 |
|||||||||||||||||||||||||||||||||
 |
|
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|||||||||||||||||
 |
|||||||||||||||||||||||||||||||||
|
Global Historical Data Solutions |
|||||||||||||||||||||||||||||||||
 |
 |
||||||||||||||||||||||||||||||||
 |
|||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||
| Survivorship
Bias-Free Database What is Survivorship Bias? Survivorship bias is the tendency for failed companies to be excluded from performance studies due to the fact that they no longer exist. This often causes the results of studies to skew higher because only companies which were successful enough to survive until the end of the period are included. Survivorship bias enters the system development process early in response to the seemingly harmless question, “Across which securities should I test my trading concepts?” This question is typically given very little thought. Often, it is resolved by acknowledging that an appropriate universe for testing is one that is defined by liquidity requirements:
The trader obtains historical price data, and perhaps supporting fundamental data, for these stocks and proceeds to test his or her trading strategy. Once refined, the models are then executed in real-time across the same static symbol set. After some period of time the trader decides to reoptimize or modify the rules of the trading strategy. More often than not, the trader goes back to the same universe, perhaps updated for current prices, and repeats the process. After many years of working with individual and institutional traders, we find the practice described above continues to be dominant. Traders define research universes based on the liquidity measures of actively traded securities. The measures most often used are market capitalization (or membership in a market index where membership itself is based on market capitalization) or some measure of money flow (i.e. price x average volume). The most common practice is to define a research universe as constituents of a market index, the S&P 500 Index® or the Russell 1000 Index ®, for example. The Solution: A Bias-Free Database Tick Data developed the Survivorship Bias-Free Database to help traders avoid this pitfall. As an add-on to our US Equities Data, this database includes the symbols that were components of the Russell 3000 at some point since January 1, 1993, but are no longer traded due to corporate actions and/or bankruptcies. Related Links
Questions? Please contact us. |
|||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||
