0 ≤ ret_asset(·) − ret_call(·) < 1 − . (16)
C0 S0 S0

3. Case : w0∗ < vT (·) ≤ w0∗∗ . Identities (12) and (13) are also satisfied. Moreover we have
 S  
0
0 < ret_call(·) ≤ w0∗∗ − w0∗ (17)
C0
ret_call(·) ≤ ret_asset(·) (18)
and  S n  C o 
0 0
0 ≤ ret_asset(·) − ret_call(·) < 1− w0∗∗ − w0∗ . (19)
C0 S0
4. Case : w0∗∗ < vT (·). Here we have
 S    S  
0 0
0< w0∗∗ − w0∗ < ret_call(·) = vT (·) − w0∗ (20)
C0 C0
ret_asset(·) < ret_call(·) (21)
and  S n  C o 
0 0 ∗∗
0 < ret_call(·) − ret_asset(·) = 1− vT (·) − w0 . (22)
C0 S0

5
 This result is a useful tool in deciding the superiority or not of the call-option investment
( in comparison with the investment in the underlying stock ) when it is combined with the
investor view on the possible level of asset relative change at the maturity T . The first step
in comparing returns related to taking a long position on a call and a long position on the
underlying stock is to compute the real numbers w0 , w0∗ and w0∗∗ as defined respectively from
(2), (6) and (7). They depend just on the call characteristics : strike K, premium C0 and the
underlying spot S0 . These numbers may be sorted in increasing order as described in (8). As
a consequence there are four cases to consider for the level of future stock relative change :
1) −1 < vT (·) ≤ w0 ,
2) w0 < vT (·) ≤ w0∗ ,
3) w0∗ < vT (·) ≤ w0∗∗ and
4) w0∗∗ < vT (·).
Only one of these situations can occur at the future time T . The investor would have a clear
idea about which interval among ] − 1, w0 ], ]w0 , w0∗ ], ]w0∗ , w0∗∗ ] and ]w0∗∗ , ∞[ should contain the
future asset value vT (·). Introducing a model for the stock dynamic, as we will do in Proposition
3, may lead to determine the probability for which this future asset relative value vT (·) belongs
to a given interval. Having historical data on the stock past values is useful to this end. The
idea of combining statistical approach and the investor views, as developed by A. Meucci [Me]
as an extension of the Black-Litterman approach, may be also an interesting way to explore
here. But we do not go further in such a direction.
Our Theorem1 describes the quantities involved in each situation. We try now to recast
this result in a more qualitative understandable form.
The case 1) : −1 < vT (·) ≤ w0 corresponds to the fact that the asset relative change has
not increased enough or probably decreased. Here we get the worst situation for the call. It
cannot be exercised and the return is equal to −100%. Nevertheless the return for the asset
position remains strictly large than that of this call, as we can read in (10). If 0 < w0 ( as in
the case of OTM-call ) and 0 < vT (·) ≤ w0 then the position in the stock leads to a profit,
which is not the case for the call position for which we loose all of the invested amount. Always
for the first case 1), as we may see from (11), the distance between the returns of the call and
underlying stock remains bounded by the fixed number 1 + w0 .
The case 2) : w0 < vT (·) ≤ w0∗ , corresponds to the fact that the call may be exercised,
but the stock growth is not sufficiently enough such that the call return remains negative (
see (14) ). Once again here, the return for the stock position remains more than that for this
call, as we can read from (15). The position in the asset leads to a profit for 0 < w0 ( as in
the case of ATM or OTM-call ). Moreover for this second case 2), as we may see from (16),
the distance between the returns of the call and underlying asset remains bounded below and
above by fixed numbers depending only on K, S0 and C0 .
The case 3) : w0∗ < vT (·) ≤ w0∗∗ , corresponds to the fact that the call may be exerciced and
the asset increase is sufficiently enough in order that the call return becomes now positive ( see
(17) ). Again here, the return for the asset position is above than that for this call, as we can
read in (18). If 0 < w0∗ ( which is already satisfied if − CS00 < w0 ) then the position in the asset
leads to a profit. Though the call also wins a profit, this last remains less than that obtained
directly with the asset position. As we may see from (19), the distance between the returns of
the call and underlying asset remains bounded above by a fixed number depending only on K,
S0 and C0 .
The case 4) : w0∗∗ < vT (·) may be viewed as the case where the asset has done a strong
increase. Not only the call may be exerciced but the stock increase is sufficiently enough in
order that the call return becomes now positive ( see (20) ) and, as can be read in (21), this
return is larger than the return linked to the stock position. It may be seen from (20) that

6
the call position leads always to a profit and a nonnegative lower bound for the corresponding
return may be given. The distance between the returns of the call and underlying stock moves
as an affine function of the return asset as described in (22).
Numerical illustrative examples and more comments are displayed in the next section. Over
these examples it may be seen that for many situations it is preferable to invest directly in
the underlying stock XX rather than in the associated call-option. So it may be interesting to
consider the strategy of covered call, which consists to sell a call and buy the stock. Selling a
call means having a bearish or neutral view on the asset movement. The purchase of the stock
is done here in order to cover an unexcepted upward asset change. The return associated to
the coverd-call strategy is given by the following.
Proposition 2 Assume that we have sold a call with the premium C0 , 0 < C0 < S0 , and
simultaneously bought the underlying stock XX at the unit price S0 . Then this covered-call
strategy leads to the profit and loss
Profit&Loss_covered_call(·) = S0 × ret_covered_call(·)
where n  C o
0
ret_covered_call(·) = vT (·) + I{vT (·)≤w0 } + w0∗ I{w0 <vT (·)} (23)
S0
where w0∗ is defined as in (6). As a consequence for 0 < w0∗ , the covered call leads always to a
profit whenever the stock relative change vT (·) remains strictly above − CS00 .
The covered-call is a winning strategy as long as the future asset change vT (·)
does not move below − CS00 nor goes above w0∗ . For vT (·) ∈] − CS00 , w0 ] the call-covered
strategy has an effect to enhance the return vT (·) into vT (·) + CS00 . For vT (·) ∈]w0 , w0∗ ] the call-
covered strategy has an effect to fix the return at w0∗ . This is also the return for the other cases
w0∗ < vT (·), corresponding to a bullish situation which is out of the call-covered strategy spirit.
Moreover in these situations the return w0∗ is less than the one obtained from an outright stock
investment.
In the framework of Black-Scholes-Merton, the underlying stock XX is assumed to follow
the Geometrical Brownian Motion ( GBM ) whose the dynamic is given by the stochastic
differential equation of the form
dSt (·) = µSt (·)dt + σSt (·)dWt (·) (24)
where µ is real number which represents the asset local tendency, and σ, with 0 < σ, is the
asset volatility.
For any w, with −1 < w, then let us introduce the quantity
 
1  1 2
x(w) = √ ln(1 + w) − µ − σ T . (25)
σ T 2

Proposition 3 Assume that the underlying asset stock XX follows the GBM process (24).
Let us consider a strategy with taking a long position on the call. Then we have the following
probabilities estimates :
h i  
P −1 < vT (·) ≤ w0 = Φ x(w0 ) (26)
h i    
P w0 < vT (·) ≤ w0∗ = Φ x(w0∗ ) − Φ x(w0 ) (27)
h i    
P w0∗ < vT (·) ≤ w0∗∗ = Φ x(w0∗∗ ) − Φ x(w0∗ ) (28)
h i  
P w0∗∗ < vT (·) = 1 − Φ x(w0∗∗ ) . (29)

7
where Φ denotes the cumulative probability function of the standard gaussian normal law ; w0 ,
w0∗ and w0∗∗ are defined respectively as in (2) (6) and (7).
Observe that x(w0 ) is well-defined since 0 < 1 + w0 . More realistic model for the underlying
asset, as with the jump diffusion model, may be considered. But we just limit here in the case of
the Black-Scholes benchmark model. Indeed our purpose in this work is to focuse on analyzing
the structure of the returnh rather than including
i a future view of the future asset change.
The computation of P a < vT (·) < b , for any real numbers a and b, remains theoretically
unknown in full generality. In practice, historical data v1 , . . . , vk , . . . , vN for the asset relative
change ( for a time-to-maturity T ) is very often available. These numbers may h be consideredi
as realizations of the random variable vT (·). Therefore a (historical) view of P a < vT (·) < b
may be taken as the proportion of asset relative changes vi , with i ∈ {1, . . . , N }, which are
contained in the interval ]a, b[.
Very recently, H. Siddiqi [Si] has been noted that some people think by analogy and take
as a basis of option valuation the feeling that a ITM call-option is similar to its underlying
stock. So rather than investing in the underlying outright, some investors prefer to buy a ITM
call, as it offers the same dollar-for-dollar increase or decrease in payoff as the underlying while
requiring only a fraction of investment. Moreover this author in [Si] is able to propose a new
option formula based on the assumption that the market consists of coarse thinkers as well as
rational investors. The behavioral option pricing formula obtained by Siddiqi relies on the idea
that coarse thinkers choose a price for the (ITM) option that equates the expected return on
the option with the expected return on the underlying stock.
Our results in Theorem 1, Proposition 3 and also the numerical examples in Section 3 lead us
to think that the returns for the call-option and the underlying asset may be very different
and there is no evidence of closeness for the associated expected returns. Indeed by (4) and
(5), first we get the pointwise equality
 S  
0
ret_call(·) = vT (·) = ret_asset(·) − w0 I{w0 <ret_asset(·)} − 1. (30)
C0
It means that for −1 < ret_asset(·) ≤ w0 then
ret_call(·) = −1.
And for w0 < ret_asset(·), it can be written that
S  S 
0 0
ret_call(·) = ret_asset(·) − w0∗ . (31)
C0 C0
Though CS00 > 1, the leverage effect associated with the use of call-option appears only if the
term w0∗ ( CS00 ) is very small compared with ( CS00 )ret_asset(·).
Second, to compare the expected returns for the call-option and its underlying asset, let us set
h i h i h i
ρ = EP ret_call(·) and η = EP ret_asset(·) = EP vT (·) .

