LOG IN SIGN UP

Blockchain and Cryptocurrencies Edited by Asma Salman and

Muthanna G. Abdul Razzaq

Asma Salman P… Tatjana Boshkov Leonardo Mataru… ‫ אראל‬Granot

Alessandra Creta… Marcelo Eduardo Ireneusz Miciuła Intech Department Ivan Šuljić

D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
 

Modeling Bitcoin Price and Bubbles13

http://dx.doi.org/10.5772/intechopen.79386

W t  r 
p  ffiffi ffi ffi ffi ffi ffi
 ¼ Zt þ 1 󿿽   r 2 V t , ½ �
t ∈   0 T 
; :   (20)

Without loss of generality, we assume that the interest rate is fixed and equal to zero. In this
setting, the discounted Bitcoin price trend and the market attention factor dynamics are
described by
( dSt
 ffiffi ffi
 ¼ μS AtStdt þ σSp  A  tStð dZt þ  dV t Þ , S0 ¼ s0
  r   r  ∈R þ ,
(21)
 ¼ μ A Atdt þ σ A AtdZt , A0 ¼ a0 þ ,
dAt ∈R

p  ffiffi 󿿽ffi ffi ffi ffi ffi
 r  ≔ 1    r 2 . The aim is to investigate the existence of asset-price bubbles in
where we have set
the underlying Bitcoin market model.
 S according to the model in Eq. (1) for several
By simulating trajectories for the asset price
values of the correlation parameter, it seems that the latter is related to the presence of bubble
 Figure  2, we plot examples of trajectories  for
effect; in fact, in r   0 ,  0 5 , 0 5 , 1, respectively : :

where higher positive values for the correlation appear to boost the asset value.  ¼  󿿽
Indeed, we will show formally that the possibility of Bitcoin speculative bubbles is related to
the sign of the correlation parameter  r .

The mathematical theory of financial bubbles is developed, among others,–25].

in [23
Precisely,
we introduce the following definition from [23].
Definition 4.1.
 The Bitcoin price process S has a bubble on the time interval ½ �
  0 T   if S is a strictF-local ;

martingale under the chosen risk-neutral measure.

D O W N LOA D P D F
The term strictF-local martingale refers to the fact that
 S is anF-local martingale, but not a true
F-martingale under the chosen risk-neutral measure. Further, S since
 is nonnegative, we must
have that  S   is an  F-supermartingale (we refer to [26] for rigorous definitions and related
concepts).
 p )

Figure 2.  Simulated trajectories with

 n = 250 daily observations for the attention process (red) and the corresponding
Bitcoin price dynamics for  ¼  ¼
 r   0 (black), r   0 5(green), and
 r   1 (blue).
:  ¼

14 Blockchain and Cryptocurrencies

ecall that the absence of arbitrage opportunitiesessentially

is  equivalent to the existence of 
 “ ”

a probability measure  Q , equivalent to the initial probability

 P , under which the discounted
rice process satisfies the martingale property.
 Note
Remark 4.2. that stock bubbles arise if S has an equivalent local martingale measure but not an
quivalent martingale measure. Arbitrage appears only if no equivalent local martingale measure exists.
 L   Lt ; t ∈
hen, to exclude arbitrage opportunities from the market, we define the process  ¼ f
½0; T 􏿽g by setting
t t t t
dQ  1 2  1 2

Lt ≔ dP F t
 ð  ð  ð 
¼ exp  󿿽 0 λudV u 󿿽 2 0 λudu 󿿽 0 γudZu 󿿽 2 0 γudu  , t  ½0; T 􏿽 ,   (22) ð   ∈

where λ ¼ f  λt ;   t  ½0; T 􏿽g and γ ¼ f  γt ;   t  ½0; T 􏿽g are -adapted processes satisfying the inte-
∈ ∈  F

grability conditions0T   λ2udu <  P-a.s. and 0T   γ2udu <  P-a.s., respectively. The (local) martin-
Р ∞
Р ∞

gale property of the discounted Bitcoin price process  S under Q implies the following condition:

 ¼ σSp  Atð  λt  þ γt Þ ,

μS At
 ffiffi ffi r  r  ½ 􏿽
t ∈   0; T  ,  󿿽
P  a:s::   (23)

o ensure that L provides the density process of a probability measure equivalent  P , weto
½ 􏿽¼
require that  E LT   1, meaning that   L   is an   F; P -martingale. The processes ð Þ
  λ   and  γ   are
nterpreted, respectively, as the risk premium and the risk perception associated to the future
direction or future possible movements of the Bitcoin market. For each choice of the γ process
 ,
he process λ is fixed by Eq. (23), that is,
 p  ffiffi ffi !
μS  At
λt  ¼ 1 r  σS
󿿽  r γt  , ½ 􏿽
t ∈   0; T  ,   (24)

 Q γ for
and we can consider the corresponding family of equivalent (local) martingale measures
S parameterized by the process
 γ. To check if there are stock bubbles in the underlying market
; P -local
model, we study under which conditions the discounted Bitcoin price is a  Fstrict ð Þ
 Q γ: By applying Girsanov
martingale with respect to an equivalent local martingale measure s ’

 γ
 Q is described by the following equations:
heorem, the dynamics of the model under
(  ¼ p  ~t
dSt  σS  A  t St   r dZ þ dV ~ t  , S0 ¼ s0 þ ,
 ffiffi ffi 󟿽  r 
􏿽 ∈R
  (25)
~ t ,
dAt ¼   μ A 󿿽 σ A γt  Atdt þ σ A AtdZ A0 ¼ a0 ∈R
þ ,

where the R2-valued process  V 

~ ; ~
Z
󟿽󟿽 󿿽􏿽󟿽 􏿽
Þ ¼   V ~ t; ~ Zt ; t  ½0; T 􏿽 ∈
� defined by ~
V t ≔ V t  þ 0t λudu,.
Р
~ t ≔ Zt
 þ 0t γudu, is anð ; Q Þ-standard Brownian motion.
Z
Р  F
 γ

 γ �is,
Now, suppose that the risk perception process is zero, that  0. Then, the change of measure
from P to Q0 is well-defined since the associated density process
 M ¼ f  Mt ; t  ½0; T 􏿽g satisfying ∈

Modeling Bitcoin Price and Bubbles15

http://dx.doi.org/10.5772/intechopen.79386

8< dMt  ¼ 󿿽
  μS p  ffiffi ffi
 A  t Mt dV t , M0 ¼ 1 ,
r σS   (26)
: dAt  ¼ μ A Atdt þ σ A At dZt , A0 ¼ a0:

is a true ðF; PÞ-martingale thanks to [27]. We have the following result, which allows to detect
the presence of bubbles in this setting.
Proposition 4.3.  (24) , the Bitcoin price process S has a bubble ½0on
 In the model outlined in Eq. ; T � if 
 r   > 0.
and only if 
The proof is based on the application of some ofsSinresults given in [27], where the existence
’

of risk-neutral measures for the Hull-White stochastic volatility model [19] and for similar
frameworks is determined by the possibility of explosion in finite time for solutions of certain
auxiliary stochastic differential equations. Precisely, it is possible to show that the martingale
 S under Q0 , given in Eq. (25) with
property of the discounted stock price  γ ¼ 0, is fulfilled if 

D O W N LOA D P D F
 r ≤ 0: Hence, a bubble arises if and only if the correlation parameter between stock
and only if 
returns and market attention is positive.

