The Library and the Road

Every Academy school ends by handing you back to your own results. This one first hands you a library — because algorithmic trading, more than any other subject here, rewards going deeper, and punishes going deeper in the wrong order.

THE LIBRARY — read in this sequence, not alphabetically:

First: Ernest Chan — "Quantitative Trading" (and later its sibling "Algorithmic Trading"). The classic honest introduction for independent traders: what a realistic retail quant operation looks like, written by someone who ran one and tells you about the failures. Chan reads like this school's Chapters 3–7 with more mathematics and war stories. Start here.

Second: Robert Carver — "Systematic Trading" (and the gentler "Leveraged Trading" if the first feels heavy). A former institutional systematic trader on rules, position sizing, and — his obsession, and Chapter 5's — not fooling yourself. Carver's framework for combining simple rules into portfolios is Chapter 8's zoo, given full engineering drawings. The single most re-readable book on this list.

Third: David Aronson — "Evidence-Based Technical Analysis". The statistician's cold shower: how to test trading claims like a scientist, why most published patterns fail honest tests, and the full machinery behind our 'beauty is evidence against you' rule. Read it when you're strong enough to enjoy being told hard truths.

Fourth: Perry Kaufman — "Trading Systems and Methods". The thousand-page encyclopedia — every classic system family, documented like a parts catalogue. Not a bedtime read; a reference you'll consult for a decade. Buy it when your zoo needs new animals.

The advanced shelf: Marcos López de Prado — "Advances in Financial Machine Learning". THE serious text on Chapter 8's first territory — and notice what it's mostly about: defence. Leakage, false discoveries, backtest overfitting at industrial scale. Read it after you're fluent in Python and statistics, not before; earlier, it will only decorate a shelf.

(And one book from outside the field that belongs on the shelf anyway: Kahneman's "Thinking, Fast and Slow" — because Chapters 2, 6 and 7 of this school were secretly about its subject, and the operator is part of every system.)

THE FREE RESOURCES — quality over quantity: the QuantStart archive (hundreds of careful articles walking the exact path of this school, with code); the freely available Quantopian lecture series (a full quant curriculum in recorded lectures, still circulating years after the platform closed); SSRN and arXiv's quantitative finance sections for the actual research papers behind every 'secret' strategy ever sold (usually free, usually more honest than the product); and your broker/API documentation — unglamorous, and the difference between Chapter 6's Wall One being a story or a surprise.

THE ROAD — a realistic first year, in five stretches:

Months 1–2: statistics before software. Distributions, sampling, why small samples lie (Chapter 5 in mathematical clothes). Any good intro statistics course works. Every hour here saves ten later.

Months 2–4: one language, learned on your own data. Python is the field's lingua franca — but learn it by doing this school's labs with real data: download prices, compute an average, mark the crossovers. Toy projects teach syntax; your own market data teaches the job.

Months 4–6: backtest ONE simple system, honestly. Your Chapter 3 shopkeeper, through Chapter 5's full checklist, splits and all. Resist every urge toward system number two — the goal of this stretch is learning what honest testing feels like, and one system teaches that better than five.

Months 6–9: paper trade it. Chapter 7, stage one, gates pre-written. You'll spend this stretch discovering plumbing — which is precisely the education.

Months 9–12: tiny real money — and the journal verdict. Thirty-plus real trades at stinging-dinner size. And here every road in this Academy converges on the same destination:

The journal is the final judge. Your algorithm's trades enter your journal like any strategy's — tagged, sampled, measured — because a system's backtest is its promise, but only its live, tagged record is its truth. Win rate, expectancy, drawdown versus design, slippage versus assumption: the verdict machine you already know from every other school doesn't retire because a machine placed the orders. It matters more — it's the instrument that catches edge decay, regime change, and your own overrides, the three quiet killers of Chapter 6.

One last thing, and then the road is yours.

Somewhere in your first year, a losing month will arrive — the machine executing faithfully, the equity curve bleeding anyway — and you'll face this field's real final exam, which isn't mathematical at all: distinguishing a system that's broken from a system that's breathing. Everything this school gave you — the pre-written gates, the failure map, the honest backtest's expected drawdowns, the journal's evidence — exists for that one night.

The traders who last in this field aren't the ones with the cleverest algorithms.

They're the ones who built, in advance, the evidence to trust — and the humility to read it.

The library is open. The road is marked.

Drive like an engineer.

🇮🇳 The India Angle

• Add the local layer to the library: your broker's API documentation and the exchanges' algo-trading circulars are required reading before deployment — frameworks have evolved rapidly through the mid-2020s and will keep doing so.
• India's systematic-trading community is unusually active — broker-run developer forums, API communities and quant meetups are where the plumbing knowledge (Wall One's local edition) actually lives.
• Months 9–12's journal verdict has a natural home: tag your system's trades as their own Strategy in QbarTrade and let the strategy-filtered reports deliver the live-vs-backtest verdict — the same machine that judges your manual playbooks judges your robots.