ArticlePDF Available

Price Prediction Line. Investment Signals and Limit Conditions Applied for the German Financial Market.

Authors:

Abstract and Figures

In the first decades of the 21st century, in the electronic trading environment, algorithmic capital investments became the primary tool to make a profit by speculations in financial markets. A significant number of traders, private or institutional investors are participating in the capital markets every day using automated algorithms. The autonomous trading software is today a considerable part in the business intelligence system of any modern financial activity. The trading decisions and orders are made automatically by computers using different mathematical models. This paper will present one of these models called Price Prediction Line. A mathematical algorithm will be revealed to build a reliable trend line, which is the base for limit conditions and automated investment signals, the core for a computerized investment system. The paper will guide how to apply these tools to generate entry and exit investment signals, limit conditions to build a mathematical filter for the investment opportunities, and the methodology to integrate all of these in automated investment software. The paper will also present trading results obtained for the leading German financial market index with the presented methods to analyze and to compare different automated investment algorithms. It was found that a specific mathematical algorithm can be optimized and integrated into an automated trading system with good and sustained results for the leading German Market. Investment results will be compared in order to qualify the presented model. In conclusion, a 1:6.12 risk was obtained to reward ratio applying the trigonometric method to the DAX Deutscher Aktienindex on 24 months investment. These results are superior to those obtained with other similar models as this paper reveal. The general idea sustained by this paper is that the Price Prediction Line model presented is a reliable capital investment methodology that can be successfully applied to build an automated investment system with excellent results.
Content may be subject to copyright.
AbstractIn the first decades of the 21st century, in the
electronic trading environment, algorithmic capital investments
became the primary tool to make a profit by speculations in financial
markets. A significant number of traders, private or institutional
investors are participating in the capital markets every day using
automated algorithms. The autonomous trading software is today a
considerable part in the business intelligence system of any modern
financial activity. The trading decisions and orders are made
automatically by computers using different mathematical models.
This paper will present one of these models called Price Prediction
Line. A mathematical algorithm will be revealed to build a reliable
trend line, which is the base for limit conditions and automated
investment signals, the core for a computerized investment system.
The paper will guide how to apply these tools to generate entry and
exit investment signals, limit conditions to build a mathematical filter
for the investment opportunities, and the methodology to integrate all
of these in automated investment software. The paper will also
present trading results obtained for the leading German financial
market index with the presented methods to analyze and to compare
different automated investment algorithms. It was found that a
specific mathematical algorithm can be optimized and integrated into
an automated trading system with good and sustained results for the
leading German Market. Investment results will be compared in order
to qualify the presented model. In conclusion, a 1:6.12 risk was
obtained to reward ratio applying the trigonometric method to the
DAX Deutscher Aktienindex on 24 months investment. These results
are superior to those obtained with other similar models as this paper
reveal. The general idea sustained by this paper is that the Price
Prediction Line model presented is a reliable capital investment
methodology that can be successfully applied to build an automated
investment system with excellent results.
KeywordsAlgorithmic trading, automated investment system,
DAX Deutscher Aktienindex.
I. INTRODUCTION
APITAL investment is a sustained activity today. An
impressive number of market participants are buying and
selling every day on thousands of free markets. Traders,
investors, different types of companies, private and public
funds are participating in financial markets in order to make a
profit from the difference between the buy and the selling
price. Once electronic trading was widespread released,
especially after 2011 when ''investment firms on both the buy-
side and sell-side were increasing their spending on
technology for electronic trading'' [1], a significant part of the
capital investments is made completely automatically. The
Cristian Păuna is with the Economic Informatics Doctoral School,
Academy of Economic Studies, 11th Tache Ionescu str. 010352 Bucharest,
Romania (phone: +407.4003.0000; e-mail: cristian.pauna@ie.ase.ro).
This paper was cofinanced by Algorithm Invest (https://algoinvest.biz).
trading orders are built and sent almost instantly by servers
running automated trading systems.
Algorithmic trading is the key in financial trading
nowadays, as long as ''large general financial volatility may
increase uncertainty about the economic environment, with
long-lasting effects as investors demand a higher risk'' [2].
With increased price volatility on markets where ''there is an
exponential over-reaction to an action'' [3], the trading
decisions must be fast indeed. Special algorithms and
methodologies are used in order to send the trading orders
before a significant change in the price level. This paper will
present one of these mathematical models specially designed
for algorithmic trading.
Thousands of trading strategies compete to find the best
way to buy and sell in order to make a profit. Related with the
capital investment strategies for the stock markets, advanced
studies and working models that can be used in algorithmic
trading can be found in [4]-[7]. For the currency and
commodities markets, reliable models and trading strategies
can be found in [8]-[10]. Original investment strategies
optimized for automated capital investment software can be
found in [11]-[15].
The theory is simple: buy cheap and sell more expensive.
To put it in practice is not easy nowadays, because of the price
volatility, unexpected news released, and unexpected behavior
of the significant market participants.
Dictionaries of trading rules and psychological practice to
enhance the capital investment performance are since a long
time on libraries. One of the essential books is [16], including
several practical strategies that can be included in algorithmic
trading in order to improve the mathematical models.
Recommendation lines from notable authors can be found as:
''buy the market after it’s dropped; not after it’s risen.'' [4] But
when the market it's dropped enough? Also, when is the price
too much higher, and it is preparing to change its direction?
How to determine all of these in order to have a supportable
risk? Moreover, how to automate the trading decisions,
including all of these considerations? The method presented in
this paper will offer answers to all these questions.
The algorithm presented in this paper is exclusively a
mathematical model based on the price time series. It can be
easily implemented in any automated trading system in order
to automate the capital investment process. The model uses
the cyclicality behavior of the time price variations to build a
price prediction trend line called “Price Prediction Line.”
Trading signals will be built using the particularities of this
function, in order to buy or to sell on the capital market
depending on the predicted price tendency. Realistic price
Cristian Păuna
Price Prediction Line, Investment Signals and Limit
Conditions Applied for the German Financial Market
C
World Academy of Science, Engineering and Technology
International Journal of Computer and Information Engineering
Vol:13, No:9, 2019
476International Scholarly and Scientific Research & Innovation 13(9) 2019 ISNI:0000000091950263
Open Science Index, Computer and Information Engineering Vol:13, No:9, 2019 waset.org/Publication/10010724
level predictions can also be made together with limit
conditions in order to exit the investment positions before the
trend changes its direction. The presented method will be
compared with other models using trading results presented in
the last part of the paper.