6. Toward a multiexchange generalization

 Let us generalize the model introduced in Eq. (1) by assuming a possible   delay τ  for the
attention factor to affect the Bitcoin price dynamics. Assume that the attention factor has been
observed or is described by a deterministic function  ½󿿽l; 0� with l ≥ τ. We get
for
 t ∈
8< dSt
¼ μS At󿿽τdt þ σS  At󿿽
p  ffiffi ffi ffi ffiffi
  τdW t , S0 ¼ s0 ∈ Rþ ,
St   (27)
: dAt  ¼ P At dt þ  A At dZt , At  ¼ φð tÞ for t ∈  󿿽l
 μ  σ ½ ; 0� ,

where φ   :  ½󿿽l; 0� ! Rþ.

Analogous results as those in Section 2 can be derived by similar computations, and model
parameters, for a fixed delay, can be estimated by means of the maximum likelihood method.

In order
[15]. to estimate
Details the delaycan
of this procedure parameter,
be foundwe maximize
in [10]. the profile
The estimation likelihood
results as defined
of model in
in Eq. (27)
on the same daily data considered in Section 2 are summed  Table
up in 3.
In Figure  3, we plot simulated trajectories of the price process in Eq. (27) by letting the delay
parameter vary.

τ    b
 μ A   b
σ  A   b
 μS   b
σ S

 A = Vol   1 day 0.4881 1.0459 0.0282 0.0924

 A = SVI    7 days 1.0964 0.9946 0.1005 0.1885

Table 3.  Parameter estimates for model in Eq. (27) fitted on daily observations from January 2015 to June 2017.

16 Blockchain and Cryptocurrencies

Figure 3.  Simulated trajectories  n

of = 250 daily observations of the attention factor (red) and the Bitcoin price according
to model in Eq. (27) when the delay parameter  τ ¼ 1
is day (black) and τ ¼ 10 days (blue).

The different delays result in a shift to the south-east between the faster and slower
reacting trajectories; in the picture, this behavior is sharp since the other model parameters
are kept constant. By looking at the picture, the idea to model the price of Bitcoin in
different exchanges by the same model in Eq. (27) but allowing different parameters natu-
rally arises.
In particular, considering for instance two exchanges, we have
1
8> dS
>> S ¼ μ  A dt þ σ p  ffiffiffiffiffiffiffiffiffiffiffi 
t 1 1 1 1 þ
1 S t󿿽τ1  A   dW  , S  ¼ s S t󿿽τ1 t 0 0 ∈ R  ,
< t
2 (28)
dS t 2
p 
 ffiffiffiffiffiffiffiffiffiffiffi 
>> S ¼ μ  A dt þ σ  A   dW  , S  ¼ s 2 2 2 þ
2 S t󿿽τ2 S t󿿽τ2 t 0 0 ∈ R  ,
t

>: dA  ¼ μ  A dt þ σ  A dZ , A  ¼ φð  tÞ for t

t
 A
t  A t t t ∈  ½󿿽l; 0� ,

where φ  :  ½󿿽l; 0� ! Rþ with l  > maxf  τ1; τ2 g and i

 μ A  , μiS , σ A
i
 >  , σiS  >
 0  0 for i = 1, 2 are constant
parameters.
Note that within this model, prices for Bitcoin traded in different exchanges are perfectly
correlated. Indeed, this is what happens in observed data; considering daily prices from
 January 2015 to June 2017 for Bitstamp, Kraken, Cex.io, Gdax, and The Rock exchanges we
get cross-correlation values larger than 0.999.

We
pricefitfrom
model in Eq. (28)
January 2015for
to the
JuneBitstamp and Gdax
2017 obtaining theexchanges
outcomeson daily observations
 Table
reported of Bitcoin
in  4, when the

Exchange   τ μ A   σ  A   μS   σ S

Bitstamp 1 0.4994 1.0461 0.0281 0.0896

Gdax 2 0.4997 1.0420 0.0326 0.1036

Table 4.  Model fitting with delay parameter: outcomes for Bitstamp and Gdax exchanges when attention is measured by
the trading volume.