Of course we make use of any probability P as perceived by the investor. Then by (30), we
have
 S  h i hn oi
0
ρ= EP vT (·)I{w0 <vT (·)} − w0 P w0 < vT (·) −1
C0
S    S  h i hn oi
0 0
= η− 1+ EP vT (·)I{−1<vT (·)≤w0 } + w0 P w0 < vT (·) . (32)
C0 C0

8
Though in (31), a sort of affine relation between the returns of the call and its underlying
asset seems clear, here in (32) we rather observe a non-linear relation between the expected
returns ρ and η. It means that the intuition used by coarse thinkers about the equality of the
call and asset expected returns, as we have mentioned above, is theoretically not clear in full
generality. However when assuming the asset price to follow the GBM dynamic as in (24), then
the situation becomes easy. Indeed for this case
h
2
 √ i
vT (·) = exp µ − 0.5σ T + σ T UT (·) − 1 with UT (·) ∼ N (0, 1)

so that h i
EP vT (·) = exp(µT ) − 1
hn oi  
P w0 < vT (·) = 1 − Φ x(w0 )
and
h i  √   
EP vT (·)I{−1<vT (·)≤w0 } = exp(µT )Φ w0 − σ T − Φ x(w0 )
 √  h i n  √   o
= Φ w0 − σ T EP vT (·) + Φ w0 − σ T − Φ x(w0 ) .

Consequently we get
 S n
0
 √ o
ρ= 1 − Φ w0 − σ T η
C0
 S  
0
  √  n  o
− Φ x(w0 ) − Φ w0 − σ T + w0 1 − Φ x(w0 ) − 1. (34)
C0
This means that the call-option return ρ is an affine function of the asset return η, whenever
the asset price is assumed to follow the GBM dynamic. Under this situation, a comparison
between the two returns may be easily done. For instance, facing to the risk related to the
call-option, a rational investor may require a high return for this instrument compared with
the associated underlying asset. To realize η ≤ ρ then, by (34), the call price should be of
limited size as
 1  n  √ o
C0 ≤ 1 − Φ w0 − σ T η
η+1
√ 
 n  o    
− w0 1 − Φ x(w0 ) + Φ x(w0 ) − Φ w0 − σ T . (35)

2.2 Results related to put-options

Throughout this subsection, we will consider european put-options, whose the underlying
asset is a stock XX, with the exercise price K and the maturity T , with 0 < K, T . Remind that
a put-option is a contract which gives to its holder the right to sell, at maturity T from the
put-seller, the underlying stock XX at the unit price K. Therefore buying a put is somewhat
corresponding to the betting strategy on the stock value decrease, whose the present time value
is denoted by S0 and its future time-T value is ST (·). As is used in the previous subsection,
vT (·) = ST (·)−S
S0
0
represents the stock relative change during the time-period [0, T ].
To characterize the nature of the put-option under consideration, we also make use of the
quantity w0 defined in (2) for the case of the call-option. Using the stock relative change vT (·)
between [0, T ] and (2), it is clear that

9
 the put-option exercise, i.e. ST (·) < K, is equivalent to the realization of
vT (·) < w0 . (36)
This inequality means that the put-option may be exercised whenever the stock relative change
vT (·) between [0, T ] is small enough. Moreover we observe that
– for an in-the-money (ITM) put ( i.e. S0 < K ), then 0 < w0 ;
– for an at-the-money (ATM) put ( i.e. K = S0 ) then w0 = 0 ;
– for an out-of-the-money (OTM) put ( i.e. K < S0 ) then w0 < 0.
Therefore intuitively, when thinking that vT (·) will be certainly strictly negatif, then we expect
that an ITM or ATM-put will be easily exercised. For the OTM-put the status seems not clear
since, by (36), it depends on the magnitude of w0 . Numerical values of w0 , for the case of
put-options, may be seen over the Tables given in the Appendix Part.
Taking a long position on a put during the time-period [0, T ] leads to the return
{K − ST (·)}I{ST (·)<K} − P0  S0  
ret_put(·) = = w0 − vT (·) I{vT (·)<w0 } − 1. (37)
P0 P0
Here P0 denotes the call premium which corresponds to the price of the right to exercise at the
maturity T . Arbitrage consideration leads to 0 ≤ P0 < S0 , and we will focuse on the sequel on
S 
0
the significant case 0 < P0 . For a ATM-put, i.e. w0 = 0, and when − vT (·) is large enough
P0
in comparison with −1 then one has
S  S 
0 0
ret_put(·) ≈ − vT (·) = − ret_asset(·).
P0 P0
Since 1 < PS00 then the leverage effect and advantage linked to option appears here. But the
other cases ( as for no ATM-put, or − PS00 vT (·) with small or moderate size, . . . ) deserve also
some analyses, we are going to perform now.
As written in (4), in absence of transaction cost, the investment return related to the stock
XX during the time-period [0, T ] is given by the corresponding relative change vT (·). This is the
case for a long position in the stock. When betting on the asset change downward movement,
the suitable position to take is the short one, which consists to sell initially the stock at the
unit price S0 , by borrowing the corresponding security, and at the future-time T to purchase
the same quantity of assets in order to return back the borrowed stock securities.
A standard approach used over the literature is to reduce the corresponding short position
return as just the opposite of the asset relative change vT (·). This is too simplistic and conse-
quently to stick more closely on the real practice, we will take into account some features
related to short selling position : as the initial collateral and borrowing interest rates. But, for
clarity and simplification, we do not incorporate into our formulation neither the transaction
fees, nor the full daily call-margin requirement involved in the short selling transaction.
In order to sell short the stock XX, at the unit price S0 , it is first required to have in
account the corresponding securities. To this end, the investor financial intermediary needs to
borrow these securities. Such a service and the corresponding interest until the future-time T
is assumed to be
ρ0 T S0 , with 0 < ρ0 < 1,
which is immediately paid at the present time 0. Here ρ0 may be considered as the borrowing
interest rate. To cushion loss, resulting from the asset adverse movement, the short selling
position is often done with a payement, at the initial time, of a collateral
λ0 S0 , with 0 < λ0 < 1,

10
as λ0 = 25% for instance. In order to prevent solicitation from daily call-margin requirement,
it is also suitable to put initially in reserve some amount
n o
B0 , with 0 ≤ B0 < (1 − λ0 ) − ρ0 T S0 .

Therefore the initial wealth to bring for a short selling operation is

 1 
W0 = λ0 + ρ0 T + B0 S0 = νS0
S0
with
1
ν = λ0 + ρ 0 T + B0 .
S0
In the sequel it is assumed that
P0
0<ν<1 and <ν
S0
in contrast with the case supposed in literature standard approach for which ν = 0.
To simplify, here the next step is taken as the same as the last step, which is the maturity
T . However in reality, things are little bit complicated since the next step is the next day, and
in general there are many days before reaching the maturity. Therefore at the time-maturity
T , first it is necessary to purchase the stock XX, at unit price ST (·) in order to return back the
securities borrowed from the initial time 0. Then the investor receives S0 , which corresponds to
the sale amount at time 0. Moreover, depending on the contract with the financial intermediary,
she also receives the interest

η0 T S0 , with 0 ≤ η0 < ρ0 < 1

in compensation with the privation of this amount S0 during the period (0, T ). Assuming that
there is no dramatic market changes between 0 and T , then the client should also retrieve the
initial reserve λS0 + B0 . Therefore the client final wealth at time T is given by
   
WT = −ST (·) + S0 + λ0 S0 + B0 + η0 T S0 = −vT (·)S0 + λ0 S0 + B0 + η0 T S0 .

