Calculate Lottery Effect in American Market & Build a simple backtest system to see factor performances
The trading strategy constructed in a relative term, that is, we long the stock with the factor value within top x% of whole data and short the last 5%, construct a zero-investment portfolio
analysis the extreme return factor, explain the possible logic base on prospect theory, present its performance analysis
A. Simple Back Test System:
three files in total. use the BackTestFrame.py to got the combined result
a. BackTest.py:
define class object back_test
initial: return, closeprice, factor三张表（横轴为公司序列，纵轴为时间timestamp）
functions: i. get trading candidates(the stock number)(see trading_candidate.csv to get a sample output）
ii. get the return yields
b. Performance.py
define class object performance
initial: the data from BackTest.py(a backtest variable)
returns: 1. get data performance (InformationRatio, Maxdrawdown, turnover, annualize yields）
2. (by choice) use the function yield_plot to get the plot of accumulated yield (***already set up the title&format***)
c. BackTestFrame.py
load the data used in backtest（make adjustments to the format of the data）
use functions in BackTest.py & Performance.py to get the backtest result
also, we define sharp ratio in this part（we didn't included it in the Performance.py because we need the data of risk free rate）
B. Output & Performance of the factor：
The related data of comps' financials
fm_regression.csv: factor data needed in fama-macbeth regression
trading_candidate(sample): the trading candidates data from backtest.py
analysize the data in statistcal methods，save the performance in .csv
since the performace result deviated in two periods (2005-2015 & 2015-2018), so we divided the data into 3 files: 2005-2018， 2005-2015， 2015-2018, record the result in each file
a. simple_sort.csv：decile sorting (one variable)
b. Mon_BM, Mon_EP, Mon_Illiq, Mon_IVol, Mon_SizeAll: bivarite sortings
c. alpha_beta.csv: alpha, beta & t values from fama-macbeth returns
d. fm_regression_result：fama-macbeth regression results
a. MAX_ts: MAX（The final factor we choice）time-serious curve of its value
b. Factor%i_tradingRatio=%n: the backtest accumulated yield curve at different trading rate