II. PRICE CYCLICALITY
The model presented in this paper will use the "Price
Cyclicality function" noted as PCY [17]. Starting with the
assumption that the price has a wave behavior with variable
wavelengths, the model can be successfully applied for any
volatile market. This hypothesis is in line with the real
behavior of the investors in financial markets. After the price
starts to increase, more market participants will open more
buy positions that will grow the price continuously by
increasing demand. After a while, some of them will start to
close those trades in order to mark the profit. In that time
interval, the price evolution slows down. At a certain time
moment, more traders and investors will close their long
positions detecting the slowdown in the price evolution. More
sell orders than the buy orders will arrive on the market. A
reversed tendency will be present sooner or later in the price
behavior. It is a cyclical phenomenon which will be seen in
the price evolution. Increasing and decreasing demand
intervals are present in each market, similar to wave
propagation in time. Because the decisions do not have
constant repeatability, the price wave will have a variable
wavelength. The model presented in this paper can be adapted
automatically to this cyclical behavior. It will use the price
time series evolution and different functional parameters
which are optimized for each capital market.
The Price Cyclicality Function (PCYi) is mathematically
defined for each time price series interval (i) by the next
recurrent formula, starting from [17]:
11
iiii PCYPCYPCY
where 0
0
PCY (1)
where
ii
ii
iminmax
max
where iii maMa
(2)
and

kk
ni
ik
imaMamin
min
(3)

kk
ni
ik
imaMamax
max
(4)
i
Ma and i
ma are two moving averages [18] with different
periods (Pma and Pma, where PMa < Pma), and (n) is the
number of the time intervals considered in the time price
series, associated with the period of the presented model.
These three parameters (PMa, Pma, and n) are functional
parameters that will be optimized for each capital market and
for each timeframe used in order to obtain the maximal capital
efficiency for a minimal capital exposure, as we will see in a
further chapter. The PCY is represented in Fig. 1 for a daily
evolution of the Deutscher Aktienindex DAX30, the main
index of the Frankfurt Stock Exchange [19].
Fig. 1 PCY and Price Prediction Line
World Academy of Science, Engineering and Technology
International Journal of Computer and Information Engineering
Vol:13, No:9, 2019
477International Scholarly and Scientific Research & Innovation 13(9) 2019 ISNI:0000000091950263
Open Science Index, Computer and Information Engineering Vol:13, No:9, 2019 waset.org/Publication/10010724
The PCY has some important properties. The (PCYi) values
are limited into the interval [0;100]. The function presents
well defined ascending and descending intervals depending on
the price tendency and most important, the function has
asymptotic behavior on the overbought and oversold price
intervals. What is missing from the PCY model is a way to
predict the price level. The PCY function predicts with enough
accuracy, only the change in the price. The model developed
and presented in this paper proposes to predict the price
behavior for the next time intervals together with a predicted
price value level. To do this we will transform the PCY back
into the price space using the next transformation function:

iiiii PminPminPmaxPCYPPL 100/
(5)
The new function obtained (PPLi) is the Price Prediction
Line, a very useful function which will be used to obtain a
price prediction level, as we will see in this paper. Once the
terms (Pmini) and (Pmaxi) represent the maximum and
minimum price values on the current monotony interval of the
Price Cyclicality Function (PCY) is given by (1), after simple
mathematical considerations we can see that the PPLi values
are defined in the price space.
In very volatile price movements, the Price Prediction Line
given by (5) needs an attenuation process in order to have a
smoothed evolution. For this step we can apply mathematical
attenuation methods as the smoothing with Spline line
interpolations [20], polynomial or trigonometric interpolations
[21] or simple, exponential or weighted moving averages [18]
with a small period. The result using trigonometric
interpolation to attenuate the PPLi values for DAX30 index
are presented n Fig. 1.
III. PRICE PREDICTION LINE
The Price Prediction Line obtained after smoothing the
transformation function (5) has ascending and descending
intervals depending on the price behavior. The monotony
intervals are determined by an increased values of Pmaxi and
Pmini on an increasing interval of the PCYi.
Due to the way the Price Prediction Line (PPL) function is
built, (5), usually the minimum values of the price are attached
to the PPL levels on the ascending intervals, and the maximum
values of the price are usually attached with the PPL levels on
the descending intervals.
Fig. 2 Clear point for the price behavior change
Intentionally the PPL is plotted in two colors depending on
the price cyclicality ascending or descending intervals. As we
can see, the monotony intervals of the prediction line are not
the same as the monotony intervals of the cyclicality function.
The maximum and minimum points of the prediction line are
delayed. The PPL is plotted in blue for the ascending periods
of the PCY function, in order to keep the advantage of the
monotony of the cyclicality function. Similarly, the PPL is
plotted in red for the descending intervals of the PCY function.
With all of these, we will have a clear point for the price
behavior change, highlighted by the changing in color or the
prediction line.
As we can see in Fig. 2, the price change behavior point is
revealed before the maximum point of the PPL. In the moment
when the PPL line changes its color, meaning the PCY
function changed its monotony, the exit decisions from the
buy positions can be met since a price descending interval is
about to arrive. The color of the PPL is associated with the
trend behavior property. In the same way, in Fig. 3, we can see
that the ascending interval is announced by the prediction line
before its minimum point. In addition, once the color of the
line is changed, the prediction line values will be under the
price level, as a minimum value for the buy price level where
the trend behavior can be changed again.
Fig. 3 Buy price level decided by the PPL
Considering the above, one can think that the PPL is only a
colored line function depending on the PCY. After detailed
studies, it was found that the PPL makes a better prediction in
comparison with other models as moving averages. In Fig. 4,
it is presented a comparison between the PPL and usually used
moving averages. In Fig. 4 are drawn together the PPL line
with a period of 20-time intervals and the simple and
exponential price moving averages for the same period, on a
time price series expressed on a daily-bases interval for a 52
days evolution of the Frankfurt Stock Exchange Deutscher
Aktienindex DAX30.
As we can see, the PPL is a better approach to real-time
price series values. Sometimes the simple or exponential
moving averages monotony can predict the price behavior.
