Skip to content
Back to the collection

On the statistical properties and tail risk of violent conflicts

Pasquale Cirillo, Nassim Nicholas Taleb

2015 · arXiv

Read the paper
HN score
0
Mentions
3
Readers
3
Months
3
What’s in a score?

From the abstract

We examine statistical pictures of violent conflicts over the last 2000 years, finding techniques for dealing with incompleteness and unreliability of historical data. We introduce a novel approach to apply extreme value theory to fat-tailed variables that have a remote, but nonetheless finite upper bound, by defining a corresponding unbounded dual distribution (given that potential war casualties are bounded by the world population). We apply methods from extreme value theory on the dual distribution and derive its tail properties. The dual method allows us to calculate the real mean of war casualties, which proves to be considerably larger than the sample mean, meaning severe underestimation of the tail risks of conflicts from naive observation. We analyze the robustness of our results to errors in historical reports, taking into account the unreliability of accounts by historians and absence of critical data. We study inter-arrival times between tail events and find that no particular trend can be asserted. All the statistical pictures obtained are at variance with the prevailing claims about "long peace", namely that violence has been declining over time.

Metadata: arXiv. Abstracts retain their original authors’ rights.

In the conversation

Original HN comments, with labels from simple text rules. A label describes the whole comment and may miss context or irony. Read the discussion to make up your own mind. About the labels

3 selected excerpts · 3 total mentions

leto_ii
Neutral mention · rule
Ok, the guy is not some Fields medalist, so it depends what you mean by math chops, but here are a few of his papers published in respectable journals that have non negligible math content: https://www.sciencedirect.com/science/article/abs/pii/S03784... (free at https://arxiv.org/pdf/1707.01370.pdf) https://www.sciencedirect.com/science/article/abs/pii/S03784... https://www.sciencedirect.com/science/article/abs/pii/S03784... (free at https://arxiv.org/pdf/1405.1791.pdf) https://www.sciencedirect.com/science/article/abs/pii/S03784... (free at https://arxiv.org/pdf/1505.04722.pdf) https://www.nature.com/articles/s41567-020-0921-x There are plenty more btw, these were the ones that I could find in more prestigious journals. if you follow him on Twitter you'll see that he also does quite a bit of recreational maths too.
concinds
Neutral mention · rule
This is a complete misunderstanding of the problem. I invite you to read Taleb (I'm not trying to sound like a fanboy; but his ideas are literally the utterly common-sense conclusions you'll reach if you study complexity science as applied to the financial system; i.e. they are objectively and necessarily true). I'll try to present a very simplified but rigorous case: 1. There's commercial banks (that take consumer/corporate deposit & give out loans), and investment banks that arbitrage the markets. If you allow banks to fulfill both commercial and investment roles, consumer deposits are at risk from bad investment decisions. Note that Glass–Steagall is still repealed,…
Read the full comment
and banks are still allowed today to put your money at risk. 2. Many investment banks used mathematically-flawed risk models, that underestimate fat-tail risk, and therefore lead to inevitable collapse. Read Jim Rickards's testimony to the House of Representatives on this; he was LTCM's general counsel, and LTCM (a who's who of finance PhD's and Nobel Prize winners) blew up due to overconfidence in these flawed models (https://moam.info/house-testimony-rickards-committee-on-scie...). 3. The problem here is systemic risk: banks only get rescued because they're systemically important. "Systemic-ness" is mainly a function of interconnectedness of markets, i.e. systemic banks cannot fail without destroying the markets and economies. Hence banks have an incentive to become systemically important, so they'll get bailed out if they make mistakes. Risk of catastrophic system failure is incentivized. 4. Small banks failing is normal. FDIC insures deposits to a certain amount; no Average Joe gets wiped out, bank investors do. The key point (covered in Taleb's Antifragile book) is that individual survival (agent) and ecosystem survival (system) are often at odds. Animal evolution requires death, since that's the filtering process; if everyone survived, no natural selection occurs, no evolution, and pathological mutations would eventually accumulate and wipe the species out. A healthy banking industry is one in which banks fail regularly, without impacting consumers (thanks to FDIC), and where bankers who worked at those banks get wiped out financially too, and don't just walk away with bonuses. Then, they'd have an incentives to stop using known-defective quant risk models. Contrary to popular opinion skin in the game isn't about incentives but about filtering (if you fail at the game, you personally suffer the consequences and cannot continue playing). Many problems in the world stem from misunderstandings of either complex systems theory (an actual scientific discipline that should be taught in high school), math (mostly dimensionality and fat-tailedness) and probability theory. Without a basic understanding of these, people simply don't have the tools to understand or discuss much of the modern world. The consequences can be seen in systemic financial failures; in ineffective pandemic responses; and in much of social science. See for example: https://arxiv.org/pdf/2007.16096.pdf https://arxiv.org/pdf/1505.04722.pdf
qwtel
Neutral mention · rule
for this is be a valid argument, the sample mean has to be good estimator for the "true" mean, which is the case only for thin-tailed random variables. unfortunately, violence in humans is verifiably fat-tailed (short bursts of violence account for the vast majority of deaths). here other techniques for estimating the mean have to be used. this has been done by cirillo and taleb [1], and the results are sobering: there is no evidence for a drop in violence. in fact, there isn't sufficient data for any kind of claim (intuition: major conflicts account for the majority of deaths; the inter-arrival time of a major conflicts is 100 years (ballpark); therefore 60 years of relative peace are not sufficient). EDIT: removed preaching, since not relevant to the parent comment. [1]: https://arxiv.org/abs/1505.04722

Keep following the thread.

More in Statistics & probability

More from the same topics.