(866)-482-2177

sales@alphatheory.com

REQUEST A DEMO

SYSTEM REQUIREMENTS


Please note the following System Requirements. Further, please limit the number of open applications (particularly price streaming applications) while logged in to Alpha Theory™.


Recommended System Specifications
Processor: Dual Core or Quad-Core 2.4GHz or faster
RAM: 4GB+
Browser: Google Chrome 30+
Screen Resolution: 1280 x 1024 or greater
Internet Access: Business Class High-Speed


Minimum System Requirements
Processor: Intel Pentium-M 2.0Ghz or equivalent
RAM: 2GB+
Browser: Google Chrome, Mozilla Firefox, Internet Explorer 9+ (without Compatibility View), Safari
Screen Resolution: 1024 x 768 or greater
Internet Access: High-Speed

Subscribe to Alpha Theory content

Alpha Theory Blog - News and Insights

« September 2017 | Main

2 posts from October 2017

October 20, 2017

American Idols

I was lucky enough to be part of a small event, The Frontier of Forecasting Conference, hosted by Good Judgment Inc. Among the participants were Phil Tetlock, Barbara Mellers, and Daniel Kahneman. For those that don’t know, Kahneman is a Nobel Laureate and considered the father of Behavioral Economics. Tetlock and Mellers are the brains behind Superforecasting.

Several of you were interested in attending but unable to make the trip. The following is a summary of the presentations from the conference. 

 

Imagejpeg_0

Left to right: Phil Tetlock, Barbara Mellers, Daniel Kahneman, Lucky Man

 

Tetlock on Gisting

Good Judgment's CEO Terry Murray opened up the morning by introducing the founder of the company, Phil Tetlock. Phil talked about a new idea that he’s working on called Gisting. The goal is to improve understanding by taking a large amount of information and having multiple people create a Gist or a shorter explanation of the information. These Gists would then be graded by peers and the best ones would be picked and synthesized into a team Gist. This leads to a deeper understanding by the “gisters” and easier understanding by readers that only have time for the gist.

Gists are important because full understanding is never just quantitative or qualitative. Superforecasting is quantitative. Supergisting attempts to provide the qualitative piece. The challenge is that time is scarce and this is a new task that will meet resistance in most organizational culture.

Gisting is a relatively new idea and it will be interesting to watch how it develops as Phil, Barbara, and Good Judgment group put more time into research. The next book, Supergisting?

Kahneman on Noise

Kahneman was next up and he spent his time talking about Noise. The concept is not new but he believes it should become a focus because it is easier to reduce than bias. He described an insurance company that he worked with to improve claims adjuster accuracy. He measured the efficacy of their claims process by having independent adjusters price the same claim. The average difference in claim value was 50%! That means that one adjuster might write a check for $1,000 and another for $1,500 for the same claim. He described how a simple algorithm would dramatically reduce noise and improve claim accuracy.

The discussion took a slightly cynical tone when he described how few of his practical ideas were actually put into practice. For example, the insurance firm, after learning of these gross miscalculations, didn’t implement the systematic approach he suggested. He gave another example of how Steven Levitt, of “Freakonomics” fame, showed a simple system of fraud detection improvement to a credit card company that would have saved many millions a year, but wasn’t implemented.

Kahneman said, “change causes winners and losers. Losers are much louder than winners, which makes reform much less likely.” And that “leaders don’t want to see their mystique questioned by systems.” Dr. Mellers had a nice rejoinder that “things will change, one funeral at a time.” For all of us Superforecasting believers, we hope it happens faster than that.

I believe the success that Ray Dalio and Bridgewater have seen by being very systematic and process-oriented may shed some light and make leaders less resistant to change. The publishing of Dalio’s “Principles” will be read by many leaders and get a conversation started about how we all can improve by being more disciplined.

Idea Exchange on Forecasting

The second half of the day was a “safe zone” event to permit free-flowing exchange of ideas due. This means that I’m not allowed to comment on the dialog but I can give a high-level recap.

I was a panelist for “Improving Probabilistic Forecasting Within Organizations.” The goal was to give real world examples of people implementing forecasting tools to improve decision making. It was exciting to see many firms experimenting with forecasting systems. In my view, shared by Good Judgment's president Warren Hatch, who chaired the panel, the challenge that most faced was getting broad adoption and keeping momentum.

The critical component for solving this challenge is getting top-level buy-in. If senior leadership asks questions and uses the output to make decisions, then people will participate. Another strategy for increasing participation was active feedback. Providing scores, leaderboards, best/worst forecasts, stats, etc. have a demonstrable impact on usage.

Better Forecasting Through Better Models

The final discussion was “Bayesian Cluster Forecasting Models for Strategic Decision-Making” lead by Dr. Kathryn McNabb Cochran. She is part of Good Judgment Inc. and is a leader in the field of better decision making through forecasting. The goal is to make better forecasts by creating better models. The models are a hybrid of pure forecasts and adjustments that lead to more accurate forecasts.

For anyone curious about how they can be better forecasters and apply that thinking to their organization, please contact the great folks at Good Judgment Inc.

Final Thought

Meeting several of my heroes in one day made me think how nice it would be if the GE ad campaign in which great scientists are treated like stars was reality. How cool would it be if my girls could grow up in a world where Kahneman and Tversky were admired as much as Brady and Gronk.

 

 

October 06, 2017

Poker: Art vs Science

 

“People describe poker as a game of art and science. Both intuition and science have merit, but the best players approach the game very quantitatively.” – Liv Boeree, Professional Poker Player

Our COO, Graham Stevens, and I met over a poker table. We’ve been playing together for many years and he was recently watching an Oxford Lecture Series video by Liv Boeree that he turned me on to.

Liv is a very successful poker player with a physics degree from the University of Manchester. She was discussing the use of Game Theory Optimal (GTO) play and the use of GTO Tables to break decisions down into ranges of hands based on different situations to aid poker players in knowing the optimal decision (Bet (big or small), Check, Fold).

She stated that the best players in the world all employ GTO. And even though all players assume their opponents are playing GTO, it is incredibly difficult to exploit those predicted decisions because they are optimal. In an interesting exchange in the video, Igor Kurganov, another very successful poker player who was in the audience, said that intuition (playing the player instead of playing the cards) factors into his decisions, but only to a small degree. He said that the best intuition can do is change a 50/50 bet to a 55/45.

The parallels to investing and Alpha Theory are clear. At Alpha Theory, we allow firms to build their own Game Theory Optimal system to figure out the “optimal” amount to bet on each position in their portfolio. And we find that firms that use intuition instead of their model lose to the hypothetical model performance about 75% of the time.

The reasons portfolio managers choose to vary from their model are numerous, but have a common theme; there is an intuition that the model isn’t capturing. Granted investing is not poker. Poker has a finite set of variables and permutations comparted to the seemingly infinite number of variables to consider in investing. But even still, just like in poker, the world-class players are going to be the ones that are following the model and only making small tweaks for intuition.

**Do I practice what I preach? A note on my own poker play. I do not play GTO because I have not memorized the tables. I know some of the shortcut rules for when to bet and fold pre-flop and I can do a rough calculation of pot odds post-flop but that’s the extent of my skills. If my buddies would let me pull out my computer while I’m sitting at the table, I would follow GTO.  If I were playing for a living, I would learn and follow the model.