Copia Macro semaforo

Peer Review

Strategy R&D

MACROPROCESS TEQUILA CAPITAL

Benchmarking

DEVELOPMENT

SALES

Development

Quality

Requirement

Release

Research Ideas

RESEARCH

DASHBOARD

DATA SERVICES

BUILDER

Paper Trading

Long Shots

Quick Hits

PRE OPERATIONS

OKRs/ KPIS

Benchmarking

OPERATIONS

Committee

PM Committee

Pre-approved models

Portfolio analytics

EXECUTION MANAGER

*

*

*

*

*

*

Committee

1- 2 moNTHS

1-2 MONTHS

3-6 MONTHS

Long Shots

Quick Hits

2-3 MONTHS

1-2 WEEKS

2-4 weeks

Blowoff valve

Approved models

Blow off valve

Knowledge base

PROCESS

Peer Review

Strategy R&D

Benchmarking

Research Ideas

Long Shots

Will usually take from 3-6 months to develop and are expected to have a meaningful impact on the business

Quick Hits

Will usually not take more than 1-3 months to implement. Defined as incremental improvements over existing components

Committee

*

Committee

Blow off valve

Approved models

Knowledge base

SALES

Middle Office/ Back Office

Customer Service

DASHBOARD

OPERATIONS

Customer Acquisition

Marketing

Portfolio Management

data services

• Alternative Data
• Sentiment and factor libraries
• Market Statistics
• Momentum Indicators
• Factor

• Sentiment Score
• Analyst Ratings
• TAM

• Data Feed
• API
• Excel Add-in

TO BE INCLUDED IN TITANIA 4

• Score AI (AI)

• Scoring Machine (AI)

Purpose:

• Build new strategies based on momentum, risk, valuation, profitability, growth, balance sheet strength, historical performance and absolute/relative sector performance.
• Potential method to determine weights for individuals stocks, stock selection
• Enhance ETF Plus method

Builder

• Elegible Universe

Defines the universe of assets that can be included. Possible filters include: 1. Eligible regions/countries 2. Market cap or liquidity based minimums 3. Asset Classes 4. Security types (common stock, leveraged ETFs, etc.)

• Constraints - AUMs

Types of constraints include: 1. Minimum or maximum sector and industry allocations 2. Maximum or minimum weight allowed per individual security 3. Risk constraints (VaR, maximum drawdown, etc.)

• Backtester

• Fractal 0 (AI)

• Cash Manager

• Fractal 1 (AI)

• Dynamic Weights

• Scenarios & Market States (AI)

• Rotator (AI)

• Dynamic Rebalancing (AI)

• Seed/Base Strategies (AI)

TO BE INCLUDED IN TITANIA 4

• PM Solutions (AI)

• Strategic Rebalancing

Execution manager

• Account Management

Allows portfolio managers to setup new client accounts and define what investment strategy will be implemented. Allows to define how orders will be sent (market, VWAP, etc.), time at which orders will be sent, broker, etc.

• Portfolio and Order Management

Gives access to portfolio managers so that manual interventions can be done to a portfolio before orders are sent (includes possibility of liquidating portfolio). Also allows PM´s to review buy/sell amounts, asset allocations, etc.

• Paper Trading

• Back / Middle Office

• Smart Trader (AI)

TO BE INCLUDED IN TITANIA 4

PORTFOLIO ANALYTICS

• Risk Dashboard

• Attribution Dashboard

• Performance Dashboard

Conciliation Titania -Broker

Corporate events Monitoring

DEPARTMENTS

OPERATIONS(BO / MO)

BROKER

AUMS

Investment Flow Complete

SALES

Send orders confirmation

Cash / Position Validation

Client Reporting

FRONT OFFICE(CIO / PM)

Titania Portfolio (Strategy Selection & Asset Allocation)

Purpose: Use or create new seed strategies to build customized base strategies with different DNA.Current Capabilities: 1. Use of all 700+ strategies currently modeled in the platform2. Run new strategies by adapting eligible universe, constraints, factor selection etc.3. Definition of transaction costs, AUMs, etc.Future: Develop AI based capabilities that will allow for automated strategy simulation/generation to maximize a given criteria (Sharpe Ratio, CAGR, minimize drawdowns, etc.)

El modelo de rebalanceo dinámico consiste en la combinación de 4 puntos principales para lograr predecir el movimiento del precio de acciones individuales en la ventana de un día con operación VWAP, los puntos son los siguientes: 1. Análisis de los datos financieros OHLCV 2. Obtención de estados y escenarios con base en el ADN de cada acción3. Entrenamiento y parametrización de los modelos de forma autónoma con base en lo obtenido en el punto anterior. 4. Combinación de los diferentes modelos para predecir o apoyar sus decisiones y medir de forma confiable la probabilidad y la magnitud del movimiento esperado en el precio

STRATEGY R&DThe quant analyst starts developing the investment model in python, excel, etc. At this point he/she will start backtesting the proposed strategy to determine whether the stated thesis is able to generate alpha on a consistent basis.

The set of variables that optimize the results are: · Momentum 6 months · Free Cash Flow & Market Cap · Gross Margin · Average momentum (50d, 200d, 30wk) · Prices to Sales Ratio · Sales Growth (YOY Revenue Growth in the Buy & Hold dataset) · Volatility

Purpose: Select the best seed strategies included in a given base strategy. Methods: Based Rule (Historical CAGR) and Markowitz Approach.

1. Historical CAGR accumulates data from the entire history of a given strategy.
2. Markowitz uses 12 month rolling timeframe to estimate expected returns.

Future: Shallow networks for estimating expected returns. Use of scenario analysis/alternative data to improve estimations.

STRATEGY R&DOnce the Research Process is finished, the deployment team is tasked with integrating the strategy code into the production line.

