Long-term oil price dynamics is shaped by the worldwide demand for oil which is highly dependent on global macroeconomic conditions, major geopolitical events, and industrial trends. Over shorter time intervals, oil price is strongly affected by market speculation involving risky financial transactions in an attempt to profit from fluctuations in the market value of the traded financial instruments.  The role of the speculation is controversial. On one hand, speculation is instrumental in as a means to reach a balanced spot price of oil consistent with its fundamental value. On the other hand, speculation can lead to detrimental effects such as economic bubbles and excessive volatility.

By 12 December 2014, both Brent and WTI crude oil prices, and the majority of the other benchmarks, reached their lowest values since 2009. While the 2014-2015 global oversupply is often considered as the leading macroeconomic cause of the overall oil price decay, the volatile dynamics observed at shorter time scales is likely a footprint of massive speculative transactions. The combined result of the two effects, the long-term supply trend and the excessive speculation, is not well understood. Can the speculative transactions destabilize the global oil market, and if so, do we need more restrictive government policies limiting the speculation? Does this market meet the requirements of the informational efficiency assumed by many existing asset theories?  Is a mathematical forecast of the ongoing oil price change theoretically possible?  These are some of the questions studying in this project.

[button1 size=”medium” color=”grey” link=”/category/project/ongoing/”] Return to Ongoing Projects [/button1]

Uritskaya O.Y., Uritsky V.М.  Predictability of price movements in deregulated electricity markets // Energy Economics, Volume 49, May 2015, Pages 72–81

Journal link

Fig.1. Time series of hourly electricity prices in Alberta (left), Ontario (center) and Mid-C (right) markets. Fromtop to bottom: all hourly prices, on-peak prices, and off-peak prices. Alberta electricity prices demonstrate significantly higher fluctuations than those in Ontariomarket, plotted on the same vertical scale. Fluctuations of electricity prices in Mid-C have twice as low amplitude as that in Ontario, and about 5 times smaller than in Alberta.

Fig.2. Dependence of the detrended variation F and the local scale-dependent DFA slope α on the time scale n for all hourly, on- and off-peak electricity prices in Alberta (left), Ontario (center), and Mid-C (right) markets. The presented statistics reveal complex correlated structure of price movements with quasi-periodic components associated with daily and weekly cycles. In all presented data sets the scale-dependent DFA exponent is significantly below the level 1.5 defining the state of informational efficiency,which provides an opportunity of forecasting the prices over wide ranges of time scales.

Fig.3.  Diagrams of aggregated price movements for several time scales n ranging from 1 to 720 h. Most of the plots have a distinctly asymmetric shape reflecting a casual relationship between the pricemovements. For the Alberta and Ontario plots (first and second columns), the asymmetry of the cloud of points assumes anti-persistent dependence which can be used for price forecasting. The Mid-C diagrams (third column) take a different formdepending on the aggregation scale,with the persistent and anti-persistent tendencies found at n=1 and n = 12, correspondingly.

Serletis А., Uritskaya O.Y., & Uritsky V.М. Detrended Fluctuation Analysis of the US Stock Market // International Journal of Bifurcation and Chaos, Vol. 18 (2), 2008 – p. 599-603.

Journal link

Fig. 1. The comparative dynamics of the Dow Jones industrial average and a simulated geometric Brownian motion time series. Inset: DFA functions of both signals in the range 4–256 days.

Uritskaya O.Y., Serletis А. Quantifying Multiscale Inefficiency in Electricity Markets // Energy Economics, Vol. 30, Issue 6, November 2008, p. 3109-3117

Journal link

Fig 1. Power spectra of electricity and natural gas prices.

Fig 2. Average and Extreme Values of DFA Exponents.

Serletis А., Uritskaya O.Y. Detecting Signatures of Stochastic Self-Organization in US Money and Velocity Measures // Physica A, Vol.385 (1), 2007. – p. 281-291.

Journal link

Fig. 1. Time series and DFA plots for sum, Divisia, and CE money measures at M1 level of monetary aggregation.

From point of view of traditional econometric modeling the critical condition in a system can be obtained under outside factor’s effect. However the reasons of critical events can be found in distinguishing property of these systems: the most economic and particular macroeconomic systems are complex nonlinear dynamical systems with a very special behavior. They can homeostatically tune their regime up in response to any new information or pressure. Using fractal methods it has been demonstrating an evidence of normal running of system’s parameters, requited scale free fluctuations in certain limits and examples suppressing of fluctuations which leads to highly unstable system’s dynamical regime and crises as a result. Based on an extended statistical analysis of exchange rate fluctuations, prove of importance this tendency plays role in the stability of the international monetary system has been provided.

O. Y. Uritskaya. The Fractals Methods of the Determination of the Criteria of Homeostatic Stability of Macroeconomic Systems // Management in the social and economic systems. // St. Petersburg: SPbSTU Press, 2006, p. 326-354.

