Algorithmic Trading Basics: Strategies & Systems

Overview of Products
& Design Methodology
An overview of our Active Trader (NQ) Package, design
specifications, correlation analysis versus the S&P 500,
post trading support methodology and product dashboard.

Contents
Contents
Introduction to Automated Trading 1
A Complete Trading System 2
Predicting Market Direction 3
Performance vs. the S&P 500 4
Mathematical Proof of Correlation 4
Correlation of Each Algorithm to the Others 5
Typical Trading Algorithm Development Cycle 6
Automated Trading System Design Specifications 6
The Five Algorithms 12
Quality Control Processes 25
Final Word 26
Risk Disclosures 27
© 2016 AlgorithmicTrading.net All rights reserved.
This publication is copyrighted by AlgorithmicTrading.net. Any reproduction or redistribution of this publication, in whole or in part,
whether in hard-copy format, electronically or otherwise to persons not authorized to receive it, without the express consent of
AlgorithmicTrading.net, is in violation of U.S. copyright law and will be subject to an action for civil damages and, if applicable,
criminal prosecution. Any questions should be directed to AlgorithmicTrading.net at 866.759.6546 or sales@algorithmictrading.net.

1Algorithmictrading.net White Paper: Overview Of Products & Design Methodology
Introduction to Automated Trading
For most automated trading system developers, the following cycle will look all too familiar. They will
start with a faulty trading strategy, only to convince themselves it is perfect, then trade live only to see
horrible performance.
Next, they rationalize the poor returns. They might say, “If only
my stop was X instead of Y, the performance would have been
amazing!” Then they modify the design — or re-optimize — only
to find that they continue to experience bad results.
When it comes to automated trading system development, it
really is a matter of art and science. As most developers know,
there is a big difference between trading systems that appear
favorably based on back-tested results, and those that perform well live.
Coding a solid trading strategy is certainly the first step, but not all automated trading systems are
created equally.
At AlgorithmicTrading.net, we have a very detailed design methodology that we follow with few
exceptions. Furthermore, we have quality control processes in place to ensure we remain on the right
track.
In this white paper we will provide details on our trading system design methodology, and the quality
control processes and details on each algorithm, and provide an in-depth analysis of the products we
offer. We will also highlight the potential pitfalls prevalent in automated trading system development, as
well as how to avoid them.
Redesign, Modify,
or Reoptimize System
Experience Poor Results
Using Too Many
Indicators
Back-Test Less
Than 10 Years
Over
Optimize
Horrible Design With
Poor Live Results

A Complete Trading System
AlgorithmicTrading.net currently offers four different trading packages.
Each package trades multiple uncorrelated algorithms. Trading
packages are allocated on a per unit basis. Each unit traded
represents 1 contract being traded on each of the algorithms
contained in that package.
As an example, consider the NQ Active Trader Package.
1. T2 Burst (NQ)
2. P2 Push-Pull (TY)
3. B2 Breakout (NQ)
4. S2 Breakdown (ES)
5. O2 Overnight Gap (NQ)
The allocation we use for our analysis assumes 1 contract is traded per $15,000 in the account.
For example, on an account of $15,000, 1 contract can be traded on each algorithm (five contracts
total); on an account of $30,000, up to 2 contracts can be traded on each algorithm (10 contracts total);
and on an account of $45,000, up to 3 contracts can be traded on each algorithm (15 contracts total).
Keep in mind that no one has the perfect system for trading. Trading futures involves substantial risk
of loss and is not for everyone. There will be days where you suffer losses or give-back gains. Our
algorithms should only be used with risk capital, that is money you can afford to lose.
Even with automated trading systems, there will be the urge to turn off the algorithms and not let them
run. It is our opinion that those periods typically produce the best opportunities, and experience shows
it’s best to permit the course of the trades to run. With that said, we do not control client accounts and
so individuals are able to modify the allocation of the algorithms.
While one algorithm might be in a drawdown, it is possible the others will be breakeven or profitable,
helping the combined system to potentially generate positive results. Of course, there are no
guarantees in trading; however, we attempt to put every odd in our favor to drive maximum probability
of success.
Predicting Market Direction
Regrettably, there is no crystal ball when it comes to trading. While we have percentage profitability
expectations for every single trade we make based on the back-testing, ultimately no one knows for
sure what the market will do at any given time.

To offset the unknown, we use what we do know. We know with certainty that the market will move
sideways, higher, or lower, in any given period. Then we develop individual algorithms that perform
uniquely based on those market movements.
For the purpose of system development, we add an additional category called “rebounding”
representing a strong move higher after a substantial down move, also called a short covering rally.
The end result is a system that is attempting to be market direction agnostic — by trading five strategies
concurrently, each with its own strengths, weaknesses and expectations for the different market
conditions.
The following diagram captures each market condition along with the expectations of positive
performance for each algorithm contained In the NQ Active Trader Package. Each algorithm has a
strongly positive expectation for one of the four conditions, along with weaker positive expectations
where the overlaps occur.
The ideal conditions for the algorithms are when the algorithms performance overlap since that implies
multiple algorithms are performing well. In fact, we have had months where all five algorithms on the
NQ Active Trader Package are profitable resulting in exceptional returns in the hypothetical account for
those periods.
CFTC RULE 4.41: Results are based on
simulated or hypothetical performance results
that have certain inherent limitations. Unlike
the results shown in an actual performance
record, these results do not represent actual
trading. Also, because these trades have not
actually been executed, these results may
have under-or over-compensated for the
impact, if any, of certain market factors, such
as lack of liquidity. Simulated or hypothetical
trading programs in general are also subject
to the fact that they are designed with the
benefit of hindsight. No representation is
being made that any account will or is likely
to achieve profits or losses similar to these
being shown.
Strong Bull
Strong Bear Run
Rebounding
Market
Sideways
Moving
Overnight Gap
(ES) Outperforms
Burst (ES)
Outperforms
Breakout (ES)
Outperforms
Push-Pull (TY)
Outperforms
Breakdown
SHORT (ES)

