Predicting ConocoPhillips and ExxonMobil stock price

Lately, I have published an article on the prediction of stock prices for COP and XOM - “Predicting ConocoPhillips and Exxon Mobil stock price”, Journal of Applied Research in Finance, v.2, 2009. (A draft version is available here.)

Abstract
Exxon Mobil and ConocoPhillips stock price has been predicted using the difference between core and headline CPI in the United States. Linear trends in the CPI difference allow accurate prediction of the prices at a five to ten-year horizon.
Key words: stock price, Exxon Mobil, ConocoPhillips, prediction, CPI
JEL classification: G1, E3

We continue modelling various stock prices and recently revisited XOM and COP. The general model can be rewritten in the following form:

sp(t)= A1S1(t+t1) + A2S2(t+t2) + A3t +A4

where sp(t) is the stock price at time t, A1 through A4 are empirical coefficients, S1 and S2 are components of headline CPI, t1 and t2 are time difference (negative or positive) between the change in the stock price and relevant changes in the CPI components. Term A3t is introduced to compensate linear trends observed in the difference between the components.

For ConocoPhillips and Exxon Mobil, more accurate empirical models are as follows:

Figure 1. Comparison of measured (open circles) and predicted (solid diamonds) stock prices of ConocoPhillips. The prediction is obtained from the following empirical relationship:

COP(t)= 3.41CF(t+2) - 7.17EC(t-5) + 7.22*(t-2000) + 145.76,

where CF(t+2) is the headline CPI less food lagged by one months behind COP(t), and EC(t-5) is the index for education and communication leading the COP(t) by five months. The explicit trend term 7.22*(t-2000) compensates the trend term in the difference between CF and EC. Standard deviation between the curves is $4.0 for the period between July 2003 and August 2009.

Figure 2. Comparison of measured (open circles) and predicted (solid diamonds) stock prices of ExxonMobil. The prediction is obtained from the following empirical relationship:

removed by author

Standard deviation between the curves is $3.27 for the period between July 2003 and August 2009.

These two models provide a good approximation of the price evolution and even predict the future price for XOM.

Crude oil at $100 in December: August update

In March 2009, we presented a prediction of crude petroleum price for 2009. Briefly, our analysis has shown that oil will overcome $100 per barrel before the end of 2009. As promised, we evaluate this prediction every month, when new readings of producer price index (PPI) with all its components become available. In this article, we report the results for August 2009.

The period between January 2008 and likely the end of 2010 is characterized by an elevated volatility in oil price, but the evolution of the price is not random. Moreover, even after the start of crisis in 2008, the price has been following a predetermined trajectory, as it has been demonstrating since 1980.

In August, the PPI reached the level of 175.1 from 172.7 in July, and the producer price index of crude petroleum (domestic production) increased to 190 after 158.9 in July. Our assumption on the average monthly increment for the crude petroleum index was +20, and for the PPI +1. So, the increase in August was larger than predicted. However, this increase just compensated the fall in the difference observed in July relative to June.

Figure 1 presents the new readings. Overall, the evolution of the difference between the PPI and the index for crude petroleum follows the predetermined path (compare the updated curve to that drawn in June) – from its peak in February 2009 to the bottom of a through, which will likely be reached by December 2009. This is a natural path for a pendulum, as discussed in the previous articles [1], [2]. Hence, there is not sign that the oil price significantly deviates from the predicted trajectory. We expect the price to hit the new (red) trend line in September 2009. This is the level of $73-77 per barrel.

Figure 2 depicts the evolution of crude oil price. Values for the period between September and December 2009 are shown by solid red circles. According to the prediction, the price should break the $100 level before the end of 2009. The next update is expected in October or November 2009.

Figure 1. Evolution of the difference between the PPI and the index for crude petroleum (domestic production). Solid circles – the readings between March and August 2009, which were anticipated in February 2009. Open circles – the predicted difference between September and December 2009. Upper panel – figure from June. Lower panel – August update.

Figure 2. The evolution of crude oil price. Red circles – oil price predicted for the period between July and December 2009. According to the prediction, the price should break the $100 level before the end of 2009.

