Hosted
by the Department of Economics, University of Peloponnese
Please
read the Disclaimer
EurOil Index, Trading-by-Rotation
and TbR logos © 2007/2008 Dimitrios D. Thomakos
|
EurOil Index Trends, Forecasts and Simulated Trading for Oil and the Euro |
|
|
|
Created and maintained by Dimitrios D. Thomakos © 2007/2008 Last updated: |
|
|
|
Forecast
and Trading updates available after Wed each week Other
content updates as indicated |
|
Figure
on right updated on:
Visit
two more simulated Trading-by-Rotation webpages: TbR with Greek banks and TbR with S&P500 and
VIX
Click
below to link on this webpage:
Latest News $, Summary of Trend Forecasts and Trading Recommendations $,
Introduction, Data, Methodology, Figures, Trends and
Forecasts $, Simulated Trading $, Future Developments,
Commentary and
additional material $
$ Indicates updated content
Summary
of Trend Forecasts and Trading Recommendations (scroll down for
more, including graphs with the strategies information ratios)
The historical
summary below until
Historical
Evaluation until |
Recommendation for
|
||
Cumulative
Returns |
Rotation #1 – no cash (24 weeks) |
32.15% |
Oil |
Rotation #2 – optimal
with cash (24 weeks) |
28.94%# |
Cash |
|
Rotation #3 – no cash,
re-optimized model (14
weeks) |
8.39% |
Euro |
|
GMV portfolio* (24 weeks) |
6.37% |
|
|
Oil (24 weeks) |
23.36% |
||
Euro (24 weeks) |
5.49% |
||
|
|||
Sign
Success Ratio |
Rotation #1 – no cash (24 weeks) |
66.66% |
|
Weeks
in Cash |
Rotation #2 – optimal
with cash (24 weeks) |
1 |
*GMV
== Global Minimum Variance
#Switch
from #2 cash strategy to #1 cash strategy
Statistics
on the 6-month performance of the TbR strategy and comparison with Oil and Euro
Summary Statistics |
Oil |
Euro |
No-cash rotation |
GMV portfolio |
Average weekly return (%) |
0.97 |
0.23 |
1.34 |
0.27 |
Weekly std. deviation (%) |
3.03 |
1.55 |
2.37 |
1.53 |
Weekly Sharpe ratio |
0.32 |
0.15 |
0.56 |
0.17 |
Minimum realized return (%) |
-7.70 |
-2.38 |
-5.06 |
-2.43 |
Maximum realized return (%) |
4.42 |
2.63 |
4.42 |
2.58 |
Trend Forecast Summary for the week of
[1.] Oil. The level of the series is still
above its upper prediction bound. The short-term outlook is for the level to
move around its optimal trend but the medium-term (4 to 6 weeks ahead) outlook
is now a reversal within the prediction bounds.
[2.] Euro. As predicted the Euro has
returned within its optimal prediction bounds. The short-term outlook is to
move around its optimal trend value.
Welcome to the EurOil
Index website! Here you can find a real-time application of some advanced
methods for modeling, forecasting and trading two widely known and highly
liquid assets, crude oil and the euro currency. This is an outgrowth of my
personal interest for an application of what I use in some of my academic
research in a real life situation. I find it very stimulating to see how
different methods work and to anticipate the actual outcomes of forecasts and
trading strategies. The current focus of this effort is neither in the usual
forecasting race approach or in technical analysis-based rules, nor in
well-known portfolio trading strategies but rather in a couple of econometric,
and possibly less well known, approaches for (1) modeling the stochastic trend
components of time series that are characterized by unit roots and (2) trading
assets by rotation. This will be an ongoing effort and updates will, hopefully,
be frequent enough to maintain your interest (and mine!) I welcome your
comments and suggestions, especially if you find that you know of related work
that I am not using or citing. Please contact me at thomakos@uop.gr.
I use publicly
available data from accessible and reliable sources. Please note that these are
not necessarily the same data, in terms of frequency, timing etc., that a real
market participant would use! Therefore all results presented below should be
interpreted accordingly. The two main series that are currently in use are:
The data start
on 01/04/1999 (euro availability) and are updated weekly for the analysis
presented here.
