Published on: 30 Dec, 2018
At the end of each week, we will share with you our favourite reads. We would be grateful if you could reciprocate. This week’s reads focus on the science of forecasting, Mckinsey’s trysts with authoritarian regimes, interactive TV and wedding excesses.
You can access previous editions of 3L & 3S at http://marcellus.in/three-longs-three-shorts-what-we-read/
We wish our readers a peaceful and prosperous 2019.
[Note: Marcellus is looking for Senior Analysts. If you know good analysts, please refer them to http://marcellus.in/jobs/]
1. Long read: Why you should care about the Nate Silver vs. Nassim Taleb Twitter war
Author: Isaac Faber
Source: Towards Data Science (https://towardsdatascience.com/why-you-should-care-about-the-nate-silver-vs-nassim-taleb-twitter-war-a581dce1f5fc)
In case you have not heard about Nate Silver, here is the bluffer’s guide. “Nate Silver is the co-founder of FiveThirtyEight. A massively popular data focused blog that gained fame for its accuracy predicting the outcomes for the U.S. elections in 2008. Silver generates predictions using a clever poll aggregating technique which accounts for biases, such as pollsters who only call people with landlines.”
The other protagonist in our story, Nassim Taleb, is a rockstar author, a successful options trader and generally a smart, outspoken guy.
The twist in the tale is that Taleb has announced that FiveThirtyEight does not know how to forecast elections. This announcement has triggered a two year Twitter war between Silver and Taleb. We can learn a lot about forecasting from this debate.
“The primary source of controversy and confusion surrounding FiveThirtyEight’s predictions is that they are ‘probabilistic.’ Practically what this means is that they do not predict a winner or loser but instead report a likelihood. Further complicating the issue, these predictions are reported as point estimates (sometimes with model implied error), well in advance of the event…
Their forecast process is to build a quantitative replica of a system with expert knowledge (elections, sporting events, etc.) then run a Monte Carlo simulation. If the model closely represents the real-world, the simulation averages can be reliably used for probabilistic statements. So what FiveThirtyEight is actually saying is:
x% of the time our Monte Carlo simulation resulted in this particular outcome
The problem is that models are not perfect replicas of the real world and are, as a matter of fact, always wrong in some way. This type of model building allows for some amount of subjectivity in construction. For example, Silver has said on numerous occasions that other competitive models do not correctly incorporate correlation. When describing modeling approaches, he also makes clear that they tune outcomes (like artificially increasing variance based on the time until an event or similar adjustments). This creates an infinitely recursive debate as to whose model is the ‘best’ or most like the real world. Of course, to judge this, you could look at who performed better in the long run. This is where things go off the rails a bit.
Because FiveThirtyEight only predicts probabilities, they do not ever take an absolute stand on an outcome: No ‘skin in the game’ as Taleb would say. This is not, however, something their readers follow suit on. In the public eye, they (FiveThirtyEight) are judged on how many events with forecasted probabilities above and below 50% happened or didn’t respectively (in a binary setting)…
The public can be excused for using the 50% rule without asking. For example, in supervised machine learning, a classification model must have a characteristic called a ‘decision boundary.’ This is often decided a priori and is a fundamental part of understanding the quality of the model after it is trained. Above this boundary, the machine believes one thing and below it the opposite (in the binary case)…
If FiveThirtyEight has no stated decision boundary, it can be difficult to know how good their model actually is. The confusion is compounded when they are crowned, and gladly accept it, with platitudes of crystal ball-like precision in 2008 and 2012, due to the implied decision boundary. However, when they are accused of being wrong they fall back to a simple quip: You just don’t understand math and probability….
What is not clear is that there is a factor hidden from the FiveThirtyEight reader. Predictions have two types of uncertainty; aleatory and epistemic. Aleatory uncertainty is concerned with the fundamental system (probability of rolling a six on a standard die). Epistemic uncertainty is concerned with the uncertainty of the system (how many sides does a die have? And what is the probability of rolling a six?). With the latter, you have to guess the game and the outcome; like an election!
Bespoke models, like FiveThirtyEight’s, only report to the public aleatory uncertainty as it concerns their statistical outputs (inference by Monte Carlo in this case). The trouble is that epistemic uncertainty is very difficult (sometimes impossible) to estimate. For example, why didn’t FiveThirtyEight’s model incorporate, before it happened, a chance that Comey would re-open his investigation into Clintons emails? Instead, this seems to have caused a massive spike in the variation of the prediction. Likely because this event was impossible to forecast….I think this is what has Taleb up in arms. The blog feels more like a slick sales pitch, complete with quantitative buzzwords, than unbiased analysis (though it may very well be). If a prediction does not obey some fundamental characteristics, it should not be marketed as a probability. More importantly, a prediction should be judged from the time it is given to the public and not just the moment before the event. A forecaster should be held responsible for both aleatory and epistemic uncertainty.
When viewed this way, it is clear that FiveThirtyEight reports too much noise leading up to an event and not enough signal. This is great for driving users to read long series of related articles on the same topic but not so rigorous to bet your fortune on.”
