The scientific method applied to markets

You have reached the top of the ladder. You can read markets, code, backtest honestly, and execute a strategy safely. This final course is about the hardest and most humbling part of all: how a quantitative trader actually finds an edge worth trading, and how to tell whether it is real. Quantitative trading is not a set of secret formulas. It is the scientific method, pointed at markets, and its first lesson is that most of what looks like an edge is not one.

From a hunch to a hypothesis

A discretionary trader makes decisions by judgement: they look at a chart, weigh the news, and decide. A quantitative trader does something different. They turn a belief into a precise, testable rule, apply it mechanically to data, and measure the result. "Cheap stocks outperform" becomes a specific way to rank stocks by a specific ratio, held for a specific period, measured over specific data. The belief is now a hypothesis, and a hypothesis can be tested, confirmed, or, far more often, rejected. That shift, from a hunch you act on to a hypothesis you test, is the whole of what makes trading quantitative.

An edge is a number

In this world, an edge is not a feeling of confidence. It is a number computed from data, and the discipline is about producing that number honestly and deciding whether to believe it. Here is what an edge looks like, reduced to its essentials.

# A quant edge is not a feeling, it is a number computed from data. Here is a set
# of monthly returns from some strategy, reduced to the numbers a quant judges it
# by: the average, the variability, and the Sharpe ratio (average return per unit
# of risk). The rest of this course is about whether to believe a number like this.
import numpy as np

# Twelve monthly returns of an illustrative strategy (fractions, so 0.02 = 2%).
returns = np.array([0.02, -0.01, 0.03, 0.00, 0.015, -0.02,
                    0.025, 0.01, -0.005, 0.02, 0.005, 0.03])

avg = returns.mean()
vol = returns.std(ddof=1)                      # sample standard deviation
sharpe_monthly = avg / vol
sharpe_annual = sharpe_monthly * np.sqrt(12)   # annualise across 12 months

print(f"Months of data:         {len(returns)}")
print(f"Average monthly return: {avg * 100:.2f}%")
print(f"Monthly volatility:     {vol * 100:.2f}%")
print(f"Monthly Sharpe ratio:   {sharpe_monthly:.2f}")
print(f"Annualised Sharpe:      {sharpe_annual:.2f}")

print("\nThis single number, the Sharpe ratio, is the 'edge'.")
print("The hard question, and this whole course, is whether it is real")
print("or whether twelve months of luck produced it.")

Months of data:         12
Average monthly return: 1.00%
Monthly volatility:     1.62%
Monthly Sharpe ratio:   0.62
Annualised Sharpe:      2.13

This single number, the Sharpe ratio, is the 'edge'.
The hard question, and this whole course, is whether it is real
or whether twelve months of luck produced it.

The code takes a strategy's returns and boils them down to the number a quant lives by: the Sharpe ratio, the average return divided by its variability, which measures return per unit of risk. Higher is better, because it rewards return and punishes the risk taken to get it. This strategy shows an annualised Sharpe of 2.13, which in the industry would be considered excellent. And here is the first and most important question of the entire course: should you believe it? It was computed from twelve months of data. As the coming chapters will show with hard arithmetic, twelve months is nowhere near enough to tell a Sharpe of 2.13 from a lucky streak. The number is real; whether the edge behind it is real is an entirely separate question, and answering it is what quant research is.

Why this is the honest course

It would be easy to make this course a tour of clever strategies, and dishonest. The truth about quantitative trading is that it invites self-deception more than any other kind, because with enough data and enough tries you can always find a rule that would have made money in the past. Most published and whispered edges are exactly that: patterns found by searching, that mean nothing about the future. So this course gives equal weight to the strategies and to the statistics that expose the false ones, and it is honest about the odds. The people who make quantitative trading pay are large firms with data, teams, and infrastructure a retail trader cannot match, and the realistic quantitative path for most individuals is not a private strategy at all but the disciplined factor investing this course ends with. None of that makes the skill worthless. Learning to tell a real edge from a lucky one is valuable no matter what you do with your money.

What to carry forward

Quantitative trading turns a belief into a testable rule and measures it, replacing judgement with evidence, and an edge in this world is a number, above all the Sharpe ratio, return per unit of risk. The hard part is not computing the number but deciding whether to believe it, because a strong-looking result from a short history is usually luck, as the 2.13 Sharpe from a single year will turn out to be. This is the most honest course in the catalogue about how rarely edges are real and how closed the professional world is. The next chapter lays out the disciplined process a quant uses to stay honest, the funnel in which most ideas rightly die.