Performance vs. the S&P 500
The equity curve to the right shows the back-
tested performance of the five merged algorithms,
in the ES Active Trader package, as compared
to the S&P 500. In order to create this
equity curve, we took the five algorithms
and applied them to the ES. This was done
so that we can do an apples to apples
comparison vs. vthe S&P 500. While this
equity curve looks quite Impressive, it is
based on back-testing which has limitations
as the CFTC RULE 4.41 explains.
As you can see, the combined automated
trading strategy performance is spectacular
during both bull and bear markets.
As the equity curve shows, there is little
correlation between our five merged algorithms
and the S&P 500. Our back-tested performance
is not tied to the performance of the S&P 500.
CFTC RULE 4.41: Results are based on simulated or hypothetical performance results that have certain inherent
limitations. Unlike the results shown in an actual performance record, these results do not represent actual trading.
Also, because these trades have not actually been executed, these results may have under-or over-compensated for
the impact, if any, of certain market factors, such as lack of liquidity. Simulated or hypothetical trading programs in
general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being
made that any account will or is likely to achieve profits or losses similar to these being shown.
Mathematical Proof of Correlation
The correlation coefficient is a percentage that represents how interrelated two data sets are. In trading
algorithm development, a designer will typically measure the correlation of their algorithms to the S&P
500 to determine how correlated an algorithm is to the broader market performance. Since the goal of
most auto trading systems is to outperform this index, it only makes sense to measure the correlation
between the trading strategy developed and the S&P. Here is a commonly accepted definition of what
different values imply:
$17,000 Starting Account Size, 1 Contract per
Algorithm (1/1/1/1/1), hypothetical account
+.70 or higher -----------------Very strong positive relationship
+.40 to +.69------------------------- Strong positive relationship
+.30 to +.39----------------------Moderate positive relationship
+.20 to +.29 -------------------------- Weak positive relationship
+.01 to +.19------------------------ No or negligible relationship
-.01 to -.19 ------------------------- No or negligible relationship
Five Algorithms Merged vs. S&P 500
ES Active Trader Package, Non-Compounded
(06.03.01– 03.01.15)

A value of 100% would imply that the two data sets are equal. A value of 0% would imply two fully
random data sets. A negative value would imply an inverse relationship.
NQ Active Trader Package (ES) Correlation Coefficient as compared to the S&P 500 = .03%.
This analysis shows our complete algorithmic trading systems back-tested performance is not driven by
the S&P 500 performance. As the merged equity curve showed on the previous page, and correlation
coefficient confirms, our automated trading system has no relationship to the performance of the broader
S&P 500 on a back-tested basis.
In our opinion, any correlation below 50% is positive news. If the goal is to outperform the S&P 500, then
anything more than 50% would defeat the purpose of implementing an algorithmic trading system since
the trader could simply buy and hold the S&P and not waste their time with trading. As our merged equity
curve demonstrates, our expectation is for continued positive returns independent of market conditions.
Correlation of Each Algorithm to the Others
Digging deeper into our correlation to the S&P 500, we can also determine the relationship between the
five different trading algorithms to one another (non-merged) on a back-tested basis.
The NQ Active Trade Package trades five separate
futures trading systems for a reason. The goal of our
system is that at all times, one trading system may be
strongly up, one slightly up, one breakeven and one
slightly down. This results in a net positive automated
trading system based on the back-testing.
Taking a close look, the Push-Pull (TY) algorithm vs. the
S&P 500 has a moderate inverse relationship (-30.70
%), while the Breakdown SHORT (ES) algorithm has
a -23.84% inverse relationship to the S&P 500. These
two algorithms work together in order to try and hedge
against downward moves on the broader market.
The Breakout (ES) has little to no correlation to the
other algorithms demonstrating its value to our trading
system.
The Burst and Overnight Gap are the most correlated
algorithms in the grouping — to each other and to the
S&P 500. That correlation is the primary reason why
in the back-testing our system is profitable during
sideways or upward moving markets.
.03%
Correlation Coefficient Results
Push-Pull (TY) vs Breakout (ES) -11.56%
Push-Pull (TY) vs Burst (ES) -15.24%
Push-Pull (TY) vs Overnight Gap (ES) -19.90%
Push-Pull (TY) vs Breakdown SHORT (ES) 15.49%
Breakout (ES) vs Burst (ES) 9.80%
Breakout (ES) vs Overnight Gap (ES) 2.52%
Burst (ES) vs Overnight Gap (ES) 29.47%
Burst (ES) vs S&P 500 31.03%
Push-Pull (TY) vs S&P 500 -30.70%
Burst (ES) vs S&P 500 27.17%
Overnight Gap (ES) vs S&P 500 21.77%
Breakdown SHORT (ES) vs S&P 500 -23.84%
All 5 Algorithms
Merged vs S&P 500
CORRELATION CORRELATION
RESULTS COEFFICIENT
Why do we care? The correlation coefficient of
.03% suggests zero correlation to the S&P 500. In other
words, our algorithms’ performance is not tied to the
performance of the S&P 500. When the market goes up,
down, or sideways, our algorithms do everything they
can to be market direction agnostic.