However, a comparison of the current price level with the
values of these moving averages in order to make a trade
decision can involve a considerable delay. The PPL values are
more close to the real price values, especially in the longer
World Academy of Science, Engineering and Technology
International Journal of Computer and Information Engineering
Vol:13, No:9, 2019
478International Scholarly and Scientific Research & Innovation 13(9) 2019 ISNI:0000000091950263
Open Science Index, Computer and Information Engineering Vol:13, No:9, 2019 waset.org/Publication/10010724
trend intervals. Besides, the specificity of the PPL can give us
more accurate information about the price change behavior
before the moving averages to start to change their monotony.
Fig. 4 Comparison between the PPL and moving averages
To trust the PPL, we also need a mathematical validation.
This confirmation is given by Pearson's correlation coefficient
[22] between two statistical series. Once our model tries to
predict the minimal value of the price for the ascending
intervals and the maximum values of the price for the
descending intervals, the correlation coefficient can be
adapted to this logic for our case with:


N
i
i
N
i
i
N
i
ii
PPLPPLqq
PPLPPLqq
r
1
2
1
2
1 (6)
where
1
if
iiii PCYPCYlowq (7)
and
1
if
iiii PCYPCYhighq (8)
in which the values (lowi) are the minimum price level and
(highi) represent the maximum price level for each (i) time
interval included in the time price series. Comparing if the
next price values are higher than the previous minimum price
interval for an ascending period of the PPL, and evaluating if
the next price values are lower than the previous high price
value for a descending period of the PPL, the correlation
coefficient has values between 0,583 and 0,999 for the
DAX30 market depending on the time frame used and the
period used to build the model. The values mentioned were
obtained for periods (N) between 10 and 50-time intervals.
The time frames used were 5 minutes (5M), 15 minutes
(15M), 30 minutes (30M), one hour (1H), 4 hours (4H) and
one day (1D). The historical time series computed were
between 01.01.2010 and 31.06.2019. Higher values for the
correlation coefficient were obtained for longer time frames.
With these values for the correlation coefficient, it can be said
that a direct and strong correlation between the price evolution
and the PPL exists. The PPL values will predict the minimal
local price values for the ascending intervals. Somewhere on
these values, the price will change the behavior to a
descending interval. Similarly, on descending price
movements, the PPL will predict the local maximal price
values.
In Fig. 5 is presented a code sample of the MetaQuotes
Language [23] for the PPL. For simplicity, the code includes
an attenuation process of the PPL values using an exponential
moving average. Also, for simplicity, the Signal vector takes 0
and 1 values depending on if the price behavior is up or down.
Sample code for the PCY can be found in [17].
IV. PRICE PREDICTION
The majority of the prediction models for algorithmic
trading are indicating a price level for the next time intervals.
After a buy trade was opened, the traders or the automated
trading software procedures are waiting the price to rich that
level in order to close the trade. Sometimes the predicted price
World Academy of Science, Engineering and Technology
International Journal of Computer and Information Engineering
Vol:13, No:9, 2019
479International Scholarly and Scientific Research & Innovation 13(9) 2019 ISNI:0000000091950263
Open Science Index, Computer and Information Engineering Vol:13, No:9, 2019 waset.org/Publication/10010724
level is never reached and the trade is turning into a loss trade.
The model presented in this paper will work differently in
order to assure the profit expectancy.
Fig. 5 MetaQuotes Language code for PPL
First of all, this prediction model will predict the next price
behavior and the moment when the price changes its tendency.
When the PPL turns from the red to the blue zone, an
ascending interval for the price is arrived. The minimal price
level where the price can change again its behavior is defined
by future values of the PPL. When the PPL turns from the blue
interval into the red values, a decreasing price interval will be
on the next scenario. The price will descend under the PPL
values, and that is the last good moment to close the long
trades. In this way, an automated trading procedure can be
organized in order to automate the capital investment process.
Similarly, for the markets where short trades can be
considered, in the presence of a significant descending trend,
sell signals can be built when the PPL turns from the blue
zone to the red one. The maximal entry price level will be
defined by the values of the PPL. The exit points will be met
when the PPL turns again from the red interval into the blue
one. Of course, the red and blue color properties were used
here for the visual presentation. In algorithmic trading, these
are properties defined as buy and sell internal properties of the
PPL class.
Fig. 6 Buy positions managed by the PPL
In Fig. 6 are displayed more consecutive buy trades
managed by the PPL. As we can see, using the exit conditions
provided only by the PPL intervals, not all trades are
profitable. Additional conditions must be met for this target, as
we will see in the next chapters. However, on a well-
capitalized market, the PPL catches early the most important
trades. The exit decisions are a separate subject later in this
paper.
V. INVESTMENT SIGNALS
By trading or investment signal, we understand a Boolean
variable with true value if a trade can be opened. Depending
on different conditions imposed with the PPL and PCY
functions, in direct correlation with the price evolution and
market behavior, these variables will permit to automate the
trading process. A buy signal is a signal that will decide if a
buy trade can be opened. A sell signal will be the signal for a
short trade. According with the PPL significance, the buy
signals can be defined as:
01 1iiiii PPLPPLPPLpBuySignal (9)
where (pi) are the price values and PPLi are the PPL values for
each (i) time interval. In (9) i
PPL are the values for the
property associated with the color of the PPL providing the
information about the PCY function monotony. These last
values can be set in many ways. Boolean variables are the
easiest way. We consider here 1 values for the ascending
periods of the PCY function and 0 values for descending
intervals of the PCY.
The trading signals made by (9) open trades when the local
price trend is changed from a descending interval to a new
increasing-price interval, in order to buy the equity near a
local minimum point and to wait for the price growth. This
scenario is for those algorithmic trading procedures that ask a
relatively small number of trades and try to keep the opened
positions as long as possible to catch the big movements of the
market. For high-frequency trading procedures, the signal
made by (9) makes a reduced number of trades, and because in
these cases the profit target is a small one. For high-frequency
trading procedures, an additional signal type is needed as long
an uptrend is present:
11 1iiiii PPLPPLPPLpBuySignal (10)
Equation (10) makes new trades even on the last section of
a trend which is not a good setup. Additional limit conditions
must be imposed to avoid these cases, as we will see in the
next chapter. In this section we have presented buy investment
signals, the most used trading signals in the stock markets. For
those markets where sell trades can be considered profitable,
especially for contract for differences (CFD) investment
procedures, the sell trading signals are built similarly with (9)
World Academy of Science, Engineering and Technology
International Journal of Computer and Information Engineering
Vol:13, No:9, 2019
480International Scholarly and Scientific Research & Innovation 13(9) 2019 ISNI:0000000091950263
Open Science Index, Computer and Information Engineering Vol:13, No:9, 2019 waset.org/Publication/10010724
and (10), considering the descending periods of the PPL.
