Algorithmic
Trading
Latest
Trends &
Developments
Dr. Lipa Roitman

Yaron Golgher

Founder

CEO
www.iknowfirst.com

© I Know First 2014. All rights reserved.
is a financial startup that
provides daily investment foresight based on
an advanced self-learning algorithm
We developed an advanced algorithm
based on artificial intelligence and machine
learning that also incorporates elements of
artificial neural networks and genetic
algorithms
Dr. Lipa Roitman, a scientist with over 20
years of experience, led our R&D team to
develop and consistently enhance the
algorithm

Our live portfolio from 2013 returned 60.66%
in 12 months beating the S&P 500 by 31.27%
A unique financial market forecasting algorithm that analyses, models and
predicts over 1,400 markets for short and long term:

Stocks
Commodities
ETF’s

Interest Rates
Currencies
World Indices

Firms that can consistently recognize the most opportunities and
overall trends has the key to the market
What is Algorithmic Trading? General
Market
Data
Market
Data

Market
Data

Algorithms
An Advanced Mathematical Model
Market
Data

Market
Data

Buy & Sell Orders
60% -70% of the US equity market volume
Two Methods of Algotrading,
High-Frequency and Quantitative Trading
1. High-Frequency Trading (HFT)
Real time intelligence, in milliseconds:
 Place and quickly cancel small orders to find at what price the
buyers and sellers are ready to trade.
 Price-volume info to catch developing trends.
 Simultaneously process volumes of information - human traders
can’t compete.
 Liquidate positions at the end of the day.

Technological costs of HFT are enormous.
High competition-low profit.
Two Methods of Algotrading,
High-Frequency and Quantitative Trading
2. Quantitative Trading, or Longer Term Trading
Algorithms analyze the structure and the trends in the market, find
predictable patterns, and trade upon the machine derived
forecasts.
Suitable for most investors
Some of the trading strategies:
Trend following vs. mean reversion:



When to use which?

Market neutral
Delta neutral
Arbitrage
Advantages of Algotrading

Human
Factors

Costs

Objective
valuation of
the stock

Psycholgical
Pressures

Risk

Algotrading

Quantitative
forecasting
the future
stock trend

Lower cost of
trading due to
the high
volumes

Volatility

Reduced
buy-sell
spreads, esp.
in most liquid
securities
Why Governments are Clamping Down on HFT?
HFT is unfair to retail investor



The HFT traders have the first choice in the trade - a form of
scalping

Level field needed to give everyone an equal chance.
Several European countries and Canada are curtailing or banning HFT
due to concerns about volatility and fairness.



In crisis the algos liquidate positions in seconds, causing huge
imbalances and price swings
HFT and Volatility
Has algotrading evolved in recent years to pose less risk on the general
market?

The risk is still present.



Notable examples:



May 6, 2010 Flash Crash. The algorithms may have
caused it, but also quickly corrected it.



January 23, 2013 AAPL plunge
Latest in Algotrading
Interpreting news and automatic trading by the
machines.
It’s all about speed!
It’s All About Speed!
Advertising campaign by Dow Jones on March 1, 2008:



Claimed that their service had beaten other news services by 2 seconds
in reporting an interest rate cut by the Bank of England.
What is Machine Learning?
Mathematics, statistics and logics are the crucial tools in studying
the markets.
They offer testable, verifiable and predictive hypothesis.
Number crunching allows finding hidden laws, not obvious to
humans.
Steps in Machine Learning

Provide
Framework
Mathematical
Tools
Programming
Tools

Give
Examples
To Learn From
Input
Output

Fitness
Function

Sequential

What should be An algorithm is a
optimized?
step-by-step
procedure
Example: Make
more good
predictions than
bad ones

Generalization
Requirement

Discover the laws
connecting the input and
output, cause and effect
Critical for forecasting
ability
Example Goal: Minimize the
Fitness Function
What is a Genetic Algorithm?
There are a number of search algorithms, from simple to complex, and
genetic is one of them.
Genetic algorithm is used for the most difficult problems, where exact
relationships are unknown, and maybe non-existent.
Many solutions are in the “gene” pool, some good, some not so.
Each solution is like a chromosome in genetics, hence the analogy.
Genetic is a circular iterative algorithm.
Genetic Algorithm