With that said, there are no guarantees that our algorithms will continue to perform well. You should
carefully consider this prior to purchasing our algorithms and trading them on live accounts. As
mentioned previously, back-testing has limitations per CFTC Rule 4.41.
Typical Trading Algorithm Development Cycle
The following is a basic overview of how an individual
algorithm might be developed.
Step 1: Create an Idea
This process begins with a simple idea, which is
subsequently coded and analyzed. It might start as an idea to
“sell or fade a gap up at the opening bell” but then changes
to see what happens if you “buy an opening gap”. After
running multiple simulations, the idea abandoned, and replaced
with new options in search of something else.
Step 2: Back Test & Optimize the Algorithm
Once a basic trading strategy is coded and looks to be promising, the developer
will optimize the algorithm’s inputs. This might be a stop, target, or some other technical indicator.
During this phase, simulations will run, changing inputs based on the granularity selected. They will also
cross-optimize the inputs to find - based on the previous history - what the most optimal inputs (stop,
target, technical indicator) would have been. Trading platforms will then produce a report indicating
those critical inputs. They will also generate back-tested performance reports indicating everything from
maximum drawdown, percent profitability, profit factors and much more.
Once the optimization is complete, the trader in theory has a mechanical trading system that could
be auto executed. However, it is our opinion that there is much more to developing a winning trading
system than just running the above outlined steps.
Automated Trading System Design Specifications
There is much more to the development of an automated trading system than just coding an algorithm,
back-testing, and optimizing. The real work comes in vetting the algorithms. The goal of any system
developer should be to thoroughly test the algorithms by attempting to find their weaknesses prior to
going live.
At first glance, the new trading system might appear to be solid, but after rigorous testing it may be
determined that it is not qualified to trade. This process is critical to the success of the trading system.
What separates AlgorithmicTrading.net from the other automated trading systems is our developer does
his best to adhere to strict development and testing guidelines, varying from these principles only if the
situation warrants due to unique special case circumstances.
Buy on
MACD
Crossover
Sell a
Gap
Buy a
Gap

One of the biggest errors an algorithm developer can make is cutting corners in the testing phase of
the development process. The developer at AlgorithmicTrading.net takes these principles seriously.
We understand that the strength of our algorithms is key to the success of our system.
As an innovative, algorithmic trading system design firm, since March of 2015 AlgorithmicTrading.net
has required the algorithms mentioned in this white paper to meet the following design criteria for our
automated trading system, with few exceptions:
Back-Test 10+ Years
When optimizing our algorithms, we back-tested starting from May 2001. We found that other
developers only back-tested four years — and even less — which avoided the 2008 crash and
market periods prior to that. To further test our algorithms, we modified the Burst and Push-Pull
so it could trade on the broader index, and tested as far back as 1984 which also showed very
good results.
Uncorrelated Algorithms
Our strategy is to combine five uncorrelated algorithms to create a complete automated
trading system. This can be easily evaluated using a correlation coefficient. Our target for the
merged value is between 0-.50, however, it depends on the goal of each algorithm. The value is
measured by comparing the algorithms’ weekly performance to the S&P 500 and determining
how correlated or uncorrelated the trading system is to the broader index. A final value of over
0.50 (a strong positive correlation) suggests that the system will simply perform as the S&P 500
does in most cases. In that case, there would be no value to use an automated trading system.
A final value closer to zero suggests that there is little to no correlation to the S&P 500.
A mentioned before, the correlation coefficient of AlgorithmicTrading.net’s five algorithms
merged is 0.03%. This means our algorithms do everything they can to be market direction
agnostic and do not appear to be dependent on the S&P 500.
Back-Test As
Far Back As
Possible
(Min. 10+ Years)
Four to Five
Uncorrelated
Algorithms
Reasonable
Profit Factors
(1.2-2.6)
Large Average
Gain Per Trade
(No Scalping)
Use Look-Inside,
Intrabar Order
Generation
Include Adequate
Slippage &
Commission
Use Three Or
Fewer Technical
Indicators
Per Algorithm
Include Monte
Carlo Simulation
Modify Inputs
+/- 10%
To Ensure
Minimal Impact
Minimum 200
Trades In
Back-Test History
Per Algorithm
Trade Live
Prior To
Offering
Drawdown
Scalable To Meet
Various Needs
Do Not
Over-Optimize
Independent
Third-Party
Evaluation
Scalable System
– Can Handle
Volume

Reasonable Profit Factors (1.2-2.6)
The Profit Factor (PF) is a ratio of total gain to total loss. A broadly accepted view is that a PF
below 1.2 is probably not profitable, and a PF above 2.6 is not realistic and likely achieved by
violating design criteria and standards. As of March 2015, the back-tested profit factors for our
system range from 1.20-2.50 depending on the algorithm.
CFTC RULE 4.41: Results are based on simulated or hypothetical performance results that have certain
inherent limitations. Unlike the results shown in an actual performance record, these results do not
represent actual trading. Also, because these trades have not actually been executed, these results may
have under-or over-compensated for the impact, if any, of certain market factors, such as lack of liquidity.
Simulated or hypothetical trading programs in general are also subject to the fact that they are designed
with the benefit of hindsight. No representation is being made that any account will or is likely to achieve
profits or losses similar to these being shown.
Large Average Gain Per Trade
By averaging all trades in the complete system, winners and losers, you can determine an
average gain per trade. It is important to have room for error, therefore the higher the average
the better.
A common mistake in developing algorithmic trading systems is the creation of a scalping
algorithm that is in and out multiple times throughout the day. These algorithms look good
back-tested, but when traded live they fall apart. More times than not, this occurs because
their average gain per trade is less than one tick on the index they are trading. While they
appear to show great equity curves, stable profit factors and great reports with low drawdowns,
the reality is that they will probably be at a loss when going live because they provided little
room for error. The truth is, no retail trader should be in and out multiple times during the day,
that is a job for HFT firms who have millions invested and dedicated design teams to monitor
their HFT algorithms.
Use Look-Inside (LIB) and Intrabar Order Generation (If Applicable)
A common mistake when developing algorithms is to turn off the look-inside bar back-testing
feature. If unchecked, back-testing results will be inaccurate showing winning trades when
in fact the actual trades were losers. This is a bigger issue for algorithms that trade on large
candles and/or algorithms that have very tight stops or very tight targets. Problems arise when
within a single candle, either the stop or target could have been hit.
For example, with LIB checked, back-testing optimizations take longer with Tradestation
because they will analyze every tick within the candle to determine which was hit first, the
stop or the target. Tradestation defaults to this being unchecked so that simulations are faster.
However, with LIB unchecked Tradestation will use its own proprietary algorithm to determine if
the stop was hit first or the target was. Unfortunately, it seems that their algorithm will more often
than not err on the side of assuming a target was hit first. During our development cycle, we
always run with LIB checked to ensure that the back-tested results are accurate.
By design, we do not run with Intrabar Order Generation (IOG) enabled, as it depends on the
intent of the trading system whether to enable it or not.

