Heisenberg Uncertainty Principle and Economic Analogues of Basic Physical Quantities
Heisenberg Uncertainty Principle and Economic Analogues of Basic Physical Quantities
Heisenberg Uncertainty Principle and Economic Analogues of Basic Physical Quantities
2
_
< ln
2
X
i
(t
n+1
)
X
i
(t
n
)
>
n,N
_
< ln
X
i
(t
n+1
)
X
i
(t
n
)
>
n,N
_
2
_1
2
. (15)
5
Pay attention that it was not necessary for us to prove the connection 12, as it was possible to postulate statement
(15) as the denition of x
i
.
It is also worth noting that the value
|v
i
(t
n
)| t
min
=
ln
X
i
(t
n+1
)
X
i
(t
n
)
,
which, accurate to multiplier t
min
coincides with |v
i
(t
n
)| (see (9)), is commonly named absolute returns, while
dispersion of a random value ln (X
i
(t
n+1
)/X
i
(t
n
)), which diers from D
vi
by (t
min
)
2
(see (13)) volatility.
The chaotic nature of real time series allows to x
i
(t
n
) as the trajectory of a certain abstract quantum particle
(observed at t
min
time spans). Analogous to (1) we can write an uncertainty ratio for this trajectory:
x
i
v
i
h
m
i
, (16)
or, taking into account (11) and (15):
1
t
min
_
< ln
2
X
i
(t
n+1
)
X
i
(t
n
)
>
n,N
_
< ln
X
i
(t
n+1
)
X
i
(t
n
)
>
n,N
_
2
_
h
m
i
, (17)
where m
i
- economic mass of an i series, h - value which comes as an economic Plancks constant.
Having rewritten the ration 17:
t
min
m
i
(t
min
)
2
_
< ln
2
X
i
(t
n+1
)
X
i
(t
n
)
>
n,N
_
< ln
X
i
(t
n+1
)
X
i
(t
n
)
>
n,N
_
2
_
h (18)
and interpreting the multiplier by t
min
in the left part as the uncertainty of an economical energy (accurate to
the constant multiplier), we get an economic analog of the ratio (4).
Since the analogy with physical particle trajectory is merely formal, h value, unlike the physical Plancks constant
, can, generally speaking, depend on the historical period of time, for which the series are taken, and the length of
the averaging interval (e.g. economical processes are dierent in the time of crisis and recession), on the series number
i etc. Whether this analogy is correct or not depends on particular series roperties.
Let us generalize the ratios (17), (18) for the case, when economic measurements on the time span T, used to derive
the series (6), are conducted with the time step t = k t
min
, where k 1 - is a certain given integer positive
number. From the formal point of view it would mean that all terms, apart from those numbered n = 0, k, 2k, 3k, ....
are discarded from the initial series (6). As a result the ratios would be the following:
1
kt
min
_
< ln
2
X
i
(t
n+k
)
X
i
(t
n
)
>
n,N
_
< ln
X
i
(t
n+k
)
X
i
(t
n
)
>
n,N
_
2
_
h
m
i
, (19)
kt
min
1
(kt
min
)
2
_
< ln
2
X
i
(t
n+k
)
X
i
(t
n
)
>
n,N
_
< ln
X
i
(t
n+k
)
X
i
(t
n
)
>
n,N
_
2
_
h
m
i
(20)
and would be dependent on k.
Let us proceed to the analysis of the acquired results, that have to be considered as intermediate.
In case of h = const, the formal analogy with the physical particle would be complete, and in this case, as appears
from (19), variance of a random i-numbered value:
ln
X
i
(t
n+k
)
X
i
(t
n
)
X
i
(t
n+k
) X
i
(t
n
)
X
i
(t
n
)
- practically coinciding with the relative increment of terms of the i initial series would keep increasing in a
linear way with kt
min
(interval between the observations) growing. Such dynamics is peculiar to the series with
statistically independent increments.
6
However, both in cases of a real physical particle and its formal economic analogue any kind of change inuences on
the result. Therefore statistic properties of the thinned series, used to create the ratio (19), have to depend on real
measurements in the intermediate points if there were any. Besides, presence of long and heavy tails increasing
along the amplitude with decreasing t on distributions of corresponding returns X/X, are in our opinion the
evidence of this thesis (see for example [31]).
