“Bots print passive income. Just switch one on.”
A bot executes a strategy. It cannot conjure one. Anyone selling a reliably profitable bot for R500 a month would earn more by running it quietly themselves.
Secure delivery. Lifetime updates.
This is a complete beginner's handbook for algorithmic trading. Every concept is explained from scratch, every key claim is cited, and every recommended tool is free. It is built around Python and Freqtrade, includes a dedicated South African chapter, and opens with the honest statistics about how most traders actually fare.
Secure delivery. Lifetime updates. Educational material, not financial advice.
Every concept explained from scratch, with diagrams, evidence, free tools and a South African roadmap.
The handbook draws on the Journal of Finance, SSRN working papers and complete exchange records. Chapter 16 lists every source.
The handbook is built around: Freqtrade, Python, Docker, VS Code, Git, Binance data, QuantConnect, TradingView, MetaTrader 5, Telegram alerts, VALR, Interactive Brokers.
Sellers of courses, signals and bots have a financial reason to make trading sound easy. Independent researchers do not. That is why chapter 1 opens with their findings. Calibrated expectations are the biggest advantage a beginner can own.
Figure 1.1: What independent research found
Day traders in Brazil who persisted for more than 300 days and still lost money
97%
Chague, De-Losso and Giovannetti (SSRN)
Day traders in Taiwan who were not reliably profitable after fees, across 15 years of records
>99%
Barber, Lee, Liu and Odean
Retail CFD accounts that lose money, according to the brokers' own mandated disclosures
65 to 85%
FCA and ESMA risk disclosures
In the United States, the most active retail traders earned less per year than the market itself.
Barber and Odean, “Trading Is Hazardous to Your Wealth”, Journal of Finance
“No honest guide can promise you profits. What this handbook can promise is competence: in about four weeks you can have a real robot trading live market data with pretend money, and understand every line of it.”
From page 2 of the handbook
There is genuinely good news too. Algorithmic trading is the one retail approach with a defensible process. Rules are written down, tested against years of history, and measured before money moves. The robot does not make a bad strategy good. It makes your process honest, and honesty is what the losing 97% never had.
“Bots print passive income. Just switch one on.”
A bot executes a strategy. It cannot conjure one. Anyone selling a reliably profitable bot for R500 a month would earn more by running it quietly themselves.
“1% a day is a modest target.”
1% a day compounds to roughly 3,700% a year. The best hedge funds on Earth celebrate 20 to 30% a year. Daily percentage promises are fiction, always.
“More indicators means a smarter strategy.”
Every indicator is recycled from the same price data. Stacking ten of them mostly adds ways to fool yourself. It does not add information.
“AI can just generate a profitable strategy.”
AI writes the code superbly and conjures no edge. Use AI to build and explain. Use statistics to judge. The handbook shows how.
Many trading courses
This handbook
The 44 pages are organised into three parts, covering the foundations, the decisions and the practice, plus a reference shelf. Here is what you actually get.
Chapter 12 builds TrendFollowerV1, a real trend following strategy with a pullback filter, layered exits and strict risk settings. Every line is explained in plain English, along with the exact commands to backtest it and run it on live data.
class TrendFollowerV1(IStrategy):timeframe = "1h" # think once per candlestoploss = -0.03 # the seatbelttrailing_stop = Truedef populate_entry_trend(self, df, meta):df.loc[(df["ema_fast"] > df["ema_slow"])& (df["rsi"] < 40), # buy the dip"enter_long"] = 1
Chapter 7 ranks the serious free frameworks and justifies the winner. The whole learning stack costs nothing. Read it from the top down.
It covers FSCA licensed venues, VALR and Luno as rand ramps, FICA, SARS and exchange controls, plus the TFSA sanity anchor that protects you from yourself.
The roadmap runs from week 1 to year 1, and each stage has a measurable milestone. The month 6 live gate sets five conditions before any real money.
Every term is defined in one place, with over 50 entries. Chapter 16 traces the key claims to their sources, including SSRN, the Journal of Finance and official documentation.