Include Adequate Slippage & Commission in Analysis
Slippage and commissions eat into any profits. Since this can vary from trade to trade, a system
designer needs to be realistic with this to err on the side of caution. If an algorithm enters at the
market and exits at a limit, then you can assume you will have at least one tick of slippage on the
buy, and potentially some slippage on the sell even though it is a limit order to exit. The reason
there is also possible slippage on the sell is that, at times, the index traded will barely hit the limit
price but it will not fill. It will then reverse. The algorithm thinks you exited the trade, even though
in the live account it did not fill. If this happens, our algorithms are programmed to exit at the
market — typically 15 seconds later — to ensure the live account is in-sync with the algorithm.
Commission rates should also factor into the performance. In our case, we add both slippage
and commission to all of our reports unless stated otherwise.
Use Three or Fewer Technical Indicators
Another broadly accepted principle in trading system development is the fewer the technical
indicators the better. We require three or fewer. In fact, in some cases our algorithms only have
one. We do use price action heavily and pattern recognition in our algorithms, which is a different
concept. The push-pull and burst algorithms also utilize a finite state machine to help define
various patterns that have occurred in order to define good entry points.
If you think of a trading system as a house of cards, the more technical the indicators the more
flimsy the house. Usually, algorithms with a large amount of technical indicators will result in
over-optimization when back-testing is performed. As results are analyzed, the developer will
add new indicators to try to avoid losses resulting in a very flimsy algorithm that will be more
likely to fall apart when traded live. This principle seems to be confirmed when walk-forward
analysis is done. Experience has shown us that the more indicators used, the less likely an
algorithm will pass the most basic walk-forward pass/fail criteria.
Our philosophy is to develop a reliable algorithm that works when traded live and accept a lower
profit factor, than one that looks good on back-testing but performs poorly after going live.
Perform Monte Carlo Simulation
Monte Carlo Simulation randomizes the back-tested trades to ensure there are no hidden
patterns that exist only due to unique market conditions. It is another way to try to break the
algorithmic trading system during the testing process and evaluate its performance with the
same trades executed randomly. This is helpful in determining a worst-case potential drawdown.
Modify Inputs +/- 10% To Ensure Minimal Impact
Once optimization is performed, we modify all inputs randomly by +/- 10% to determine how
flimsy the algorithm is. For example, after optimizing an algorithm we might determine that the
most optimal target is 10 points. We will then go back and modify the target to be 9 points and
11 points to ensure that the algorithm still looks acceptable. If it falls apart at that point, that is a
warning sign that it has been over-optimized.

At Least 200 Trades in Back-Testing History
In general, the bigger the data set the better when analyzing an algorithm. Our complete system
has over 3,300 trades as of the time of this paper. We believe if an algorithm has less than 200
trades there is not enough data to make a case for that algorithm’s performance going forward.
Trade Live Prior To Offering To Public
Any algorithm should be traded live prior to making any strong conclusions about it and offering.
Drawdown Scalable To Meet Various Customer Needs
The drawdown should be scalable to meet an individual’s needs. Our algorithms can be
scaled by adjusting the number of contracts traded times amount of dollars in the account.
For example, if someone is uncomfortable with a 30% drawdown potential, they could trade
1 contract per $30,000 in the account instead of 1 contract per $15,000 in the account.
This cuts the expected maximum back-tested drawdown from 30% to 15%, but would also cut
the potential gains in half.
AlgorithmicTrading.net is not a registered CTA and does not provide any risk management
services. The ability to scale the number of contracts traded simply refers to the fact that
someone can trade less than the maximum allowed contracts (1 per $15,000). This decision has
to be the customers and we will not provide any custom advice tailored to your specific needs.
The maximum drawdown numbers we present are based on back-testing and actual losses can
exceed those numbers.
Do Not Over Optimize
Once an algorithm is coded, it is optimized to determine the best possible values for each input.
These values can be optimized with as much granularity as a developer wants. While we could
optimize down to 0.0001 points or lower for any give input, we choose to use a much higher
granularity to further reduce the risk our algorithms are over-optimized.
Independent Third Party Evaluation
Ideally, a third party should evaluate any algorithm or complete trading system prior to a final
seal of approval. The intention is simple, get one more set of eyes on the product.
Some of our algorithms were evaluated by an independent design firm. We received a report
that gave our algorithms very high marks. The evaluator spent over one month trying to break
our algorithms, to no avail. The final report was extremely positive and recommended forward
testing the algorithms. The entity funding execution of the report decided not to proceed with
this final step since the previous optimizations showed live trades over-performed over the
previous 6 months. In our opinion, this was better than conducting any forward testing as the
analysis would have required a fresh optimization — changing the inputs to the algorithms as
they are now — and re-run on the same data set that we already had live returns for. While it
would provide another angle at examining each algorithm, it was deemed unnecessary given
that we already had live returns for the existing optimizations. To view the third-party evaluation
visit http://bit.ly/1H5CeF6.