The collateral λS0 is fixed by the financial intermediary in order to ensure that this wealth WT
should be positive, that is
1
vT (·) ≤ λ0 + B0 + η0 T.
S0
From this last inequality, we can understand that roughly speaking we should have vT (·) ≤ λ0 .
It means that if vT (·) < 0, in accordance with the stock downward movement betting related to
the short selling operation, then the collateral does not play any role. In contrast if 0 < vT (·),
then the collateral part is useful in order to absorb any adverse asset change increases of order
up to λ0 , ( with for instance λ0 = 25% for various practical situations ). The fact that the final
wealth WT is positive does not imply necessarily that the investor gets a positive profit&loss.
Since the investment initial is W0 = λ0 S0 + B0 + ρ0 T S0 then the return linked to our stock
investment via a short-selling operation is given by
1 1 
ret_asset_SS(·) = P&L_asset_SS(·) = − vT (·) + (ρ0 − η0 )T (38)
W0 ν

11
with ν = λ0 + ρ0 T + S10 B0 . It should be stressed here that this identity has only a sense under
the assumption
1
vT (·) ≤ λ0 + η0 T + B0 (39)
S0
0 < λ0 < 1 (40)
0 ≤ η0 < ρ0 < 1 (41)
and
1 n o
0 < ν = λ0 + ρ0 T + B0 < 1 with 0 ≤ B0 < (1 − λ0 ) − ρ0 T S0 . (42)
S0
Formulation of the return related to a short-selling operation is in general more involved than
(38) when we take into consideration the various daily call-margins during the time-period
(0, T ). However when the daily asset changes are roughly small in comparaison with λ0 and
when the initial cash B0 is large enough, then we are reduced to a formula which remains
essentially the same as (38). Moreover the short-selling operation has really a full economical
meaning whenever the maturity period remains not too long.
In order to state a general result related to the returns associated with the put-option and
short-selling operation, we introduce the following two quantities
P0 1
w0•• = w0 − = (K − P0 ) − 1 (43)
S0 S0
and
θw0•• + (ρ0 − η0 )T S 
0 W0
w0• = where θ = ν = > 1. (44)
θ−1 P0 P0
Recall that the put can be exercised whenever the future asset change vT (·) goes under w0 .
Here w0•• may be seen as the put-break-even point, that is the put return becomes nonnegative
only when vT (·) is below w0•• . As will seen below in the proof of Theorem 4, the put return is
less important than with the associated asset short position whenever the stock relative change
remains above w0• . Observe that
w0•• < w0
but there is no obvious order when considering w0• . However, for any put with given characte-
ristics ( i.e. strike K, premium P0 and spot S0 ), only one of the following five assumptions is
satisfied
w0•• + (ρ0 − η0 )T < 0 (45)
w0•• + (ρ0 − η0 )T = 0 (46)
P 
0
0 < w0•• + (ρ0 − η0 )T < (θ − 1) (47)
S0
P 
0
(θ − 1) = w0•• + (ρ0 − η0 )T (48)
S0
and P 
0
(θ − 1) < w0•• + (ρ0 − η0 )T. (49)
S0
These five assumptions arise naturally since
−1 n •• o
w0•• − w0• = w0 + (ρ0 − η0 )T
(θ − 1)

12
and  
1 P0  ••
w0• − w0 = −(θ − 1) + w0 + (ρ0 − η0 )T .
(θ − 1) S0
Under the assumption (45) it is clear that

w0• < w0•• < w0 . (50)

Here we have the symmetric situation as seen for the case of a call-option long position.
Assumption (45) means that w0•• = w0 − PS00 is strongly negative. This may be happened when
w0 ≤ 0 ( that is for ATM or OTM-put ).
While with the assumption (46) we have

w0• = w0•• < w0 . (51)

Under the assumption (47) it appears that

w0•• < w0• < w0 . (52)

Up to here, the situation is new and has not an equivalent for the call-option case. As an
explanation is the short selling operation, for which the associated return depends not only on
the asset relative change but also on the characteristic of the contract under consideration.
Assumptions (48) and (49) respectively imply

w0•• < w0 = w0• . (53)

and
w0•• < w0 < w0• . (54)

We are now ready to state the main result for the put-option and short-selling operation.
In view of the possible situations to consider as written in (50) to (54), it should be suitable to
present five separated results. But for shortness we choose to state just one main result which
encompasses all the arisen eighten cases.
Theorem 4 We assume that all assumptions (39) to (42) are sastisfied. As seen in (38),
observe that
1 
ret_asset_SS(·) = − ret_asset(·) + (ρ0 − η0 )T .
ν
1
1. Case : w0 ≤ vT (·) ≤ λ0 + η0 T + S0
B and under hypothesis (45) Here we have

ret_put(·) = −1 = −100% (55)

such that
ret_put(·) < ret_asset_SS(·) (56)
and
1n o
0 < ret_asset_SS(·) − ret_put(·) ≤ − w0 + (ρ0 − η0 )T + 1. (57)
ν
2. Case : w0•• ≤ vT (·) < w0 and under hypothesis (45). Here we have
 S  
0
ret_put(·) = −vT (·) + w0•• (58)
P0

13
  θ − 1 n o
ret_asset_SS(·) − ret_put(·) = vT (·) − w0• (59)
ν
such that
−1 < ret_put(·) ≤ 0 (60)
ret_put(·) < ret_asset_SS(·) (61)
and
 θ − 1 n o  θ − 1 n o
0< w0•• − w0• ≤ ret_asset_SS(·) − ret_put(·) < w0 − w0• . (62)
ν ν
3. Case : w0• ≤ vT (·) < w0•• and under hypothesis (45). Here the put-return and the diffe-
rence between the two returns ( put and asset short-selling ) are also given respectively by (58)
and (59). Moreover we have
 S n o
0
0 < ret_put(·) ≤ w0•• − w0• (63)
P0
ret_put(·) ≤ ret_asset_SS(·) (64)
and  θ − 1 n o
•• •
0 ≤ ret_asset_SS(·) − ret_put(·) < w0 − w0 . (65)
ν
4. Case : −1 < vT (·) < w0• and under hypothesis (45). Here the put-return and the diffe-
rence between the two returns ( put and asset short-selling ) are also given respectively by (58)
and (59). Moreover we have
 S n o  S n o
0 •• • 0 ••
0< w0 − w0 < ret_put(·) < 1 + w0 (66)
P0 P0
ret_asset_SS(·) < ret_put(·) (67)
and  θ − 1 n o
0 < ret_put(·) − ret_asset_SS(·) < 1 + w0• . (68)
ν
1
5. Case : w0 ≤ vT (·) ≤ λ0 + η0 T + S0
B and under hypothesis (46) The results stated in Case
1. also hold ; that is (55), (56) and (57) are both satisfied.
6. Case : w0• = w0•• ≤ vT (·) < w0 and under hypothesis (46). Here the put-return and the
difference between the two returns ( put and asset short-selling ) are also given respectively by
(58) and (59). Moreover we have

inequalities (60) are satisfied (69)

inequality (64) is satisfied (70)

and  θ − 1 n o
0 ≤ ret_asset_SS(·) − ret_put(·) < w0 − w0• . (71)
ν
7. Case : −1 < vT (·) < w0• = w0•• and under hypothesis (46). Here the put-return and the
difference between the two returns ( put and asset short-selling ) are also given respectively by
(58) and (59). Moreover we have
 S n o
0
0 < ret_put(·) < 1 + w0•• (72)
P0
14
 inequality (67) is satisfied (73)
and  θ − 1 n o
0 < ret_put(·) − ret_asset_SS(·) < 1 + w0•• . (74)
ν
1
8. Case : w0 ≤ vT (·) ≤ λ0 + η0 T + S0
B and under hypothesis (47) The results stated in Case
1. also hold ; that is (55), (56) and (57) are both satisfied.
9. Case : w0• ≤ vT (·) < w0 and under hypothesis (47). Here the put-return and the diffe-
rence between the two returns ( put and asset short-selling ) are also given respectively by (58)
and (59). Moreover we have
 S n o
0
−1 < ret_put(·) ≤ w0•• − w0• < 0 (75)
P0
inequality (64) is satisfied (76)
and  θ − 1 n o
•
0 ≤ ret_asset_SS(·) − ret_put(·) < w0 − w0 . (77)
ν
10. Case : w0•• ≤ vT (·) < w0• and under hypothesis (47). Here the put-return and the diffe-
rence between the two returns ( put and asset short-selling ) are also given respectively by (58)
and (59). Moreover we have
 S n o
0
−1 < − w0• − w0•• < ret_put(·) ≤ 0 (78)
P0
inequality (67) is satisfied (79)
and  θ − 1 n o
0 < ret_put(·) − ret_asset_SS(·) ≤ w0• − w0•• . (80)
ν
11. Case : −1 < vT (·) < w0•• and under hypothesis (47). Here the put-return and the diffe-
rence between the two returns ( put and asset short-selling ) are also given respectively by (58)
and (59). Moreover we have

inequalities (72) are satisfied (81)

inequality (67) is satisfied (82)

and
 θ − 1 n o  θ − 1 n o
0< w0• − w0•• < ret_put(·) − ret_asset_SS(·) < 1 + w0• . (83)
ν ν
1
12. Case : w0 ≤ vT (·) ≤ λ0 + η0 T + S0
B and under hypothesis (48) The results stated in
Case 1. also hold ; that is (55), (56) and (57) are both satisfied.
13. Case : w0•• ≤ vT (·) < w0 = w0• and under hypothesis (48). Here the put-return and the
difference between the two returns ( put and asset short-selling ) are also given respectively by
(58) and (59). Moreover we have

inequality (60) is satisfied (84)

