Quant Quest is a hunt for the best quantitative trader. To participate, all that is required is an interest in financial markets and enough data science/ programming experience to allow you to translate your ideas into code.

What to do?

Solve our trading problems using any logic you like to develop a trading algorithm, back-test against historical data and submit your strategy.

We will provide you with a toolbox to develop and backtest your strategy, as well as pre-cleaned data and pre-built feature set, so you can rapidly iterate and backtest a variety of ideas.

At the end of the competition, we will run your code against test/live data to evaluate your performance

In the meantime if you want to practice, you can practice with the Auquan Toolbox, where you can write your own trading strategies and backtest with real stock data.

What can I win?

  • Get recruited by Optiver, Europe
  • Guaranteed investments and profit share in strategies that top the leaderboard:
    • Strategies that top the leaderboard (and meet the minimum score criteria) will receive a monetary allocation and traded by us in live markets.
    • Creator of these strategies will be awarded prize money in the form of performance fees of ten percent (10%) of Net New Profits based on the performance of their Strategy for a period of six months.

How will I win?

Broadly, we are looking for good prediction algorithm which generate stable returns with low risk and low correlation with the benchmark index. Our toolbox will provide the following metrics to judge your algorithm:

  1. Alpha: Measure of how your algorithm performs against the benchmark. Higher the alpha, the better a strategy is.
  2. Beta: Correlation of your algorithm’s performance with the benchmark. Lower the beta (the strategy is immune to swings in the market), the better a strategy is.
  3. Sharpe Ratio: Returns/Risk, measure of risk adjusted returns. Higher Sharpe is better.
  4. Annualized volatility: Lower volatility is better.
  5. Max Drawdown: The greatest loss suffered from a peak in returns to its subsequent low. Lower values are better.

If you want a detailed explanation of what these parameters are, follow our tutorial series!

The exact judging metric will be updated shortly, stay tuned!

The leaderboard is updated when you make a submission. If you submit several trading algorithms, we will take the one with the highest score. We will announce the winners at the end of the competition.

Where is the toolbox?

The details of the toolbox for the competition, how to use and sample code snippets will be available when the competition starts.

In the meantime if you want to practice, you can practice with the Auquan Toolbox, where you can write your own trading strategies and backtest with real stock data.

What can I trade?

We will make a list of datasets available shortly. Your algorithm must trade any of the products which are available in the dataset ONLY.

How far back does the data go?

We will make the range of datasets available shortly.

Anything else?

A few other considerations:

  • Your algorithm must evaluate in less than 10 minutes
  • Your algorithm must return the same result if it is run twice (deterministic)
  • Your algorithm should not use ex post common knowledge (e.g. ‘don’t trade in 2008’)
  • Your algorithm must complete a full backtest for the entire range of data available
  • Your algorithm must not violate or infringe on any applicable law or regulation or third-party rights. Any violation would mean immediate disqualification.

How can I prepare?

Why don’t you check out our tutorials?