Links for Later 9-4-15

1. If we regulated firearms like abortions.
2. Why the world is getting weirder.
3. Moog stopping production of the Voyager.
4. Shakespeare's father appears to have been a wealthy wool merchant & tax evader with dodgy connections. Shakespeare himself may have been the family's representative in London who did theatrical producing gigs on the side.
5. Anne Sexton, "The Truth the Dead Know".
6. Paul Romer's article "Endogenous Technological Change" introduced the concept of nonrival and excludable goods 25 years ago. Joshua Gans has a review of its impact. Paul Romer looks back.

Decision Making in a Nutshell

Justin Fox has a nice review article in the Harvard Business Review which reviews the three major schools of decision-making (decision analysis, heuristics and biases, and "we're not as stupid as we look"). Particularly fascinating, in an inside baseball sort of way, is the bit about how Gerd Gigerenzer "fatigued" Daniel Kahneman with his arguments:

During an academic year at Stanford’s Center for Advanced Study in the Behavioral Sciences, in 1989–1990, he gave talks at Stanford (which had become Tversky’s academic home) and UC Berkeley (where Kahneman then taught) fiercely criticizing the heuristics-and-biases research program. His complaint was that the work of Kahneman, Tversky, and their followers documented violations of a model, Bayesian decision analysis, that was itself flawed or at best incomplete. Kahneman encouraged the debate at first, Gigerenzer says, but eventually tired of his challenger’s combative approach. The discussion was later committed to print in a series of journal articles, and after reading through the whole exchange, it’s hard not to share Kahneman’s fatigue.

A Gentle Risk Assessment

Just a reminder: No matter how many people terrorists kill, the current Ebola outbreak will kill a thousand times more. Where should those dollars and our attention be directed? No matter how many people the current Ebola outbreak kills, AIDS will kill a thousand times more. Diabetes will kill a thousand times more. Which one are you most worried about? No matter how many people all three of these diseases will kill, global warming could kill everyone. Where do you put your attention?

1% Feeling Put-Upon

Max Abelson reports some solipsistic quotes from members of the 1% who are quite unhappy with their lot in life. Income inequality will do that to a person, especially at the top. The more you make, the more insecure you are that you're going to be able to keep making it--your income volatility goes through the roof, and small changes in your relative ranking have big impacts on how well you do.

Nassim Taleb's fragility of the power law hits hard no matter where you sit. Joshua Brown (Reformed Broker) sees this other side of inequality. Being smart about investing (or anything) is no longer any guarantee of success. The market will eat your lunch.

This common problem of the rich and poor under conditions of high inequality is an idea I'll be working with over the coming year.

Weird Algo

Last night, someone's trading system reversed the buy/sell signals in the natural gas market in a play to either a) trigger a shedload of limit orders in the market, b) generate increased volatility and/or c) screw with people's heads.

Zero Hedge's Tyler Durden has three articles up on this:

Yesterday, just after 8 pm Eastern we presented a very curious move in NatGas trading on the NYMEX when under very light volume, the NG performed something akin to a sine wave expansion, with about 12 peaks and troughs with ever increasing amplitude, until ultimately it triggered a major sell off when it appeared to touch off an avalanche of limit orders about 3% from the prevailing price, leading to an almost instantaneous 8% drop in Natgas which was promptly recovered. We dubbed this a fractal pattern, and after a follow up with the trade forensics experts at Nanex, it appears this was a very spot on designation, as zooming into the pattern indicates increasing levels of self-similarity and complexity. Yet aesthetic observations aside, this latest algo appears to be nothing more than a limit order-busting market manipulation device, whose sole purpose is to destabilize and generate volatility for the creator of the algo.

Take a look at the market graphs in the linked article(s). There are some weird people programming these things. That or we're seeing rogue AI's, and Skynet really did go live.

