Research Notes: Four Levels of Learning in Forex

Trading

Source: [Link] ([Link]

psychology-4-levels-of-competence-in-forex-trading-2024-08-09)

Origin: Based on Martin Broadwell's "four levels of teaching" (1969), adapted for
psychology and self-development.

Level 1: Unconscious Incompetence * Definition: You don’t know that you don’t
know. Ignorance of one's own lack of skill. * Trader Experience: Hear about Forex, think
it's easy (like stocks). Open account, make beginner's luck trades, then lose significantly.
Try random indicators (like SMA) without understanding. Realize wins are luck, not skill,
potentially ending in the red. No real learning occurs initially.

Level 2: Conscious Incompetence * Definition: You know that you don’t know.
Awareness of the skill gap. * Trader Experience: Realize Forex is difficult. Actively try to
learn: fundamental/technical analysis, market conditions. Experiment with indicators,
time frames. Use a trading journal to track results. Profile currencies and reactions to
news. Test different strategies to find one fitting personality and risk tolerance. Expect
many failures and learning from mistakes.

Level 3: Conscious Competence * Definition: You know that you know. Skill requires
conscious effort. * Trader Experience: Have profitable strategies but need concentration
and effort for consistency. Consult trading journal to choose plays. Have risk
management rules but struggle with emotional discipline (fear/greed). Focus shifts from
winning trades to consistent execution, understanding profitability despite losses.
Trading journal remains crucial for tracking execution consistency.

Level 4: Unconscious Competence * Definition: You don’t know that you know (skill
becomes automatic). Mastery, intuitive application. * Trader Experience: Trading feels
automatic, second nature. Pattern recognition is fast. Risk management (position sizing,
cutting losses) is effortless, almost subconscious. Deep understanding of the chosen
system, knowing when it needs adjustment. Rules become habits; execution guided by
"feel" or intuition. Consistent profitability is achieved naturally.

Important Note: Competence is not mastery. Reaching Level 4 is a significant step, but
becoming an elite trader requires continuous challenge, faster learning, and deliberate
practice.
Source: Optimus Futures ([Link]
of-futures-trading-competence/)

General Idea: Learning a skill often reveals more complexity than initially perceived.
Gauging competence in trading isn't just about P/L; it involves knowledge, skills, and
experience.

Hierarchy of Competence (based on Broadwell, 1969): * Level 1: Unconscious

Incompetence (Wrong Intuition) * Level 2: Conscious Incompetence (Wrong Analysis) *
Level 3: Conscious Competence (Right Analysis) * Level 4: Unconscious Competence
(Right Intuition)

Stage 1: Unconscious Incompetence * Definition: Not realizing how much you don't
know. Overconfidence despite lack of skill. * Trader Experience: May have beginner's
luck or learned a simple, non-robust method. Don't realize the vastness of required
knowledge or the hidden risks. Vulnerable because they "can't see what's often in front
of their faces." Analogy: "You wouldn't recognize a [fill in the blank] if it bit you in the
behind."

Stage 2: Conscious Incompetence * Definition: Seeing the entire field of knowledge

from the outside; realizing the gap. * Trader Experience: Often triggered by losses or
frustration. Willingness to go back to the drawing board. Involves multiple iterations,
failure ("fail forward"). See an "ocean of limitations" but start exploring specific areas
(e.g., focusing on swing trading while acknowledging other approaches exist). Aware of
what they don't know and what's left unknown.

Stage 3: Conscious Competence * Definition: Acting on what works and building on it.
Skill requires deliberate effort. * Trader Experience: Know how to execute well-
identified trades, contextualize setups, and act properly. Understand the boundaries and
limitations of their chosen methodology (e.g., using Fibonacci requires measuring
trends, calculating risk/return, following strict sequence). Know what they are good at
and actively expand skills. Execution is still somewhat mechanical, not fully internalized
or intuitive.

Stage 4: Unconscious Competence * Definition: Skill becomes second nature; intuitive

flow state. Baseline competence is far above average. * Trader Experience: Experience
accumulates to the point of automaticity. Can enter an intuitive flow. May still learn, but
baseline skill is high. Performance is superior; may follow procedures, but the sense of
the process is intuitive. Can focus on multiple angles (technical, fundamental, risk
management) simultaneously. Develop a sense for appropriate strategies (defensive/
offensive) and switch seamlessly. Execute trades without mechanically burdensome
effort; trading becomes "second nature." Finding what resonates with the trader's
natural inclination is key to reaching this level.

Key Point: Not every trader reaches Level 4. Finding what resonates personally is
crucial.