VI. LIMIT CONDITIONS
Automated trading systems involve different types of
capital investment strategies and methodologies. One of the
most used algorithmic trading procedures opens a trade and
wait for the market as long as possible if the buy signal is
present. For these, the PPL gives proper entry and exit signals.
The entries will be made when the PPL turns in blue and will
exit the trades when the PPL turns into the red zone. The price
level for entry and exists are given by the PPL values. These
trading signals are made by (9). Not all trades will be
profitable as we will see in the last chapter, but the
methodology tries to catch some big trades in order to cover
all the small losses and to make a profit.
For high-frequency trading procedures, which make a
significantly large number of tiny trades, the trading entries
are made by the trading signals (10). Because these signals
open trades even on the last part of the trend, when the price
prepares to turn into the descending zone, additional limit
conditions must be imposed in order to ensure that the entry
point is good enough in order to make that small profit and to
avoid essential losses.
A. Entry Limit Conditions
In practice there are cases when the high price volatility can
produce exception cases. One of these cases is presented in
Fig. 7. This is the case of the DAX30 Index market at 7 June
2016 when the price made a significant correction in the up
direction in a very small time interval. The PPL turns form the
red zone into the blue at very high values of the price. The
trading signals made by (9) produced an entry point on the top
level price figured in Fig. 7. After that, the market reversed
strongly in down direction, fact signalized by the PPL which
produced a small gradient evolution and turned into the red
zone after a while, without the price to recover the losses on
those days. The strange evolution is due significant volatility
on the currency market following the Brexit news [24].
Fig. 7 Volatility exceptions with the PPL
To avoid cases like the one presented above, a limit
condition imposed for the distance of the price from the PPL
are used:



iii
iiii
PPLpPPL
PPLPPLpBuySignal
0
1
1
(11)
and for high-frequency trading:


iii
iiii
PPLpPPL
PPLPPLpBuySignal
1
1
1
(12)
World Academy of Science, Engineering and Technology
International Journal of Computer and Information Engineering
Vol:13, No:9, 2019
481International Scholarly and Scientific Research & Innovation 13(9) 2019 ISNI:0000000091950263
Open Science Index, Computer and Information Engineering Vol:13, No:9, 2019 waset.org/Publication/10010724
where δ is the distance from the PPLi accepted to be viable for
opening a trade into the risk level accepted. The functional
parameter δ can be optimized for each market, depending on
the exposure capital level, using the historical time price series
for a long interval of time. The limit conditions imposed by

ii PPLp will filter the trading signals made by (9) and
will allow only those trades with the price level under a limit.
From this reason, we call these conditions as to be limit
conditions. A more convenient graphic representation of these
types of limit conditions is presented in Fig. 8.
Looking Fig. 8, we can see that, on a long trend, the price
has the tendency to go up and to make new maximum points
together with the PPL. Due to the construction of the PPL, the
minimal values of the price will be in the zone of the PPLi
values, and the price tries to exceed these values and to make
new local picks. On a normal trend zone, after a new
minimum in the narrow of the PPLi values, a new maximum
will be made in one of the next time intervals if the price does
not change its behavior. Using this idea the trading signals
made by (12) work.
Fig. 8 Limit conditions depending on the PPL and PCY functions
When the price decreases under the δ distance from the
PPLi, the high-frequency trading procedure will open a new
trade. This position will be closed on one of the next time
intervals when the price makes a new local point, once the
profit target is small. The high-frequency trading procedure
will wait for the price to turn back into the δ distance from the
PPLi and will open a new trade, and so on, until the behavior
of the price will be changed. With a good optimization for the
δ, the strategy works for a large number of markets.
The second type of limit conditions imposed with the PP
line is regarding the gradient of this function. As we have
resented in Fig. 7, evolution with a small grading of the PPL
indicates a weakness of the price trend. A limit condition in
the gradient of the PPL function is imposed to avoid opening
trades in cases like these:



11 1
1
iii
iiii
PPLPPLPPL
PPLPPLpBuySignal (13)
where ξ is the minimum gradient value for the PPL function to
be considered strong enough in order to open a trade at a
specified risk level. The ξ parameter can be considered in
direct correlation with the power of the price trend. The filter
in the power of the trend is imposed by the term
1ii PPLPPL . This can be used with good results in
order to filter any other trading signal. Trading results for
high-frequency trading procedures made with these formulas
will be presented in the last chapter.
For high-frequency trading procedures, limit conditions can
be imposed using the PCY function as presented in [17]. To
exit from the trades and to stop making new trades in high-
frequency trading, the PCY function is limited in values as it is
displayed in Fig. 8. However, these conditions cut a
significant number of trades, especially for those cases when a
long-time trend is present when the asymptotic section of the
PCY function is long.
More limit conditions can be described with additional
algorithms based on the direct of inverse Fished transform of
the price, presented in [11] and [12]. In these cases also the
limits are too restrictive for long trends because they are
imposed only in the asymptotic stage of those indicators. The
PPL function permits a new kind of limit conditions which can
World Academy of Science, Engineering and Technology
International Journal of Computer and Information Engineering
Vol:13, No:9, 2019
482International Scholarly and Scientific Research & Innovation 13(9) 2019 ISNI:0000000091950263
Open Science Index, Computer and Information Engineering Vol:13, No:9, 2019 waset.org/Publication/10010724
be combined with all the others mentioned.
B. Exit Limit Conditions
Based on this PPL behavior and the price movement near
the prediction function levels, the exit conditions can be
imposed as:

0
1
1
i
iiii
PPL
PPLpPPLBuyExit
(14)
where φ is a functional parameter that can be optimized for
each traded capital market using a repetitive procedures with a
large historical time price series interval. The parameter φ is
the maximal distance of the price from the PPL function
where the price is considered high enough in order to wait the
next local minimum point. The limit conditions (14) can be
used with good results to automate the exit trading decisions
for buy trades made with any other trading strategy both for
algorithmic trading and high-frequency trading procedures.
For those market cases in which sell trades can be considered,
the limit conditions in order to exit the short trades can be
expressed with:

1
0
1
i
iiii
PPL
PPLpPPLSellExit
(15)
The limit conditions imposed by (14) and (15) on daily or
four-hour time frames can also be combined with the limit
conditions presented in [17], [11], and [12] imposed in a
smaller time frame as one hour or even lower, in order to filter
all overbought and oversold trades.