Thus, generalizing everything said above, h/m
i
ratio on the right side of (19) (or (20)) has to be considered a certain
unknown function of the series number i, size of the averaging windowN, time n (centre of the averaging window),
and time step of the observation (registration) k.
To get at least an approximate, yet obvious, formula of this function and track the nature of dependencies, we
postulate the following model presentation of the right side (19):
h
m
i
( n, N
) H
i
(k, n, N
H
)
t
min
m
i
, (21)
where
1
m
i
= <
i
(n, 1) >
(0nN2)
, (22)
m
i
is a non-dimentional economic mass of an i-numbered series,
( n) =
<
i
(n, 1, N
) >
( nN
) >
( nN
2 <n< n+N
H
/2
<
i
(n, 1, N
H
) >
nN
H
2 <n< n+N
H
/2
; k = 1, 2, ...k
max
(24)
- non-dimentional coecient of the order of unit, which indicates dierences in the dependence of variance D
xi
(see (13) taking into account the case of k 1) on the law D
xi
k for the given i and n.
i
_
n, k,
N
_
=
1
k
_
ln
2
X
i
(t
n+k
)
X
i
(t
n
)
_
< ln
X
i
(t
n+k
)
X
i
(t
n
)
>
n,
N
_
2
_
(25)
(index
N = N, N
, N
H
in the last formula indicates the averaging parameters according to n and formulae
(22),(23),(24), averaging windows N
, N
H
are chosen with thew following the conditions taken into consideration:
k
max
< N
< N
H
< N. (26)
According to the denitions (23),(24) for coecients ( n) and H
i
(k, n) following conditions of the normalization
take place:
< ( n) >
n,N
= 1; H
i
(1, n) = 1, (27)
and the multiplier 1/t
min
on the right side (21) can be considered as an invariant component of an economic
Plancks constant h:
h = 1/t
min
, (28)
As you can see,
h has a natural dimension time to the negative rst power.
It is also worth noting that average economic mass of the whole set of series (or any separate group of the series)
can be introduced with the help of the following formula:
1
m
=
1
M
M
i=1
1
m
i
. (29)
7
Acquired with the help of series (6) ratios (7,(19)-(28) also allow dierent interpretations. For example, it can be
considered that normalized series (7) depict the trajectory of a certain hypothetical economic quantum quasi-particle
in an abstract M-dimensional space of economic indices, and m
1
i
are the main components of inverse mass tensor
of the quasi-particle (the analogy with quasi-particles as free carriers of the electric charge in semiconductors), which
has already been used in the previous chapter.
In the nal part of this chapter we would like to pay attention to the chosen variant of the theory, which is probably
the simplest one, because of the following reasons.
Carrying out various n (discrete time) and i (series number) averagings, we didnt take into account at least two
fairly important factors: 1) amounts of nancial and material resources (their movement is reected by each series)
and 2) possible correlation between the series.
However generalization of the theory and introduced notions is not so dicult in this case. It is enough to form a
row (
1
,
2
, ...
M
) of positive weight coecients with the following condition of normalization:
M
i=1
i
= M, (30)
with each of them taking into account the importance of separate series in terms of a certain criterion, while for
random values
i
(n, k) =
_
i
k
ln
X
i
(t
n+k
)
X
i
(t
n
)
, i = 1, 2, ...M (31)
instead of a one-dimensional massive (25) we should introduce a covariance matrix:
= [
ij
] , (32)
where
ij
=
ij
_
k,
N
_
= <
_
i
(n, k)
i
(n, k)
_
j
(n, k)
j
(n, k)
_
>
n,
N
, (33)
i
_
j, k,
N
_
= <
i
(n, k) >
n,
N
(34)
(with
i
= 1 and absence of correlations
ij
=
i
ij
). Using a standard algorithm of characteristic constants
i
, i = 1, 2, ...M and corresponding orthonormal vectors C
i
= (c
i1
, c
i2
, ...c
iM
) search in matrix, we proceed to
the new basis, where renormalized series y
i
(t
n
) =
M
j=1
c
ij
x
j
(t
n
) (new basis vectors) arent correlated any more.