Methodology
(including
references)
Trends and Forecasts
Most price series in financial markets can be
characterized as having stochastic (and possibly deterministic) trends,
implying that can be adequately modeled by a process that has a unit root. It
is well known that such processes cannot be predicted, the last observed value
of the series being the best forecast using all available past information. Peter Phillips at Yale is a pioneer in the analysis of
such processes and has made significant contributions to their understanding.
In a strand of papers ([1], [2] and [3] - all available at this link) he shows how we can understand unit
root processes via the Karhunen-Loeve decomposition
and what the implications are for what is known as spurious regression and
modeling. This line of his work has gone largely unnoticed although it provided
us with a solid theoretical background for understanding and predicting trend
components of unit root processes. I use the relevant prediction formulas from
[3] to construct forecasts for the trend
components of the two prices series and then to construct their empirical
predictive distribution. For more information on this part of the methodology
please consult the following references:
[1.] Spurious Regression Unmasked,
CFDP 1135.
[2.] Descriptive Econometrics for
Nonstationary Time Series with Empirical Illustrations, reprinted in Journal of Applied Econometrics (2001),
16: 389-413, CFDP
1219.
[3.] Challenges of Trending Time
Series Econometrics, reprinted in Mathematics
and Computers in Simulation (2005), 68: 401-416, CFDP
1472.
Simulated Trading
Forecasting the price or the return of a financial
time series can be fun! That is, if we move beyond the realm of purely
statistical modeling and the difficulties associated with making statistically accurate
predictions and get into trading action. The two assets I use in this project
are highly liquid and tradable across the world and also highly related. It
therefore makes sense to try to examine whether there is a profitable trading
strategy for them. I use a modification of the rotation ideas found in [4]
to construct and update a strategy that rotates, on a weekly basis, an initial
investment between oil and the euro. The strategy gives long only signals. The
results that are presented below do not take into account trading costs
and should be interpreted accordingly.
[4.]
Market timing and cap rotation, D. Thomakos, T. Wang and J. Wu, Mathematical and Computer Modeling
(2007), 46: 278-291. (http://www.sciencedirect.com)
You can also visit the Trading-by-Rotation for Greek banks and Trading-by-Rotation for S&P500 and VIX webpages
that I am also maintaining for an additional real-time example of simulated
trading.
Figures,
Trends and Forecasts $ scroll
down for updated content
The first set of
figures below present the full sample sojourn densities for the oil and the
euro prices. The sojourn densities are the analogue to a standard kernel
density for the case of unit root processes and they measure the, so called,
sojourn time: this is the time spent by the process (as a fraction of an
appropriately scaled measure of total time) in various regions for a range of
spatial values. We see below that both densities are bimodal and therefore the
two underlying processes have spent a considerable amount of time around the
corresponding modal values.
Figures updated:
The figures
below give one week ahead rolling predictions of three trend components for the
two series, evaluated for the last 200 weeks of the current sample. The shape
of each trend depends on the fineness of the decomposition we are looking into,
so the lower is the value of the parameter K the smoother will be the trend
component we are predicting. It is important to stress that these trends are not
moving averages and do not correspond to moving averages!!! They are exact,
forward looking predictions of underlying components that form the original
series, assuming that the series has a unit root (this is the basic,
empirically viable assumption we make when analyzing these price series.) As
such we can use them to infer something about the potential future trend
behavior of the series (this is important: unless K is very large the
corresponding predictions will not be about the actual level of the series!)
Note that the analysis of such trend predictions offers an advantage in that it
is plausible to assume that the underlying structure of the series will not
change in the future, i.e. we will still have a unit root present. This is not
the case in many other modeling instances that are plagued by problems of
structural breaks inside and outside the estimation sample.
Figures updated:
Selecting an optimal value for K $ updated content
A question of
immediate practical interest is whether one can chose K, the number of trend
components, in some optimal fashion. The notion of optimality here is very
subjective and cannot be tied to the idea of making accurate point forecasts.
In a standard forecasting setting making optimal point forecasts would be
accomplished by minimizing the forecast error variance. Here, however, this
would lead to a choice of K that is relatively large: the theory in [3] indicates that as K increases the series is reproduced with
greater accuracy. To maintain the idea of trend, rather than point, prediction
K should be kept relatively low and an appropriate value should be selected
based on a different objective function than the forecast error variance. A
suitable alternative could be the following. Assume that you would like to
predict the correct trend on average. You would then expect to make about an
equal amount of positive and negative errors: sometimes the price would be
higher than the trend and sometimes it would be lower. Minimizing an
appropriate function we can not only obtain a meaningful optimal value of K but
we can also compute appropriate prediction bands using historical simulation.