VII. TRADING RESULTS
In this section we will present some trading results obtained
with the trading signals presented in above. The exit
conditions presented in previous chapter were also included in
the algorithmic trading procedures. The results were obtained
using TheDaxTrader [25], an automated trading system that
uses the PPL in order to generate buy trades for DAX30 index
[19].
The results presented in Table I were obtained for the time
period between 1 June 2015 and 31 May 2018. An additional
condition was imposed for the entry trades regarding the
hourly intervals between 8:00 and 16:00 coordinated universal
time (UTC), Monday to Friday, in order to ensure the liquidity
on the market. The DAX30 index was traded as a CFD with a
spread of 1 point. For simplicity, the trades were made into an
account with no leverage. The capital risk was managed using
the “Global Slot Loss Method” [26]. The trading signals in
Table I were assembled into four-hour time frame interval.
TABLE I
TRADING RESULTS FOR THE PPL METHODOLOGY
Signal Target Positions Profit Drawdown RRR
(9) 10 points 203 1,230 1,495 1:0.82
(11)+(14) 10 points 179 1,939 845 1:2.29
(11)+(14) 20 points 122 2,683 846 1:3.17
(11)+(14) 50 points 60 3,318 1,032 1:3.21
(13)+(14) 10 points 101 1,199 796 1:1.50
(13)+(14) 20 points 101 2,423 796 1:3.04
Fig. 9 Capital evolution with and without PPL limit conditions
The results presented in Table I are obtained using (11) for
δ=10 and (14) for φ=100. In Fig. 9 is represented the capital
variation for the trading case with (9) without the limit entry
conditions and the same variation for the case for (11) using
the limit conditions. As we can see, the limit conditions have
successfully filtered the exception case presented in Fig. 7
which produced a heavy loss in the trading case with (9)
without limit conditions. Even the number of trades is lower
using the limit conditions, the profit is higher and the risk to
reward ratio (RRR) is considerably improved for all cases
World Academy of Science, Engineering and Technology
International Journal of Computer and Information Engineering
Vol:13, No:9, 2019
483International Scholarly and Scientific Research & Innovation 13(9) 2019 ISNI:0000000091950263
Open Science Index, Computer and Information Engineering Vol:13, No:9, 2019 waset.org/Publication/10010724
using the limit conditions for the trading entries.
Trading with (11) and (14) for targets higher than 20 points
we do not obtain a significant increase in RRR. In addition,
using longer targets will produce longer trades. Values higher
than 20 pips can be used for automated investment system
instead high-frequency trading systems.
Working with signals made by (13) and (14) on volatile
markets need sometimes additional conditions. In a volatile
market, after a new trend was established and the PPL changes
its color, entries can be made on points too far away the PPL
or in the last part of the trend where the price is near the
turning point. Additional limit conditions to entry and to exit
the trades can be imposed. In this section are presented results
obtained using the same PPL strategy applied in two time
frames in the same time. The results in Table I were obtained
applying (13) and (14) on one and four-hour time frames.
When the buy trading signals were true for the both time
frames, a new trade was opened with a ten points target.
VIII.
CONCLUSIONS
The PPL can be built based on the local minimum and
maximum points of the price time series using the PCY and a
transformation function into the price space. The strong and
direct correlation between the PPL function and the price
movement indicates that the monotony of the PPL function
can give us a strong indication about the price direction for the
next intervals. Besides, the PPL has an additional property
which is given by the monotony of the PCY function used to
build the transformation. Reading the evolution of the PPL
function together with the significance of the PCY function,
trend decision can be made automatically by building logical
trading signals.
The PPL function predicts the direction of the next price
evolution and the moment when the trend will be changed.
The values of the PPL function are more close to the price
level than the moving averages and the PPL can be used to
estimate the next possible price level where the new trend will
start. The gradient of the PPL function is also a good indicator,
being in strong relation with the power of the existing price
trend and with the amplitude made by the price movements.
High values for the gradient will indicate a strong price
tendency. Low values of the gradient will indicate weak price
movements. An exit signal can also be automated based on the
PPL function gradient.
Limit conditions in order to filter the entry trades can be
built using the PPL function, which can be used as price level
reference. A higher value of the price, too far away the PPL
levels can indicate a local maximum point and is a reasonable
limit signal for exiting the trades. The limit conditions and the
exit conditions based on the PPL function can be automated
using functional parameters that are optimized for each traded
market in order to reduce capital exposure and to maximize
the profit. All PPL limit conditions can be used in order to
filter any other type of trading signal. The model presented
can be applied in any time frame for any financial market. The
PPL can be easily automated in order to be included in any
automated trading system, the model being exclusively a
mathematical model based on the time price series.
REFERENCES
[1] ISITC, International Securities Association for Institutional Trade
Communication, 2011 Member Survey, March 12, 2012. Available on
http://www.isitc.org
[2] Funke, N., Goldstein, A., Financial Market Volatility, Intereconomics
Journal, volume 31, Issue 5, 1996, pp. 215-220. DOI:
10.1007/BF02927152
[3] Peters E., Chaos and Order in Capital Markets. A New View of Cycles,
Prices and Market Volatility, 1996, John Wiley & Sons Inc., ISBN 0-
471-13938-6
[4] Connors L., Alvarez C., Short term trading strategies that work. A
Quantitative Guide to Stock Market Behavior, US: Trading Markets
Publishing Group, 2009, ISBN 978-0-9819239-0-1
[5] Connors L., Sen C., How Markets Really Work – A Quantitative Guide
to Stock Market Behavior, US: Connors Research Group, 2004, ISBN
978-0-9755513-1-8
[6] Connors L., Alvarez C., High Probability ETF Trading – 7 Professional
Strategies to Improve Your ETF Trading, US: Connors Research, 2009,
ISBN: 978-0-615-29741-5
[7] Connors L.A., Best Trading Patterns, The Best of the Professional
Traders Journal, US: M. Gordon Publishing Group, 1999, ISBN: 0-
9650461-0-2
[8] Lien K., Day Trading and Swing Trading the Currency Market.