How big are bank problems in the USA?

Bloomberg cites Joseph Stiglitz, who has evaluated bank problems as bigger than pre-Lehman. This is a questionable statement because it is based on a qualitative evaluation. Our stock price model allows to evaluate the difference between 2008 and 2009 in a quantitative format - adjusted share price.

First, we show some examples of poor banks which have approached the level of bankraptcy:

1. Citigroup (C) reached the bottom in the beginning of 2009. Was bailed out. On an upward trend right now - no problems in the near future, as the model says. I consider the possibility to invest right now.
(N.B. here and below we do not provide actual empirical relationships behind the predicted curves because they are confidential.)

2. Colonial Bank (CNB) reached the bottom in March 2009. Is a part of BBT now. Why isn't it bailed out?

3. CIT Group Inc. Must fail in April, but was bailed out.


4. E*TRADE financial corporation (ETFC) definitely has problems in 2009. Our model is highly unceratain because of the uncertainty in underlying data. ETFC will likely have a problem by the end of 2009, but then will recover quickly.


5. SLM Corp. (SLM) seems to have hard time in the near future. This is the only problem bank we have found among those in S&P 500. Will watch its evolution




Several positive examples of large importance:

6. Fifth Third Bancorp (FITB) approached the zero line in the beginning of 2009. has been successfully recovering since. Demonstrates sustainable growth, but still its share price is at the level of October 2008. I wonder that Stiglitz confused the down- and upgoing branches. It is worth noting that S&P 500 banks are chiefly on rise with large positive trends.



7. Bank of America dropped very low but recovered fast. The model predicts relatively quick growth.


8. Goldman Sachs has only slight problems compared to those associated with CIT and CNB. Currently is closing its 2007 level. No problems are foreseen.


9. JPMorgan Chase is very similar to GS. Stays good for investment.


Conclusion.
Stiglitz and Co. look to be mistaken about problems in 2009. It might be another decline in the second part of 2010, however.

Does economics need a scientific revolution?

David Altig at macroblog has joined the fierce discussion of the problems in theoretical economics related to the complete failure in prediction of the current crisis. The economic profession is very reluctant to recognize (not admit) the inconsistency of the mainstream ideas and tools.

Briefly, our approach has been formulated in the article "Does economics need a scientific revolution?" as a reaction to the article by J.-P. Bouchaud "Economics need a scientific revolution" in Nature.

Our research, as a whole, is a quantitative alternative to the mainstream economics. This blog highlights relevant results.

S&P 500 in August 2009

I have found a short paper citing our article on the prediction of S&P 500 returns as an example of " how statistics can be manipulated to show relationships where none exist". I have to admit that these people did not give any effort to read our article before writing own. As a consequence, they misinterpreted the essence and details of our study.

For a quantitative prediction, the best way to proceed is to continue presenting current results, as we have been doing from March 2009. In August 2009, S&P 500 jumped to 1020 from 987 in July 2009, i.e. by 33 points. As in our previous posts, we depict an updated graph for observed and predicted S&P 500 - Figure 1. All in all, our predictions from March 2009 are accurate to the extent of the number of 9-year-olds is uncertain.

Figure 1. Observed (red) and predicted (black) S&P 500. The leg from March 2009 (bottom point of the curve) was also predicted as a straigt line continuing the black portion. Considering the 6-month horizon of the prediction and the time when it was done (near the bottom of S&P 500), so far the prediction is accurate, but likely we have to calibrate it a bit down because the change in the number of 9-year-olds is overestimated for 2008 through 2010. In any case, we expect another nine months of growth. And the level of 1200 to 1300 by the end of 2009.

Figure 2. Observed and predicted 12-month returns of S&P 500. The model linking S&P 500 to the number of 9-year-olds and real GDP is described in our article and previous posts.

We will continue comparing our prediction and actual S&P 500 and presenting update figures.

As to the paper mentioned in the beginning, these people will never be convinced that they do not understand how the economy and financial market work. It is against their interests.

