Skip to main content

Learn quant finance by building the projects firms look for.

QuantDXB teaches probability, programming, markets and data through short, rigorous lessons with interactive charts and code you can run. Each lesson ends with exercises, and some grow into projects you can show an interviewer. It's made in the UAE, where good material for this is hard to find.

Each line is one simulated year of a stock's price. The bars count where the paths finished. That lognormal spread is what you integrate over to price an option, the subject of the option pricing lesson.

What you'll learn

Four areas, taught in order, because each one leans on the one before it.

Probability and statisticsAvailable now
Averages and confidence intervals, conditional expectation, random walks, Monte Carlo option pricing, Bayesian updating and false discoveries.
  1. How sure is an average?Free
  2. Conditional expectation
  3. Random walks and Brownian motion
  4. Pricing an option with Monte Carlo
  5. Bayesian updating
  6. Hypothesis tests and false discoveries
Programming for quantAvailable now
Python with NumPy and pandas, floating-point pitfalls and how to test numerical code. A C++ lesson for the parts that need to be fast is planned.
  1. Thinking in arrays with NumPyFree
  2. Floating point and testing numerical code
  3. Time series with pandas
Markets and tradingAvailable now
Order books and how trades happen, market making, options and their Greeks, delta hedging, execution costs.
  1. The limit order bookFree
  2. Market making
  3. Options and the Greeks
  4. Delta hedging
  5. Execution costs
Data and machine learningAvailable now
Fat tails and volatility clustering, backtesting without look-ahead bias, overfitting, pairs trading, and validating models on time series.
  1. Returns, volatility and fat tailsFree
  2. Backtesting without look-ahead bias
  3. Overfitting
  4. Pairs trading and cointegration
  5. Validation for financial machine learning

Projects you'll build

More projects from the lessons

  • Market-making simulator

    Quotes around a fair value, manages inventory risk, and measures whether you actually had an edge. It starts in the market making lesson.

  • Pairs trading strategy

    Tests two assets for cointegration, fits the hedge ratio on past data only, and trades the spread. It starts in the pairs trading lesson.

Monte Carlo option pricer

Simulate thousands of possible futures for a stock, average the option's payoff, and check your number against the Black–Scholes formula. The lesson walks through the maths and the code, then lists ways to extend it into something worth putting on a CV.

Start this project
The white line is the running Monte Carlo estimate and the shaded band its 95% confidence interval. Each tenfold increase in paths shrinks the band by about √10 ≈ 3.2.

Try the free lessons

The first lesson of every track is free. Each one has a piece here to play with, and the lesson explains what you're seeing.

Is your edge real?

A strategy that wins 55% of its trades sounds great, but a coin can do that for a while. Pick the true win rate and watch the error bar shrink as trades come in. The smaller the edge, the longer it takes to prove it isn't luck.

Open the free lesson: How sure is an average?
55%

Not yet distinguishable from a coin flip.

Trades
0
Observed win rate
–
95% interval
–
Trades a typical run needs
381

A loop, or one line

A daily return is each price divided by the one before it, minus one. Python can loop over the prices one index at a time. NumPy does it in one expression, dividing two shifted views of the same array. Run both and check they agree.

Open the free lesson: Thinking in arrays with NumPy
r = p[1:] / p[:-1] - 1

p the prices

100.00102.00101.00104.50103.20106.00

p[1:] every price but the first

102.00101.00104.50103.20106.00

p[:-1] every price but the last

100.00102.00101.00104.50103.20

r the returns

+2.00%-0.98%+3.47%-1.24%+2.71%

One expression computed all 5 returns at once, in NumPy's compiled code.

A market made of random orders

Traders post orders to buy or sell, cancel them, or trade at the best price on offer. This simulation does only that, with no strategy at all, and a spread, a moving price and market impact appear on their own. Send an order of your own and watch it eat through the book.

Open the free lesson: The limit order book

The chart shows the mid price, with each trade as a dot (green buys, grey sells, white yours). The order book shows offers to sell above and bids to buy below. The order flow is a simplified version of the zero-intelligence model of Farmer, Patelli and Zovko (2005).

When the risk model breaks

Risk desks set a daily loss limit that should be crossed only rarely, often assuming returns are normal. Real returns have fat tails. Here an AED 1,000,000 portfolio lives through fat-tailed years: compare how often the normal model's limit is broken with how often it promised.

Open the free lesson: Returns, volatility and fat tails
Confidence

The normal model says losses beyond AED 23,263 happen on about 2.5 days a year. This year so far: 0 in 0 days.

Normal model's limit
AED 23,263
Fat-tailed model's limit
AED 26,216
Breaches, all years
–
the normal model promised 0

Why QuantDXB exists

Trading and investment firms in the Gulf are growing, but the way into quant roles is still hard to find from here. Most guides assume you're in London or New York, at a university with a quant society down the hall and alumni already on a desk.

QuantDXB is the resource we wanted: structured, honest about the maths, and focused on the kind of work that gets you an interview.