These statistical charts shows the directional distributions for Tesla Motors, for up and down moves in its daily closing price, as well as the non-parametric autocorrelation in the same. In the initial chart below, for example, there were three trading days where the price change was <-15%. And there were four trading days where the price change was >15%.
The average for Losing tail is -2.4% (shows as 0.024 in the chart below), while the average for Gaining tail is 2.8% (shows as 0.028 in the chart below). Note that in order to balance the distributions, the median daily change of positive 0.1% was subtracted from all data so there is a slight adjustment for the net change.
We then reinforce this daily performance analysis with the non-parametric serial correlation of the equity market price, in addition to that of the S&P 500 to show how this is related. We see the low conditional directional autocorrelation for Tesla, across various streak levels. A non-parametric distribution fit was performed on the underlying daily count data, and the results are highly statistically significant.
The average for Losing tail is -2.4% (shows as 0.024 in the chart below), while the average for Gaining tail is 2.8% (shows as 0.028 in the chart below). Note that in order to balance the distributions, the median daily change of positive 0.1% was subtracted from all data so there is a slight adjustment for the net change.
We then reinforce this daily performance analysis with the non-parametric serial correlation of the equity market price, in addition to that of the S&P 500 to show how this is related. We see the low conditional directional autocorrelation for Tesla, across various streak levels. A non-parametric distribution fit was performed on the underlying daily count data, and the results are highly statistically significant.
Hey this is a cool post.
ReplyDeleteAlthough I was lost on the first image.
Is there another way you can depict this or perhaps an easier verbal breakdown?
Found this site on Bloomberg hooked to some of the posts. I will be sure to check back.
Thanks
Hi Srinivas, I appreciate your comment. The empirical data for the top chart reflects that the Tesla price pattern, over its history, is that they have had a large upward skew thusfar. This is reflected in their typical average daily move of +0.1%. That's a high median price change per day (e.g., imagine your savings account providing you this daily rate of return). And on days where the performance is greater than this, it is typically +2.8%. In contrast, on days where the performance is less then this, it is typically just -2.4%.
ReplyDeleteIt is clear that we are also still in this upward skew phase of Tesla's price pattern. For example, in the past month we have seen strings of trading days where the price change was consistently in the 2% to 4% range, each day. And very few days where we have seen a daily stock price fall in the -2% to -4% range.
This recent pattern will not last forever. At some point, perhaps later this year, it would invert to balance the more symmetrical pricing patterns that we should theoretically expect. You will recognize this at a time you begin to see a string of trading days where the price change is consistently -2% or -3% or worse, each day.
I also suggest that you visit these two blog articles: Autocorrelation in financial markets (http://statisticalideas.blogspot.com/2013/05/autocorrelation-in-financial-markets.html), and Conditional implied volatility (http://statisticalideas.blogspot.com/2013/06/conditional-implied-volatility.html). Thanks much.
The above comment is timely. Sounds like you were looking for a drop in the stock.
ReplyDeleteDo you think that the options traders are right in Elon Musk's companies? Options traders in Solar City see the stock MUCH lower than what it is.