These bands can be very useful in devising trading strategies. We are adding a
new section of trend forecasts and forecast error bands based on this
methodology for the week starting in 12/10/2007. The figures and discussion
below illustrate the application of this method using current data.
Figures updated:
The figures
above were computed using an evaluation period of the last 200 weeks in the
sample. The red lines indicate one standard deviation prediction bands. The
selected optimal values for K were K=12 for Oil and K=8 for Euro. The coverage
of these prediction bands (% of times that the actual observations were within
the bands) was 71% for Oil and 73% for Euro.
Trend Forecasts for the week of
Series |
K |
Last Prediction |
Current
Prediction |
Error
Band Current Prediction |
Oil |
Optimal |
108.11 |
109.32 |
+/- 5.60 |
Oil |
10 |
105.60 |
106.90 |
+/- 5.49 |
Oil |
15 |
113.20 |
114.60 |
+/- 4.46 |
Euro |
Optimal |
1.55 |
1.56 |
+/- 0.03 |
Euro |
10 |
1.56 |
1.56 |
+/- 0.03 |
Euro |
15 |
1.65 |
1.65 |
+/- 0.02 |
Notes:
[1.] Trend forecasts based on past 288 weeks.
[2.] Optimal K may change from week to week , see discussion above.
[3.] Error bands based on the standard error of trend
deviations for the past 200 weeks.
[4.] Monday data used in calculations.
Trend Forecast Summary for the week of
[1.] Oil. The level of the series is still
above its upper prediction bound. The short-term outlook is for the level to
move around its optimal trend but the medium-term (4 to 6 weeks ahead) outlook
is now a reversal within the prediction bounds.
[2.] Euro. As predicted the Euro has
returned within its optimal prediction bounds. The short-term outlook is to
move around its optimal trend value.
Click here for past forecasts.
Simulated
Trading $ scroll down for updated content
Rotation Strategy with no
Cash
The two figures below show the end-of-period wealth
of one US dollar invested in the rotation strategy and in each asset separately
(left) and the corresponding information ratios (right). Results are shown for
evaluation periods ranging from the latest 25 weeks and up to the past 400
weeks and are color coded according to the legends. Information ratio is weekly
(multiply by 7.2 to have an approximate annual ratio).
Figures updated:
Trading
recommendation for Monday, strategy
without cash (based on calculations using data of
week before) |
Asset |
Realized
Return % Oil
(for last trade) |
Realized
Return % Euro
(for last trade) |
Success
in Direction |
Cumulative
Value of $1 |
Cumulative
Value of $1 for
GMV portfolio |
week
of |
Oil |
3.53 |
1.38 |
+ |
1.0353 |
1.0149 |
week
of 12/03/2007 |
Euro |
-7.70 |
-1.39 |
+ |
1.0214 |
0.9978 |
week
of 12/10/2007 |
Oil |
-0.59 |
0.39 |
- |
1.0155 |
1.0012 |
week
of 12/17/2007 |
Oil |
3.37 |
-2.37 |
+ |
1.0492 |
0.9804 |
week
of 12/24/2007 |
Euro |
1.40 |
0.28 |
- |
1.0520 |
0.9837 |
week
of 12/31/2007 |
Oil |
2.26 |
1.34 |
+ |
1.0746 |
0.9975 |
week
of 01/07/2008 |
Oil |
0.54 |
0.55 |
- |
1.0800 |
1.0030 |
week
of 01/14/2008 |
Euro |
-1.72 |
1.30 |
+ |
1.0930 |
1.0145 |
week
of 01/21/2008 |
Oil |
-5.06 |
-1.82 |
- |
1.0424 |
0.9947 |
week
of 01/28/2008 |
Oil |