However the presence of zero characteristic constants or
i
, which are distinguished with relatively low values in
absolute magnitude, will mean that the real dimension of the set of series (7) is in fact less than M (initial series (7)
or their parts are strongly correlated). In this case renormalized series y
i
(t
n
) with zero or low characteristic constants
have to be discarded. The remaining renormalized series will undergo all above-listed procedures.
4. EXPERIMENTAL RESULTS AND THEIR DISCUSSION
To test the suggested ratios and denitions we have chosen 9 economic series with t
min
in one day for the period
from April 27, 1993 to March 31, 2010. The chosen series correspond to the following groups that dier in their origin:
1) stock market indices: USA (S&P500), Great Britain (FTSE 100) and Brazil (BVSP);
2) currency dollar cross-rates (chf, jpy, gbp);
3) commodity market (gold, silver, and oil prices).
On Fig. 1-3 normalized plots of the corresponding series, divided by groups, are introduced, while t
min
is taken
equal to the unit.
As you can see from the Fig. 1-3, all time series include visually noticeable chaotic component and obviously dier
from each other, which allows us to hope for the successful application the afore-mentioned theory to the interpretation
and analysis of real series. Let us conne to its elementary variant.
As an example on g. 4 we suggest absolute values of immediate speeds (or absolute returns according to the
general terminology used in literature), calculated with the help of the formula evaluation (9), and their variance
8
(volatility), calculated with the help of the formula evaluation (13) for the series of Japanese yen (jpy) US dollar
cross-rates.
As we can see from the plots, the dependence of immediate speed or returns on time is of chaotic nature, while the
dependence of volatility is smooth but not monotonous. For the rest of initial series, the dependencies of volatility
and returns are similar to the depicted on the g. 4 ones.
Fig. 5 shows averaged coecients of time (t) compression-expansion (formula (23)) for three groups of incoming
series: currency (forex), stock, and commodity markets.
The formulae (11),(23),(25) show that (t) exists in proportion to the averaged square speed (according to the
chosen time span and series), i.e. average energy of the economical particle (as it is in our analogy), and can be
thus interpreted as the series temperature. Crises are distinguished with the intensication of economic processes
(the temperature is rising), while during the crisis-free period their deceleration can be observed (the temperature
is falling), what can be interpreted as the heterogenous ow of economic time. (t) dependences shown on the g. 5
illustrate all afore-mentioned. Note that local time acceleration-deceleration can be rather signicant.
Transition to heterogenous economic time allows to make the observed economic series more homogenous, which
can simplify both analysis and prediction [42].
In table we give the values of a non-dimentional economic mass of the m
i
series, calculated using (22) for all 9
incoming series, as well as average masses of each group (formula (29)).
Table. Economic series masses
Incoming series Economic mass Average economic
mass of the group
Commodity market
gold 2, 816 10
4
4, 983 10
3
silver 4, 843 10
3
oil 2, 777 10
3
Currency market
jpy 2, 148 10
4
2, 499 10
4
gbp 3, 523 10
4
chf 2, 180 10
4
Stock market
S&P 500 6, 251 10
3
4, 748 10
3
FTSE 100 6, 487 10
3
BVSP 1, 507 10
3
As you can see from the table, the stock market is distinguished with the lowest mass value, while the currency one
shows the maximum number. Oil price series has the lowest mass on the commodity market, gold the highest one.
As for the currency market, British pound (gbp) have the highest value and Japanese yen rates (jpy) demonstrates
the minimum mass of the group, although the dispersion is lower than that of the commodity market. The smallest
spread is peculiar to the currency market. Dynamic and developing Brazilian market (BVSP) has the lowest mass,
while the maximum value, just like in the previous case, corresponds to Great Britain (FTSE 100). It is explained by
the well-known fact: Britain has been always known for its relatively closed economy as comraped with the rest of
the European and non-European countries.
The last group of experimental data corresponds to the dependence of Plancks economic constant (calculated for
dierent series) on time t = kt
min
(time between the neighbouring registered observations), which is characterised
by H
i
(k, n) coecient (see formula (24)).
On g. 6-8 integral dependencies H
i
(k) are depicted. The following are averaged on the whole period of time 1993-
2010 and calculated for commodity, stock, and currency markets. As you can see there are no obvious regularities,
which can be explained by various crises and recessions of the world and national economies that took place during
the investigated period.