Reject

Gene
Pool

Select

Genetic
Algorithm

Mutate
Steps in Genetic Algorithm
Genetic algorithm uses these ways to improve the gene (solutions) pool:

Combination:
 Combine two or more solutions in hope of producing a better solution.
Mutation:
 Modify a solution in random places in hope of producing a better
solution.
Crossover:
 Import a solution from a similar problem
Selection:
 Survival of the fittest
A unique financial market forecasting algorithm that analyses, models and
predicts over 1,400 markets for short and long term:

Stocks
Commodities
ETF’s

Interest Rates
Currencies
World Indices

Firms that can consistently recognize the most opportunities and
overall trends has the key to the market
Loyal and Growing Client Base
Larger Institutions
Hedge Funds

Family Offices
Investment managers

– Fund manager & I Know First subscriber

Financial advisors

Professional investors

Hundreds Of Clients
Worldwide

I Know First grew 400% in
2013 from all over the world
Academic Cooperation
Dr. Roitman lecture in Tel-Aviv
University (View it Here)
Projects with Harvard Business
School
Partner with international
Universities-internship programs,
lectures
Market Trends

Transparency

S&P 500

Competition

There is more transparency than ever of fund
performance

To retain and attract new investors as well as other
mutual funds, a firm should be able to beat the S&P
500 on a regular basis

• Competition amongst investment firms is higher
than before
• To stay competitive investment banks are looking
for the most advanced tools to enhance their
performance
Customer Challenges

Complex

Market Evolution

The market is evolving beyond
previously established theories
however customers still expect
strong and consistent returns

of Traditional Tools and
Fundamental Analysis

Investment Firms

Traditional tools and fundamental
analysis are not enough to stay
competitive in the contemporary
market

Investment firms need to stay one
step a head in order to be the first
to recognize trends and take
advantage of opportunities
The Algorithm
Tracks and predicts
the flow of money
from one market or
investment channel
to another

Artificial
Intelligence
(AI)

The system is a
predictive model based
on Artificial Intelligence,
Machine Learning, and
incorporates elements of
Artificial Neural Networks
and Genetic Algorithms

I Know First
predicts 1,400+
investment
channels daily

Artificial
Neural
Networks

The results are constantly improving as the algorithm learns from its
successes and failures
Synopsis of the algorithm

Daily data is
added to our 15
years historical
file

Run a learning & prediction
cycle with new combined data.

Daily predictions for
each stock, currency,
commodity, etc..
Daily Market Heat Map

Two indicators:
Signal – Predicted movement of the
asset

Predictability Indicator – Historical
correlation between the prediction and the
actual market movement
XOMA returned 61.45% in
1 month from this forecast

Two indicators:
Signal – Predicted movement of the asset
Predictability Indicator – Historical correlation between the
prediction and the actual market movement
Forecast vs. Actual
I Know First Sample Portfolio

60.66% Return in 1-year beating
the S&P by over 30%

Click To View
Main Features of the Algorithm
Identifies The Best Market Opportunities Daily

6 Time Frames
Tracks Over 1,400 Markets
Self-Learning
Adaptable
Always Learning New Patterns
Scalable
A Decision Support System (DSS)
Predictability Indicator
Strong Historical Performance – 60.66% gain in 2013

The algorithm becomes more and more accurate with every prediction as it
constantly tests multiple models in different market circumstances
Algorithmic Trading Strategies
To Implement With Mutual Funds

Top
Stocks
Forecast

Currencies
Prediction

Interest
Rates
Forecast

Industry
Forecast
Customized
Algorithmic
Forecast

Dividend’s
Forecast

Gold
Prediction
Commodities
Prediction

Conservative
Stock
Forecast

ETF’s
Forecast

World
Indexes
Forecast

Aggressive
Stock
Forecast

European
Stock
Forecast
Algorithmic Trading
Strategies To Implement With
Mutual Funds
Assess Risk
Aggressive Stock Forecasts
Conservative Stock Forecasts
Assets That Carry A Dividend
Aggressive Dividend Forecasts
Conservative Dividend Forecasts
Recognize Top Performers In Each Industry
Bank Stocks Forecasts
Best Tech Stocks
International Opportunities
European Stock Forecast
Custom Forecasts
Customized Algorithmic