3.24 |
1.22 |
+ |
1.0748 |
1.0079 |
week
of 02/04/2008 |
Oil |
0.20 |
0.29 |
- |
1.0768 |
1.0107 |
week
of 02/11/2008 |
Oil |
3.07 |
-2.22 |
+ |
1.1075 |
0.9911 |
week
of 02/18/2008 |
Oil |
3.25 |
1.61 |
+ |
1.1400 |
1.0080 |
week
of 02/25/2008 |
Oil |
0.41 |
0.54 |
- |
1.1441 |
1.0133 |
week
of 03/03/2008 |
Oil |
4.42 |
2.48 |
+ |
1.1883 |
1.0391 |
week
of 03/10/2008 |
Oil |
3.38 |
1.14 |
+ |
1.2221 |
1.0516 |
week
of 03/17/2008 |
Euro |
-0.88 |
2.54 |
+ |
1.2475 |
1.0753 |
week
of 03/24/2008 |
Oil |
-3.39 |
-2.38 |
- |
1.2136 |
1.0510 |
week
of 03/31/2008 |
Euro |
1.38 |
2.63 |
+ |
1.2399 |
1.0767 |
week
of 04/07/2008 |
Euro |
3.50 |
-0.58 |
- |
1.2340 |
1.0729 |
week
of 04/14/2008 |
Oil |
2.18 |
0.72 |
+ |
1.2558 |
1.0808 |
week
of 04/21/2008 |
Oil |
2.76 |
0.34 |
+ |
1.2834 |
1.0854 |
week
of |
Oil |
3.83 |
-1.48 |
+ |
1.3217 |
1.0733 |
week
of |
Oil |
-0.02 |
-1.01 |
+ |
1.3215 |
1.0637 |
week
of |
Oil |
|
|
|
|
|
Notes:
[1.] The data quotes have different reporting times within
the day and the trading strategy assumes that the trade is placed for the
reporting time of each asset.
[2.] The realized return for each asset is the weekly return
(from Monday to Monday) from the week before to the week of the trade – return
measured as difference in log prices.
[3.] See the contributions section
for a discussion on the GMV (Global Minimum Variance) portfolio weights; the
weight assignment is 95% for Euro and 5% for Oil.
Rotation Strategies including Cash $ updated content
We introduce two
additional strategies for our simulated trading exercise. The strategies rotate
between Oil and the Euro but also they include a cash component so that there
are periods that no trade is recommended. The wealth accumulated up to that
point can be invested in a risk-free rate, which is taken to be the
(appropriately scaled for a one-week horizon) one-month certificate of deposit
series CD1M
from the Federal Reserve Bank
of St. Louis database (FRED.).
Strategy with Cash #1
The figures
below give the end-of-period wealth and the corresponding information ratios,
for the first strategy #1 that involves cash. The proportion of total time that
the strategy stayed in cash, depending on the length of the evaluation period,
was between 84% (25 weeks) and 41% (400 weeks). Note that the strategy still
outperforms the two assets and has the highest information ratio for the
longest evaluation period.
Figures updated:
Trading
recommendation for Monday, strategy
#1 with cash , signal from past 88 weeks (based on calculations using data of
week before) |
Asset |
Realized
Return % Oil
(for last trade) |
Realized
Return % Euro
(for last trade) |
Success
in Direction |
Cumulative
Value of $1 |
Cumulative
Value of $1, cash
earns risk free rate |
week
of |
Oil |
3.53 |
1.38 |
+ |
1.0353 |
1.0353 |
week
of 12/03/2007 |
Euro |
-7.70 |
-1.39 |
+ |
1.0214 |
1.0214 |
week
of 12/10/2007 |
Oil |
-0.59 |
0.39 |
- |
1.0155 |
1.0155 |
week
of 12/17/2007 |
Oil |
3.37 |
-2.37 |
+ |
1.0492 |
1.0492 |
week
of 12/24/2007 |
Cash |
1.40 |
0.28 |
- |
1.0492 |
1.0501 |
week
of 12/31/2007 |
Oil |
2.26 |
1.34 |
+ |
1.0748 |
1.0757 |
week
of 01/07/2008 |