Source: Investopedia ([Link]

trading/092114/strategies-and-secrets-high-frequency-trading-hft-
[Link])

Overview: * HFT uses powerful computer algorithms for high-speed order execution
(fractions of a second). * Accounts for a significant portion of trading volume (e.g.,
~50-60% in the US). * Relies heavily on speed, low latency connections, and access to
fast market data (e.g., Nasdaq TotalView-ITCH, NYSE OpenBook). * Strategies and
methods are often considered trade secrets.

Types of HFT Firms: * Independents (Proprietary Trading): Trade with the firm's own
capital for its own profit. Most common type. * Subsidiaries of Broker-Dealers: Have
dedicated prop trading desks separate from client business. * Hedge Funds: Focus on
arbitrage and exploiting pricing inefficiencies across assets. * (Historically, investment
banks had HFT desks before the Volcker Rule).

HFT Strategies & How They Make Money: * Directional Trading (Very Short-Term):
Taking brief long/short positions based on anticipated price moves. Requires high
liquidity and advanced analytics to predict moves faster than others (e.g., order
anticipation). * Providing Liquidity (Market Making): Creating bid-ask spreads,
especially in high-volume, low-priced stocks. Earn fees/rebates from exchanges/ECNs for
providing liquidity. Helps narrow spreads and increase market efficiency. * Arbitrage
(Statistical Arbitrage): Exploiting temporary price discrepancies of the same or related
assets across different exchanges or markets. Relies on speed to capture fleeting
opportunities before the market corrects. * Types: Market-neutral, cross-asset, cross-
market, ETF arbitrage, latency arbitrage. * Structural Strategies: Capitalizing on market
structure weaknesses or speed advantages over slower participants (latency arbitrage). *
Momentum Ignition (Controversial/Potentially Illegal): Initiating trades to create a
brief price spike, attracting other algorithms, then profiting from the expected reversion.
Considered market manipulation if intent is deceptive. * Other Tactics (often part of
broader strategies): * Sniffing, Pinging, Sniping: Used to gain information on market
trends or other traders' intentions. * Quote Matching: Ensuring orders are competitive
for fast execution. * Spread Capturing: Profiting from the bid-ask spread.
Key Players (Examples): * Citadel Securities * Virtu Americas * G1 Execution Services
(G1 Trading) * Two Sigma Securities * Wolverine Securities * Jane Street Capital * UBS
Securities * Goldman Sachs & Co.

Risks in HFT: * Software Anomalies/Glitches: Can lead to massive, rapid losses (e.g.,
Knight Capital 2012 incident - $440M loss in 45 mins). * Market Volatility: Rapid price
moves can cause unexpected losses if algorithms don't perform as expected under
stress. * Compliance & Regulation: Certain practices (like manipulative momentum
ignition) are illegal.

Pros & Cons: * Pros: Detect more opportunities, speed, increased liquidity (adds
buyers/sellers, reduces spreads). * Cons: Potential unfair advantage (access to tech/
speed), increased volatility (algorithms triggering cascades, e.g., Flash Crash 2010),
trading errors.

Skills Needed for HFT Roles: * Advanced Quantitative Abilities: Strong background in
math, statistics, physics, computer science, engineering. * Programming Experience:
Crucial for developing/working with algorithms (Python often mentioned, C++ for speed,
potentially AI/LLM knowledge becoming important). * Analytical Skills: Analyze large
datasets quickly, derive conclusions. * Fast but Precise: Milliseconds matter; ability to
work quickly and accurately. * Risk Management: Understanding potential downside in
high-volume, rapid trading. * Often requires advanced degrees (M.A., Ph.D.) and relevant
experience.

Source: [Link] ([Link]

what-is-quantitative-trading/)

What is Quantitative Trading? * Uses mathematical computations and statistical

models to find opportunities. * Intakes large amounts of data to identify opportunities
faster than manual analysis. * Requires building a quantitative trading system (computer
models). * Needs institutional knowledge in finance, mathematics, and programming. *
Heavily relies on backtesting models on historical data to validate strategies. * Can
utilize High-Frequency Trading (HFT) for rapid execution.

Quantitative vs. Algorithmic Trading: * Quant: Uses more datasets, advanced math;
execution can be automatic or manual. * Algo: Typically relies on traditional technical
analysis, exchange data; execution is always automatic.

Why Use Quantitative Trading? * Potentially highly lucrative (but demanding, high
burnout rate). * Objective and data-driven compared to manual technical analysis. * Can
use diverse datasets beyond traditional exchange data. * Faster and more accurate way
to test models. * Can be used by retail traders (e.g., via APIs like REST) but often requires
institutional resources (computational power).