Forecast
Algorithmic Trading
FIVE Strategies To Implement
With Mutual Funds
Buy All Assets In The Forecast
Of Equal Weights
* Live Portfolio is based on this strategy *

Only Buy Stocks With High Predictabilities
* A predictability of .2 is good but .5 is excellent *

Buy Stocks That Have A Strong
Signal In Each Time Horizon
Multiply the Signal And The
Predictability Indicator Together
Identify New Opportunities and
Double-Check Your Analysis

Optimize Returns
&
Reduce Risk
Algorithmic Trading Tactical
Approach
The first appearance of a stock does not mean buy it at any price that
same day



Put it in a watch list, unless there is significant discount of at least
3%

Recognize the general color of the heat map
Consider the forecasts for major indexes to get an overall picture of the
market trend



We advise not to trade against
the general market trend

When analyzing stocks, review the specific industry
forecast as well
Algorithmic Output Chart: ALU
Apple Inc. AAPL Bubble Crash
Financial Bubble Detection
Algorithmic Output Chart: NOK
Two different types of algorithmic outputs
 Heat maps
 Charts
Algorithmic trading is becoming more popular as it has
proven more effective than traditional forms of analysis
alone.
 Algorithm’s are the future of financial analysis
Network Virtualization

S&P 500

Level of Confidence
 The self-learning algorithm not only gives you a
prediction but its level of confidence as well

Investing

Track record of regularly beating the S&P 500
 I Know First beat the S&P 500 by over 30% in
2013

Exchanges

1,400+ assets are forecasted

Forecasting

Key Advantages of I Know First:
Algorithmic Trading

39
Recent Publications
How Can We Predict The Financial Markets By
Using Algorithms? Tel-Aviv University Lecture –
Dr. Roitman
Seeking Alpha articles – Dr. Roitman
Seeking Alpha articles –I Know First Research
Algorithmic Trading With