New Strategy proposal

Strategy presentation

Research Phase

Data upload (if required)

Research & Development Strategy

Results presentation

Code Integration in production line

DEPARTMENTS

RESEARCH

DEVELOPMENT

INVESTMENT COMMITEE

Research Idea

Strategy ( development)

DATA TEAM

Peer Review

El modelo de rebalanceo dinámico consiste en la combinación de 4 puntos principales para lograr predecir el movimiento del precio de acciones individuales en la ventana de un día con operación VWAP, los puntos son los siguientes: 1. Análisis de los datos financieros OHLCV 2. Obtención de estados y escenarios con base en el ADN de cada acción3. Entrenamiento y parametrización de los modelos de forma autónoma con base en lo obtenido en el punto anterior. 4. Combinación de los diferentes modelos para predecir o apoyar sus decisiones y medir de forma confiable la probabilidad y la magnitud del movimiento esperado en el precio

Management and Research Committee

DEPARTMENTS

ANALYSIS

QUALITY

ResearchDataBusiness

DEVELOPMENT

Paper Trading Testing

Quality Assurance

Implementation

Release

DEVELOPMENT

DEVELOPMENT

Purpose: Privilige stocks with a higher score and expected returns vs a simple equal weighted approachMethods: Scoring Machine or Score AI (Presage)

Purpose: Improve success rate and generate alpha under any market conditions. Initiatives:

1. Stock Sentiment Scores (Market Psych)
2. Alpha Factor Library (S&P Global)
3. Market cap vs equal weight across sectors
4. Sector performance relative to SPY Relative and absolute momentum historical distributions

Purpose: Take advantage of tactical ETF´s (tech, semis, healthcare, countries/regions, etc) when their expected return is higher than that of the base strategyMethods: Based Rule (Momentum based) and Markowitz Approach.

1. Momentum based rule uses relative momentum between base strategy and tactical ETF´s.
2. Markowitz uses 12 month rolling timeframe to estimate expected returns.

Future: Shallow networks for estimating expected returns. Use of scenario analysis/alternative data to improve estimations.

Investment Commitee:

The research team presents the results using the following files and formats.

• Strategy Presentation Template

• Strategy Scoring Template

At this point the investment committee will decide on whether the strategy should be handed over to the IT team for implementation, require further tests or reject the idea from moving forward

Purpose: Allow quant analysts to test any given strategyCurrent Capabilities: 1. Not available for quants. Needs to be runned internally by IT team2. Transaction costs, scalability considerations, constraints, AUM definition, netting of trades, sub-account method, multi-day trading, etc3. Incorporates all available methods (fractal 0, fractal 1, ETF method, Score AI, Scoring Machine, Dynamic Weights, etc.)Future: Allow quants to test any Python based strategies by providing a transparent tool that includes modules for all current capabilities so that they can be rigorously tested as realistically as possible

Purpose: Preserve alpha, minimize drawdowns and determines points of entry to use leverage. Main variables: Historical VaR, metrics around short/medium term VaR, moving averages, cost of leverage, etc. Methods: Cash Manager V3 Future: Use of machine learning to determine oversold/overbought conditions.

Back Office: Handles the administrative and support tasks, crucial for operational functionalities like settlements and record maintenance. Middle Office: Risk management, paying and receiving functions, ensure market data is always available and compliance duties

Purpose: Customize portfolios for new clients and backtest a mix of given methods/strategies in order to estimate past and future performanceCurrent Capabilities: 1. Can be constructed on a manual basis.Future: Define an investment objective based on the client's return/risk profile and allow for Titania to optimize the best possible combination of strategy/methods in order to achieve such a goal

10 Stages of Full Life Cycle Use Case

PEER REVIEW The strategy will be subjected to a second opinion from a qualified peer to objectively validate the model, methodology or results before they are submitted to the investment committee for approval.

RESEARCH IDEASGenerate new research ideas based on predefined priorities such as:1. Fractal Algorithms2. Scenario Analysis3. Risk Management / Alpha Perservation4. New Seed and Base Strategies5. Artificial Intelligence

Purpose: Allocate capital to best performing agents given a defined utility function. Equivalent to fractal methods using in the modern Titania. Features:

1. Daily evaluation of agents to determine who should receive the most allocation (ranking)
2. Several critera method (aggressive, balance, low vol) can be used to determine the ranking of a given agent

Customer registration

Customer Account

Add Subaccount

Order creation

Sending orders

Benchmarking

DEPARTMENT

RESEARCH

COMMITEE (PMs, CIO,CEO)

Portfolio Creation

OPERATIONS IT

Strategy Go /No Go

Strategy Monitoring

c) Graduationd) Re allocatione) Decomission

Approved & monitored Strategy

Purpose:

1. Build new strategies based on momentum, risk, valuation, profitability, growth, balance sheet strength, historical performance and absolute/relative sector performance.
2. Potential method to determine weights for individuals stocks, stock selection
3. Enhance ETF Plus method

INVESTMENT COMMITTEEThe applicant submits his/her idea (using the research framework) to the commitee which evaluates the proposal based on the following criteria:1. Estimation of time and economic resources2. Alignment with pre-established priorities, as well as an estimate of impact and viability of the initiative

Investment Commitee:

The research team presents the results using the following files and formats.

• Strategy Presentation Template

• Strategy Scoring Template

OUT OF SAMPLE BENCHMARKINGThe strategy is run by a third independent party on data observed after the end date of the backtest. If performance is consistent with backtest results, the strategy is promoted to the next stage. Will take place for as long as it is needed to gather enough evidence that the strategy performs as expected vs benchmark.

Investment Commitee:

The research team presents the results using the following files and formats.

• Strategy Presentation Template

• Strategy Scoring Template