Scalable System – Can Handle Volume
Any successful trading system should be able to handle a large account size and have the
ability to scale higher with the success of the system (i.e. increasing contract size as the system
performs). The key is to only trade markets with the greatest amounts of liquidity. Our system
trades the Emini S&P 500 Futures (ES), Emini NASDAQ Futures (NQ) and the 10-Year Note (TY),
which are some of the most liquid futures instruments traded. While futures trade 24-hours per
day, we ensure the algorithms can handle volume at all hours by limiting our trading to only when
the equity markets are open. This helps to ensure when a trade is triggered, there will be enough
liquidity to ensure our slippage is minimized.
According to the CME Group, the average daily volume (ADV) on the ES is almost 2 million
contracts. At an initial margin rate of $5,000 per contract traded, this amounts to approximately
$10 billion in trades on the S&P every day. The TY has an ADV of almost 1.5 million contracts,
which is equal to approximately $2.2 Billion worth of shares traded every day. Averaged over a
24-hour period, it is our opinion that this allows for plenty of liquidity to handle our algorithms
traded with very large accounts across multiple customers.
Final Sanity Check
This final step is slightly less structured and difficult to quantify, so we do not list it as an actual
design requirement. Simply put, the concept or principles behind the automated trading strategy
should make sense and pass a basic sanity check.
For example, it is not sufficient to stumble upon a random pattern and justify it as a reliable
basis for an algorithmic trading system. Algorithms must have reasons behind their expectation
for success.
For the Breakout algorithm, we are capturing short covering rallies and buying when it is difficult
(i.e. on a gap up). When most day traders are shorting the large gap up and expecting it to fill the
gap, we will typically buy the breakout. Once it has made a large up move from our entry and
most daytraders will feel it has moved to far and get out, the back-testing data suggests that
you should hold until the end of the day, so our algorithm holds.
Our Breakdown SHORT algorithm is similar, however instead of buying into strength it will sell
into weakness. When most retail traders are buying a gap down thinking that the market has
gone too far and will rally, the best trade in our opinion is once again the harder trade, namely
shorting into the weakness.
The logic behind the Burst is to buy breakouts within range bound or sideways moving markets
but exit quickly in case they are false breakouts. The Burst also buys the bottom of the range
in sideways trading markets, allowing for a larger target, and exit once the futures trade back
towards the top of the range. The Push-Pull is similar to the Burst, except that we hold longer
and typically only buy on dips.
The principle behind the Overnight Gap is equally straightforward. It buys into strength during
upward trending markets, attempting to exit the following morning when the equity markets
open. This tendency to gap up is, in our opinion, due to the ramp in futures that tends to happen
in strong markets during the overnight light volume trading session.

At AlgorithmicTrading.net, our number one principle when designing algorithms is to think in terms of
“Why do most day traders fail?” We believe it is because they make comfortable trades instead of the
difficult ones. They are reluctant to buy breakouts because they feel it has already moved too far, so they
sit around waiting for the pullback to happen. When it does happen, they tend to get scared and will not
buy out of fear that the market will finally crash. If the pullback intensifies, they will finally feel like the
market has moved too far down and cannot go further and then they buy, which is typically the exact
wrong time. They take the comfortable trade instead of the right trade.
At AlgorithmicTrading.net, we determine “What’s the hardest trade to make?” and execute on that trade.
We focus our time, energy and resources on developing the trading strategy so we have a confident and
trust it once it goes live, and just let the trades play out without any emotions being involved. We simply
let the robotic trading system run our trading.
Algorithmic Trading Packages
The S&P Crusher Package
This package is our flag ship trading system, designed to maximize gain while also attempting to
minimize losses. This package is a combination of the ES Weekly Options & The Gambler packages.
The combined result (based on the back-testing) is what appears to be an extremely robust system. The
strength of this package lies in it’s ability to potentially out-perform in bull, bear and sideways moving
market conditions. When the market goes higher, the F1 Bull-Fire will place well timed swing trades on
the ES and the O1-Onightgap_sPuts algo will sell out-of-money weekly puts on the S&P Futures. When
the market is rebounding in a short covering rally, the B3-Breakout shines placing a day-trade in the
morning then ex-its at the close. During market sell-offs, the F1-BullFire & O1-Onightgap_sPuts algo are
de-signed to get on the sidelines while the S3-Breakdown places short day-trades, S2-Breakdown_
sCalls sells out of money calls and P2-PushPull takes a longer term bearish po-sition. During periods
of sustained sideways movement, the S2-Breakdown_sCalls and O1-Onightgap_sPuts algo sells
weekly options potentially adding to gains seen in the preceding directional periods. By combining
the ES Weekly Options package with The Gambler, the equity curve smooths out substantially without
sacrificing potential gains (lower back-tested drawdown). Trades the highly liquid ES and TY futures
markets (lower slippage) as well as weekly call and put options on the ES. At most, it could be long 1 ES
contract, 1 ES weekly call or put (not both) and 1 TY contract per $30k traded. A fully automated trading
system designed with the highest standards, this might be the best algorithmic trading system we have
ever designed. With that said, trading futures and options does involve substantial risk of loss and is not
appropriate for all investors. You should only trade our algorithms with risk capital. Read on to become
familiar with each of the six Algorithmic Trading systems, traded in this package.

F1 Bull-Fire
Key Features
• Trades the Emini S&P 500
Futures (ES) on 385 minute
increments
• Extremely effective during up
moving markets
• Can place swing and short
term trades, depending on
market conditions
Entry-Exit Points
• Enters long at 3:55 AM EST
if certain market conditions are
present
• Exits when either the stop or limit (target) is hit
Example Trade (Image, Above): This sequence shows 6 trading days (between 3/1/2016 and 3/8/2016).
During this period, we closed out five winning trades (the blue dotted line indicates a winning trade) in
the live account. It placed a swing trade first, followed by four shorter term trades. Past performance not
indicative of future performance.
P2 PUSH-PULL BOND
Key Features
• Trades the 10 Year
Note (TY) on 120 minute
increments
• Extremely effective
during down moving
markets (best back-tested
year was 2008.)
• Performs very well during all other market conditions
Entry-Exit Points
• Potentially enters at closure of 120 minute candles (10 AM EST, 12 PM EST, 2 PM EST, 4PM EST
or 4:59 PM EST) if certain market conditions are present.
• Exits when either stop or target is hit. (Can hold overnight.)
Example Trade (Image, Above): While the market was selling off in the early part of January-February

2016, the Push-Pull TY algorithm had an incredible run. This algorithm compliments the others very well
(i.e. while the equity markets are dropping, the Push-Pull algorithm will typically be hitting it out of the
park with winning trades, profiting during down moving markets).
B3 BREAKOUT
Key Features
increments
• Extremely effective during
down moving markets (captures
short covering rallies)
• Profitable during most other
market conditions as well
• This is a very low risk day
trade — in at the morning and
out at the close with a very tight stop). Uses a trailing stop once a certain price level is reached.
Entry-Exit Points
• Enters at 9:48 AM EST if certain market conditions are present
• Exits at the market close, unless stopped out
During this period, we closed out 1 winning trade on a short covering rally (the blue dotted line indicates
a winning trade) in the hypothetical account.