Technical and Fundamental Strategies to Profit the Market Moves. US:
John Wiley & Sons, 2009, ISBN: 978-0-470-37736-0
[9] Lien K., The Little Book of Currency Trading. How to Make Big Profits
in the World of Forex, US: John Wiley & Sons, 2011 ISBN: 978-0-470-
77035-1
[10] Cheng G., 7 Winning Strategies for Trafing Forrex. Real and actionable
techniques for profiting from the currency markets. GB: Hariman
Trading, 2007 ISBN: 978-0-857190-90-1
[11] Păuna C., Reliable Signals Based on Fisher Transform for Algorithmic
Trading. Romania: Timișoara Journal of Economics and Business.
Volume 11, Issue 1/2018 ISSN: 2286-0991. West University of
Timișoara. DOI: 10.2478/tjeb-2018-0006 Available at: https://tjeb.ro
[12] Păuna C., Reliable Signals and Limit Conditions for Automated Trading
Systems. Romania: Review of economic and Business Studies. Volume
XI, Issue 2/2018. ISSN: 1843-763X. Alexandru Ioan Cuza University
Press. DOI: 10.1515/rebs-2018-0070 Available at:
http://rebs.feaa.uaic.ro
[13] Păuna C., Smoothed Heikin-Ashi Algorithms Optimized for Automated
Trading Systems, Austria: Proceeding of the 2nd International Scientific
Conference on IT, Tourism, Economics, Management, and Agriculture.
ITEMA 2018. Graz University of Technology. Available at:
https://itema-conference.com
[14] Păuna C., Low risk trading algorithm based on the price cyclicality
function for capital markets. Romania: 13th International Conference on
Business ExcellenceAt: FABIZ, ASE, Bucharest DOI: 10.2478/mmcks-
2019-0006
[15] Păuna C., Additional Limit Conditions for Breakout Trading Strategies.
Romania: Informatica Economica Journal. Volume 23, Issue 2/2019.
DOI: 10.12948/issn14531305/23.2.2019.03 Available at:
http://revistaie.ase.ro
[16] Ward S., High Performance Trading. 35 Practicaal Strategies and
Techniques to enhance Your Trading Psychology and Performance. GB:
Hariman House 2010 ISBN: 978-1-905641-61-1
[17] Pauna C., Lungu, I., Price Cyclicality Model for Financial Markets.
Reliable Limit Conditions for Algorithmic Trading, Romania: Economic
computation and economic cybernetics studies and research, Volume 52,
Issue 4/2018. DOI: 10.24818/18423264/52.4.18.10 Available at:
http://ecocyb.ase.ro
[18] Cox, D.R. Sir, Prediction by Exponentially Weighted moving Averages
and Related Methods, 1961, Journal of the royal Statistical Society,
Series B, Vol. 23, No. 2, pp. 414-422
[19] Börse, Frankfurt, Frankfurt Stock Exchange Deutscher Aktienindex
DAX30 Components, 2019. Available on: http://www.boerse-
frankfurt.de/index/dax
[20] Reinsch, C., Smoothing by Spline functions, Numerische Mathematik,
Volume 10, Issue 3, pp 177–183, ISSN 0945-3245, 1967, DOI
https://doi.org/ 10.1007/BF02162161
[21] Berbente, C., Mitran, S., Zancu, S., Metode Numerice, romania: Editura
World Academy of Science, Engineering and Technology
International Journal of Computer and Information Engineering
Vol:13, No:9, 2019
484International Scholarly and Scientific Research & Innovation 13(9) 2019 ISNI:0000000091950263
Open Science Index, Computer and Information Engineering Vol:13, No:9, 2019 waset.org/Publication/10010724
Tehnică,1997, ISBN 973-31-1135-X
[22] Andrei T., Statistică și econometrie, Romania: Editura Economică,
București, 2003. ISBN: 973-590-764-X
[23] MetaQuotes Language, 2019, Available at: https://www.
metatrader4.com/en/automated-trading/mql4-programming
[24] Brexit, The prospective withdrawal of the United Kingdom (UK) from
the European Union (EU), 2016, Available at:
https://en.wikipedia.org/wiki/Brexit
[25] Păuna C., TheDaxTrader automated trading system, online software
presentation. 2010, Available at: https://pauna.biz/ thedaxtrader
[26] Păuna, C., Capital and Risk Management for Automated Trading
Systems, Proceedings of the 17th International Conference on
Informatics in Economy, May 2018, pp 183-188. Available at:
https://pauna.biz/ideas
World Academy of Science, Engineering and Technology
International Journal of Computer and Information Engineering
Vol:13, No:9, 2019
485International Scholarly and Scientific Research & Innovation 13(9) 2019 ISNI:0000000091950263
Open Science Index, Computer and Information Engineering Vol:13, No:9, 2019 waset.org/Publication/10010724
... For the currency and commodity markets, reliable trading models that can be automated can be found in [8], [9], and [10]. Original investment models and methodologies especially optimized for automated capital investment software can be found in [11], [12], [13], [14], [15], and [16]. Essential psychological strategies that can be used for better adaptation to the context of actual markets can be found in [17], [18], and [19]. ...
... In addition, price behavior can be considerate a periodic one. The price can be associated with "a wave with variable wavelengths" [16], and the price cyclicality will be well approximated by the trigonometric interpolation, better than the linear interpolation, especially for very volatile markets, where fast changes in the price direction are usually met. ...
... In Figure 5, it is presented a time price series of DAX30 Frankfurt Stock Exchange Deutscher Aktienindex [22] between June and September 2019. Over the price, with blue and red is drawn the Price Prediction Line introduced by [16], and noted here with PPL. Also over the price, with pink and purple is drawn the Trigonometric Price Line, introduced in this article in chapter 3, noted with TPL. ...