Oil |
0.54 |
0.55 |
- |
1.0802 |
1.0811 |
week
of 01/14/2008 |
Cash |
-1.72 |
1.30 |
- |
1.0810 |
1.0819 |
week
of 01/21/2008 |
Oil |
-5.06 |
-1.82 |
- |
1.0304 |
1.0313 |
week
of 01/28/2008 |
Oil |
3.24 |
1.22 |
+ |
1.0628 |
1.0637 |
week
of 02/04/2008 |
Oil |
0.20 |
0.29 |
- |
1.0648 |
1.0657 |
week
of 02/11/2008 |
Oil |
3.07 |
-2.22 |
+ |
1.0955 |
1.0964 |
week
of 02/18/2008 |
Oil |
3.25 |
1.61 |
+ |
1.1280 |
1.1299 |
week
of 02/25/2008 |
Cash |
0.41 |
0.54 |
- |
1.1280 |
1.1305 |
week
of 03/03/2008 |
Oil |
4.42 |
2.48 |
+ |
1.1722 |
1.1747 |
week
of 03/10/2008 |
Oil |
3.38 |
1.14 |
+ |
1.2060 |
1.2085 |
week
of 03/17/2008 |
Cash |
-0.88 |
2.54 |
- |
1.2065 |
1.2090 |
week
of 03/24/2008 |
Oil |
-3.39 |
-2.38 |
- |
1.1726 |
1.1751 |
week
of 03/31/2008 |
Euro |
1.38 |
2.63 |
+ |
1.1989 |
1.2014 |
week
of 04/07/2008 |
Cash |
3.50 |
-0.58 |
- |
1.1994 |
1.2019 |
week
of 04/14/2008 |
Oil |
2.18 |
0.72 |
+ |
1.2212 |
1.2237 |
week
of |
Oil |
2.76 |
0.34 |
+ |
1.2488 |
1.2513 |
week
of |
Oil |
3.83 |
-1.48 |
+ |
1.2872 |
1.2896 |
week
of |
Oil |
-0.02 |
-1.01 |
- |
1.2870 |
1.2894 |
week
of |
Cash |
|
|
|
|
|
Trading
recommendation for Monday, strategy
#1 with cash , signal from past 288 weeks (based on calculations using data of
week before) |
Asset |
Realized
Return % Oil
(for last trade) |
Realized
Return % Euro
(for last trade) |
Success
in Direction |
Cumulative
Value of $1 |
Cumulative
Value of $1, cash
earns risk free rate |
week
of |
Cash |
3.53 |
1.38 |
- |
1.0000 |
1.0009 |
week
of 12/03/2007 |
Euro |
-7.70 |
-1.39 |
- |
0.9861 |
0.9870 |
week
of 12/10/2007 |
Cash |
-0.59 |
0.39 |
- |
0.9861 |
0.9880 |
week
of 12/17/2007 |
Oil |
3.37 |
-2.37 |
+ |
1.0198 |
1.0217 |
week
of 12/24/2007 |
Cash |
1.40 |
0.28 |
- |
1.0207 |
1.0226 |
week
of 12/31/2007 |
Cash |
2.26 |
1.34 |
- |
1.0216 |
1.0235 |
week
of 01/07/2008 |
Cash |
0.54 |
0.55 |
- |
1.0224 |
1.0243 |
week
of 01/14/2008 |
Euro |
-1.72 |
1.30 |
+ |
1.0355 |
1.0373 |
week
of 01/21/2008 |
Cash |
-5.06 |
-1.82 |
+ |
1.0362 |
1.0380 |
week
of 01/28/2008 |
Cash |
3.24 |
1.22 |
- |
1.0368 |
1.0386 |
week
of 02/04/2008 |
Oil |
0.20 |
0.29 |
- |
1.0388 |
1.0406 |
week
of 02/11/2008 |
Oil |
3.07 |
-2.22 |
+ |
1.0695 |
1.0713 |
week
of 02/18/2008 |
Oil |
3.25 |
1.61 |
+ |
1.1012 |
1.1038 |
week
of 02/25/2008 |
Cash |
0.41 |
0.54 |
- |
1.1012 |
1.1044 |
week
of 03/03/2008 |
Cash |
4.42 |
2.48 |
- |
1.1018 |
1.1050 |
week
of 03/10/2008 |
Cash |
3.38 |
1.14 |
- |
1.1024 |
1.1056 |
week
of 03/17/2008 |
Euro |
-0.88 |
2.54 |
+ |
1.1278 |
1.1310 |
week
of 03/24/2008 |
Cash |
-3.39 |
-2.38 |
+ |
1.1283 |
1.1315 |
week
of 03/31/2008 |
Euro |
1.38 |
2.63 |
+ |
1.1547 |
1.1578 |
week
of 04/07/2008 |
Cash |
3.50 |
-0.58 |
- |
1.1552 |
1.1583 |
week
of 04/14/2008 |
Cash |
2.18 |
0.72 |
- |
1.1557 |
1.1588 |
week
of |
Oil |
2.76 |
0.34 |
+ |
1.1833 |
1.1865 |
week
of |
Oil |
3.83 |
-1.48 |
+ |
1.2216 |
1.2248 |
week
of |
Cash |
-0.02 |
-1.01 |
+ |
1.2221 |
1.2253 |
week
of |
Cash |
|
|
|
|
|
Notes:
[1.] The data quotes have different reporting times within
the day and the trading strategy assumes that the trade is placed for the reporting
time of each asset.
