Prerequisite: Quant 201.

Quant 101 and 201 were guided. This project is yours.

Sample student outputs (one possible question — yours will differ): Worked examples — use it to compare structure and tone, not to copy the hypothesis.

Module 1 — Your Own Research Project

Manager email

From: Elena Vasquez
Subject: One question — your choice

You have completed the guided modules. Now I want to see how you work when nobody picks the hypothesis for you.

Choose one narrow question about market data. Download or reuse prices from quant101/data/ (or add one new ticker if you need it). Produce:

1. A notebook with your analysis
2. One chart worth showing a client
3. A short memo in report/quant301_memo.md (400–600 words)

I am not grading cleverness. I am looking for clear thinking, honest limitations, and evidence you enjoyed the process — or did not.

— Elena

Why this matters

Junior quants spend much of their time on assigned tasks. The ones who advance learn to find good questions themselves: specific enough to test, important enough to matter, humble enough to answer with available data.

This project mirrors that moment.

Step 1 — Pick one question

Start narrow. Bad questions are too vague:

• "Will AI stocks keep going up?" (not testable with two years of data)
• "Can you beat the market?" (too broad)

Better examples:

Write your question at the top of notebooks/quant301_project.ipynb before writing analysis code.

Step 2 — State a hypothesis and a null hypothesis

Write them in this fixed form — not a free-text paragraph of hopes. (If you have done Data Science 201, it is the same habit; you do not need that course to continue here.)

Hypothesis

A precise claim about the data that you expect to be true, written so a colleague could disagree with it. Include the population / sample and the comparison.

Bad (vague expectation):
"I think NVIDIA will bounce after bad days because it is volatile."

Good (testable claim):
"After NVDA daily returns of −3% or worse, the mean next-5-day NVDA−SPY excess return is higher than on other days."

Null hypothesis

The default claim you test against — usually "no difference" or "no effect" in the quantity you measured.

Matching null:
"After NVDA daily returns of −3% or worse, the mean next-5-day NVDA−SPY excess return equals the mean on other days."

You may add one or two sentences of motivation after the formal statements (why the claim is interesting). Motivation is not a substitute for the hypothesis and null.

Your conclusion must say whether the evidence supports the hypothesis, rejects it (typically by rejecting the null at a stated threshold), or is inconclusive — and separate statistical significance from practical size.

Step 3 — Follow the workflow

Use the same structure as Quant 201 Module 5, but you define every step:

1. Question — one sentence (the human question).
2. Hypothesis — one precise testable claim (see above).
3. Null hypothesis — the no-effect / no-difference claim you test against.
4. Data — tickers, date range, source (Yahoo Finance CSV).
5. Method — exact rules or statistics you will compute.
6. Analysis — pandas/numpy code, clearly commented.
7. Chart — save to outputs/charts/quant301_main.png.
8. Conclusion — support / reject / inconclusive; quote numbers.
9. Limitations — sample size, one regime, no costs, etc.

Step 4 — Suggested notebook outline

Question

What are you testing?

Hypothesis

Write the claim in one or two sentences.

Null hypothesis

Write the no-difference claim in one or two sentences.

Code — load data

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

def load_returns(path):
    df = pd.read_csv(path)
    df["Date"] = pd.to_datetime(df["Date"])
    col = "Adj Close" if "Adj Close" in df.columns else "Close"
    df = df.sort_values("Date")
    df["return"] = df[col].pct_change()
    return df[["Date", "return"]].dropna()

# Example — change paths to match your question
series_a = load_returns("../data/NVDA.csv")
series_b = load_returns("../data/SPY.csv")

What the code does

Reuse the loading pattern from Quant 101. Consistency saves time and reduces bugs.

Add your own cells below — filters, groupbys, regressions, backtests, whatever your question requires.

Code — one summary chart

# Replace with a chart that answers YOUR question
plt.figure(figsize=(8, 4))
# ... your plot ...
plt.title("Your question in plain English")
plt.tight_layout()
plt.savefig("../outputs/charts/quant301_main.png", dpi=150)
plt.show()

Step 5 — Write the memo

Create report/quant301_memo.md:

To: Elena Vasquez
Subject: Research note — [your question in ~8 words]

1. Executive summary
2. Question
3. Hypothesis
4. Null hypothesis
5. Method (data + technique)
6. Results (numbers and chart reference)
7. Conclusion — support, reject, or inconclusive
8. Limitations
9. Would you want to investigate this kind of problem for a living?

Reflection questions

Add a final notebook cell or short section in your memo:

• Did you change your question halfway through? Why?
• What was harder — choosing the question or executing the analysis?
• Did the result match your intuition?
• After 101, 201 and this project, do you still find quant work appealing?

Manager feedback

From: Elena Vasquez

The best submissions I see at intern level are not the flashiest strategies. They are the ones where the analyst knows what they do not know.

If you finished this project curious to try another question — that is a good sign. If you finished relieved that nobody will ask you to look at markets again — that is useful information too.

— Elena

AQA Mathematics links

• Problem solving — defining and refining a question
• Statistics — hypothesis, null hypothesis, choosing an appropriate test
• Communication — structured technical writing

Beyond A-Level

Look up pairs trading and cointegration. Many student projects start with "Do these two stocks move together?" — the professional version uses formal statistical tests before betting money.

Module 2 — What You Have Learned

When you finish Quant 301, you will have learned to:

1. choose a narrow, testable question about market data;
2. state a precise hypothesis and null hypothesis (not a vague expectation);
3. work through the quant workflow without step-by-step guidance;
4. load and analyse price data using patterns from Quant 101 and 201;
5. produce a chart and memo that a portfolio manager could actually read;
6. conclude whether the hypothesis was supported or rejected, with honest limitations;
7. reflect on whether this career path fits you.

You have finished the Aurora Capital quantitative finance track:

• Quant 101 — Could I be a quant?
• Quant 201 — Becoming a quant
• Quant 301 — Your own research project

If this style of work felt motivating, explore mathematics, finance or computer science at university — and keep building notebooks that measure honestly.