« What happens after the math is done? | Main | Bias, variance, random errors and honest mistakes »


Feed You can follow this conversation by subscribing to the comment feed for this post.

John S.

I get your point, but Satyajit Das seems to be saying exactly that about investment science in this podcast.


Prof. Zvi Bodie of Boston University has been saying for a long time that if you invest in stocks, risk increases with time. See his website (www.zvibodie.com).
Also, read here the section on price volatility: http://www.retailinvestor.org/risk.html#terminal



SB: Sure, not all economists agree that stocks deliver in the long run although the majority probably do. I'm making a guess that Mr. O'Leary is someone who believes in the existence of an equity premium in the long run, and if that's true, then he believes the possibility of predicting something in the long run while not being to predict the same thing in the short term. Thanks for the link.

John: Heard the Das podcast, good stuff. Das is talking about the belief that no trader can have a long-term performance above the market return. Siegel, Malkiel and others are talking about the equity premium, which is the long-term performance of the aggregate stock market above return of some other class of assets (say bonds). I think these two beliefs can co-exist.

Jon Peltier

It's a stretch to say that not believing climate scientists equates to not believing economists. Both topics are politically charged, but they are unrelated.


I think this post deserves a more straightforward rendering since Jon and others are reading it the wrong way.

I'm not making a categorical comparison between economists and climate scientists. The reason given by O'Leary for not believing climate science is very specific: it is the apparent paradox that one can predict the value of something in the long run but not in the short run.

However, this is not only not a paradox but in statistics, it is absolutely natural that one has more confidence in long-term trend predictions than in short-term predictions, and especially so if the data series is subject to large natural variations.

I'm appealing to an analogy in finance: I believe most of us (not all) agree that in the long run, in aggregate and with an average starting time, stocks as an asset class will outperform bonds. However, most of us will also agree that it is almost impossible to predict whether next Tuesday stock prices would be up or down. If you agree with this, then you should believe that it is possible to predict long-term temperatures without being able to predict short-term temperatures.

This belief does not need to have anything to do with whether the long-term temperature predictions are correct or not - in other words, the fact that short-term temperature predictions are wrong does not imply that long-term predictions must be wrong.

And I'm writing about this at length because this belongs to the "unintuitive" parts of statistical thinking.

The comments to this entry are closed.

Get new posts by email:
Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR, Wired.

See my Youtube and Flickr.


  • only in Big Data
Numbers Rule Your World:
Amazon - Barnes&Noble

Amazon - Barnes&Noble

Junk Charts Blog

Link to junkcharts

Graphics design by Amanda Lee

Next Events

Jan: 10 NYPL Data Science Careers Talk, New York, NY

Past Events

Aug: 15 NYPL Analytics Resume Review Workshop, New York, NY

Apr: 2 Data Visualization Seminar, Pasadena, CA

Mar: 30 ASA DataFest, New York, NY

See more here

Principal Analytics Prep

Link to Principal Analytics Prep