S3 BREAKDOWN
Key Features
• Trades the Emini S&P 500 Futures (ES)
on 9 minute increments
• Extremely effective during longer term
bear markets
• Great hedge against a sustained bear
market.
Entry-Exit Points
• Enters short at 9:48 AM EST if certain market conditions are present
During this period, we closed out a two winning trades (the blue dotted line indicates a winning trade)
in the hypothetical account. As demonstrated, shorting into the weakness was the correct trade. These
gains were huge, which contributed greatly to our incredible run in August 2015 when the equity markets
were selling off huge.
O1 OVERNIGHTGAP_sPUTS
Key Features
• Sells the weekly Puts on the S&P 500
Emini-Futures (ES)
• Extremely effective during up and
sideways moving markets
• Typically sells the puts trading 10-20
points out-of-money (approximately 1%)
Entry-Exit Points
• Enters at 3:59 PM EST if certain market
conditions are present, Monday – Thursday
• Attempts to buy back the option at 0.15 points
• Holds until options expiration on Friday. If option expires in-the-money, it will execute an order to
offset the option
On Mon-day (2/29/2016) this algorithm sold the 1910 Put collecting $475 in premium. This diagram
shows the full profit zone, partial profit zone and loss zone. The market rallied in our favor and we bought

back the option at 0.15 points just prior to expiration. Total gain on this trade factoring commission was
$443 per $20,000 traded.
S2 BREAKDOWN_sCALLS
Key Features
• Sells the weekly Calls on the S&P 500
Emini-Futures (ES)
• Extremely effective during down and
• Typically sells the calls trading 10-20
Entry-Exit Points
• Enters at 9:50 AM EST if certain market
conditions are present, Monday – Thursday
• Attempts to buy back the option at 0.15
points
offset the option
On Monday (2/8/2016) this algorithm sold the 1860 Call collecting $750 in premium. This diagram shows
the full profit zone, partial profit zone and loss zone. The market traded sideways and we bought back
the option at 1.00 points just prior to expiration. Total gain on this trade including commission was $675
per $20,000 traded or approximately 3.38%.

The Gambler Package
The strength of this package lies in it’s ability to potentially outperform in both bull and bear market
conditions. When the market goes higher, the F1 Bull-Fire will place well timed swing trades on the ES.
When the market is rebounding in a short covering rally, the B3-Breakout shines placing a day-trade in
the morning before exiting at the close. During market sell-offs, the F1-BullFire is designed to get on
the sidelines while the S3-Breakdown places short day-trades and the P2-PushPull takes a longer term
bearish position. It trades the highly liquid ES and TY futures markets (lower slippage) and does not
place any options trades. At most, it could be long 1 ES contract and 1 TY contract over the weekend
(per unit traded) which is an attempt to minimize overnight exposure. A fully auto-mated trading system
designed with the highest standards. With that said, trading futures does involve substantial risk of loss
and is not appropriate for all investors. You should only trade our algorithms with risk capital. Read on to
become familiar with each of the four Algorithmic Trading systems, which are traded in this package.
P2 PUSH-PULL BOND
Key Features
• Trades the 10 Year Note (TY)
down moving markets (best
back-tested year was 2008.)
• Performs very well during all other market conditions
Entry-Exit Points
• Potentially enters at closure of 120 minute candles (10 AM EST, 12 PM EST, 2 PM EST,
4PM EST or 4:59 PM EST) if certain market conditions are present.
Example Trade (Image, Above): While the market was selling off in the early part of January-February
2016, the Push-Pull TY algorithm had an incredible run. This algorithm compliments the others very well
(i.e. while the equity markets are dropping, the Push-Pull algorithm will typically be hitting it out of the
park with winning trades, profiting during down moving markets).

B3 BREAKOUT
Key Features
• Trades the Emini S&P 500 Futures (ES) on 9
minute increments
• Extremely effective during down moving
markets (captures short covering rallies)
• Profitable during most other market conditions
as well
• This is a very low risk day trade — in at the morning and out at the close with a very tight stop).
Uses a trailing stop once a certain price level is reached.
Entry-Exit Points
During this period, we closed out 1 winning trade on a short covering rally (the blue dotted line indicates
a winning trade) in the hypothetical account.
S3 BREAKDOWN
Key Features
increments
longer term bear markets
• Great hedge against a sustained bear market.
Entry-Exit Points
• Enters short at 9:48 AM EST if certain market conditions are present
During this period, we closed out a two winning trades (the blue dotted line indicates a winning trade)
in the hypothetical account. As demonstrated, shorting into the weakness was the correct trade. These
gains were huge, which contributed greatly to our incredible run in August 2015 when the equity markets
were selling off huge.

F1 BULL-FIRE
Key Features
• Trades the Emini S&P 500 Futures
(ES) on 385 minute increments
• Extremely effective during up moving
markets
• Can place swing and short term
trades, depending on market conditions
Entry-Exit Points
• Enters long at 3:55 AM EST if certain
market conditions are present
• Exits when either the stop or limit
(target) is hit
During this period, we closed out five winning trades (the blue dotted line indicates a winning trade) in
the live account. It placed a swing trade first, followed by four shorter term trades. Past performance not
indicative of future performance.