[2.] The realized return for each asset is the weekly
return (from Monday to Monday) from the week before to the week of the trade –
return measured as difference in log prices.
Strategy with Cash #2
The figures
below give the end-of-period wealth and the corresponding information ratios,
for the second strategy #2 that involves cash. The proportion of total time
that the strategy stayed in cash, depending on the length of the evaluation
period, was between 4% (25 weeks) and 16% (400 weeks), significantly lower than
that of strategy #1. This is reflected in the higher performance of the
strategy, compared to the first strategy with cash discussed above.
Figures updated:
Trading
recommendation for Monday, strategy
#2 with cash , signal from past 88 weeks (based on calculations using data of
week before) |
Asset |
Realized
Return % Oil
(for last trade) |
Realized
Return % Euro
(for last trade) |
Success
in Direction |
Cumulative
Value of $1 |
Cumulative
Value of $1, cash
earns risk free rate |
week
of |
Euro |
3.53 |
1.38 |
- |
1.0138 |
1.0138 |
week
of 12/03/2007 |
Euro |
-7.70 |
-1.39 |
- |
1.0000 |
1.0000 |
week
of 12/10/2007 |
Oil |
-0.59 |
0.39 |
- |
0.9941 |
0.9941 |
week
of 12/17/2007 |
Oil |
3.37 |
-2.37 |
+ |
1.0278 |
1.0278 |
week
of 12/24/2007 |
Euro |
1.40 |
0.28 |
- |
1.0306 |
1.0306 |
week
of 12/31/2007 |
Oil |
2.26 |
1.34 |
+ |
1.0532 |
1.0532 |
week
of 01/07/2008 |
Oil |
0.54 |
0.55 |
- |
1.0586 |
1.0586 |
week
of 01/14/2008 |
Cash |
-1.72 |
1.30 |
- |
1.0594 |
1.0594 |
week
of 01/21/2008 |
Oil |
-5.06 |
-1.82 |
- |
1.0088 |
1.0088 |
week
of 01/28/2008 |
Oil |
3.24 |
1.22 |
+ |
1.0412 |
1.0412 |
week
of 02/04/2008 |
Oil |
0.20 |
0.29 |
- |
1.0432 |
1.0432 |
week
of 02/11/2008 |
Oil |
3.07 |
-2.22 |
+ |
1.0739 |
1.0739 |
week
of 02/18/2008 |
Oil |
3.25 |
1.61 |
+ |
1.1064 |
1.1064 |
week
of 02/25/2008 |
Oil |
0.41 |
0.54 |
- |
1.1105 |
1.1105 |
week
of 03/03/2008 |
Oil |
4.42 |
2.48 |
+ |
1.1547 |
1.1547 |
week
of 03/10/2008 |
Oil |
3.38 |
1.14 |
+ |
1.1885 |
1.1885 |
week
of 03/17/2008 |
Cash |
-0.88 |
2.54 |
- |
1.1890 |
1.1890 |
week
of 03/24/2008 |
Oil |
-3.39 |
-2.38 |
- |
1.1551 |
1.1551 |
week
of 03/31/2008 |
Cash |
1.38 |
2.63 |
- |
1.1556 |
1.1556 |
week
of 04/07/2008 |
Euro |
3.50 |
-0.58 |
- |
1.1497 |
1.1497 |
week
of 04/14/2008 |
Oil |
2.18 |
0.72 |
+ |
1.1715 |
1.1715 |
week
of |
Oil |
2.76 |
0.34 |
+ |
1.1991 |
1.1991 |
week
of |
Euro |
3.83 |
-1.48 |
- |
1.1843 |
1.1843 |
week
of |
Oil |
-0.02 |
-1.01 |
- |
1.1841 |
1.1841 |
week
of |
Oil |
|
|
|
|
|
Trading
recommendation for Monday, strategy
#2 with cash , signal from past 288 weeks (based on calculations using data of
week before) |
Asset |
Realized
Return % Oil
(for last trade) |
Realized
Return % Euro
(for last trade) |
Success
in Direction |
Cumulative
Value of $1 |
Cumulative
Value of $1, cash
earns risk free rate |
week
of |
Oil |
3.53 |
1.38 |
+ |
1.0353 |
1.0353 |
week
of 12/03/2007 |
Euro |
-7.70 |
