External Data and Marking criteria

Alphathon
#1

Hi, I just have 2 questions regarding the S&p500 competition.


First one is about external data:

From our understanding of the data we’re given, we’re only given price of the companies, and some information on a few indexes, but not the financial information of the companies, such as P/E, ROE etc. Are we able to obtain the data externally and use those data to determine our strategy? If so, is there anywhere showing us how we can import data (such as an excel) into the alphine system?


Secondly, we want to ask more specifically about the marking criteria, from our understand we’re marked 50% based on how much we outperform the Sp500 (meaning that the team with the highest % return will be ranked the highest?). We’re also want some clarification on the diversification part and the robustness part of the criteria. How is that part marked and what do they measure exactly.


Thank you.

#2

l>

  • The dataset for the S&P 500 competition is simple: 6 fields - “open_price”, “high_price”, “low_price”, “close_price”, adjusted close (we call it “bb_live”), and “volume” of stock data that were in the S&P 500 between 2007 to end of 2016. You can get these stocks identifiers (tickers) thanks to the function getTickersSP500(). It is in the data package in Python, or directly callable in R. You can also use exogenous data that reflect the macro environment. Use getTickersSP500Data() to retrieve these tickers. The only field available for them is “close_price”. No other data is allowed. You cannot import other data. Rather, you should try to create useful features from the dataset that is given.
  • 50% is based on your information ratio. Search for information ratio to learn more about it; it’s a standard metric for investment strategies / fund. Diversification will assess how far your portfolio is from a diversified portfolio. For example, if you decide to select 50 stocks amongst only 1 sector, your diversification score will likely be low. As per robustness, it assesses how likely your payout is to fail: how does it react when a stock drops out from the index? Can it handle a stock addition well? What if a stock price gets discontinued for any reason? Robustness reflects whether your portfolio is truly systematic, i.e. that it will generate allocations in the future, even if market structure or composition change. We’ll disclose some examples about these scores in the coming week.
    • </ol