The ES Weekly Options Package
This package places at most one trade per week. Once we sell either a put or call, we wait for options
expiration on Friday. Trades can be placed on Monday, Tuesday, Wednesday or Thursday at either
9:50 AM EST or 3:59 PM EST. Options sold are typically 10-20 ES points out-of-money. The strength
of this package lies in it’s ability to potentially outperform in bull and bear market conditions as well as
sideways moving markets. When the market goes higher, the O1-Onightgap_sPuts algo will sell puts on
the ES Weeklies. During market sell-offs, the sPuts algo will attempt to get on the sideline while the B2-
Breakdown_sCalls algo begins selling calls. It trades the highly liquid ES weekly options (lower slippage)
and does not place any futures trades except to offset an assigned in-the-money option at Fridays
close. At most, it could be short either a call or put at any given time (never both). It will not hold an
option over the weekend in an attempt to minimize “Black Swan Event” exposure. Trade this package
as-is or in addition to any other packages we offer. Remember, trading options does involve substantial
risk of loss and is not appropriate for all investors. You should only trade our algorithms with risk capital.
Read on to become familiar with each of the two Algorithmic Trading systems, which are traded in this
package.
O1 OVERNIGHTGAP_sPUTS
Key Features
• Sells the weekly Puts on the S&P 500
Emini-Futures (ES)
• Extremely effective during up and
• Typically sells the puts trading 20
Entry-Exit Points
• Enters at 3:59 PM EST if certain
market conditions are present, Monday
– Thursday
• Attempts to buy back the option at
0.15 points
offset the option
On Mon-day (2/29/2016) this algorithm sold the 1910 Put collecting $475 in premium. This diagram
shows the full profit zone, partial profit zone and loss zone. The market rallied in our favor and we bought
back the option at 0.15 points just prior to expiration. Total gain on this trade factoring commission was
$443 per $20,000 traded.

S2 BREAKDOWN_sCALLS
Key Features
• Sells the weekly Calls on the S&P 500 Emini-
Futures (ES)
• Extremely effective during down and sideways
moving markets
• Typically sells the calls trading 20 points out-of-
money (approximately 1%)
Entry-Exit Points
• Enters at 9:50 AM EST if certain market conditions
are present, Monday – Thursday
• Attempts to buy back the option at 0.15 points
offset the option
On Mon-day (2/8/2016) this algorithm sold the 1860 Call collecting $750 in premium. This diagram shows
the full profit zone, partial profit zone and loss zone. The market traded sideways and we bought back
the option at 1.00 points just prior to expiration. Total gain on this trade including commission was $675
per $20,000 traded or approximately 3.38%.

The NQ Active Trader Package
The strength of this package lies in it’s ability to potentially outperform in both bull and bear market
conditions. When the market goes higher, the O2-OvernightGap will place long trades on the NQ
(NASDAQ Emini Futures). When the market is rebounding in a short covering rally, the B2-Breakout
places a day-trade in the morning before exiting at the close. During market sell-offs, the O2-
OvernightGap algo is designed to get on the sidelines while the S2-Breakdown places short day-trades
and the P2-PushPull takes a longer term bearish position. During sideways moving markets, the T2-
Burst algo will place trades during brief pull-backs and as the market rallies towards the upper end of
it’s range. It trades the ES (short), NQ (long) and TY futures markets and does not place any options
trades. At most, it could be long 2 NQ contracts and 1 TY contract over the weekend (per unit traded).
With that said, trading futures does involve substantial risk of loss and is not appropriate for all investors.
You should only trade our algorithms with risk capital. Read on to become familiar with each of the four
Algorithmic Trading systems, which are traded in this package.
UPDATE: This algorithmic trading package has reached it’s subscriber limit and is not available to
new users at this time. Feel free to visit the S&P Crusher, ES Weekly Options or The Gambler product
pages (all of which are still available).
Read the third party evaluation of our algorithms
T2 BURST
Key Back-Tested Features
• Trades the Emini NASDAQ Futures
(NQ) on 120 minute increments
sideways & upward drifting market
conditions
• Outperforms during down moving
markets
Entry-Exit Points
• Potentially enters at closure of 120 minute candles (11:30 AM EST, 1:30 PM EST, 3:30 PM EST or
4:59 PM EST) if certain market conditions are present.
Example Trade (Image, Above): This sequence shows a period where the market traded sideways with a
slight bias to the upside (10/13/2015-10/17/2015). The Burst algorithm timed the entries and exits very
well. We had 2 winners and no losers in this 7 day period in the hypothetical account (The blue dotted
line indicates a winning trade).

P2 PUSH-PULL BOND
Key Features
• Trades the 10 Year Note (TY)
down moving markets (best
back-tested year was 2008.)
• Performs very well during all
other market conditions
Entry-Exit Points
• Potentially enters at closure of 120 minute candles (10 AM EST, 12 PM EST, 2 PM EST, 4PM EST
or 4:59 PM EST) if certain market conditions are present.
Example Trade (Image, Above): While the market was selling off in the early part of January-February 2016,
the Push-Pull TY algorithm had an incredible run. This algorithm compliments the others very well (i.e.
while the equity markets are dropping, the Push-Pull algorithm will typically be hitting it out of the park with
winning trades, profiting during down moving markets).
B2 BREAKOUT
Key Features
• Trades the Emini S&P 500 Futures
(ES) or the Emini NASDAQ Futures (NQ)
• Extremely effective during down
moving markets (captures short
covering rallies)
• Profitable during most other market
conditions as well
• This is a very low risk day trade — in
at the morning and out at the close with a very tight stop). Uses a trailing stop once a certain price level
is reached.
Entry-Exit Points

During this period, we closed out 1 winning trade on a short covering rally (the blue dotted line indicates a
winning trade) in the hypothetical account.
S2 BREAKDOWN
Key Features
• Trades the E-Mini S&P 500
Futures (ES) or the Emini NASDAQ
Futures (NQ) on 389 minute
increments
• Extremely effective during up moving market conditions
• Outperforms during down moving markets
Entry-Exit Points
• Enters one minute before the market closes (3:59 PM EST) if certain market conditions are present
Example Trade (Image, Above): This sequence shows the month of October 2015. During this period, we
closed out 12 winning trades and only two losers (the blue dotted line indicates a winning trade, red dotted
line a losing trade) in the hypothetical account. As the markets rallied in October 2015, this algorithm hit it
out of the park; contributing to an amazing month with the algorithms.
CFTC RULE 4.41: Results are based on simulated or hypothetical performance results that have certain inherent limitations. Unlike
the results shown in an actual performance record, these results do not represent actual trading. Also, because these trades have
not actually been executed, these results may have under-or over-compensated for the impact, if any, of certain market factors,
such as lack of liquidity. Simulated or hypothetical trading programs in general are also subject to the fact that they are designed
with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profits or losses similar to
these being shown.
O2 OVERNIGHT GAP
Key Features
• Trades the E-Mini S&P 500 Futures (ES)
or the Emini NASDAQ Futures (NQ) on 389
minute increments
• Extremely effective during up moving
market conditions
• Outperforms during down moving
markets