-1.39 |
+ |
1.0214 |
1.0214 |
week
of 12/10/2007 |
Oil |
-0.59 |
0.39 |
- |
1.0155 |
1.0155 |
week
of 12/17/2007 |
Oil |
3.37 |
-2.37 |
+ |
1.0492 |
1.0492 |
week
of 12/24/2007 |
Euro |
1.40 |
0.28 |
- |
1.0520 |
1.0520 |
week
of 12/31/2007 |
Oil |
2.26 |
1.34 |
+ |
1.0746 |
1.0746 |
week
of 01/07/2008 |
Oil |
0.54 |
0.55 |
- |
1.0800 |
1.0800 |
week
of 01/14/2008 |
Cash |
-1.72 |
1.30 |
- |
1.0808 |
1.0808 |
week
of 01/21/2008 |
Oil |
-5.06 |
-1.82 |
- |
1.0302 |
1.0302 |
week
of 01/28/2008 |
Oil |
3.24 |
1.22 |
+ |
1.0626 |
1.0626 |
week
of 02/04/2008 |
Oil |
0.20 |
0.29 |
- |
1.0646 |
1.0646 |
week
of 02/11/2008 |
Oil |
3.07 |
-2.22 |
+ |
1.0933 |
1.0933 |
week
of 02/18/2008 |
Oil |
3.25 |
1.61 |
+ |
1.1258 |
1.1258 |
week
of 02/25/2008 |
Oil |
0.41 |
0.54 |
- |
1.1299 |
1.1299 |
week
of 03/03/2008 |
Oil |
4.42 |
2.48 |
+ |
1.1741 |
1.1741 |
week
of 03/10/2008 |
Oil |
3.38 |
1.14 |
+ |
1.2079 |
1.2079 |
week
of 03/17/2008 |
Euro |
-0.88 |
2.54 |
+ |
1.2333 |
1.2333 |
week
of 03/24/2008 |
Oil |
-3.39 |
-2.38 |
- |
1.1994 |
1.1994 |
week
of 03/31/2008 |
Euro |
1.38 |
2.63 |
+ |
1.2257 |
1.2257 |
week
of 04/07/2008 |
Euro |
3.50 |
-0.58 |
- |
1.2198 |
1.2198 |
week
of 04/14/2008 |
Oil |
2.18 |
0.72 |
+ |
1.2416 |
1.2416 |
week
of |
Oil |
2.76 |
0.34 |
+ |
1.2692 |
1.2692 |
week
of |
Euro |
3.83 |
-1.48 |
- |
1.2544 |
1.2544 |
week
of |
Oil |
-0.02 |
-1.01 |
- |
1.2542 |
1.2542 |
week
of |
Oil |
|
|
|
|
|
Notes:
[1.] The data quotes have different reporting times within
the day and the trading strategy assumes that the trade is placed for the
reporting time of each asset.
[2.] The realized return for each asset is the weekly
return (from Monday to Monday) from the week before to the week of the trade –
return measured as difference in log prices.
[1.] Trading based
on smoothed trends.
[2.] Combination
of long-short signals & alternative trading strategies.
[3.] Additional
measures of trading performance.
[4.]
Introduction of transaction costs.
Acknowledgements:
I would like to
thank, without implicating, my colleagues Timotheos Angelidis, Kostas Nikolopoulos, Greg Siourounis and Viki Skintzi for
stimulating discussions on the theoretical and practical aspects of financial
markets and real-time trading.
Disclaimer:
The
contents of this webpage are provided for information purposes only. Prices
shown in this webpage are indicative and the author is not offering to buy or
sell or soliciting offers to buy or sell any financial instrument. The views in
this publication are those of the author alone and are subject to change at any
time, and he has no obligation to update the opinions or the information in
this webpage. The author of this webpage does not accept any liability
whatsoever for any direct or consequential loss arising from any use of the
information provided. The information in this webpage is not intended to
predict actual results, which may differ substantially from those presented.