2. Long read: How McKinsey has helped raise the stature of authoritarian Governments?
Author: Walt Bogdanich and Michael Forsythe
Source: New York Times (https://www.nytimes.com/2018/12/15/world/asia/mckinsey-china-russia.html)
This year McKinsey’s annual retreat for its busy consultants was in Kashgar in China, a country which McKinsey, like many other MNCs, believes is important for its future. “About four miles from where the McKinsey consultants discussed their work, which includes advising some of China’s most important state-owned companies, a sprawling internment camp had sprung up to hold thousands of ethnic Uighurs — part of a vast archipelago of indoctrination camps where the Chinese government has locked up as many as one million people. One week before the McKinsey event, a United Nations committee had denounced the mass detentions and urged China to stop….For a quarter-century, the company has joined many American corporations in helping stoke China’s transition from an economic laggard to the world’s second-largest economy. But as China’s growth presents a muscular challenge to American dominance, Washington has become increasingly critical of some of Beijing’s signature policies, including the ones McKinsey has helped advance.”
However, China isn’t the only authoritarian regime which can pride itself on being a McKinsey client.
“…clients have included Saudi Arabia’s absolute monarchy, Turkey under the autocratic leadership of President Recep Tayyip Erdogan, and corruption-plagued governments in countries like South Africa. In Ukraine, McKinsey and Paul Manafort — President Trump’s campaign chairman, later convicted of financial fraud — were paid by the same oligarch to help burnish the image of a disgraced presidential candidate, Viktor F. Yanukovych, recasting him as a reformer….Inside Russia itself, McKinsey has worked with Kremlin-linked companies…In Malaysia, the company laid out the case for one of Asia’s most corrupt leaders to pursue billions of dollars from China at a time when he was suspected of funneling vast sums of public money into his own pocket, drawing tens of thousands into the streets to protest against him.”
The NYT article highlights how projects/authoritiarian leaders advised by McKinsey often get into trouble (we are putting it politely; the article is less polite). Here is an example:
“In a confidential PowerPoint report, McKinsey told Malaysian officials that the rail line could increase economic growth in parts of the country by as much as 1.5 percent. It was a figure that the prime minister at the time, Najib Razak, who now has been charged with corruption, liked to cite. In bullet points, McKinsey also said the project would help improve ties with China — “build the nation-to-nation relationship” — because of its importance in China’s Belt and Road Initiative. And McKinsey endorsed the idea of heavy borrowing from China, referring to it as a “game changer” elsewhere in the region.”
Hopefully, Indian politicians – who often use McKinsey to prop up their reformist credentials – will read the NYT article.
3. Long read: How the Surprise New Interactive Black Mirror Came Together
Author: Peter Rubin
Source: WIRED (https://www.wired.com/story/black-mirror-bandersnatch-interactive-episode/)
For the uninitiated, Black Mirror is a futuristic series on Netflix about the dark side of mankind’s advances in technology. In reality, the biggest horror is from the fact that what was considered futuristic in its earlier seasons are already playing out in the real world. But this article is about television’s next step – interactive TV. It is only apt that such an experience is first seen with Black Mirror – just in case you weren’t fascinated enough by the content in the first place. The script of the episode isn’t reality a script but more of a software program written in the programming language “Twine” used for video games suggesting the first step towards convergence of gaming and television (Gaming is already a popularly viewed ‘sport’). It is also no surprise that the producers will ‘use’ the viewer’s choices during the interactive show to customise content, advertising, product placement, etc.
“Black Mirror: Bandersnatch sits squarely in the dystopian anthology show’s tradition of chilling tech parables. It has elements of horror, science fiction, and ’80s nostalgia. It has British actors you know you know, just can’t remember where you know them from. It has moments you’re not sure you’ve ever seen on TV before. One thing it doesn’t have, though, is a run time. You might watch it in 50 minutes, or it might be closer to 70. Hell, it might take you two hours to watch. Because Bandersnatch isn’t just any episode. It’s an “interactive film”—one you steer as you watch, choosing the way you want the story to unfold.”
“It registers your choice via “state tracking,” a technology Netflix developed to accomplish exactly what Brooker and Jones were hoping for, and saves it to be deployed later—in this case, when a television set in the background of a scene plays a commercial for the cereal you chose.”
“The magic of combinatorial math means that there are technically more than a trillion paths through the story, though in reality the number is much smaller. But “much smaller” is still pretty huge: There are five main endings, with multiple variants of each—though upon reaching an ending, Netflix will also helpfully bring you back to pivotal decision points so that you can ease your FOMO and try the path not travelled”
4. Short read: People who exercise have different proteins moving through their bloodstreams than those who are generally sedentary
Author: Gretchen Reynolds
Source: New York Times
We know that people who exercise are healthier and live longer. What we don’t know is why is there a link between exercise and a longer life. “…for the new study, which was published in November in the Journal of Applied Physiology, researchers at the University of Colorado, Boulder, set out to look at various people’s proteins.
