Hypothesis eps china abhas - kitzz03/WorldQuant-Alphas GitHub Wiki
Alpha
x =rank(ts_delta(mdl175_spe, 63) / ts_std_dev(ts_delta(mdl175_spe, 63), 126))ts_rank(sales,63)(ts_rank(-debt_lt,63));
y= regression_neut(regression_neut(x,ts_delta(close,378)),ts_mean(log(volume),20));
signed_power(y-0.5,5)
Hypothesis
Companies with increasing earnings per share adjusted to their volatility will continue to generate outstanding profits and outperform the market.
Datasets used:
mdl175_spe: Earnings per share, a measure of company's profitability
sales: net sales per turnover
debt_lt: lost term total debt
Initial Alpha
rank(ts_delta(mdl175_spe, 63) / ts_std_dev(ts_delta(mdl175_spe, 63), 126))
taken change of eps over a quarter and normalized it using its standard deviation taking into account half a year. This causes us to select stocks with growing profits and lower volatility in terms of generating the profits
Improved Alpha
to make the idea more robust we tried amplifying the signal by selecting good companies growing in terms of sales and lowering their standing long term debts.
also we observed that stocks with increasing eps could be generating income because of increase in the stock market as a whole and also due to large volume of investors trading in the particular stock which could influence the overall sentiment regarding the company in the market boosting its income. therefore to remove the linear effects of long-term price changes and recent trading volume and isolate the effect of the fundamental momentum factor we neutralize the alpha with this market volatility. neutralised with price change of 1.5 years and mean log volume (log of volume so as to sober the effects of volume changes) over a month.
rank has range from 0 to 1 so to amplify the difference between good and bad companies we will apply signed power which further makes the alpha more robust, making stronger signals more pronounced while preserving the direction.
Answers to questions by judges:
why market neutralization - because we wanted to reduce the effects of a growing market
why decay so large = 15 - reduce the turnover from 15% to 8% which also improved fitness by a huge margin
why 1.5 years and not 1 year - ???
Links:
https://pubsonline.informs.org/doi/full/10.1287/mnsc.2023.4904
https://www.investopedia.com/terms/e/eps.asp
Backstory:
(optional)
some quote
something related to chess
some book