Tuesday, May 26, 2015

New Views for a Short Market Week

*  In one branch of my analytic work, I break the market down into five factors that potentially drive near-term returns: momentum, breadth, sentiment, correlation, and volatility.  The above regression model is derived from 2013 - present data and predicts next five-day returns in SPY.  Interestingly, volatility and momentum have been the significant drivers of the model.  While the other factors are correlated with future returns, they are not uniquely so--i.e., they overlap one another quite a bit.  I find it interesting that the model is more bearish now than at any time since the start of 2015 and has been deteriorating through the year.  When the model has been negative, the next five-day return in SPY has been a loss of -.25%.  When the model has been positive, the next five days have averaged a gain of +.57%.  That being said, the model accounts for far less than half of the total variance in forward prices, which means that there is more price movement *not* accounted for in the model than actually predicted.  This kind of work gives one a healthy respect for the role of randomness in near-term returns.

*  One consequence of the gender skew in trading:  Many of the traits we think are correlated with trading success actually predict sub-par performance and we fail to cultivate the traits that actually are associated with positive performance.

*  Ivanhoff Capital takes a look at a candidate to be the next Chipotle.

*  Changing your brain patterns and more great reads from Abnormal Returns.

*  Steve Burns on focusing on signal and tuning out noise.

*  See It Market tracks unusual options activity among stocks.

Have a great holiday-shortened week!

Brett