They first gathered 31 healthy young men and women, about half of whom exercised regularly, while the rest did not. They also recruited an additional group of 16 healthy middle-aged and older men, half of whom were physically active and half of whom were sedentary….
In five of the modules, in fact, levels of certain proteins varied, sometimes substantially, if someone exercised compared to if he or she were sedentary. Many of the differences were apparent both in the younger participants and those in middle age.
Perhaps most important, the researchers also found correlations between the makeup of people’s modules and their health. People who exercised and had similarities in various protein levels also tended to have desirable blood pressures and insulin responses, with the opposite true for the inactive volunteers.
These data suggest that changes in protein levels are likely to be integral to the complex process by which a workout becomes wellness.”
5. Short read: Wedding excess has reached giddy new heights
Author: John Gapper
Source: Financial Times (https://www.ft.com/content/f57f7a40-02b0-11e9-9d01-cd4d49afbbe3)
John Gapper helps us understand why intelligent, rich people like Isha Ambani and Priyanka Chopra spend so much money on their marriage. For answers he looks at the customs of Native Indian tribes in Canada. These tribes would mark a wedding with something called “Potlatch”.
“Their display is reminiscent of potlatch, the traditional ceremony of feasting and gift giving among Native Americans, which was banned by Canada in 1885. The winter feast, meaning “gift”, was a way not only of celebrating status and family bonds at events such as weddings, but of keeping the rest of the community close with largesse.
Potlatch could be extravagant, particularly after the arrival of European fur traders and their goods: some chiefs would burn canoes and pieces of shields to show off their wealth. It was compared by the US economist Thorstein Veblen to conspicuous consumption by the Victorian leisure class at balls, to which guests were invited to “witness the consumption of that excess of good things” owned by the host.”
Gapper reminds us that Veblen had told us over a century ago that “It is not sufficient merely to possess wealth or power. [It] must be put in evidence,”
Combine Potlatch and Veblen’s insights and you can rationalise the Ambani and Chopra weddings and those of other billionaires in China and America.
“The Ambani wedding shows how globalisation, entertainment and luxury create an even bigger splash. It is part of the shift towards “experiential luxury”, in which the rich covet experiences, such as at the 2013 wedding of the technology billionaire Sean Parker in a Californian redwood forest — guests passed through an imposing iron gate forged with the betrothed couple’s names…Weddings have become ferociously expensive because they involve a heady mix of family and status (as well as being, at least in theory, one-off events that parents feel compelled to make memorable).“
6. Short read: My life as an oracle
Author: Gideon Rachman
Source: Financial Times (https://www.ft.com/content/c0072220-044d-11e9-99df-6183d3002ee1)
An introspective piece by Gideon Rachman, arguably one of the best commentators on geopolitics. The introspection brings about two key aspects which most commentators might relate to – first, how one or two key guiding ideas underpin most of your opinions as a commentator and second, how one tends to find more success in doomsday predictions compared to your more optimistic forecasts. Not quite the happy thought we would have liked to part 2018 with….
“If anything, I think my bigger mistakes came when I ditched my habitual pessimism, and tried to look on the bright side. In a column written at the beginning of this year, I suggested that 2018 might see positive change in Iran and Saudi Arabia. Spectacularly wrong. Blinded by optimism, I also predicted that England would win the World Cup. Wrong again. Going further back, I also failed to anticipate the aggressive actions of Russia in Georgia in 2008 and Ukraine in 2014, guessing that the Kremlin would behave more “reasonably”. By contrast, imagining the worst has usually proved to be quite a good strategy. In November 2015, a year before the US presidential election and before Mr Trump had won a single primary, I wrote a column urging readers to take seriously the idea that Mr Trump and Marine Le Pen could be elected as presidents of the US and France. Of course, Ms Le Pen failed. But Mr Trump did indeed make it all the way to the White House. (In case you haven’t noticed). My longstanding predictions of a trade war — reiterated in the FT’s end-of-year predictions for 2018— were also finally vindicated.”
Note: the above material is neither investment research, nor financial advice. Marcellus is not authorized to provide either. Marcellus does not seek payment for or business from this email in any shape or form. Marcellus Investment Managers is regulated by the Securities and Exchange Board of India as a provider of Portfolio Management Services.
Copyright © 2018 Marcellus Investment Managers Pvt Ltd, All rights reserved.
This communication is confidential and privileged and is directed to and for the use of the addressee only. The recipient if not the addressee should not use this message if erroneously received, and access and use of this e-mail in any manner by anyone other than the addressee is unauthorized. If you are not the intended recipient, please notify the sender by return email and immediately destroy all copies of this message and any attachments and delete it from your computer system, permanently. Any opinions or advice contained in this email are subject to the terms and conditions expressed in a duly executed contract or written agreement between Marcellus Investment Managers Private Limited and the intended recipient. No liability whatsoever is assumed by the sender as a result of the recipient or any other person relying upon the opinion unless otherwise agreed in writing. The recipient acknowledges that Marcellus Investment Managers Private Limited may be unable to exercise control or ensure or guarantee the integrity of the text of the email message and the text is not warranted as to its completeness and accuracy.