Backtesting was used to identify the traits of stocks that are likely to outperform or underperform. Although historical performance does not guarantee future performance, it is always good to know what kinds of stocks are good buys and what stocks are better to avoid.
What is the data source for our analysis?
Reuters Datastream
What is the return measurement?
We use total shareholder return (TSR) which considers both capital appreciation and dividend income (reinvested).
What is the currency of your analysis?
U.S. dollar.
Have you considered investment delay?
Yes, we allow for one month investment delay.
How do you treat illiquid stocks?
For our analysis, we have already excluded those stocks that are ranked in the bottom 15% by market capital or whose past 1 year average trading volume is less than 10 thousand shares in any particular month.
How do you treat outliers in your backtesting?
If a stock's prior 6 month TSR is greater than 200%, we set the return equal to 200% for 6-month horizon analysis, 400% for 1-year horizon analysis, and 800% for 3-year horizon analysis.
Do you consider rebalancing?
No.
Have you considered potential transaction fees?
No.
Have you considered survival bias?
Yes, we have included dead companies in our historical analysis as long as they have positive trading volume and market capital as captured by the database.
How do you deal with possible data errors with the data source?
As shown by other analysts, the errors with Reuters Datastream normally happen to penny stocks. And we have excluded mini-caps and illiquid stocks from our analysis and backtesting.
0 comments:
Post a Comment