A Forecast for Netflix (NFLX)….and the Weather

C-3PO: “But sir, the possibility of successfully navigating an asteroid field is approximately 3,720 to one.
HAN SOLO: “Never tell me the odds.”

There is probably no profession on planet Earth more criticized than meteorology.  If they say it’s going to be 80 degrees and sunny on Sunday, and we plan a beach day and it ends up being cloudy and rainy, we are not happy campers.  We don’t consider the last 5 forecasts that the weather man might have nailed, it’s just the ones they don’t get right which stick in our craw.

Forecasting the weather has in fact become a much more precise science, but the modeling of it is challenging based on all of variables especially for longer range forecasts.  Per Wikipedia:

“The chaotic nature of the atmosphere, the massive computational power required to solve the equations that describe the atmosphere, error involved in measuring the initial conditions, and an incomplete understanding of atmospheric processes mean that forecasts become less accurate as the difference in current time and the time for which the forecast is being made (the range of the forecast) increases.”

Ok, so modelling all of the factors that go into a weather forecast is challenging to no surprise.  They are also opaque.  The average person has no idea what goes into a weather forecast and therefore we can’t apply our own opinion to validate or discredit the forecast.  We basically have no insight into the process which we generally do not like.

Although the process is not clear, Meteorologists are generally pretty good with their forecasts given their current limitations.  In my opinion the biggest issue is that they are not effective at managing our expectations!  They are taking a forecasting model built to generate a RANGE of likely temperatures and weather conditions and they are trying to provide a precise temperature and forecast to the public.

Wouldn’t it be great if the forecast told you, “This Sunday there’s 85% probability that the weather will be between 77 and 82 degrees and Partly Sunny.”?  That would allow us to apply our own logic about whether or not it made sense to pack up the car with the beach gear, or what to pack on a trip, etc.  Basically, we could add value based on how we wanted to use the forecast.

At EidoSearch we have simplified the modelling of forecasts for all “Big Data” problems like weather forecasts.  Our ground breaking technology doesn’t need models and data scientists, we simply search by example to provide a range of forecasts based on similar environments or patterns historically, e.g. weather patterns.  For now we’re focused on applying this technology and applying it to price and economic time series data to provide investors with probabilities to support their fundamental investment process.

We’ve validated the accuracy of predicting return probabilities going forward using actual return distributions historically.  Literally based on tens of millions of predictions.  Our clients, most of whom are fundamental to the core, are using it to enhance their process and profitability.  A few examples of questions we’re quantifying for clients through probabilities:

1.    We like the fundamental story, but is this a good time to build a position?

2.    The stock is up 30% in two months and at our original forecast price.  Do we lock in gains?

3.    What positions in the book have the biggest downside risk?

4.    Where are opportunities in our sector with a ton of upside potential we should do some work on?

Let’s provide a live example with Netflix (NFLX).  The stock was at $484 on September 10^th, closed at $448 on October 15^th, and dropped 20% a couple weeks ago on a bad quarter.  The price now sits at around $380.

We found 97 historical instances of the current trading pattern in Netflix across Consumer Cyclical stocks historically.  Per the chart below, the average return in the next 3 months is 10.8% and based on the RANGE of historical return distributions, there’s more than 3x the upside to downside and 85% probability the price will be above $351.30 in three months.
 

The below chart shows the skew of the 3 month forward returns of all 97 historical instances.  There are only 3 instances down more than 25% in the next 3 months, and 25 instances that are up more than 25%.

A range of historical outcomes and probabilities in the hands of Fundamental Investors that can incorporate them.  Have a great week!