15
 inequality (67) is satisfied (85)
and  θ − 1 n o
0 < ret_put(·) − ret_asset_SS(·) ≤ w0• − w0•• . (86)
ν
14. Case : −1 < vT (·) < w0•• and under hypothesis (48). Here the situation is the same as
in Case : −1 < vT (·) < w0•• and under hypothesis (47).
15. Case : w0• ≤ vT (·) ≤ λ0 + η0 T + S10 B and under hypothesis (49) Here we have that

identity (55) is satisfied (87)

inequality (61) is satisfied (88)

and
1n • o
−1 + w0 + (ρ0 − η0 )T ≤ ret_asset_SS(·) − ret_put(·) < 0. (89)
ν
16. Case : w0 ≤ vT (·) < w0• and under hypothesis (49) Here we have that

identity (55) is satisfied (90)

inequality (61) is satisfied (91)

and
1n o 1n • o
−1 + w0 + (ρ0 − η0 )T ≤ ret_asset_SS(·) − ret_put(·) < −1 + w0 + (ρ0 − η0 )T < 0.
ν ν
(92)
••
17. Case : w0 ≤ vT (·) < w0 and under hypothesis (49). Here the put-return and the diffe-
rence between the two returns ( put and asset short-selling ) are also given respectively by (58)
and (59). Moreover we have

inequalities (60) are satisfied (93)

inequality (67) is satisfied (94)

and
 θ − 1 n o  θ − 1 n o
0< w0• − w0 < ret_put(·) − ret_asset_SS(·) ≤ • ••
w0 − w0 . (95)
ν ν

18. Case : −1 < vT (·) < w0•• and under hypothesis (49). Here the put-return and the diffe-
rence between the two returns ( put and asset short-selling ) are also given respectively by (58)
and (59). Moreover we have

inequalities (72) are satisfied (96)

inequality (67) is satisfied (97)

and
inequalities (83) are satisfied. (98)

For the given characteristics of put-option ( strike K, spot S0 , premium P0 ), only one of the
five assumptions (45) to (49) is satisfied. Theorem 4 appears to be a useful investment decision

16
tool when it is combined with the investor view about the possible level of asset relative change
at the maturity T . To this end, first one should compute the real numbers w0 , w0• and w0•• as
defined respectively from (2), (44) and (43). Next one has to identify which of hypotheses (45)
to (49) is satisfied. Then the description of returns corresponding to the most probable ( or
forecasted ) future situation may be read from the eighteen cases described in Theorem 4. Of
course, introducing a model for the asset dynamic, as we will do in Proposition 6, may lead to
determine the probability of a required event.
Though our Theorem 4 describes the quantities involved in each situation, we try now to
give a more qualitative understandable form of this result under the assumption (45), which
appears to be satisfied in the numerical situations we consider below and resulting from market
data.
The case 1) : w0 ≤ vT (·) ≤ λ0 + η0 T + S10 B, corresponds to the fact that the asset relative
change has increased enough such that we get the worst situation for the put. It cannot be
exercised and the return is equal to −100%. Nevertheless the return for the asset short-selling
position remains more than that of this put, as we can read in (56). If w0 ≤ vT (·) < −(ρ0 −η0 )T
then the the asset short-selling position leads to a profit, which is not the case for the put
position for which we loose all of the amount invested. Always for the first case 1), as we may
see from (57), the distance between the returns of the put and underlying asset short-selling
remains bounded by the fixed number − ν1 {w0 + (ρ0 − η0 )T } + 1.
The case 2) : w0•• ≤ vT (·) < w0 , corresponds to the fact that the put-option may be
exercised, but the stock decrease is not sufficiently enough such that the put return remains
negative ( see (60) ). Once again here, the return for the stock short-selling position remains
strictly more than that for this put, as we can read in (61). Moreover for this second case 2), as
we may see from (62), the distance between thereturns  of the put and underlying asset short
position remains bounded by the fixed number ν {w0 − w0•• }.
θ−1

The case 3) : w0• ≤ vT (·) < w0•• , corresponds to the fact that the put may be exerciced and
the asset decrease is sufficiently enough in order that the put return becomes now positive (
see (63) ). Again here, the return for the stock short-selling position is above than that for
this put, as we can read in (64). Though the put position wins a profit, this last remains less
than that obtained directly with the stock short-selling position. As we may see from (65), the
distance between  the returns
 of the put and stock short-selling position remains bounded by
the fixed number ν {w0 − w0• }.
θ−1 ••

The case 4) : −1 < vT (·) < w0• may be viewed as the case where the asset has done a strong
decrease. Not only the put may be exerciced but the stock decrease is sufficiently enough in
order that the put return becomes now positive ( see (66) ) and, as can be read in (67), this
return is larger than the return linked to the stock short-selling position. It may be also seen
from (66) that the put position leads always to a profit and a nonnegative lower bound for the
corresponding return may be given. Though the put return may be important, the distance
between the returns of the put and underlying asset short-selling remains bounded by the fixed
number as given in (68).
Numerical illustrative examples and further comments are displayed in the next section.
Over these examples it appears that for many situations it is preferable to take a stock short-
selling position rather than to invest in the associated put-option. As a consequence, it may be
interesting to consider the strategy of covered put, which consists to sell a put and sell short
the stock. Selling a put means having a bullish or neutral view on the asset movement. Selling
short the stock is done in order to cover an unexpected downward asset change. The return
associated to the covered-put strategy is given by the following.

17
Proposition 5 Assume that we have sold a put with the premium P0 , 0 < P0 < S0 , and
simultaneously sold short the underlying stock XX at the unit price S0 . For this last operation
to make a sense, we suppose that all assumptions (39) to (45) are sastisfied. Then the covered-
put strategy leads to the profit
Profit&Loss_covered_put(·) = (νS0 )ret_covered_put(·)
where

1 n P
0
o
ret_covered_put(·) = −vT (·) + − (ρ0 − η0 )T I{w0 ≤vT (·)}
ν S0
n o 
••
− w0 + (ρ0 − η0 )T I{vT (·)<w0 } . (99)

As a consequence under the technical assumption (45), which the same as w0•• + (ρ0 − η0 )T < 0,
the covered put leads always to a profit whenever the asset relative change vT (·) remains strictly
P 
0
under − (ρ0 − η0 )T .
S0
The covered-put is a winning strategy as long as the future asset change vT (·) does not move
above PS00 − (ρ0 − η0 )T nor goes under w0•• . For vT (·) ∈ [w0 , PS00 − (ρ0 − η0 )T [ the put-covered
n o
strategy has an effect to enhance the return − ν1 vT (·) into − ν1 vT (·) + ν1 PS00 − (ρ0 − η0 )T . For
n o
vT (·) ∈ [w0•• , w0 [ the put-covered strategy has an effect to fix the return at − ν1 w0•• +(ρ0 −η0 )T .
Though this last is also the return for vT (·) < w0•• , this case should correspond to a bearish
view which is out of the strategy spirit.
Observe that here the initial wealth invested is νS0 , with 0 < ν < 1, which is smaller than S0
required for a long position on the stock.
We end with the analogous result for the put-option as stated in Proposition 3.
Proposition 6 Assume that the underlying asset stock XX follows the GBM process (24). Let
us consider a short position on the underlying asset and a long position on the put. Suppose that
assumption (45) is satisfied so that w0• < w0•• < w0 . Then we have the following probabilities
estimates :
h i  
P w0 ≤ vT (·) = 1 − Φ x(w0 ) (100)
h i    
P w0•• ≤ vT (·) < w0 = Φ x(w0 ) − Φ x(w0•• ) (101)
h i    
P w0• ≤ vT (·) < w0•• = Φ x(w0•• ) − Φ x(w0• ) (102)
h i  
P −1 < vT (·) < w0• = Φ x(w0• ) . (103)

where Φ denotes the cumulative probability function of the standard gaussian normal law ; w0 ,
w0• and w0•• are defined respectively as in (2), (44) and (43). Remind that x(w) is defined as
in (25).