Unemployment in Japan is approaching 6.0%

In June and August 2009, we published two articles at Seeking Alpha devoted to the evolution of unemployment in Japan [A1] and [A2], which predicted the rate of 6.0% in August 2009. On August 28, a new reading for July 2009 was reported by the Statistics Bureau of Japan. The (seasonally adjusted) rate for July was measured at 5.7%, i.e. 0.3% higher than in June. It is also 0.3% left to match our prediction of 6.0% in August, which is based on the model linking the rate of unemployment, UE(t), to the rate of change of labor force, dLF(t)/LF(t):
UE(t)= -1.5*dLF(t)/LF(t) +0.045 (1)
According to the results of labor survey presented on August 28, the level of labor force in for July was 66280000, i.e. 200,000 lower than in June. The number of unemployed people increased by 11,000 from June.
Figure 1 updates the observed and predicted curves for 2009. Because the accuracy of short-term estimates provided by labor surveys is not high, the monthly estimates of unemployment and labor force are prone to large measurement errors. So, the discrepancy between the observed and predicted curves likely manifests the problems with measurements, because in the long run the curves fit much better, as presented in Figure 2.
From 1, one can conclude that the rate of unemployment in August still may rise to 6.0%. On the other hand, this likely to be the peak of the unemployment growth and it will be declining to the long-term level between 4.5% and 5% [1,2] in 2009 and 2010.

[1] Kitov, I., (2006). The Japanese economy, MPRA Paper 2737, University Library of Munich, Germany, http://ideas.repec.org/p/pra/mprapa/2737.html
[2] Kitov, I., (2007). Exact prediction of inflation and unemployment in Japan, MPRA Paper 5464, University Library of Munich, Germany, http://ideas.repec.org/p/pra/mprapa/5464.html



Figure 1. Observed and predicted rate of unemployment in Japan in 2009.


Figure 2. Observed and predicted rate of unemployment in Japan between 1998 and 2008.

Figure 3. Prediction of the evolution of unemployment rate in Japan between 1990 and 2050.

Looking into the (gold) crystal ball: the evolution of the iron and steel price index

Introduction

The presence of long-term trends, both linear and nonlinear, in the differences between various expenditure categories of producer and consumer prices has been reported in several papers [1-7]. These trends allow predicting prices of many commodities at time horizons of several years. Such predictions are of fundamental importance for all stock market participants because of their anchoring and stabilizing effect. A striking feature associated with the trends is high volatility during the transition periods between adjacent trends. At first glance, the transient processes in the differences between price indices are driven by some stochastic forces, and thus, can not be predicted. It was found, however, that during the transition to new trends, some differences mimic trajectories of a pendulum-like oscillation [3,4]. Therefore, these trajectories are predictable ones and have been actually used in forecasting crude oil and gold ore prices [4,6]. The next natural step is to compare individual trajectories in order to evaluate their similarity in shape and timing of major peaks and troughs. If a price difference lags behind others, it would be possible to predict its evolution at a time horizon of corresponding lag.

1. The model
The model derived in [1,4] implies that the difference between the headline PPI, PPI, and the index for an individual commodity, iPPI, can be described by a linear time function over time intervals of several years:

iPPI(t) – PPI(t) = A + Bt (1)

where A and B are empirical constants, and t is the elapsed time. Therefore, the “distance” between the PPI and the studied index is a linear function of time, with a positive or negative slope B.
As an example, Figure 1 presents the difference between the iron and steel index and the PPI (all PPI indices are retrieved from the BLS web-site: http://www.bls.gov/data on 20.08.2009). There is a transition period between 2000 and 2001. This turning point is characterized by a slightly elevated volatility. Since 2008, the difference has been also passing a turning point, but this time with a very high volatility caused by the uncertainty in the characteristics of the following trend. A naive assumption about the future trend is that it will repeat the predecessor but with an opposite sign. The green line in Figure 1 represents the hypothetical new linear trend.
From Figure 1, the fundamental feature of the difference is that all deviations from the trends were only short-term ones. This implies that the current or future deviations from the new trend, which has been under development since 2008, must be promptly compensated. This feature allows a short-term (months) price prediction along the trend. The principal task of this study is different - to predict the evolution of the difference during the current transition period.