Buy the first edition once and receive every future revision free. Markets, tools and regulations move, and your handbook moves with them.
These pages are drawn from the first edition and recreated for your browser. The tone, the diagrams and the code are what you will get.
Every diagram was drawn specifically for this handbook, from order books and candlesticks to drawdown maths and the bot's decision loop.
Plain English boxes translate every piece of jargon the moment it appears.
The full 44 page PDF is delivered securely after purchase, personalised and licensed to you.
Every concept explained from scratch, with diagrams, evidence, free tools and a South African roadmap.
About this handbook
This handbook answers one question as honestly as the evidence allows: what is the fastest realistic path into algorithmic trading for someone starting from absolute zero? Every term is defined the moment it appears, every claim that matters carries a citation, every tool recommended is free.
PART I
The foundations
ch. 1 to 5
PART II
The decisions
ch. 6 to 10
PART III
The practice
ch. 11 to 14
Reality check
One promise up front, because everything else depends on it: no honest guide can promise you profits. What this handbook can promise is competence, the skills that quant firms and fintechs pay for.
2.3: Where prices come from
The order book for 1 Bitcoin (example numbers)
The spread, a cost you pay
There is no committee that sets Bitcoin's price. Each buyer states the most they will pay, which is a bid. Each seller states the least they will accept, which is an ask. The gap between them is the spread, your first and unavoidable trading cost.
12.1: The plan in English before the code
“Only ever buy when the tide is rising. Within a rising tide, buy the dips and never the euphoria. Take profit at +4%. If the trade goes 3% against us, cut it instantly, no discussion.”
PART B: exits and the seatbelt
# the non-negotiable seatbelt stoploss = -0.03 minimal_roi = {"0": 0.04, "1440": 0.02} trailing_stop = True # winners can't become losers
Every strategy you will ever build is these organs, rearranged: a regime filter, a timing filter, layered exits and a seatbelt. Chapter 12 walks through all of it, then hands you the exact commands to backtest, audit and dry-run it.
39 more pages await.
Unlock the full first edition. It is delivered securely, licensed to you and updated for life.
Cover (1 of 5)
Chapters 1 to 5 build the mental model. Chapters 6 to 10 make the decisions. Chapters 11 to 14 do the work, and the reference shelf holds every definition and source. Expand any chapter to see inside it.
Most plans give you a schedule. This one gives you milestones you must actually meet before advancing, including the month 6 gate that decides whether real money ever enters the picture.
Week 1: a stage
Read chapters 1 to 5 twice, install the complete lab, and run your first backtest.
The milestone that must be real
A backtest table has printed on your machine, and you can define spread, slippage, expectancy and drawdown from memory.
Week 2: a stage
Take a Python for trading crash course, then build, backtest and audit your first real strategy.
The milestone that must be real
You can change an entry rule and roughly predict the backtest's response before running it.
Week 3: a stage
Read community strategies critically and split your data into in sample and out of sample sets.
The milestone that must be real
You can name the four horsemen of fake backtests and show how you checked each one.
Week 4: a gate
Start the dry-run with Telegram alerts, and start the trade journal.
The milestone that must be real
The bot runs unattended for 7 days and you can explain every trade it took.
Month 2: a stage
Build a second strategy family and learn hyperopt with walk forward guardrails.
The milestone that must be real
You have an experiment log with at least ten hypothesis driven tests, mostly failures you can explain.
Month 3: a stage
Complete the QuantConnect Boot Camp and port your trend logic to stocks or forex.
The milestone that must be real
You have tested the same idea on two asset classes and written one page on the differences.
Month 6: a gate
There are five conditions: three months of clean simulation, results inside the backtest envelope, a drawdown calmly survived, pessimistic cost modelling, and a signed plan.
The milestone that must be real
All five, or no live money. Failing the gate is the system working, not you failing.
Year 1: a gate
Run a portfolio of small, uncorrelated systems with monitoring and clean tax records.
The milestone that must be real
Your capital is preserved, your process is professional, and you can honestly answer where, if anywhere, your edge is.