3 Numerical Results
3.1 Examples for the call position
Table1 is related to the quote for options on Agricole Credit (AC3) as seen from the website
of the Nyse-Euronext ( www.euronext.com/trader/priceslistderivatives/derivativespriceslists-

18
1930-FR.html ) on December 24, 2009. Here we consider european options with the time-
maturity September 2010. In columns 1 and 2 of Table 1 the strikes and premia of the considered
call are reported respectively. Moreover, the computed values of w0 , w0∗ and w0∗∗ as defined in
(2), (6) and (7) respectively are presented in columns 3 to 5. Following our result in Theorem1,
these values and the investor view on the asset relative change level for the maturity T = 276
days can be mixed in order to shed light the best investment to perform ( call-option or
underlying stock ).
For instance, we observe here that for a OTM-call with the strike-price 14 euros, the call-
exercise may be done only if the asset increase is above w0 = 15.05%, and the investment in
call-option is better than that in the underlying stock only for a level of asset change above
w0∗∗ = 24.45%. Clearly this last situation should not be realized unless the market has done
an exceptional upward jump. Having a bullish market view for the underlying asset at the
maturity, the best investment to perform here is to take a long position on the stock.
For an almost ATM-call, i.e. with the strike-price 12 euros, the call-exercise arises if at least
the asset relative change level is above −1.45%. So it seems to be not too difficult to reach
as long as the market remains slightly bullish at the maturity T . However, by part4 of our
Theorem 1, the investment in the call is superior to that in the underlying stock only if the
asset change increase level is above w0∗∗ = 14.65%. Such a requirement is not so easy to realize
in a neutral or slightly bullish market, so here the investment in stock may be preferable to
choose. If the investor expects to get at least a stock relative increase of 3.5% then she may
get a best return by investing in a ITM-call with the strike 8.8 euros ( or 8 euros ) rather than
investing in the underlying stock.
The remaining examples presented from Table 2 are not directly drawn from real market
data. But we have introduced them in order to better grasp how do the various related factors
( as spot price, strike-level, . . . ) affect the decision investment. For this purpose, the call prices
are generated by using the ( practical ) standard Black-Scholes-Merton approach for a given
choice of implied volatility surface. As introduced in [Al], this volatility surface is defined from
the following formula
 
σ ≡ σ τ, S, K, r = β0 + β1 x + β2 x2 + β3 τ + β4 xτ (104)

with  
  S exp(rτ )
x ≡ x τ, S, K, r = ln
K
where S is the underlying spot level, K is the strike-price, r is the risk-free interest-rate and
τ is the remaining time-to-maturity. The constants β0 , . . . , β4 are inferred from the available
option prices. In our example we take β0 = 0.0990, β1 = 0.0408, β2 = 0.4671, β3 = 0.0272 and
β4 = 0.2762.
Tables 2 and 3 are related to call-options associated with the spot S0 = 12 euros, maturity
T = 30 days, and interest rate r = 2.85%. It is clear from Table 2 that for any ITM-call with
the strike-price level less than 11.5 euros then the investment with the call is interesting if the
investor expects that for the maturity T , the asset increase level is at least above 0.34%. For
the OTM-call with the strike-price 12.5 euros, the investment with the underlying stock seems
better than with the call whenever the investor forsees an asset level which cannot go beyond
4.20%. Similarly the call-investment with the strike price 13 euros is only interesting whenever
one expect an asset growth of at least 8.35% for the remaining 30 days time-to-maturity.
Then these results seen in Table 2 would lead us to think that things are easy. However, in
real exchange market situation any call-option with a given strike level may be not quoted.
Moreover it is important to feel the asset change level at maturity.

19
 To localize the interval which would contain the final asset change value vT (·) is useful. This
end is our target with Table 3. In columns 3 to 6, we can read respectively the probabilities
h i h i h i h i
P −1 < vT (·) ≤ w0 , P w0 < vT (·) ≤ w0∗ , P w0∗ < vT (·) ≤ w0∗∗ and P w0∗∗ < vT (·)

computed from (26) to (29) with the drift µ = r = 2.85%.

For the ITM-call with the strike-price 11.5 euros, we have seen in Table 2 that the call-
investment is preferable if one expects at the maturity T = 30 days the asset increase level
to be at least 0.34%. With Table 3, h we see that such i a situation may be realized with the
∗
probability 48.15%. Observe that P w0 < vT (·) ≤ w0 = 44.5271% such that the more likely
situation is either w0 < vhT (·) ≤ w0∗ or w0∗∗i < vT (·). The event w0∗∗ < vT (·) seems little bit
difficult to obtain since P −1 < vT (·) ≤ w0∗∗ ≈ 51.85%.
For the OTM call with the strike-price 13.5, by Table 2, the stock investment seems to preferable
if for the remaining time-to-maturity T = 30 days the asset increase level does not exceed
12.50%. Probabilities
h values given
i in Table 3 go in this direction since for the same strike-price
13, one has P −1 < vT (·) ≤ w0 = 99.9969%.
Observe that in all of our Tables the values are rounded up to four digits. For instance with
the strike 13.5, the call-price is 2.7086 × 10−6 but not zero as we present in this Table.
Tables 4 and 5 are given with the same attention and purpose as for Tables 2 and 3. The
difference is that here we consider the time-maturity 180 days, but with the same spot price
12 and interest rate 2.85%.
Table 6 is realized under the same conditions as Table 2, that is for the spot S0 = 12
euros, maturity T = 30 days, and interest rate r = 2.85%. Our purpose here is to understand
the call-covered strategy as presented in Proposition 2. In the third column of this Table 6 is
displayed the value of the threshold − CS00 such that the strategy leads to a gain when the asset
relative change is above this quantity. In the fourth column we show the probability that the
asset relative change is over the threshold, when assuming the asset dynamic follows a GBM
process. It appears here that such a probability is high for ITM calls. As we have noted after
our Proposition 2, the call-covered strategy is only interesting if one excpects that the asset
relative change at maturity is of moderated size, otherwise it is better to bet on an outright
call. Therefore in the fifth column of Table 6, we have presented the probability that the asset
relative change lies nside the interval [− CS00 , w0∗ ].
As noted in the introduction, the numerical examples from Tables 1 to 5 are given in order to
clarify how interesting is the Theorem 1 in taking the investement decision. But these examples
do not necessarily reflect any given real market situation. Theorem 1 allows the investor to
have a first sight of the consequence of her investment choice between the call or its underlying
stock. However, as we have illustrated with the use of probabilities in Proposition 3, the other
determing aspect in the investment decision is the future asset change level localization inside
some interval. This is a matter of forecasting. A theoretical approach is to introduce a dynamic
model for tthe stock price, as the Black-Scholes one presented in (24). This last is suitable for
getting explicit expressions of the needed probabilities, though empirical studies show that the
real market situation is not captured by such a benchmark model.

3.2 Examples for the put position

Table 7 is related to the quote for options on Agricole Credit (AC3) as seen from the website
of the Nyse-Euronext ( www.euronext.com/trader/priceslistderivatives/derivativespriceslists-
1930-FR.html ) on December 24, 2009. Here we consider european options with the time-

20
maturity September 2010. In columns 1 and 2 of Table 7, the strikes and premia of the consi-
dered puts are reported respectively. The maturity considered is T = 276 days and the spot-
price is 12.17 euros. The short-selling operation is assumed to be possible under the following
assumptions : λ = 25%, cash level of 5 euros, borrowing interest rate of 5.80%, lending interest
rate of 2.50%. Moreover, the computed values of w0• , w0•• and w0 , as defined in (44), (43) and
(2), respectively are presented in columns 3 to 5. Here we are in the case w0• < w0•• < w0 which
holds under assumption (45). Following our result in Theorem 4, these values and the investor
view on the asset relative change level for the maturity T = 276 days can be mixed in order to
shed light the best investment to perform ( put-option or short-selling the underlying stock ).
For instance, we observe here that for a OTM-put with the strike-price 8.8 euros, the call-
exercise may be done only if the asset relative change at maturity is under w0 = −27.70%,
and the investment in put-option is better than that in the underlying stock only for a level
of asset change under w0• = −36.06%. Clearly this last situation should not be realized unless
the market has done an exceptional downward jump. Having a bearing market view for the
underlying asset at the maturity, the best investment to perform here is to take a short position
on the stock.
For an almost ATM-put, i.e. with the strike-price 12 euros, the put-exercise arises if at least the
asset relative change level is under −1.40%. So it seems to be not too difficult to reach as long
as the market remains slightly bearish at the maturity T . However, by part4 of our Theorem
4, the investment in the put is superior to that in the underlying stock only if the asset change
increase level is under w0• = −21.53%. Such a requirement is not so easy to realize in a neutral
or slightly bearish market, so here the investment in stock may be preferable to choose. If the
investor expects to get at least a stock relative decrease of −11.72% then she may get a best
return by investing in a ITM-put with the strike-price 16 euros rather than investing in the
underlying stock.
The remaining examples presented from Table 8 are not directly drawn from real market
data. But we have introduced them in order to better grasp how do the various related factors
( as spot price, strike-level, . . . ) affect the decision investment. The put prices are generated
by using the ( practical ) standard Black-Scholes-Merton approach associated with the same
implied volatility surface as for the call-option situation as in the previous subsection with
always β0 = 0.0990, β1 = 0.0408, β2 = 0.4671, β3 = 0.0272 and β4 = 0.2762.
Tables 8 and 9 are related to put-options associated with the spot S0 = 12 euros, maturity
T = 30 days, interest rate r = 2.85% and again with λ = 25%, cash level of 5 euros, borrowing
interest rate of 5.80% and lending interest rate of 2.50%.
From Table 8, it may be observed that for OTM-puts one has w0• < w0•• < w0 , and in contrast
w0•• < w0•• < w0 for ITM-puts. By our Theorem 4, for any ITM-put with the strike-price level
more than 12.5 euros then the investment with the put is interesting if the investor expects
that for the maturity T , the asset relative change level is at under 0.12%. For the OTM-put
with the strike-price 11 euros, the investment with the underlying stock seems better than
with the put whenever the investor forsees an asset level which cannot go under −8.33%. The
put-investment with the strike price 10.5 euros is only interesting whenever one expect an asset
relative change level under −12.50% for the remaining 30 days time-to-maturity.
To localize the interval which would contain the final asset change value vT (·) is useful.
This end is our target with Table 9. For puts, with strike prices ranging from 10.5 to 12 euros,
in columns 2 to 5, we can read respectively the probabilities
h i h i h i h i
P −1 < vT (·) ≤ w0• , P w0• < vT (·) ≤ w0•• , P w0•• < vT (·) ≤ w0 and P w0 < vT (·) .