Entry-Exit Points
• Enters one minute before the market closes (3:59 PM EST) if certain market conditions are present
Example Trade (Image, Above): This sequence shows the month of October 2015. During this period, we
closed out 12 winning trades and only two losers (the blue dotted line indicates a winning trade, red dotted
line a losing trade) in the hypothetical account. As the markets rallied in October 2015, this algorithm hit it
out of the park; contributing to an amazing month with the algorithms.
CFTC RULE 4.41: Results are based on simulated or hypothetical performance results that have certain inherent limitations. Unlike
the results shown in an actual performance record, these results do not represent actual trading. Also, because these trades have
not actually been executed, these results may have under-or over-compensated for the impact, if any, of certain market factors,
such as lack of liquidity. Simulated or hypothetical trading programs in general are also subject to the fact that they are designed
with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profits or losses similar to
these being shown.
Quality Control Processes
At AlgorithmicTrading.net we have implemented the following quality control mechanisms to monitor the
performance of the automated trading system and ensure its integrity to the best of our ability.
This includes the following cycle that continually repeats itself:

Monitor Live Returns
As time goes on and more live trades are placed, we continue to monitor the performance of the
algorithms and constantly compare profit factors, drawdowns and equity curves on each of the
trading packages. Live returns are posted on our website, normalized to a per unit basis.
Monitor Slippage
Slippage is monitored closely across our live accounts.
Monitor Auto-Execution Service
We closely monitor the live trading accounts that are setup to ensure trades are properly
executed (best-efforts).
Monitor Actual Returns Seen by Our Customers
The auto-execution brokers issue buy/sell orders such that the majority of the fills are at the same
price. We have noticed times where a fill might be slightly different. In general, the fills our customers
see appear to be the same (within reason). Tradestation customers are more likely to see different
fills, however even in these cases the fills are in our opinion well within reason. The exception to this
rule is if someone turns off an algorithm, or gets out early by manually overriding the algorithms.
Once our customers sign up, they have access to our online trading room where they can watch
each package trade in real-time in the tradestation simulated account. They can also monitor the
trades in their own account using the OEC iBroker smart phone app. This app alerts you every
time a new trade is placed. As you see trades getting executed in the trading room, you can cross
check with the actual trades in your own account.
Make Adjustments if Needed
When needed, we will provide updates to the algorithms. Updates are included as part of our
maintenance agreement. Updates are determined by the walk-forward analysis which uses an
out of sample period of approximately 1 year. This means that once per year, we may reoptimize
the algorithms and upload them to our customers’ tradestation accounts and the auto-execution
brokers.
Final Word
AlgorithmicTrading.net is a leading provider of high quality Automated Trading Systems to not only
professional CTA’s, but also retail traders. Our customers receive our full attention and we devote and
pride ourselves with customer service while sticking to our core competency of developing high quality
algorithmic trading systems. Our team is dedicated to providing our customers with the best algorithmic
trading system we can.
By using our automated trading system, our customers are able to remove their emotions from trading
allowing the algorithms to excel and potentially capitalize on short-term market inefficiencies to reap
profits. Since going live with the NQ Active Trader package (v2) back in March of 2015, we have done
very well. However, always remember that past performance is not indicative of future performance and

trading futures Involves substantial risk of loss and is not for everyone.
While no system is perfect and we cannot guarantee continued success, it is our expectation that we will
continue to do well moving forward and would love to answer any questions you might have.
Risk Disclosures
Futures trading has large potential rewards, but also large potential risk. You must be aware of the risks and be willing to accept
them in order to invest in the futures markets. Do not trade with money you cannot afford to lose. This is neither a solicitation nor
an offer to Buy/Sell futures. No representation is being made that any account will or is likely to achieve profits or losses similar to
those discussed on this website or on any reports. The past performance of any trading system or methodology is not necessarily
indicative of future results.
CFTC RULE 4.41 — Hypothetical or simulated performance results have certain limitations. Unlike an actual performance record,
simulated results do not represent actual trading. Also, since the trades have not been executed, the results may have under-or-
over compensated for the impact, if any, of certain market factors, such as lack of liquidity. Simulated trading programs in general
are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any account
will or is likely to achieve profit or losses similar to those shown.
This strictly is for demonstration purposes. AlgorithmicTrading.net does not make buy, sell or hold recommendations. Unique
experiences and past performances do not guarantee future results. You should speak with a CTA or financial representative
(broker dealer or financial analyst) to ensure that the software/strategy that you utilize are suitable for your investment profile,
before trading in a live brokerage account. All advice and/or suggestions given hereto are intended for running automated
software in simulation mode only. Trading futures is not for everyone and does carry a high level of risk. AlgorithmicTrading.net is
NOT registered as an investment adviser (nor any of its principles). All advice given is impersonal and not tailored to any specific
individual.
For more information or to schedule a live demo, visit
AlgorithmicTrading.net or see our full contact information below:
AlgorithmicTrading.net
702 W. Idaho Street
Suite 1100
Boise, ID 83702 USA
866.759.6546
Email: sales@algorithmictrading.net
Facebook: https://www.facebook.com/algorithmictrading.net
Twitter: https://twitter.com/Algos_Trading
Google+: https://plus.google.com/117873234813080912511/posts
YouTube: https://www.youtube.com/channel/UCvwEENi0gOJTWhOM1i2445w

Algorithmic Trading Basics: Strategies & Systems

More Related Content

What's hot (6)

Viewers also liked (13)

Similar to Algorithmic Trading Basics: Strategies & Systems (20)

Recently uploaded (20)

Algorithmic Trading Basics: Strategies & Systems