Algorithmic Trading Latest Trends & Developments Lecture By Dr. Lipa Roitman

  • 1.
    Algorithmic Trading Latest Trends & Developments Dr. LipaRoitman Yaron Golgher Founder CEO www.iknowfirst.com © I Know First 2014. All rights reserved.
  • 2.
    is a financialstartup that provides daily investment foresight based on an advanced self-learning algorithm
  • 3.
    We developed anadvanced algorithm based on artificial intelligence and machine learning that also incorporates elements of artificial neural networks and genetic algorithms Dr. Lipa Roitman, a scientist with over 20 years of experience, led our R&D team to develop and consistently enhance the algorithm Our live portfolio from 2013 returned 60.66% in 12 months beating the S&P 500 by 31.27%
  • 4.
    A unique financialmarket forecasting algorithm that analyses, models and predicts over 1,400 markets for short and long term: Stocks Commodities ETF’s Interest Rates Currencies World Indices Firms that can consistently recognize the most opportunities and overall trends has the key to the market
  • 5.
    What is AlgorithmicTrading? General Market Data Market Data Market Data Algorithms An Advanced Mathematical Model Market Data Market Data Buy & Sell Orders 60% -70% of the US equity market volume
  • 6.
    Two Methods ofAlgotrading, High-Frequency and Quantitative Trading 1. High-Frequency Trading (HFT) Real time intelligence, in milliseconds:  Place and quickly cancel small orders to find at what price the buyers and sellers are ready to trade.  Price-volume info to catch developing trends.  Simultaneously process volumes of information - human traders can’t compete.  Liquidate positions at the end of the day. Technological costs of HFT are enormous. High competition-low profit.
  • 7.
    Two Methods ofAlgotrading, High-Frequency and Quantitative Trading 2. Quantitative Trading, or Longer Term Trading Algorithms analyze the structure and the trends in the market, find predictable patterns, and trade upon the machine derived forecasts. Suitable for most investors Some of the trading strategies: Trend following vs. mean reversion:  When to use which? Market neutral Delta neutral Arbitrage
  • 8.
    Advantages of Algotrading Human Factors Costs Objective valuationof the stock Psycholgical Pressures Risk Algotrading Quantitative forecasting the future stock trend Lower cost of trading due to the high volumes Volatility Reduced buy-sell spreads, esp. in most liquid securities
  • 9.
    Why Governments areClamping Down on HFT? HFT is unfair to retail investor  The HFT traders have the first choice in the trade - a form of scalping Level field needed to give everyone an equal chance. Several European countries and Canada are curtailing or banning HFT due to concerns about volatility and fairness.  In crisis the algos liquidate positions in seconds, causing huge imbalances and price swings
  • 10.
    HFT and Volatility Hasalgotrading evolved in recent years to pose less risk on the general market? The risk is still present.  Notable examples:  May 6, 2010 Flash Crash. The algorithms may have caused it, but also quickly corrected it.  January 23, 2013 AAPL plunge
  • 11.
    Latest in Algotrading Interpretingnews and automatic trading by the machines.
  • 12.
  • 13.
    It’s All AboutSpeed! Advertising campaign by Dow Jones on March 1, 2008:  Claimed that their service had beaten other news services by 2 seconds in reporting an interest rate cut by the Bank of England.
  • 14.
    What is MachineLearning? Mathematics, statistics and logics are the crucial tools in studying the markets. They offer testable, verifiable and predictive hypothesis. Number crunching allows finding hidden laws, not obvious to humans.
  • 15.
    Steps in MachineLearning Provide Framework Mathematical Tools Programming Tools Give Examples To Learn From Input Output Fitness Function Sequential What should be An algorithm is a optimized? step-by-step procedure Example: Make more good predictions than bad ones Generalization Requirement Discover the laws connecting the input and output, cause and effect Critical for forecasting ability
  • 16.
    Example Goal: Minimizethe Fitness Function
  • 17.
    What is aGenetic Algorithm? There are a number of search algorithms, from simple to complex, and genetic is one of them. Genetic algorithm is used for the most difficult problems, where exact relationships are unknown, and maybe non-existent. Many solutions are in the “gene” pool, some good, some not so. Each solution is like a chromosome in genetics, hence the analogy. Genetic is a circular iterative algorithm.
  • 18.
  • 19.
    Steps in GeneticAlgorithm Genetic algorithm uses these ways to improve the gene (solutions) pool: Combination:  Combine two or more solutions in hope of producing a better solution. Mutation:  Modify a solution in random places in hope of producing a better solution. Crossover:  Import a solution from a similar problem Selection:  Survival of the fittest
  • 20.
    A unique financialmarket forecasting algorithm that analyses, models and predicts over 1,400 markets for short and long term: Stocks Commodities ETF’s Interest Rates Currencies World Indices Firms that can consistently recognize the most opportunities and overall trends has the key to the market
  • 21.
    Loyal and GrowingClient Base Larger Institutions Hedge Funds Family Offices Investment managers – Fund manager & I Know First subscriber Financial advisors Professional investors Hundreds Of Clients Worldwide I Know First grew 400% in 2013 from all over the world