In contrast, for puts, with strike prices ranging from 12.25 to 14 euros, in columns 2 to 5, we

21
can read respectively the probabilities
h i h i h i h i
P −1 < vT (·) ≤ w0•• , P w0•• < vT (·) ≤ w0• , P w0• < vT (·) ≤ w0 and P w0 < vT (·) .

For the computation we make use of GBM process with the drift µ = r = 2.85%.
For the ITM-put with the strike-price 12.5 euros, we have seen in Table 8 that the best return
for the put-investment at the maturity T = 30 days corresponds to the asset level under
0.12%. With Table i a situation may be realized with the probability 49.00%.
h 9, we see that such
•
Observe that P w0 < vT (·) ≤ w0 = 41.8103% such that the more likely situation is either
−1 < vT (·) ≤ w0•• or w0• < vT (·) < w0 .
Tables 10 and 11 are given with the same attention and purpose as for Tables 8 and 9. The
difference is that here we consider the time-maturity 90 days, and the other parameters remain
the same as in Tables 8 and 9.

4 Conclusions and perspectives

1. The main issue raised and analyzed in this paper is to decide about the superiority of the
investment in option compared with its underlying stock, as claimed by various invest-
ment banks and brokerage firms when promoting their derivative products. To cope with
this situation, in Theorems 1 and 4, we determine the return structures corresponding to
a long position on the option and a long or short position on the underlying stock. We
have also estimated the distance between the two returns. The analysis we have perfomed
here is general since it does not assume any specific situation or particular data. Applying
our results to real market and synthetic option prices, it appears that for many situations
the option investment may be worse than the investment in the underlying stock. This
happens when the future asset change level at the investment maturity is not sufficiently
high ( resp. low ) in the case of a call-option ( resp. put-option ). The threshold level
may be determined previously before investing and from our main results in Theorems 1
and 4. However, there is no definitive answer about the superiority or not of investing in
option. It is up to the investor to decide the best investment support to choose depending
on her views on the future asset relative change and on the option characteristics which
are available on the markets.
2. The returns of the call-option position and the associated underlying stock position are
compared in our Theorem 1. It is seen here that only when the stock relative change
has done a strong growth that the investment in call-option is superior. Moreover in this
case, the option leverage amplifies the asset relative change. It would be also interesting
to introduce the leverage effect with the stock position by borrowing one part of the
amount required to buy the corresponding asset. Then the analysis is closed to the one
used for the case of short-selling the underlying stock as presented in Subsection 2.2.
3. This paper is restricted to the case of european options and we have assumed that the
investment takes end only at the option maturity. It may be also interesting to raise
and investigate the question about the superiority or not of the investment in option
when the investment maturity is shorter than the option maturity. This situation seems
to be difficult to handle since specific models for the underlying asset process and the
corresponding option pricing are needed.
4. To get easy readable results and to simplify the analysis, in this work we have not consi-
dered the related transaction costs. Of course to get more reliable investment decision,

22
 fees and liquidity factors should be included into the analyses. Technics and ideas pre-
sented in [Ra1] may be used for this purpose, nevertheless further details are presented
in 5.6 of the next Section 5.
5. It may be noted over this work, that considering a dynamic model for the asset evolution
is not so useful when analyzing and comparing the returns related to the investment in
option and the corresponding underlying stock. Indeed a clear separation between the
return structure and forecasting of the future asset change level should be made. An asset
model and the investor view/feeling may be helping for this last point. Understanding
the return structures, as we perform in this paper, is another thing which is at the core
and sucess of the investment decision. Our call-option return analysis is used to derive an
affine relationship between the returns of the call and its underlying stock when the asset
price follows a GBM process. As an application we get a high bound for the call-option
price in order that the expected return for the call is higher than that of its underlying
asset.
6. As will seen over the proofs of our results, the option premium is essential when compa-
ring the returns between the investment in option and the correponding investment in
the associated underlying stock. Conversely, when raising and solving incoherence and
drawbacks related to the Black-Scholes-Merton pricing, I. Gikhman [Gi] has lead to com-
pare the returns for the two investments in order to define the good notion of option price
( in stochastice sense ). One question, underlying the issue we consider in this paper, is
to understand whether the considered option is correctly priced by the market or the
model. If it is not the case, we may also ask about the kind of arbitrage to perform.

5 Proofs of results
5.1 Proof of Theorem 1
We need to make some preliminaries which are useful in the sequel of the proof. Assuming
that w0 < vT (·) and making use of (5) then the call return is given by
 S  
0
ret_call(·) = vT (·) − w0 − 1
C0
 S n  C o  S  
0 0 0
= vT (·) − w0 − = vT (·) − w0∗ . (105)
C0 S0 C0
Using this last expression we get

ret_asset(·) − ret_call(·) =
 S n o
0
= vT (·) − vT (·) − w0∗
C0
 S n C  o 
0 0 ∗
= − 1 vT (·) + w0
C0 S0
 S n  C o 
0 0
= 1− −vT (·) + w0∗∗ . (106)
C0 S0
Identities (105) and (106) are the main keys to derive Theorem 1. Particularly from (106) for
w0 < vT (·) it appears that

ret_asset(·) < ret_call(·) if and only if w0∗∗ < vT (·).

23
1. Case : −1 < vT (·) ≤ w0 .
Using (5) with the fact that vT (·) ≤ w0 then we get immediately (9). Moreover we have

v(·) + 1 = ret_asset(·) − ret_call(·) ≤ 1 + w0

which leads to (10) and (11).

2. Case : w0 < vT (·) ≤ w0∗ .
For w0 < vT (·), the call return written in (5) is just reduced to (12) because of identity
(105). Obviously the returns difference in (13) is obtained from (106).
Using identity (105) associated with w0 < vT (·) ≤ w0∗ and the fact that w0 − w0∗ = − CS00
then we get −1 < ret_call(·) ≤ 0 which is (14).
Since −w0∗ ≤ −vT (·) < −w0 , and using the identity (106), then the inequalities written in
(16) appear.
The fact that the put return is less than the asset return, as announced in (15), is an
immediate consequence of the first two inequalities from (16).
3. Case : w0∗ < vT (·) ≤ w0∗∗ .
Using identity (105) associated with w0∗ < vT (·) ≤ w0∗∗ then the call return estimates in
(17) appear immediatly.
Since −w0∗∗ ≤ −vT (·) < −w0∗ then, using the identity (106), we obtain the inequalities
written in (19).
The fact that the put return is less than the asset return as announced in (18) is also an
obvious consequence of the first inequality in (19).
4. Case : w0∗∗ < vT (·).
The low bound of the call return written in (20) arises by using identity (105) and the fact
that w0∗∗ ≤ vT (·).
Also from this last inequality and (106) then both (22) and (21) are satisfied.

5.2 Proof of Proposition 2

Using the notations as above then we have
n o n o
Profit&Loss_covered_call(·) = ST (·) − S0 − ST (·) − K I{K<ST (·)} + C0
n o
= vT (·)S0 + S0 vT (·) − w0 I{w0 <vT (·)} + C0
 n o  C 
0
= S0 vT (·) − vT (·) − w0 I{w0 <vT (·)} +
S0
n  C o 
0 ∗
= S0 vT (·) + I{vT (·)≤w0 } + w0 I{w0 <vT (·)} .
S0

Since the amount invested for the call-covered strategy is reduced to S0 then clearly the cor-
responding return ret_covered_call(·) is reduced to (23).