Conference Paper
Full-text available
ABSTRACT Algorithmic capital investment procedures became the essential tools to make a profit in the volatile price markets of the 21st century. A large number of market participants, private traders, companies, or investment funds are buying and selling on thousands of markets every day to make a profit. After the 2010 year, algorithmic trading systems became a significant part of the capital investment environment. The price evolution is analyzed today in real-time by powerful computers. To buy cheap and to sell more expensive is a simple idea, but to put it on practice is not easy today in very volatile price markets. The orders are built and set almost instantly today by artificial intelligence software using special mathematical algorithms. These procedures automatically decide the best moments to buy and to sell on different financial markets depending on the price real-time movements. This paper will present a specific methodology to analyze the time price series of any capital market. The model will build reliable trading signals to enter and to exit the market to make a profit. The presented method uses trigonometric interpolation of the price evolution to build a significant trend line called here the Trigonometric Trend Line. It will be mathematically proved that this function is in a positive and direct correlation with the price evolution. The Trigonometric Trend Line will be used to build and automate capital investment signals. Besides, the introduced function will be used in order to qualify the actual price trend and to measure the trend power in order to decide if the price makes an important evolution or not. Limit conditions will be imposed in the financial market to avoid trading in non-significant price movement and to reduce the risk and capital exposure. Comparative trading results obtained with the presented methodology will be included in the last part of this paper to qualify the model. Each trading signal type presented in the paper was traded separately to have a qualitative image. Also, all capital trading signals built with the Trigonometric Trend Line were traded together in order to obtain a better risk to reward ratio. To classify the presented methodology, the presented results were compared with real trading profits obtained with the other three well-known capital investment strategies. With all of these, it was found that using the Trigonometric Trend Line reliable automated trading procedures can 7 th International Symposium "SOCIO-ECONOMIC ECOSYSTEMS " January 22-24, 2020 University of Alicante , Spain Please send to: abstract-submission@bslab-symposium.net be made and optimized for each financial market to obtain good results in the capital investment. Being exclusively a mathematical model, the Trigonometric Trend Line methodology presented in this paper can be applied with good results for any algorithmic trading and high-frequency trading software. The functional parameters can be optimized for each capital market and for each timeframe used in order to optimize the capital efficiency and to reduce the risk. The optimization methods will use the historical time price series in order to catch the price behavior and specificity of each market. The reduced number of parameters and the simplicity of the presented method recommend the Trigonometric Trend Line model to be used in any advanced algorithmic trading software.
... A particular approach for risk management techniques that can be used in automated decision-making systems for capital investments can be found in (Vince, 1992). Mathematical models especially designed and optimized for algorithmic trading with proved and sustained results in real capital investments can be found in (Păuna & Lungu, 2018), (Păuna, 2018a), (Păuna, 2018b), (Păuna, 2018c), (Păuna, 2019a), (Păuna, 2019b), (Păuna, 2019c), (Păuna, 2019d), (Păuna, 2019e), (Păuna, 2019f), and (Păuna, 2020). ...
... The Price Prediction Line is a function introduced in (Păuna, 2019d). It represents the back transformation into the price space of the Price Cyclicality Function, using the formula: ...
Conference Paper
Full-text available
Capital investment is a sustained activity nowadays. After the worldwide release of the electronic trading systems, automated decision-making investment software is the new trend in financial speculation. A significant part of capital trading is fully computerized today. The buying and selling orders are made and sent automatically, almost in real-time. The price evolution is analyzed by servers using advanced mathematical algorithms. This paper will present one of these models named Price Probability Predictor. It is a method to build a probability field based on the price history and the real-time price action. The revealed function will generate the current probability of a price growth in the next time intervals. Automated entry and exit signals and market limit conditions will be built using the new indicator, in order to automate the whole investment process. Capital investment results will also be included in the current paper to qualify the presented trading methodology and to compare it with other similar models. In conclusion, it was found that the Price Probability Predictor is a reliable mathematical algorithm that can assist any trading decisions, in both ways, manual or automatic capital investments.
... investment positions in the current market will be closed if the PCY function exceeds a particular value: conditions can be made using the Inverse Fisher RSI (IFR) introduced in [16]: type of limit conditions to avoid trading near-maximal price levels can be made limiting the distance between the current price level and an adequate price trend line. Very good results are obtained using the Price Prediction Line (PPL) introduced in[11] and Trigonometric Price Line (TPL) introduced in[10]. The limit conditions are made using: ...
... No new investments trades will be opened if the PCY function overtakes a specified limit value: (4) A particular type of limit conditions to avoid trading near-maximal price levels can be made limiting the distance between the current price level and an adequate price trend line. Very good results are obtained using the Price Prediction Line (PPL) introduced in [11] and Trigonometric Price Line (TPL) introduced in [10]. The limit conditions are made using: ...
Conference Paper
Full-text available
Using automated capital investment software systems is a common task today. At the beginning of the third millennium, modern investors are using artificial intelligence resources and methods to find the best investment opportunities on capital markets and to process the trading orders. One of the most important aspects of this activity, besides the buying and selling decisions, is to stay away from the market risk in specific conditions. For this purpose, in the current doctoral research, the notion of limit conditions in capital markets was introduced by the authors. On the high price volatility markets, when the economic or geopolitical background is changing fast, real-time decisions for earlier investment closing, or filtering decision not to open new positions in specific market states, will contribute together to the risk reduction and will provide a higher capital efficiency a the long time run. In the real-time investment software systems, the limit conditions method's implementation presumes particular aspects in order not to introduce additional time delays for the trading orders. This paper will present the way how to include additional limit conditions procedures into automated algorithmic trading software systems. It was found that any investment strategy can be improved by using the limit conditions methods presented in this paper. Based on particular data-mining methods applied to real-time price series of any market, these methods can be automated and included in any capital investment informatics systems in order to improve the results and to reduce the allocated capital risk.
Thesis
Full-text available
After several attempts to publish my Ph.D. thesis with different prestigious publishers, I have decided to make this work public and free of charge for anyone. Enjoy! Cristian Păuna
Article
Full-text available
One of the most popular trading methods used in financial markets is the Turtle strategy. Long-time passed since the middle of 1983 when Richard Dennis and Bill Eckhardt disputed about whether great traders were born or made. To decide the matter, they recruited and trained some traders (the Turtles) and give them real accounts and a complete trading strategy to see which idea is right. That was a breakout trading strategy, meaning they bought when the price exceeded the maximum 20 or 50 days value and sold when the price fell below the minimum of the same interval. Since then many changes have occurred in financial markets. Electronic trading was widespread released and financial trading has become accessible to everyone. Algorithmic trading became the significant part of the trading decision systems and high-frequency trading pushed the volatility of the financial markets to new and incredible limits nowadays. The orders are built and sent almost instantly by smart computers using advanced mathematical algorithms. With all these changes there are many questions today regarding the breakouts strategies. Are the Turtle rules still functional? How can the Turtle strategy be automated for algorithmic trading? Are the results comparable with other modern trading strategies? After a short display of the history and the system's rules, this paper will find some answers to all these questions. We will reveal a method to automate a breakout strategy. More different trading strategies originating from the Turtle rules will be presented. A mathematical model to build the trading signals will be described in order to automate the trading process. It was found that all of these rules have a positive expectancy when they are combined with modern limit conditions. The paper will also include trading results obtained with the methods presented in order to compare and to analyze this capital investment methodology adapted especially for algorithmic trading.