  • 22.
    Academic Cooperation Dr. Roitmanlecture in Tel-Aviv University (View it Here) Projects with Harvard Business School Partner with international Universities-internship programs, lectures
  • 23.
    Market Trends Transparency S&P 500 Competition Thereis more transparency than ever of fund performance To retain and attract new investors as well as other mutual funds, a firm should be able to beat the S&P 500 on a regular basis • Competition amongst investment firms is higher than before • To stay competitive investment banks are looking for the most advanced tools to enhance their performance
  • 24.
    Customer Challenges Complex Market Evolution Themarket is evolving beyond previously established theories however customers still expect strong and consistent returns of Traditional Tools and Fundamental Analysis Investment Firms Traditional tools and fundamental analysis are not enough to stay competitive in the contemporary market Investment firms need to stay one step a head in order to be the first to recognize trends and take advantage of opportunities
  • 25.
    The Algorithm Tracks andpredicts the flow of money from one market or investment channel to another Artificial Intelligence (AI) The system is a predictive model based on Artificial Intelligence, Machine Learning, and incorporates elements of Artificial Neural Networks and Genetic Algorithms I Know First predicts 1,400+ investment channels daily Artificial Neural Networks The results are constantly improving as the algorithm learns from its successes and failures
  • 26.
    Synopsis of thealgorithm Daily data is added to our 15 years historical file Run a learning & prediction cycle with new combined data. Daily predictions for each stock, currency, commodity, etc..
  • 27.
    Daily Market HeatMap Two indicators: Signal – Predicted movement of the asset Predictability Indicator – Historical correlation between the prediction and the actual market movement
  • 28.
    XOMA returned 61.45%in 1 month from this forecast Two indicators: Signal – Predicted movement of the asset Predictability Indicator – Historical correlation between the prediction and the actual market movement
  • 29.
  • 30.
    I Know FirstSample Portfolio 60.66% Return in 1-year beating the S&P by over 30% Click To View
  • 31.
    Main Features ofthe Algorithm Identifies The Best Market Opportunities Daily 6 Time Frames Tracks Over 1,400 Markets Self-Learning Adaptable Always Learning New Patterns Scalable A Decision Support System (DSS) Predictability Indicator Strong Historical Performance – 60.66% gain in 2013 The algorithm becomes more and more accurate with every prediction as it constantly tests multiple models in different market circumstances
  • 32.
    Algorithmic Trading Strategies ToImplement With Mutual Funds Top Stocks Forecast Currencies Prediction Interest Rates Forecast Industry Forecast Customized Algorithmic Forecast Dividend’s Forecast Gold Prediction Commodities Prediction Conservative Stock Forecast ETF’s Forecast World Indexes Forecast Aggressive Stock Forecast European Stock Forecast
  • 33.
    Algorithmic Trading Strategies ToImplement With Mutual Funds Assess Risk Aggressive Stock Forecasts Conservative Stock Forecasts Assets That Carry A Dividend Aggressive Dividend Forecasts Conservative Dividend Forecasts Recognize Top Performers In Each Industry Bank Stocks Forecasts Best Tech Stocks International Opportunities European Stock Forecast Custom Forecasts Customized Algorithmic Forecast
  • 34.
    Algorithmic Trading FIVE StrategiesTo Implement With Mutual Funds Buy All Assets In The Forecast Of Equal Weights * Live Portfolio is based on this strategy * Only Buy Stocks With High Predictabilities * A predictability of .2 is good but .5 is excellent * Buy Stocks That Have A Strong Signal In Each Time Horizon Multiply the Signal And The Predictability Indicator Together Identify New Opportunities and Double-Check Your Analysis Optimize Returns & Reduce Risk
  • 35.
    Algorithmic Trading Tactical Approach Thefirst appearance of a stock does not mean buy it at any price that same day  Put it in a watch list, unless there is significant discount of at least 3% Recognize the general color of the heat map Consider the forecasts for major indexes to get an overall picture of the market trend  We advise not to trade against the general market trend When analyzing stocks, review the specific industry forecast as well
  • 36.
  • 37.
    Apple Inc. AAPLBubble Crash Financial Bubble Detection
  • 38.
  • 39.
    Two different typesof algorithmic outputs  Heat maps  Charts Algorithmic trading is becoming more popular as it has proven more effective than traditional forms of analysis alone.  Algorithm’s are the future of financial analysis Network Virtualization S&P 500 Level of Confidence  The self-learning algorithm not only gives you a prediction but its level of confidence as well Investing Track record of regularly beating the S&P 500  I Know First beat the S&P 500 by over 30% in 2013 Exchanges 1,400+ assets are forecasted Forecasting Key Advantages of I Know First: Algorithmic Trading 39
  • 40.
    Recent Publications How CanWe Predict The Financial Markets By Using Algorithms? Tel-Aviv University Lecture – Dr. Roitman Seeking Alpha articles – Dr. Roitman Seeking Alpha articles –I Know First Research
  • 41.