To decide whether it is possible for local regularities of Plancks economic constant dependence on t to appear, we
have chosen relatively small averaging fragments, N = 500, which approximately equals two years. Corresponding
results for some of these fragments on commodity, currency, and stock markets are given on g. 9-11. Evidently,
all three gures show clear tendencies of H
i
(k, n) recession and rise for each type of the market (unlike integral
dependences H
i
(k)).
5. CONCLUSIONS
We have conducted methodological and philosophical analysis of physical notions and their formal and informal
connections with real economic measurements. Basic ideas of the general relativity theory and relativistic qantum
mechanics concerning spacetime properties and physical dimensions peculiarities were used as well. We have suggested
9
procedures of detecting normalized economic coordinates, economic mass and heterogenous economic time. The afore-
mentioned procedures are based on socio-economic time series analysis and economical interpretation of Heisenbergs
uncertainty principle. The notion of economic Plancks constant has also been introduced. The theory has been tested
on real economic time series, including stock indices, currency rates, and commodity prices. Acquired results indicate
the availability of further investigations.
[1] V. N. Soloviev and V. M. Saptsin. Quantum econophysics - a physical substantiation of system concepts in socio-economic
processes modeling. In Analysis, modeling, management, development of economic systems Proceedings of the II Inter-
national Symposium of the School-AMUR-2008 Sebastopol Sept. 12-18. 2008 / ed. O. Koroleva, A. Segal., Simferopol,
2008.
[2] V. M. Saptsin and V. N. Soloviev. Relativistic Quantum Econophysics. New paradigms of Complex systems modeling:
Monograph. http://kafek.at.ua/sol sap monogr.rar Brama-Ukraine, Cherkassy, 2009.
[3] V. Saptsin and V. Soloviev. Relativistic quantum econophysics - new paradigms in complex systems modelling.
arXiv:0907.1142v1 [physics.soc-ph].
[4] B. E. Baaquie. Quantum Finance. Cambridge University Press, Cambridge, 2004.
[5] V. Maslov. Quantum Economics. Science, Moscow, 2006.
[6] E. Guevara. Quantum Econophysics. arXiv:physics/0609245 , September 2006.
[7] W.-X. Zhou and D. Sornette. Self-organizing Ising model of nancial markets. Eur. Phys. J. B, 55(2):175181, 2007.
[8] C. P. Goncalves. Quantum Financial Economics - Risk and Returns. arXiv:1107.2562 , July 2011.
[9] R. N. Mantegna and H. E. Stanley. An Introduction to Econophysics: Correlations and Complexity in Finance. Cambridge
Univ. Press, Cambridge UK, 2000.
[10] M. Yu Romanovsky and Yu. M. Romanovsky. An introduction to econophysics. Statistical and dynamical models. IKI,
Moscow, 2007.
[11] V. N. Soloviev. Mathematical economics. A textbook for self-study (in Ukrainian). CHNU, Cherkassy, 2008.
http://kafek.at.ua/Posibnyk Soloviev.rar
[12] V. D. Derbentsev, A. A. Serdyuk, and V. N. Solovievand O. D. Sharapov. Synergetical and econophysical methods for
the modeling of dynamic and structural characteristics of economic systems. Monograph (in Ukrainian). Brama-Ukraine,
Cherkassy, 2010. http://kafek.at.ua/Monogr.pdf
[13] V. M. Saptsin. Genetically complex Markov chains a new neural network approach for forecasting of socio-economic and
other poorly formalized processes. In Monitoring, modelyuvannya i management emerdzhentnoyi ekonomiki: Conference
Proceedings., pages 196 198, Cherkassy, 13-15 may 2008. Brama.
[14] V. M. Saptsin. Experience of using genetically complex Markov chains for the neural network technology forecasting.
Visnyk Krivorizkogo ekonomichnogo institutu KNEU, 2 (18):56 66, 2009.
