Share New Research 
I got the idea from this article: https://quantdare.com/elo-system/ and decided to work on it in my free time.
For those of you who don't know, the Elo system was invented by Arpad Elo as a way to rank chess players. The idea behind it being that there is no true way to know the skill of a player, but that each player has a skill level. The information we do have access to is the games that the players have played.
This situation bears similarities to stocks in the stock market, we know that there are stocks that are "better" than others, but we don't have a real way of knowing. This algorithm sets each of the sectors against each other every week in a "matchup" in which the winner is decided by who had the higher return adjusted for volatility (I'm unsure on my implementation of this, definitely open to review there). These matchups are used to calculate the Elo rating of each sector. I use a low K factor here (2, much lower than the minimum 10 used by the FIDE for experienced players), and this is because I'm exposing all the sectors to a repeated amount of games every week and they have no choice in opponent, so I wanted to minimize the volatility of the ratings.
When you have the Elo ratings, you can calculate the expected value of a matchup, or as I understand it, the probability of victory. For this algorithm, I go long the highest rated sector, and short the lowest rated sector, and adjust the weights for volatility.
A few further things I want to look into with this system is the use of it with low or uncorrelated stocks/funds, using the ratings to build a weighted portfolio based on the ratings, and the idea of using other similar rating systems such as the glicko system.
Alexandre Catarino
Very interesing stuff!
It would make a great portfolio construction model.
Chris J.T. Auld
Looks interesting and I need to go and have a bit of a play. Have used Elo previously in automated sports wagering situations. Is a novel use. You might like to consider that Elo is predicated on 1:1 player matchups. When dealing with correlated assets it might be interesting to think of these as multi-palyer games and use something similar to TrueSkill. See
William Patterson
That looks really interesting Chris J.T. Auld , definitely going to take a look into that!
William Patterson
The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.
To unlock posting to the community forums please complete at least 30% of Boot Camp.
You can continue your Boot Camp training progress from the terminal. We hope to see you in the community soon!