Full text (in Russian)

Fig.  1. An example of a of currency from group H. Fractal indexes a before the crisis take values above the upper limit of norm 1.75. After the crisis, they are normalized and remain in the interval [1.25, 1.75] as it observed in time series from group N.

Fig.  2. Examples of time series of logarithmic increments r_t of daily average exchange rate for the Japanese yen and Turkish lira (rates against the USD).

Fig.  3. Expanding the range of values of logarithmic increments during the active phase of the crisis (Russian ruble, currency from group H). Before the crisis most fluctuations are suppressed, that does not provide any chances for homeostatic regulation. After the crisis the range stays close to values of logarithmic increments from group N.

O. Y. Uritskaya. Stability of the Open Macroeconomic Systems // Management in the Social and Economic Systems. // St. Petersburg: SPbSTU Press, 2006, p. 304-326.

 Full text (in Russian)

Fig. 1. Stability diagram of the dynamics of national currencies: the intensity σr of and statistical temperature T of currency fluctuations as a function of the fractal index a2 . The dashed lines correspond to the levels of deviation parameters N group for the value of three standard deviations (±3s) from the mean values.  N – Economically developed countries: Great Britain, Greece, EU, Canada, New Zealand, Norway, USA, Swiss, Japan, Australia;  D – Developing countries with relatively stable monetary systems: Israel, Columbia, Chili, South Africa;  H and L – Unstable Developing countries, prior to crises: Bulgaria, Brazil, India, Kazakhstan, Mexico, Russia, Rumania, Turkey, Ecuador;  А – Unstable Developing Asian countries before the 1997 monetary crisis: Indonesia, Malaysia, Singapore, Thailand, Taiwan, Philippines, South Korea;  М – Marginally stable, Countries from groups Н, L and А after crises.

O. Y. Uritskaya. Forecasting of Magnitude and Duration of Currency Crises Based on Analysis of Distortions of Fractal Scaling in Exchange Rate Fluctuations // Noise and Fluctuations in Econophysics and Finance. Eds. D.Abbott, J.-Ph.Bouchaud, X.Gabaix. – Proc. SPIE Vol.5848, 2005. – p. 17-26 (Third SPIE International Symposium on Fluctuations and Noise (Austin, USA, 2005)).

Journal link

Fig.  1. Example of unstable currency dynamics (group Н, developing countries during periods preceding large-scale monetary crashes: Bulgaria, Brazil, India, Kazakhstan, Mexico, Russia, Rumania, Turkey, Ecuador) including a period of large-scale crisis. Before this event, DFA exponents had systematically increased values above the limit 1.75 of normal DFA exponent variations. After the crisis, both exponents returned to the interval [1.25, 1.75] signaling a normalization of dynamical stability of studied monetary system.

Fig.  2. Normalized crisis magnitude as a function of cumulative fractal indices characterizing the degree of fractal distortions in time series of exchange rate fluctuations with ₂ < 1.25 (left) or ₂ > 1.75 (right)

Fig.  3. Crisis duration (in days) as a function of the same parameters as in Fig. 2.

O. Y. Uritskaya. Making Decisions under Risk and Limited Information Conditions // St. Petersburg: SPbSTU Press, 2005, 108 p.

Fig. 1. Process transformation of data to the information. Preceding throw the physical, semantic and pragmatic filters significant part of important information can be lost with statistical and semantic noise or is not recognized as useful one.

Fig. 2. Antagonistic Game. If participants have the opposite goals, they can not negotiate and they have to hide information form each other, because in this case the gaincan be increased only if the opponent makes a mistake. The solution shown is most cautious one. In literature it calls pessimistic approach. But if your opponent is your enemy you have no other choice. The classic example of this game at the practice is gaining a new part of stable market. If you get it, somebody has to loose. But in real life this happens very rare, markets are use to grow and the opponents are not your real enemy, just partners with different goals and negotiation always is good solution.

Fig. 3. Bimatrix Game. For more reasonable choice than just guessing we usually ask for help from professionals. They are people, so have their own goals and interests in the situation. Result – the goals are not opposite, but just different and model become Bimatrix game. This name is used for 2 person games, if more is consider more – it just become too complicated to demonstrate in matrix form, but approach is good for any number of participants. Here everyone is interested in max gain, but have to choose depend on moving of other participant. In this particular example they probably never choose right strategy if only will not negotiate. In this case they can share they profit or some benefits. Here we still have some option to increase the gain with new inf from our opponent. What why the always pays off.

O. Y. Uritskaya. Organizing and planning at the company // St. Petersburg: SPbGTU Press, 2005, 236 p.

Annotation and contents (in Russian)

Fig.1 The Blake Mouton model.