24
5.3 Proof of Theorem 4
We need to make some preliminaries which are useful in the sequel of the proof. Assuming
that vT (·) < w0 and making use of (37) then the put return is given by
 S  
0
ret_put(·) = w0 − vT (·) − 1
P0
 S n P  o  S  
0 0 0
= w0 − − vT (·) = −vT (·) + w0•• . (107)
P0 S0 P0
Using this last expression we get

ret_asset_SS(·) − ret_put(·) =
1n o  S n
0
o
= − vT (·) + (ρ0 − η0 )T − −vT (·) + w0••
ν P0
 S  n o 1n o
0 ••
= vT (·) − w0 − vT (·) + (ρ0 − η0 )T
P0 θ
S 1 n o
0 ••
= (θ − 1)vT (·) − θw0 + (ρ0 − η0 )T
P0 θ
 S  θ − 1  1 n •• o
0
= vT (·) − θw0 + (ρ0 − η0 )T
P0 θ (θ − 1)
 θ − 1 n o
= vT (·) − w0• . (108)
ν
Identities (107) and (108) are the main keys to derive Theorem 4. Particularly from (108) for
vT (·) < w0 it appears that

ret_asset_SS(·) < ret_put(·) if and only if vT (·) < w0• .

Under assumption (45) it appears that w0• < w0•• < w0 and consequently we have to study
the following four cases :
C.1) w0 ≤ vT (·)
C.2) w0•• ≤ vT (·) < w0
C.3) w0• ≤ vT (·) < w0••
and
C.4) − 1 < vT (·) < w0• .
With assumption (46) then w0• = w0•• < w0 and we are reduced to consider the following three
cases :
C.5) w0 ≤ vT (·)
C.6) w0• = w0•• ≤ vT (·) < w0
and
C.7) − 1 < vT (·) < w0• = w0•• .
Under assumption (47) it appears that w0•• < w0• < w0 and consequently we have to study the
following four cases :
C.8) w0 ≤ vT (·)
C.9) w0• ≤ vT (·) < w0

25
 C.10) w0•• ≤ vT (·) < w0•
and
C.11) − 1 < vT (·) < w0•• .
With assumption (48) then w0•• < w0• = w0 and we are reduced to consider the following three
cases :
C.12) w0• = w0 ≤ vT (·)
C.13) w0•• ≤ vT (·) < w0• = w0
and
C.14) − 1 < vT (·) < w0•• .
Under assumption (49) it appears that w0•• < w0 < w• and consequently we have to study the
following four cases :
C.15) w0• ≤ vT (·)
C.16) w0 ≤ vT (·) < w0•
C.17) w0•• ≤ vT (·) < w0
and
C.18) − 1 < vT (·) < w0•• .

Under one of cases C1, C.2, C.8, C12, C.15 and C. 16 one has w0 ≤ vT (·) and consequently
it is clear that ret_put(·) = −1 = −100%.
Under one of cases C.2, C.3, C.4, C.6, C.7, C.9, C.10, C.11, C13, C.14, C.17 and C.18 one
has vT (·) < w0 and consequently the value of put-return is given by expression (107).
The remaining details part by part for the proof of Theorem 4 may be now left to the
reader. The main points are essentially (107) and (108) by using the low and high bounds for
vT (·) and −vT (·) depending on the case under consideration.

5.4 Proof of Proposition 5

Using the notations as above then we have

Profit&Loss_covered_put(·)
n o n o
= −ST (·) + S0 − (ρ0 − η0 )S0 T − K − ST (·) I{ST (·)<K} + P0
n o n o
= −vT (·)S0 − (ρ0 − η0 )S0 T − S0 −vT (·) + w0 I{vT (·)<w0 } + P0
 n o  P 
0
= S0 −vT (·) − (ρ0 − η0 )T − −vT (·) + w0 I{vT (·)<w0 } +
S0
n P o
0
= S0 −vT (·) + − (ρ0 − η0 )T I{w0 ≤vT (·)}
S0
n o 
••
− w0 + (ρ0 − η0 )T I{vT (·)<w0 } .

Since the amount invested for the put-covered strategy is reduced to νS0 then clearly the
corresponding return ret_covered_put(·) is reduced to (99).

26
5.5 Proofs of Propositions 3 and 6
For the GBM process (24) and with the underlying probability P, the Itô lemma leads to
the future asset value
√
 
1 2
ST (·) = S0 exp µ − σ T + σ T UT (·)
2
where UT (·) is a standard gaussian normal law under the probability P.
The main point to get results in Propositions 3 and 6 is to observe that for −1 < a < b < ∞
n o n o
a < vT (·) ≤ b = x(a) < UT (·) ≤ x(b) .

Now identities (27) and (28) can be easily derived by taking respectively a = w0∗ , b = w0 and
a = w0∗∗ , b = w0∗ since
hn oi    
P a < vT (·) ≤ b = Φ x(b) − Φ x(a) .

Identities (26) and (29) appear since

   
lim Φ x(a) = lim Φ(y) = 0 and lim Φ x(b) = lim Φ(y) = 1.
a→−1+ y→0 b→∞ y→∞

Similar arguments may be used to get identities in Proposition 6.

5.6 The main issue under transaction costs

As promised in the previous Section 5, we outline in this part the solution to the problem
about the superiority or not of the investment in the stock compared with the corresponding
investment in the associated derivative when the transaction costs are taken into consideration.
For shortness, we just focuse on the case of long positions on call and the underlying stock XX.
As we have explained in [Ra1], when the transaction costs come into play, it is important
to consider the blocks of traded quantities as N S0 , N C0 , N ST and so on, since the transaction
cost functions used in ( at least european ) markets are essentially piecewise affine functions
but not just linear as assumed in various academic papers.
For the long position on N stocks bought at the unit price S0 , the initial wealth required is

Wa = N S0 + Φ(N S0 )

where Φ(x) is some piecewise affine function defined on [0, ∞[ with Φ(0) = 0, ( see [Ra] for
examples and full definition of the function Φ ). The corresponding profit&loss, when selling
all of the stocks at the unit price ST (·) = {1 + vT (·)}S0 is then
 
P&Lasset (·) = {1 + vT (·)}N S0 − Φ {1 + vT (·)}N S0 − Wa .

Obviously the return corresponding to such a long position on the stocks XX is just W1a P&Lasset (·).
On the other hand, when taking a long position on N calls having the stock XX as under-
lying asset, and K as a strike-price then the required initial wealth is

Wc = N C0 + Ψ(N C0 )

where Ψ(x) is some function given by the sum of a piecewise affine transaction cost function as
Φ and another function depending both on the traded quantity N C0 and the security number

27
N . The corresponding profit&loss, when buying such a call and assuming the exercise taking
place at the time-maturity T is given by
 
P&Lcall (·) = {1 + vT (·)}N S0 − Φ {1 + vT (·)}N S0 − (We + Wc )

where
We = N K + Φ(N K)
is the amount required to buy the N stocks XX at the unit strike-price K. It should be noted
here that most of time Wc < Wa since the premium C0 paid for the call is generally small
in compared with the stock unit price S0 . Of course the return corresponding to such a long
position on the call on XX is just W1c P&Lcall (·).
Stating that the investment in the stock is superior to the investment in the corresponding
call is saying that
1 1
P&Lcall (·) < P&Lasset (·)
Wc Wa
which is equivalent to
 
Wa
  Wc
(We + Wc ) − Wa
{1 + vT (·)}N S0 − Φ {1 + vT (·)}N S0 < Wa
. (109)
Wc
− 1

It means that we are reduced to solve this inequality for the vT (·) variable. Since the functions
Φ and Ψ are not linear, an approach is to make use a suitable cut-off of the semi-line ]−1, ∞[ as
we have performed in [Ra]. It appears from here that the problem becomes difficult whenever
the transaction costs are taken into consideration. Unfortunately nonlinear inequalities as (109)
must be handled if we are willing to deal with the real market situations.

6 References
[Al] : A. Alentorn (2004). Modelling the implied volatility surface : an empirical study for
FTSE options. See the website : www.theponytail.net/CCFEA
[Gi] : I. Gikhman (2009). Remarks on basics of financial modeling. See the website :
www.ssrn.com
[Me] : A. Meucci (2009). Enhancing the Black-Litterman and related approach : views and
stress-test on risk factors. Journal of Asset Management Vol. 10 (2). See also the website :
www.ssrn.com
[Os] : J. Osbonne (2009). Why buy a call instead of the stock ? See the website :
www.optionetics.com/market/articles/22149
[Ra1] : Y. Rakotondratsimba (2009). Effect of the asset change on the portfolio return in
presence of transaction costs. See the website : www.ssrn.com
[Ra2] : Y. Rakotondratsimba (2010). Is it really interesting to invest in option rather than
in its underlying stock ? Accepted the International J. Economic and Finance.
[Si] : H. Siddiqi (2010). Coarse thinking, implied volatility, and the price of call and put
options . See the website : www.ssrn.com

28
7 Appendix
7.1 Tables for the call-options
7.1.1 Table 1
strike call-price w0 (%) w0∗ (%) w0∗∗ (%)

