A Budding Romance: Finance and AI

Imagining the Future without The Past

History is not only a guide to the future, there literally is no imaginable future without it. Amnesic patients show just how profoundly our predictions are rooted in memory. When the psychiatrist Endel Tulving asked one patient, N.N., “What are you doing tomorrow,” he recorded the patient’s response:

The Upside of FedEx Has A Probability of X

"FedEx shares surged last week.... They’ve got another 15% to 20% upside in a year."

-Jack Hough, Barron's on March 19, 2016

Now before you go rushing off to buy FedEx shares, try filling in the blank:

There is a __% chance FedEx will rise another 15% to 20% in the next 1 year.

A.10% chance
B. 30% chance
C. 75% chance
D. I’m drawing a blank

The Future of Yahoo!?

"The path to success is never linear, right?"

-Marissa Mayer, Yahoo CEO

Marissa Mayer’s defense of Yahoo!’s track record: non-linearity.

That might not resonate with shareholders in a proxy-fight, but if you look at pretty much any stock data, you’d see Marissa is absolutely correct. Stock valuations are event-driven, complex, hard-to-model…non-linear systems.

Tesla’s Model 3

"I’m always to some degree wrong. And the aspiration is to be less wrong…trying to minimize the wrongheadedness over time. "

-Elon Musk

Elon Musk plans to unveil the every-man Model 3 on March 31st, and your reaction alone won’t move Tesla’s stock price. What really matters is how the market might react. Our forecast for Tesla shows a wide range of potential returns. One year out the stock has a 70% probability of landing within the range of up 42.5% and down -36.8%.

The ABC’s of Forecasting

"Weather forecast for tonight: dark. "

-- George Carlin

A forecast has 3 essential components:

A. Time horizon
B. Range of outcomes
C. Probability that an outcome will fall within the range

A lot of statements sound like forecasts because they use the future tense. But they are not meaningful statements unless you explicitly hear A, B and C. The silence can be deafening.

Don’t Sweat Crude Unless These 3 Things Happen

"We don't choose experiences, we choose memories. "

-Daniel Kahneman

Sorry for the click-bait subject line. To get to the point: the price of Crude is likely going to rip to the upside over the next year, unless we’re in a period like Jan 1985, Aug 1997 or Feb 2015.

EidoSearch shows 13 periods in history with similar conditions matching today’s environment. 9 of the 13 matches were up in the next year with an average return of 23%, a max return of positive 80% and a min return of negative 43%. The chart below shows the time periods and returns following these 13 matches.

LinkedIn and The Biggest Losers

"Unless you puke, faint, or die, keep going!"

-Jillian Michaels, Red Team trainer on The Biggest Loser

The word Loser sounds so harsh. Tableau, Linkedin, Splunk and Workday – all companies that we greatly admire – earned that dubious distinction on Friday. We’re just repeating what Bloomberg said. While most financial software vendors serve up backwards-looking comparative stats, EidoSearch looks forward.

Below is our projected price performance for Linkedin showing an average return of positive 1.6% in the next five days, with a 70% likelihood of landing within the range of up 8.0% to down -5.4%.

Search For Probabilities

Thinking in terms of probability is a central tenet of being a Superforecaster. Yet, when when it comes to projections, all too often forecasts are delivered in 30-second sound bites or a tweet. We reached out to EidoSearch, a forecasting firm that uses a patented pattern matching technology, to project event outcomes and find relationships in Big Data.

When examining data, EidoSearch President and Co-founder David Kedmey says investors should take a disciplined approach to discover a better understanding of future events.

The Tail Ends of the S&P

"I've experienced the highest of highs and lowest of lows. I think to really appreciate anything you have to be at both ends of the spectrum."

-John Elway

Super Bowl 50 is the Broncos’ eighth trip to the big game, all of those with Elway’s help, first as QB and now as GM. In January 1987, Elway, in a moment known in NFL lore as “The Drive,” led a 98-yard, game-tying touchdown drive against the Cleveland Browns. Following that game, Elway and the Broncos lost in Super Bowl XXI to the New York Giants. Talk about extreme highs and lows.

This week, we’d like to spend some time discussing the extreme tails of the S&P. While we typically focus our analysis on the projected average return and the probable range around it (e.g., the 5-day projected cone in the chart below), occasionally it helps to look beneath the aggregate statistics to study the improbable ends of the spectrum.

We Were Right! Right?

"The Most Ominous Snow Forecasts For New York City Were Right"

-Harry Enten, senior analyst for FiveThirtyEight

Central Park’s weather station recorded 26.8 inches of snow, surpassing the blizzard of 1888, and just 1/10th off from breaking the record set in Feb 2006. So if you’re under the age of 10, this was a storm of a lifetime.

With most of the city shut down, we had a little extra time this weekend to think about forecasting accuracy. As FiveThirtyEight commented, “New York City is getting far more snow than the forecast of 7 to 12 inches from the National Weather Service, which was issued less than 24 hours before the storm hit the city. It’s the perfect example of why forecasts that vary based on probability are so important, and why outlier predictions should not be dismissed.”