[15] Vladimir Soloviev, Vladimir Saptsin, and Dmitry Chabanenko. Prediction of nancial time series with the technol-
ogy of high-order Markov chains. In Working Group on Physics of Socio-economic Systems (AGSOE), Drezden, 2009.
http://www.dpg-verhandlungen.de/2009/dresden/agsoe.pdf
[16] D. M. Chabanenko. Financial time series prediction algorithm, based on complex Markov chains. Visnyk Cherkasskogo Uni-
versitetu (in Ukrainian), 1(173):90 102, 2009. http://www.nbuv.gov.ua/portal/Soc Gum/Vchu/N173/N173p090-102.pdf
[17] D. M. Chabanenko. Detection of short- and long-term memory and time series prediction methods of complex Markov
chains. In Visnyk Natsionalnogo tehnichnogo universitetu Kharkivsky politehnichny institut. Zbirnik Naukovyh pratz.
Tematichny vypusk: Informatika i modelyuvannya (in Ukrainian), number 31, pages 184 190. NTU KHPI, Kharkov,
2010. http://www.pim.net.ua/ARCH F/V pim 10.pdf
[18] V. Soloviev, V. Saptsin, and D. Chabanenko. Financial time series prediction with the technology of complex Markov
chains. Computer Modelling and New Technologies, 14(3):63 67, 2010. http://www.tsi.lv/RSR/vol14 3/14 3-7.pdf
[19] V. N. Soloviev, V. M. Saptsin, and D. N. Chabanenko. Prognozuvannya sotsialno-ekonomichnih protsesiv: suchasni pidhody
ta perspektivy. Monograph (in Ukrainian), chapter Financial and economic time series forecasting using Markov chains and
Fourier-continuation, pages 141155. 2011.
[20] V. N. Soloviev, V. M. Saptsin, and D. N. Chabanenko. Modern methods of analysis and prediction of oscillatory processes
in complex systems. The scientic heritage of Simon Kuznets and prospects of the global and national economies in the
XXI Century, 2011. CD-ROM.
[21] V. N. Soloviev, V. M. Saptsin, and D. N. Chabanenko. Fourier-based forecasting of economic dynamics. In Informatsiyni
tehnologiyi modelyuvannya v ekonomitsi: Conference proceedings (in Ukrainian), page 204, Cherkassy, 19-21 may 2009.
Brama-Ukraine.
[22] V. M. Saptsin and D. N. Chabanenko. Fourier-based forecasting of low-frequency components of economical dynamics time
series. In Problemy ekonomichnoyi kibernetiki: Tezy dopovidey XIV Vseukrayinskoyi Naukovo-praktichnoyi konferentsiyi.(in
Ukrainian), page 132, Kharkiv, Oct 8-9, 2009 2009. KhNU imeni VN Karazina.
[23] D. M. Chabanenko. Discrete Fourier-based forecasting of time series. Sistemni tehnologii. Regionalny mizhvuzivsky zbirnik
naukovyh pratz (in Ukrainian), 1(66):114 121, 2010. http://www.nbuv.gov.ua/portal/natural/syte/2010 1/15.pdf
[24] V. M. Saptsin. Two-phase nonlinear discrete second-order system of logistical type: the local stability condition and clas-
sication. In Kompyuterne modeluvannya ta informatsiyni tehnologii v nautsi, ekonomitsi i osviti. Conference procceedings
10
(in Ukrainian), pages 194 195, Krivy Rig, 2005. KEI KNEU.
[25] Y. A. Olkhovaya and V. M. Saptsin. The relationship of concurence, cooperation and dominance in network models with
discrete-time. In Electromechanical and energetic systems, simulation and optimization methods: Conference proceedings
(in Russian), pages 302305., Kremenchuk, Apr 8-9 2010. KDU.
[26] V. M. Saptsin and A. N. Chabanenko. Analysis of the stability limit of the nonlinear two-agent model with dominance
in the space of parameters. In Kompyuterne modelyuvannya i informatsiyni tehnologiyi v nautsi, ekonomitsi i osviti:
Conference proceedings (in Ukrainian), pages 143146, Cherkassy, 2011. Brama.
[27] V. M. Saptsin, V. N. Soloviev, and A. Batyr. Nonlinear concurence in two-agent systems. In 2 volumes, editor, Tezy
dopovidey VII Vseukryinskoyi Naukova-praktichnoyi koferentsiyi Informatsiyni tehnologiyi in osviti, nautsi i tehnitsi
(ITONT 2010)., volume 1, page 101, Cherkassy, 2010. Cherkassy State Technological University.