Figure 1. Illustration of linear trends in the difference between the headline PPI and the producer price index of iron and steel in the U.S. There are two quasi-linear segments with a turning point near 2000. Currently, the difference passes second transition period. Two linear trends with relevant linear regression lines are also shown. A tentative future trend is shown by a green line.

2. Price prediction

Relationship (1) describes the evolution of a given difference. The absolute value of such a difference may vary between commodities due to variation in start years. This might complicate the cross-commodity comparisons. So, it is instructive to use the differences normalized to the PPI: (iPPI(t)-PPI(t))/PPI(t). In a sense, the normalized differences represent the evolution of the rate of deviation from the PPI over years. For historical and logical reasons, we have chosen the following commodities: iron and steel, crude petroleum (domestic production), and gold ores. Figure 2 depicts corresponding time histories of the normalized deviation from the PPI. The initial inspection reveals the following apparent features: the (normalized deviation from the PPI of the) index for gold ores is characterized by a higher volatility; the index for iron and steel lags by several months behind the other two indices; the deviation of the crude oil index is comparable to the PPI itself.

Figure 2. The deviation of the iron and steel price index, the index of crude oil, and the index for gold ores from the PPI, normalized to the PPI.

In order to refine the pattern we additionally normalized the curves in Figure 2 to their peak values between 2005 and 2009:

(iPPI(t)-PPI(t))/[PPI(t)*max{iPPI-PPI)}]

The normalization allows a direct comparison of corresponding shapes. In Figure 3, we display the normalized index for iron and steel shifted by six and eight months back in the past for the synchronization of its peak with that observed in the normalized index for crude petroleum (upper panel) and gold ores (lower panel), respectively. The (normalized) index for crude petroleum demonstrates larger discrepancies from the index of iron and steel in the overall shape and timing of the current trough. On the other hand, despite its higher volatility, the index of gold ores is very similar to that of iron and steel. Simple smoothing with a weighted MA(3) makes the curves resemblance even better. As an invaluable benefit of the resemblance, one can use the eight-month lag to predict the future of the iron and steel price index.


Figure 3. The deviation of the iron and steel price index from the PPI, normalized to the PPI and the peak value after 2005 as compared to the normalized deviations of the index for crude petroleum (upper panel) and gold ores (lower panel). The normalized index for iron and steel is shifted six and eight months back in the past, respectively.

Conclusion
Between 2006 and 2010, the deviation of the price index for iron and steel from the PPI in the USA repeats the trajectory of the deviation of the index of gold ores with an eight-month lag. Therefore, the prediction of price for iron and steel at this horizon is straightforward – it will be growing during the next six to eight months. It is likely that in 2010 the index of iron and steel will approach closely the level attained in August 2008. From this level, it will be declining in the long run following the new trend, as shown in Figure 1.

References
1. Kitov, I., Kitov, O., (2008). Long-Term Linear Trends In Consumer Price Indices, Journal of Applied Economic Sciences, Spiru Haret University, Faculty of Financial Management and Accounting Craiova, vol. 3(2(4)_Summ), pp. 101-112.
2. Kitov, I., (2009). Apples and oranges: relative growth rate of consumer price indices, MPRA Paper 13587, University Library of Munich, Germany.
3. Kitov, I., Kitov, O., (2009). A fair price for motor fuel in the United States, MPRA Paper 15039, University Library of Munich, Germany,
4. Kitov, I., Kitov, O., (2009). Sustainable trends in producer price indices, Journal of Applied Research in Finance, Spiru Haret University, Faculty of Financial Management and Accounting Craiova, vol. I(1(1)_ Summ), pp. 43-51
5. Kitov, I., Kitov, O., (2009). PPI of durable and nondurable goods: 1985-2016, MPRA Paper 15874, University Library of Munich, Germany
6. Kitov, I., (2009). Predicting gold ores price, MPRA Paper 15873, University Library of Munich, Germany
7. Kitov, I., (2009). Predicting the price index for jewelry and jewelry products: 2009-2016, MPRA Paper 15875, University Library of Munich, Germany