There are no promises about markets here, because those would be dishonest. These are skills, and they are the kind that quant firms and fintechs pay for.
You can read candlesticks, order books, spreads and market structure without the mystique, and without anyone's secret setup.
You can define expectancy, slippage, drawdown and walk forward testing from memory, because every term was taught the moment it appeared.
TrendFollowerV1 dry-runs live market data on your own machine. Every line of it is yours, and every decision is logged.
You can spot the four horsemen of fake backtests, split data honestly, and hold your own best results in proportionate suspicion.
You size positions so any single outcome is boring, backed by kill switches, hard stops and a written plan you sign.
You know the FSCA licensed venues, the rand ramps, FICA and SARS, with a TFSA investing anchor running quietly alongside.
A handbook that promises competence instead of profits self-selects its readers. That's deliberate.
This is for you if
Skip it if
If that's you, keep your R799. Chapter 1 would only annoy you, because it names who profits from those promises.
Before
Trading content is a wall of jargon and contradictory YouTube advice.
After
You can explain spread, expectancy and drawdown with a pen and a napkin.
Before
Bots feel like magic, something rented from a Telegram group.
After
A bot you built yourself trades live data with pretend money, and you understand every line.
Before
There is no way to tell a real edge from a lucky backtest.
After
You test like a researcher, with out of sample splits, pessimistic costs and counted attempts.
Before
Risk is a vibe. You are either too scared to act or not scared enough.
After
You have a signed one page plan with hard risk rules, kill switches and a wellbeing clause.
There are no tiers and no subscription. You buy one complete handbook.
You buy the handbook once. There is no subscription, no upsell ladder and no advanced course waiting behind it.
Everything else in the stack is free. Python, Freqtrade, Docker and the market data cost nothing. The handbook is the only thing you will be asked to buy.
Read the preview before you decide. The preview section above shows pages from the first edition, so you can judge the standard for yourself.
The digital handbook
First editionDelivered to your inbox straight after checkout. Licensed to you. Updated for life.
Secure checkout hosted by Shopify. We never see your card details.
Clear expectations are part of responsible education. If your question is not here, email us and you will get a considered answer, not a funnel.
Yes. It assumes nothing. Chapters 1 to 5 build the vocabulary and the mental model before any tooling appears, and every term is defined the moment it first shows up. If you can follow a recipe, you can follow this handbook.
No. Python is one of the friendliest first languages, and chapter 12 walks through the entire first strategy line by line, in plain English. If you already code in any language, you will simply move faster.
You need nothing beyond the handbook for months. The data is free, every recommended tool is free, and Freqtrade's dry-run mode trades a pretend wallet against live prices without even needing an exchange account. Real money is optional, months away, and sits behind the month 6 gate.
No honest resource can promise that, and chapter 1 shows the published statistics on exactly this question. What the handbook promises is competence: a real bot you built and understand, an honest testing process, and risk discipline. Whether that becomes profit depends on finding a genuine edge, which most people never do, and the handbook says so plainly.
Freqtrade is free at every stage, its simulation needs no account, it ships tools that detect the classic ways backtests lie, and it has a large and active community. As for paid bots, chapter 1 explains the economics: anyone with a genuinely profitable bot would earn more running it quietly than renting it out.
No. One chapter is South Africa specific and covers licensed venues, FICA, SARS, the TFSA anchor and exchange controls. The other fifteen chapters apply anywhere. If you are in South Africa, that one chapter can save you weeks of confused research.
You get the complete 44 page first edition PDF. Checkout is handled by Shopify, which emails your download link as soon as payment clears, usually within a minute. Lifetime updates are included, so every future revision is free.
No. It is educational material only, and it is not financial, investment, tax or legal advice. Trading involves risk, including loss of capital, and the handbook is unusually blunt about that.
Four weeks from now
That is the honest promise. Not riches, but competence. Start with the map instead of the hype, and let the evidence do the selling.
Questions first? Write to hello@zerotoalgorithmictrading.co.za and you will get a clear, considered reply. There is no mailing list pressure.