What is a Quant? * Trader basing strategy entirely on quantitative analysis. * Develops

and backtests systems. * High expertise in math, stats, programming. * Often work for
large financial institutions. * Roles: Traders (build/maintain models) and Researchers
(analyze data, find signals, repair models). * Career Path: Often higher degrees (Financial
Engineering, Quant Modelling), start as data analysts.

Quantitative Trading Strategies: * Mean Reversion: Assumes prices revert to long-term

trends. Identify divergences and trade the correction. Can apply to single assets or
correlated groups. * Trend Following: Identify and follow price patterns/trends. Buy
rising, sell falling. Can use various definitions of trend (e.g., price action, sentiment
analysis, momentum indicators like volatility/volume). * Statistical Arbitrage (Stat
Arb): Focuses on correlated assets (e.g., Coke vs. Pepsi). Determine a fair price
relationship/spread and trade deviations using HFT. Combines computer-calculated
analysis with rapid execution. Requires significant computing power. * Algorithmic
Pattern Recognition: Identify patterns indicating large institutional orders (e.g.,
detecting disguised orders spread across exchanges/dark pools) and trade ahead or
alongside them. Requires high computing power and HFT. * Sentiment Analysis: Uses
external data (social media, news reports) to gauge market sentiment towards assets
and trade based on predicted short-term price movements. Requires models (often
machine learning) to interpret and rank sentiment (e.g., bearish/bullish phrasing).

How to Build a Quantitative Trading System (General Steps): 1. Identify Strategy:

Define the core logic (e.g., mean reversion, trend following, arbitrage, or based on
hypothesis/data analysis). Determine parameters (stop-loss, take-profit, position size,
entry/exit). 2. Backtest Strategy: Simulate trades using historical data to see
performance. Crucial for validation. Use a solid platform, include costs, use accurate
data, avoid bias. If results are bad, alter or reject strategy. If good, optimize further. 3.
Implement Strategy: Need a system to automatically send signals to the broker (e.g.,
via API) or allow manual/semi-automatic execution. Often requires programming (C/C+
+, R, Python, MATLAB mentioned). 4. Risk Management: Critical element. Consider
capital allocation, position sizing relative to holdings. Manage human bias in model
development/management. Rigorous backtesting helps identify risks.

Pros & Cons of Quantitative Trading: * Pros: Analyze large data amounts quickly,
remove emotion/bias from execution, potentially faster execution, rigorous backtesting
possible. * Cons: Models only as good as the quant making them, requires significant
expertise/resources, risk of unforeseen events/market changes affecting models,
potential for large losses if models fail (e.g., due to overfitting or flawed logic).
Source: QuantInsti Blog ([Link]
library/)

Overview: Python is widely used for algo trading due to its extensive library ecosystem.

Categories of Libraries Covered: * Fetching data * Data manipulation * Technical

analysis * Plotting and visualization * Backtesting * Machine learning

Fetching Data Libraries: * yfinance: Fetches data (historical price, fundamental, real-
time) from Yahoo Finance. Easy access for traders/researchers. * Alpha Vantage: Fetches
historical price and fundamental data via API (requires free API key). Also offers technical
indicator data (SMA, EMA, MACD, Bollinger Bands). * Pandas-DataReader: Extracts
economic data (FRED, Fama French, World Bank). List of sources available. * IBridgePy:
Wrapper for Interactive Brokers API, allowing Python to call IB's C++ API directly for data
and trading.

Data Manipulation Libraries: * NumPy (Numerical Python): Efficient numerical

computing, handles large datasets, math operations, multi-dimensional arrays/matrices
(N-dimensional arrays, math functions, vectorized ops, broadcasting, random numbers,
linear algebra). * Pandas: Data manipulation and analysis for structured data. Provides
DataFrame and Series. Handles missing data, data manipulation, vectorized operations.

Technical Analysis Libraries: * TA-Lib: Open-source library for technical analysis using
indicators (RSI, Bollinger Bands, MACD, etc.). Helps in strategy creation.

Plotting and Visualization Libraries: * Matplotlib: Plots 2D structures (graphs, charts,

histograms, scatter plots). Functions include scatter , pie , stackplot , colorbar . *
Plotly: Interactive data visualization, builds on Matplotlib. Creates interactive charts/
plots. Packages: plotly (main), graph_objs (figure templates), matplotlib (supports
matplotlib figures). Often used with cufflinks for seamless plotting from Pandas
DataFrames.

Backtesting Libraries: * Backtrader: Open-source library for backtesting, strategy

visualization, and live trading. Simplifies complex backtesting components. * vectorbt:
Designed for backtesting, optimizing, and analyzing strategies. Leverages NumPy/
Pandas for efficiency, suitable for large data/complex strate (Content truncated due to
size limit. Use line ranges to read in chunks)