Modeling Bitcoin Price and Bubbles17

http://dx.doi.org/10.5772/intechopen.79386

Exchange   τ μ A   σ  A   μS   σ S

Bitstamp 7 days 1.0934 0.9946 0.0992 0.1782

Gdax 7 days 1.0964 0.9946 0.1160 0.2087

Table 5.  Model fitting with delay parameter: outcomes for Bitstamp and Gdax exchanges when attention is measured by
the SVI index.

D O W N LOA D P D F
 Table
attention is measured by the trading volume, and in  5, when attention is measured by
the Google SVI index.
It is evident from the outcomes Table
in 4 that the model parameters are not significantly
 different while the delay might be quite different as if the reaction to the attention factor is
faster for some exchanges and slower for others. On the contrary, when attention is measured
 by the Google SVI Index, the delay is unchanged, but the difference between estimated
parameters for the price dynamics is nonnegligible.
By analyzing the outcomes and considering the shift effect as depicted   Figure
in  3, it is
tempting to conjecture that the faster reaction determines the leader exchanges and that the
slower exchange will then follow. If we could forecast that the next day price of the slower
exchange will reach the price today for the faster one, we could obtain a profit by suitably
investing in the two exchanges. However, it is worth noticing that the estimation of the delay
parameter is obtained by maximizing the likelihood over a whole time series and is a product
of averaging so arbitrage cannot be achieved in a direct way.
Nevertheless, in a multivariate setting as ours, the theory guarantees that arbitrage opportuni-
ties are ruled out if the market price of risk in the market is unique. Without entering technical
details and assuming  r 0 for the sake of simplicity, this is true if and only if 
¼

μ1S p 
 ffiffiffiffiffiffiffiffiffiffiffi  μ2S p  ffiffiffiffiffiffiffiffiffiffiffi 
 At   τ1 ¼  At   τ2 , t ≥ 0:   (29)
σ1S σ2S
� �

It is evident that these values are not equal if we plug parameter estimates in Eq. (30); hence,

arbitragethis
address opportunities are not ruled
issue more precisely out research.
in future at least from a theoretical point of view. We will

7. Conclusion

In this chapter, we have introduced a model in continuous time in order to describe the dynam-
ics of Bitcoin price depending on an exogenous stochastic factor, which represents market
attention on the Bitcoin system. Market attention is measured either by the total trading volume
in Bitcoins or by means of the Google Search Volume Index, which, as suggested in [16], is a
direct measure of the revealed attention for uniformed retail investors. More precisely, the
attention factor affects directly the instantaneous mean and volatility of logarithmic returns; in
addition, it may be also correlated with the price changes. An estimation procedure to fit the

18 Blockchain and Cryptocurrencies

model to observed data is also suggested and, under the assumption of no correlation, a closed
formula for standard European option prices on Bitcoin is provided.
By applying outcomes within the mathematical theory of bubbles –25,[23
27], we are able to show
hat Bitcoin boosts in a bubble if and only if there is a positive correlation between changes in the
rice and in the attention factor. This finding is reasonable and claims that a stronger positive
dependence between the two processes in Eq. (21) may result in an explosion of the price process.
inally, we allow for a delay on the effect of market attention on the Bitcoin price, and, based

on this generalized
o take into accountmodel, we introduce
the special a multivariate
feature of setting for
multiple exchanges our model
where (Eq. (28))
it is possible in order
to trade in
Bitcoins. Preliminary results indicate that arbitrage opportunities may arise in two exchanges
hat are characterized by different delays.

cknowledgements

We gratefully acknowledge Marco Patacca for having provided the routines in Matlab® to
develop the empirical sections of the paper and for useful suggestions and comments. We also
acknowledge funding from Fondazione Cassa di Risparmio di Perugia (Grant 2015:0459013)
and Bank of Italy (Grant 407660/16).