Conference Paper
Full-text available
Buy cheap and sell more expensive is one of the basic ideas of trading the capital markets for hundreds of years. To apply it in practice has become difficult nowadays due to the high price volatility. The uncertainty in the price movements often leads to high-risk allocation. One main question is when the price is low enough for a low-risk entry? Once established an entry point, the second question is how long to keep the open trades in order to optimize the investment efficiency? This article will answer these questions. A general trading algorithm based on the price cyclical behavior will be revealed. The mathematical model is developed using the Price Cyclicality Function combined with other computational techniques in order to establish low-risk intervals. The algorithm will use multiple entry points in order to catch the best price opportunities. A simple empirical exit algorithm will be optimized in order to maximize the profit for a certain capital exposure level. The presented model uses a low number of functional parameters which can be optimized with a reasonable computational effort for any financial market. Trading results obtained for several markets will also be included in this paper in order to reveal the efficiency of the presented methodology. It was found that the Low-Risk Trading Algorithm can be used with good results for algorithmic trading in any financial market. With the right parameters set, this methodology can be wide range applied in the stock markets, currency and cryptocurrency markets, commodities and raw materials markets and even for the real estate investments. The simplicity of the presented model and the good efficiency level obtained will recommend it. This methodology can be used by any investor in order to manage the investment plan with multiple capital markets.
Article
Full-text available
Trading and investment on financial markets are common activities today. A very high number of investors, companies, public or private funds are buying and selling every day with a single purpose: the profit. The common questions for any market participant are: when to buy, when to sell and when is better to stay away from the market risk. In order to answer all these questions, many trading strategies are used to establish the best moments to entry or to exit the trades. Due to the large price volatility, a significant part of the trades is set up automatically today by computers using algorithmic trading procedures. For this particular field, special aspects must be met in order to automate the trading process. This paper presents one of these mathematical models used in automated trading systems, a method based on the Fisher transform. A general form of this method will be presented, the functional parameters and the way to optimize them in order to reduce the risk. It will be also suggested a method to build reliable trading signals with the Fisher function in order to be automated. Three different trading signal types will be explained together with the significance of the functional parameters in the price field. A code sample will be included in this paper to prove the simplicity of this method. Real results obtained with the Fisher trading signals will be also presented, compared and analyzed in order to show how this method can be implemented in algorithmic trading.
Article
Full-text available
Automated trading software is a significant part of the business intelligence system in a modern investment company today. The buy and sell orders are built and sent almost instantly by computers using special trading and computational strategies. The trading decisions are made by automated algorithms. In this paper it will be presented one of these mathematical models which generate trading signals based only on the time price series. The algorithm combines several known computing techniques to build a trading indicator to automate the trades. With this method, buy decisions on oversold intervals and sell decisions on overbought price values can be built. Limit conditions in order to close the long and short trades can be also automatically generated. More trading signal types based on this model will be revealed. Trading results obtained with all these signals will be presented in order to qualify this methodology developed especially for algorithmic trading.
Article
Full-text available
Trading the financial markets is a common idea nowadays. Millions of market participants, individuals, companies or public funds are buying and selling different equities in order to obtain profit from the buy and sell price difference. Once the equity was established, the main question marks are when to buy, when to sell and how long to keep the opened positions. This paper will present a mathematical model for the cyclicality of the price evolution. The model can be applied for any equity in any financial market, using any timeframe. The method will gives us information about when is good to buy and when is better to sell. The price cyclicality model is also a method to establish when the price is approaching to change its behavior in order to build limit conditions to stay away the market and to minimize the risk. The fundamental news is already included in the price behavior. Being exclusively a mathematical model based on the price evolution, this method can be easily implemented in algorithmic trading. The paper will also reveal how the cyclicality model can be applied in automated trading systems and will present comparative results obtained in real-time trading environment.
Conference Paper
Full-text available
Heikin-Ashi is the Japanese term for "average bar". This methodology is well known as one of the methods to identify and follow the trends using a price time series in financial markets. Nowadays, in the first decades of the 21st century, in the electronic trading environment, with very volatile price market conditions, using the Heikin-Ashi method gets new and special connotations especially when it is about the high-frequency trading. It was found that combining the classical Heikin-Ashi candlesticks with modern limit conditions reliable trading algorithms can be generated in order to produce a good trading return with automated trading systems. This paper will present several trading algorithms based on Heikin-Ashi method for algorithmic trading especially adapted for high-frequency trading systems. It will be revealed how the trading signals can be automatically built and used in order to automate the trading decisions and orders. Exit signals will also be discussed. Trading results obtained with the presented algorithms for Frankfurt Stock Exchange Deutscher Aktienindex Market will be displayed in order to qualify the methods and to compare them with any other trading strategies for high-frequency trading. As conclusions, Heikin-Ashi combined with special limit conditions can generate reliable trading models for algorithmic trading.
Conference Paper
Full-text available
The most important part in the design and implementation process of automated trading systems in any financial investment company is the capital and risk management solution. Starting from the principle that the trading system must run fully automated, the design process gets several particular aspects. The global stop loss is a special approach for the risk management strategy that will ensures a positive expectancy in algorithmic trading. A case study based on an already optimized trading algorithm will be used to reveal how important the risk level optimization is, in order to improve the efficiency of the trading software. The main optimal criteria are as usual the profit maximization together with the minimization of the allocated risk, but these two requirements are not enough in this case. This paper will reveal an additional optimization criterion and the main directions to build a reliable solution for an automated capital and risk management procedure. Keywords: automated trading software (ATS), business intelligence systems (BIS), capital and risk management (CRM), algorithmic trading (AT), high frequency trading (HFT). (Available at: https://pauna.biz/Capital_and_Risk_Management)
Article
The mean square error of prediction is calculated for an exponentially weighted moving average (e.w.m.a.), when the series predicted is a Markov series, or a Markov series with superimposed error. The best choice of damping constant is given; the choice is not critical. There is a value of the Markov correlation ρ0 below which it is impossible to predict, with an e.w.m.a., the local variations of the series. The mean square error of an e.w.m.a. is compared with the minimum possible value, namely that for the best linear predictor (Wiener). A modified e.w.m.a. is constructed having a mean square error approaching that of the Wiener predictor. This modification will be of value if the Markov correlation parameter is negative, and possibly also when the Markov parameter is near ρ0.