7.38 4.7200 -39.3591 -0.5752 -0.9396

8 4.2400 -34.2646 0.5752 0.8827
8.80 3.6300 -27.6910 2.1364 3.0445
9.60 3.1100 -21.1175 4.4371 5.9603
10 2.8500 -17.8307 5.5875 7.2961
11.07 2.2200 -9.0386 9.2030 11.2563
12 1.7000 -1.3969 12.5719 14.6132
13 1.3700 6.8200 18.0772 20.3704
14 0.9200 15.0370 22.5965 24.4444
14.76 0.7100 21.2818 27.1159 28.7958
16 0.4600 31.4708 35.2506 36.6354
17 0.3100 39.6878 42.2350 43.3390
18 0.2100 47.9047 49.6302 50.5017
20 0.0900 64.3385 65.0781 65.5629

7.1.2 Table 2
strike volatility call- w0 (%) w0∗ (%) w0∗∗ (%)
(%) price

10.5000 11.8567 1.5249 -12.5000 0.2076 0.2378

11.0000 11.0703 1.0264 -8.3333 0.2198 0.2404
11.5000 10.5077 0.5382 -4.1667 0.3180 0.3329
12.0000 10.1421 0.1547 0 1.2890 1.3059
12.5000 9.9504 0.0147 4.1667 4.2894 4.2947
13.0000 9.9128 0.0004 8.3333 8.3363 8.3366
13.5000 10.0123 0.0000 12.5000 12.5000 12.5000
14.0000 10.2341 0.0000 16.6667 16.6667 16.6667

29
7.1.3 Table 3
strike call- proba1 proba2 proba3 proba4
price (%) (%) (%) (%)

10.5000 1.5249 0.0038 50.3278 0.3514 49.3169

11.0000 1.0264 0.2709 50.1426 0.2560 49.3305
11.5000 0.5382 7.1297 44.5271 0.1955 48.1477
12.0000 0.1547 47.3497 17.1177 0.2111 35.3214
12.5000 0.0147 91.1948 0.6371 0.0265 8.1415
13.0000 0.0004 99.6819 0.0009 0.0001 0.3171
13.5000 0.0000 99.9969 0.0000 0.0000 0.0031
14.0000 0.0000 100.0000 0.0000 0.0000 0.0000

7.1.4 Table 4
strike volatility call- w0 (%) w0∗ (%) w0∗∗ (%)
(%) price

10.5000 14.9239 1.6916 -12.5000 1.5968 1.8588

11.0000 13.5505 1.2372 -8.3333 1.9767 2.2040
11.5000 12.4271 0.8228 -4.1667 2.6900 2.8880
12.0000 11.5244 0.4780 0 3.9837 4.1490
12.5000 10.8176 0.2316 4.1667 6.0968 6.2168
13.0000 10.2852 0.0898 8.3333 9.0814 9.1499
13.5000 9.9085 0.0273 12.5000 12.7274 12.7564
14.0000 9.6714 0.0066 16.6667 16.7217 16.7309

7.1.5 Table 5
strike call- proba1 proba2 proba3 proba4
price (%) (%) (%) (%)

10.5000 1.6916 8.8888 43.8157 0.9706 46.3249

11.0000 1.2372 15.6508 38.4701 0.9206 44.9585
11.5000 0.8228 27.3400 29.9535 0.8582 41.8483
12.0000 0.4780 44.6653 18.8395 0.7301 35.7652
12.5000 0.2316 65.0114 8.4113 0.4825 26.0948
13.0000 0.0898 82.6652 2.3164 0.2004 14.8181
13.5000 0.0273 93.4826 0.3582 0.0445 6.1146
14.0000 0.0066 98.1233 0.0314 0.0052 1.8401

30
7.1.6 Table 6
strike call-price threshold proba win proba
(%) (%) good
appl(%)

10.5000 1.5249 -12.7076 99.9971 50.3288

11.0000 1.0264 -8.5531 99.7856 50.1991
11.5000 0.5382 -4.4847 94.2461 45.9029
12.0000 0.1547 -1.2890 69.4841 33.9516
12.5000 0.0147 -0.1228 54.4225 46.2545
13.0000 0.0004 -0.0030 52.7800 52.4628
13.5000 0.0000 -0.0000 52.6999 52.6968
14.0000 0.0000 -0.0000 52.6159 52.6159

7.2 Tables for the put-options

7.2.1 Table 7
strike put-price w0• (%) w0•• (%) w0 (%)

7.38 0.3600 -44.059 -42.317 -39.359

8 0.5000 -40.5900 -38.3730 -34.265
8.80 0.6900 -36.056 -33.361 -27.691
9.60 0.9600 -32.3400 -29.006 -21.118
10 1.0900 -30.316 -26.787 -17.831
11.07 1.5200 -25.6170 -21.5280 -9.0386
12 1.9300 -21.5270 -17.2560 -1.3969
13 2.5900 -19.6180 -14.4620 6.8200
14 3.1300 -15.3610 -10.6820 15.0370
14.76 3.6800 -13.7790 -8.9565 21.282
16 4.6500 -11.7120 -6.7379 -6.7379
17 5.5000 -10.812 -5.5053 39.688
18 6.3900 -10.6360 -4.6015 47.905
20 8.2600 -29.1380 -3.5333 64.339

31
7.2.2 Table 8
strike put-price w0• (%) w0•• (%) w0 (%)

10.5 0 0.0000 -12.5000 -12.5000 -12.5000

11 0.0003 -8.3360 -8.3357 -8.3333
11.25 0.0020 -6.2682 -6.2667 -6.2500
11.5 0.0109 -4.2627 -4.2573 -4.1667
11.75 0.0433 -2.4558 -2.4443 -2.0833
11.85 0.0689 -1.8373 -1.8243 -1.2500
12 0.1262 -1.0636 -1.0518 0.0000
strike put-price w0•• (%) w0• (%) w0 (%)

12.25 0.2772 -0.2266 -0.2237 2.0833

12.5 0.4851 0.1243 0.1520 4.1667
12.75 0.7225 0.2293 0.2822 6.2500
13 0.9695 0.2540 0.3308 8.3333
13.5 1.4680 0.2668 0.3949 12.500
14 1.9668 0.2768 0.4655 16.6667
14.5 2.4656 0.2866 0.5488 20.8333

7.2.3 Table 9
strike proba1 proba2 proba3 proba4
(%) (%) (%) (%)

10.5 0.0038 0.0000 0.0000 99.9962

11 0.2701 0.0001 0.0007 99.7291
11.25 1.6001 0.0020 0.0232 98.3747
11.5 6.6905 0.0239 0.4153 92.8703
11.75 18.3768 0.1059 3.4953 78.0220
11.85 24.4022 0.1416 6.6108 68.8454
12 33.2973 0.1485 13.9040 52.6503
strike proba5 proba6 proba7 proba4
(%) (%) (%) (%)

12.25 44.1975 0.0398 29.8114 25.9513

12.5 49.0004 0.3841 41.8103 8.8052
12.75 50.4530 0.7357 46.7919 2.0194
13 50.7960 1.0679 47.8180 0.3181
13.5 50.9766 1.7595 47.2608 0.0031
14 51.1141 2.5350 46.3509 0.0000

32
7.2.4 Table 10
strike put-price w0• (%) w0•• (%) w0 (%)

10.5 0 0.0040 -12.5394 -12.5337 -12.5000

11 0.0174 -8.4940 -8.4779 -8.3333
11.25 0.0351 -6.5664 -6.5423 -6.2500
11.5 0.0680 -4.7654 -4.7335 -4.1667
11.75 0.1247 -3.1567 -3.1227 -2.0833
11.85 0.1561 -2.5822 -2.5506 -1.2500
12 0.2139 -1.8056 -1.7826 0.0000
strike put-price w0•• (%) w0• (%) w0 (%)

12.25 0.3414 -0.7620 -0.7549 2.0833

12.5 0.5075 -0.0622 -0.0052 4.1667
12.75 0.7061 0.3659 0.4878 6.2500
13 0.9280 0.6002 0.7952 8.3333
13.5 1.4070 0.7750 1.1278 12.500
14 1.9009 0.8258 1.3554 16.6667
14.5 2.3971 0.8577 1.5958 20.8333

7.2.5 Table 11
strike proba1 proba2 proba3 proba4
(%) (%) (%) (%)

10.5 1.6583 0.0041 0.0245 98.3130

11 5.8864 0.0343 0.3159 93.7635
11.25 10.3104 0.0802 1.0041 88.6053
11.5 16.6675 0.1494 2.7864 80.3968
11.75 24.5032 0.2024 6.5556 68.7387
11.85 27.8086 0.2015 8.7822 63.2076
12 32.677 0.1587 12.9033 54.2613
strike proba5 proba6 proba7 proba4
(%) (%) (%) (%)

12.25 39.8617 0.0529 21.2881 38.7973

12.5 45.0137 0.4403 29.7946 24.7514
12.75 48.2744 0.9556 36.8071 13.9628
13 50.0923 1.5405 41.4161 6.9511
13.5 51.4684 2.7987 44.5173 1.2156
14 51.8736 4.1675 43.8114 0.1475

33