Conflict of interest

he authors declare no conflict of interest.

hanks

Gianna Figà-Talamanca gratefully acknowledges interesting discussions on the topic of this

chapter with colleagues of the Department of Finance and Risk Engineering where she was
visiting professor while preparing this research. The authors also wish to thank Stefano Bistarelli
for having introduced them to the intriguing and worth to explore world of cryptocurrencies.

uthor details
1 2
lessandra Cretarola
* and Gianna Figà-Talamanca
*Address all correspondence to: alessandra.cretarola@unipg.it
1 Department of Mathematics and Computer Science, University of Perugia, Perugia, Italy
Department of Economics, University of Perugia, Perugia, Italy

Modeling Bitcoin Price and Bubbles19

http://dx.doi.org/10.5772/intechopen.79386
D O W N LOA D P D F
References

[1] Nakamoto S. Bitcoin: A Peer-To-Peer Electronic Cash System. 2009. http://www.Bitcoin.org

 [1] Nakamoto S. Bitcoin: A Peer To Peer Electronic Cash System. 2009. http://www.Bitcoin.org
[2] Urquhart A. The inefficiency of bitcoin. Economics Letters.148
2016;
:80-82
–A garch volatility analysis. Finance Research
[3] Dyhrberg AO. Bitcoin, gold and the dollar
16:85-92
Letters. 2016;
[4] Katsiampa P. Volatility estimation for bitcoin: A comparison of garch models. Economics
158:3-6
Letters. 2017;
[5] Kristoufek L. BitCoin meets Google trends and Wikipedia: Quantifying the relationship
 between phenomena of the internet era. Scientific Reports.32013;
(3415):1-7
[6] Kristoufek L. What are the main drivers of the bitcoin price? Evidence from wavelet
10(4):e0123923
coherence analysis. PLoS One. 2015;
[7] Bukovina J, Marticek M. Sentiment and bitcoin volatility. Technical report, Mendel Uni-
versity in Brno, Faculty of Business and Economics; 2016
[8] Figà-Talamanca G, Patacca M. Does market attention affect Bitcoin returns and volatility?
SSRN Electronic Journal. 2018:11
[9] Blau BM. Price dynamics and speculative trading in bitcoin. Research in International
41
Business and Finance. 2017;:493-499
[10] Cretarola A, Figà-Talamanca G, Patacca M. Market Attention and Bitcoin Price Modeling:
Theory, Estimation and Option Pricing. Submitted; 2018
[11] Kim YB, Lee SH, Kang SJ, Choi MJ, Lee J, Kim CH. Virtual world currency value fluctua-
tion prediction system based on user sentiment analysis. PLoS One.102015;
(8):e0132944
[12] Yermack D. Is bitcoin a real currency? An economic appraisal. In: Handbook of Digital
Currency. USA: Academic Press; 2015. pp. 31-43

[13] and
LevyFinance.
E. Pricing European average rate currency options. Journal of International Money
11(5):474-491
1992;
[14] Davison AC. Statistical Models, Volume 11. USA: Cambridge University Press; 2003
[15] Pawitan Y. In all Likelihood: Statistical Modelling and Inference Using Likelihood. UK:
Oxford University Press; 2001
662011;
[16] Da Z, Engelberg J, Gao P. In search of attention. The Journal of Finance. (5):1461-1499
[17] Föllmer H, Schweizer M, Hedging of contingent claims under incomplete information. In
M.H.A Davis and R.J.Elliot, editors, Applied Stochastic Analysis. New York, Gordon and
5:389-414
Breach. 1991;
[18] Föllmer H, Schweizer M. Minimal martingale measure. In: Encyclopedia of Quantitative
Finance. Wiley Online Library; 2010

D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
D O W N LOA D P D F
Find new research papers in:

Physics

Chemistry

Biology

Health Sciences

Ecology

D O W N LOA D P D F
 Earth Sciences

Cognitive Science

Mathematics

Computer Science

D O W N LOA D P D F