Market History

Using the past to rationally predict the future.

U.S. Stock Market in 2014

The following is the S&P 500 chart in 2014.

The January 2014 small correction (6.1%) had no fundamental reason or bearish news. However, our model still predicted this small correction. Small corrections (6-7%) often occur without any news/bearish fundamentals. The slightly bigger correction (8-12%) tend to coincide with some bearish news.

In March 2014, there was a sudden downturn in U.S. economic data. This is why David Tepper went on CNBC in March and said “don’t be too heavily long stocks”. However, the S&P did not go down at all. U.S. economic data turned upbeat again very soon and the S&P started rallying. Tepper quickly turned bullish again on the U.S. stock market.

The S&P made a slightly bigger correction (9.8%) from September 19 – October 15. The initial leg of the S&P’s decline was tied to oil’s decline because energy is a sector of the S&P 500. However, there was no way to predict the S&P’s decline by predicting oil’s decline (i.e. a rapid fall in oil prices will lead to a small S&P correction). By September and October 2014, crude oil prices had only fallen moderately! Historically, the S&P will often go up when oil prices fall moderately.

In addition, the relationship between the S&P and oil is unpredictable. Sometimes, fast 10-20% declines in oil prices will coincide with S&P corrections. Sometimes, there will be no correlation between oil and the S&P (the S&P will ignore oil price drops).

After the S&P bottomed on October 15, oil prices started to crash even though the S&P surged. This leads to another problem with using correlation: you have no idea when the correlation will break.

However, there was a tighter correlation in September-October between the S&P 500 and XLE (energy sector ETF). The entire S&P correction coincided with XLE’s crash. XLE bottomed in mid-October (coinciding with the S&P’s bottom) even though oil kept falling after October 15. This demonstrates that even the correlation between XLE (energy sector ETF) and oil prices isn’t predictable.

Here we can see an interesting point. If you can predict a big XLE decline, you can predict a medium sized S&P correction. For example, XLE crashed in July 2008, August 2011, and September 2014. However, we have no way of predicting an XLE decline, so our model does not incorporate this factor.

Our model predicted this medium sized correction.

Oil prices kept crashing from December 2014 to January 2015. During this period, the S&P did not go down but instead swung wildly (volatility) in a big sideways range. Once oil stopped crashing in early February, the S&P started to make new all time highs.


Get blog posts delivered to your email

* indicates required