[28] V. N. Soloviev and V. M. Saptsin. Dynamical network mathematics - a new look at the problems of complex systems
mathematical modeling. In Analysis, modeling, management, development of economic systems: collection of scientic
papers of the IV International Symposium school - AMUR-2010, (in Russian), pages 340342, Simferopol, Sep 13-19 2010.
TNU named after V. I. Vernadsky.
[29] V. N. Soloviev, V. M. Saptsin, and A. N Chabanenko. Dynamical network mathematics - a new look at the problems
of complex systems mathematical modeling. In Visnyk Cherkaskogo universitetu. Seriya pedagogichni nauki, Vypusk 191.
Part 1., pages 121127. CHNU, 2010.
[30] V. M. Saptsin. Econophysical analysis of the network nature of the low-frequency noise in socio-economic systems. In
Monitoring, modelyuvannya i management emerdzhentnoyi ekonomiki, conference proceedings (in Ukrainian), pages
184189, Cherkassy, 2010. Brama-Ukraine.
[31] V. M. Saptsin. Econophysical analysis of network nature of low-frequency noise in socio-economic systems. Visnyk
Cherkaskogo universitetu, 187:108115., 2010.
[32] V. N Soloviev, V. M. Saptsin, and Shokotko L. N. The Heisenbergs principle and nancial markets. In Systemny analiz.
Informatika. Upravlinnya (SAIU-2011): Conference proceedings, pages 197198, Zaporizhzhya, 2011. Classical Private
University.
[33] V.N. Soloviev, V. M. Saptsin, and L. N. Shokotko. Heisenberg uncertainty principle and nancial markets. In The 9-th
International conference Information technologies and management 2011, pages 135136, Riga, Latvia, April 14-15 2011.
Information Systems Management Institute.
[34] L.D. Landau and E.M. Lifshitis. The classical theory of elds. Course of theoretical physics. Butterworth Heinemann,
1975.
[35] John G. Cramer. The transactional interpretation of quantum mechanics. Rev. Mod. Phys., 58:647 687, Jul 1986.
[36] M. Kaku. Introduction to superstrings and M-theory. Graduate texts in contemporary physics. Springer, 1999.
[37] V. Balasubramanian. What we dont know about time. Foundations of Physics, page 139, sep 2011.
[38] Y. S. Vladimirov. A relational theory of space-time interactions. Part 1. MGU, Moscow, 1996.
[39] Y. S. Vladimirov. A relational theory of space-time interactions. Part 2. MGU, Moscow, 1996.
[40] M. Kaku. Parallel worlds: a journey through creation, higher dimensions, and the future of the cosmos. Anchor Books,
2006.
[41] M. Kaku. Physics of the impossible: a scientic exploration into the world of phasers, force elds, teleportation, and time
travel. Doubleday, 2008.
[42] V. M. Saptsin and D. M. Chabanenko. The complexity problem and non-linear time in socio-economic processes forecasting.
In Problemy ekonomichnoyi kibernetiki: Tezi dopovidey XIV Vseukrayinskoyi Naukova-praktichnoyi konferentsiyi, pages
130131, Kharkiv, 2009. KNU imeni V. N. Karazina.
11
Figure 1: USA (S&P500), Great Britain (FTSE 100), and Brazil (BVSP) daily stock indices from April 27, 1993 to
March 31, 2010.
12
Figure 2: Daily currency dollar cross-rates (chf, jpy, gbp)
13
Figure 3: Commodity market. Daily gold, silver, and oil prices.
14
Figure 4: Absolute values of immediate speeds (abs returns) and their dispersions (volatility).
15
Figure 5: Coecients of time compression-expansion, market temperature. The explanation is in the text.
16
Figure 6: Integral coecient H
i
(k) dependences for commodity market.
17
Figure 7: Integral coecient H
i
(k) dependences for stock market.
18
Figure 8: Integral coecient H
i
(k) dependences for currency market.
19
Figure 9: Local coecient H
i
(k, n) dependeces for commodity market (averaging time span from 27.04.1993 to
12.06.1995, 500 daily values).
20
Figure 10: Local coecient H
i
(k, n) for currency market (averaging time span from 12.06.1995 to 15.07.1997, 500
daily values).
21
Figure 11: Local coecient H
i
(k, n) dependences for stock market (averaging time span from 12.06.1995 to
15.07.1997, 500 daily values).