Alphien
  • LAUNCH A COMPETITION
    • QUANTITATIVE RESEARCHERS
      • What We Offer
      • How it Works
      • Your Benefits
      INVESTMENT MANAGERS
      • What We Offer
      COMMUNITY
      • Forum
      • Blog
      • Careers
  • |
  • LOGIN
  • SIGN UP
    • Register with LinkedIn
    • Researcher
    • Student
    • Investment Manager

Beat the S&P 500

Competition

Terms Of Use Privacy Policy Cookie Policy
Alphien Twiter Alphien Twiter Alphien Twiter Alphien Twiter
© 2022 Alphien All Rights Reserved

Beat the S&P 500

Only Registered users can join the competition

Competition Background

BACKGROUND


Introduction

In this challenge, your task is to create a long-short strategy on the S&P 500 futures that times the market correctly: use down-turns to your advantage and catch rebounds at the right time to outperform the market. You will have to demonstrate your data science skills by using one of the popular machine learning models in your payout. Will you be able to outperform the S&P 500 ?

OBJECTIVES


Description

Build a market timing long-short strategy on S&P 500 futures using only the S&P 500 futures building block series (ES), S&P 500 index levels (Open / High / Low / Close) and the VIX index. Your strategy has to use at least one data science model amongst gradient boosting trees, random forests, genetic algorithms, neural networks, clustering or any regularised regressions. You can add features to price series by using technical indicators (such as EMA, RSI, Stochastics, Bollinger Bands, etc), statistical models or any series you can derive from available inputs. Your strategy can rebalance on a daily basis at most and should not use any leverage (signals between -1 and 1). There are no transaction costs for this challenge.

Investment universe

Invest in S&P 500 futures building block ("ES"). Your payout can use S&P 500 index levels (Open / High / Low / Close) and the VIX index, as well as all features that can be computed from these series: technical indicators (such as RSI, EMA, Stochastics, Boolinger Bands, etc), statistics, time series models to name a few.

Backtesting period

All series are available from 1998. Start your backtest from 1 Jan 1998. You are allowed to use up to two years of data for the initial calibration of your model. Consequently, your first allocation should be on 3 Jan 2000 at the latest. In different words, your payout function has to output allocations that starts on 3 Jan 2020 at the latest.

Scoring criteria

Strategies will be ranked on its Sharpe ratio assuming a risk-free return of zero. We'll also test strategies are robustness by computing the minimum of the 3Y-rolling Sharpe Ratio over the backtest. Final rankings are an average of these two metrics.




AWARDS


Collect Alphien Points in order to unlock special features that can help you win a competition, redeem prizes, increase your codespace computing power, and reach the highest recognition amongst our quant community.

Rules regarding points:

  • Users can join as individuals or with a team.
  • Users can participate in any number of challenges at the same time.
  • Points will be distributed to all users (including all team members who joined as team) on a bi-monthly basis based on the final ranking at
    • The 15th day of the month. Use these points to increase your chances of winning at the end of the month by redeeming special features.
    • The final day of the month.
  • Participants can submit up to 2 (two) submissions for this challenge. They can make an extra submission with the submission wildcard (redeemed with points).
  • Only the best strategy of each user/team will be rewarded.

Points distribution schedule

User/Team Ranking Points awarded
Mid-Month End of Month
A 1 15 50
B 2 10 25
C 3 5 15


DATA


You can only invest in the S&P 500 index future building block (ES).
On top of this, you can use the VIX level at close and the S&P index levels to help you spot market turns:

Symbol Description
ES S&P 500 futures building block
SPX Index S&P 500 index levels (Open / High / Low / Close - fields are respectively called "open_price", "high_price", "low_price", "close_price")
VIX Index Implied 1M volatility on S&P 500 futures options (close_price)

Use the tutorial notebook to see how to retrieve the data set for this challenge.


GUIDELINES


Support tools

  • After you have joined the challenge, click the "get started" button. You will be redirected to your dashboard. In "Public Notebooks", filter notebooks by competition tag using the "Beat the S&P 500" tag.
  • Read about how to avoid look ahead bias to maximise your chances to be at the top of the leaderboard.
  • Use the forum to find help from the community or from Alphien's support team.

Tips and advice

  • Use AlphienStudio or AlphienLab to conduct your research.
  • Once you have fine-tuned your model, document it in a notebook and submit the notebook for our review.
  • We highly recommend you use the forum as much as you can to find help.

RULES


Terms and Conditions

[1] Prizes are non-transferable to other users or accounts.

[2] Alphien reserves the right at its sole discretion to alter, amend, modify, suspend or terminate this Competition, or any aspect of it, including but not limited to changing the contest duration or backtesting parameters, at any time and without prior notice. In such event Alphien will make reasonable efforts to notify all Participants by email.

[3] Alphien will make all reasonable efforts to support the acceptance of entries and the execution of backtests and paper trading. Entries may not be successfully processed or executed because of errors or failures. Those errors or failures may be caused by Alphien, the Participant, third parties, or a combination of parties. Alphien will not be liable or otherwise responsible for any entry that cannot be processed or executed, regardless of the cause of the problem.

[4] Alphien will make all reasonable efforts to execute the Winner's algorithm. The algorithm may not be successfully processed or executed because of errors or failures. Those errors or failures may be caused by Alphien, the Winner, third parties, or a combination of parties. Alphien will not be liable or otherwise responsible for any failure of the algorithm to be processed or executed, regardless of the cause of the problem. Furthermore, Alphien is explicitly not liable for any delays in the acceptance or transmission of orders due to a breakdown, failure of transmission, communication facilities or for any other technical error or cause. If the Winner's algorithm cannot be executed as planned as a result of errors, Alphien shall incur no liability and the prize will not be awarded.

[5] Alphien will have complete discretion over interpretation of the Competition Rules, administration of the Competition and selection of the Winners. Decisions of Alphien as to the administration of the Competition, interpretation of the Competition Rules and the selection of the Winners will be binding and final.

[6] Participant agrees and give his/her express consent for Alphien to use or publish without additional compensation in any medium (including, without limitation, in print, via television, via the internet, via email and or/other electronic form) and/or share with its agents, business partners and successors during and after the Competition period, information for publicity purposes including photographs, videotape or digital recordings that Alphien takes of Participant, Participant's Alphien username, public written statements, and algorithm performance for all or part of the Competition, without compensation. Participant hereby waives any rights, claims or interests that Participant may have to control the use of any or all of the publicity material in whatever medium used.

[7] The Winner is solely responsible for the payment of any and all taxes that may apply on their Generated Profits. Alphien shall have the right, but not the obligation, to make any deductions and withholdings that Alphien deems necessary or desirable under tax laws, rules, regulations, codes or ordinances.

LEGAL


Please refer to Alphien Terms of Use.