Xerxes and Artaboanos on Risk

49. Then he made answer and said: "O king, neither with this army would any one who has understanding find fault, nor with the number of the ships; and indeed if thou shalt assemble more, the two things of which I speak will be made thereby yet more hostile: and these two things are--the land and the sea. For neither in the sea is there, as I suppose, a harbour anywhere large enough to receive this fleet of thine, if a storm should arise, and to ensure the safety of the ships till it be over; and yet not one alone ought this harbour to be, but there should be such harbours along the whole coast of the continent by which thou sailest; and if there are not harbours to receive thy ships, know that accidents will rule men and not men the accidents. Now having told thee of one of the two things, I am about to tell thee of the other. The land, I say, becomes hostile to thee in this way:--if nothing shall come to oppose thee, the land is hostile to thee by so much the more in proportion as thou shalt advance more, ever stealing on further and further, for there is no satiety of good fortune felt by men: and this I say, that with no one to stand against thee the country traversed, growing more and more as time goes on, will produce for thee famine. Man, however, will be in the best condition, if when he is taking counsel he feels fear, reckoning to suffer everything that can possibly come, but in doing the deed he is bold."

50. Xerxes made answer in these words: "Artabanos, reasonably dost thou set forth these matters; but do not thou fear everything nor reckon equally for everything: for if thou shouldest set thyself with regard to all matters which come on at any time, to reckon for everything equally, thou wouldest never perform any deed. It is better to have good courage about everything and to suffer half the evils which threaten, than to have fear beforehand about everything and not to suffer any evil at all: and if, while contending against everything which is said, thou omit to declare the course which is safe, thou dost incur in these matters the reproach of failure equally with him who says the opposite to this. This then, I say, is evenly balanced: but how should one who is but man know the course which is safe? I think, in no way. To those then who choose to act, for the most part gain is wont to come; but to those who reckon for everything and shrink back, it is not much wont to come. Thou seest the power of the Persians, to what great might it has advanced: if then those who came to be kings before me had had opinions like to thine, or, though not having such opinions, had had such counsellors as thou, thou wouldest never have seen it brought forward to this point. As it is however, by running risks they conducted it on to this: for great power is in general gained by running great risks. We therefore, following their example, are making our march now during the fairest season of the year; and after we have subdued all Europe we shall return back home, neither having met with famine anywhere nor having suffered any other thing which is unpleasant. For first we march bearing with us ourselves great store of food, and secondly we shall possess the corn-crops of all the peoples to whose land and nation we come; and we are making a march now against men who plough the soil, and not against nomad tribes."

49. [1] ὃ δ᾽ ἀμείβετο λέγων «ὦ βασιλεῦ, οὔτε στρατὸν τοῦτον, ὅστις γε σύνεσιν ἔχει, μέμφοιτ᾽ ἂν οὔτε τῶν νεῶν τὸ πλῆθος· ἢν δὲ πλεῦνας συλλέξῃς, τὰ δύο τοι τὰ λέγω πολλῷ ἔτι πολεμιώτερα γίνεται. τὰ δὲ δύο ταῦτα ἐστὶ γῆ τε καὶ θάλασσα. [2] οὔτε γὰρ τῆς θαλάσσης ἐστὶ λιμὴν τοσοῦτος οὐδαμόθι, ὡς ἐγὼ εἰκάζω, ὅστις ἐγειρομένου χειμῶνος δεξάμενός σευ τοῦτο τὸ ναυτικὸν φερέγγυος ἔσται διασῶσαι τὰς νέας. καίτοι οὐκὶ ἕνα αὐτὸν δεῖ εἶναι τὸν λιμένα, ἀλλὰ παρὰ πᾶσαν τὴν ἤπειρον παρ᾽ ἣν δὴ κομίζεαι. [3] οὔκων δὴ ἐόντων τοι λιμένων ὑποδεξίων, μάθε ὅτι αἱ συμφοραὶ τῶν ἀνθρώπων ἄρχουσι καὶ οὐκὶ ὥνθρωποι τῶν συμφορέων. καὶ δὴ τῶν δύο τοι τοῦ ἑτέρου εἰρημένου τὸ ἕτερον ἔρχομαι ἐρέων. [4] γῆ δὲ πολεμίη τῇδέ τοι κατίσταται· εἰ θέλει τοι μηδὲν ἀντίξοον καταστῆναι, τοσούτῳ τοι γίνεται πολεμιωτέρη ὅσῳ ἂν προβαίνῃς ἑκαστέρω, τὸ πρόσω αἰεὶ κλεπτόμενος· εὐπρηξίης δὲ οὐκ ἔστι ἀνθρώποισι οὐδεμία πληθώρη. [5] καὶ δή τοι, ὡς οὐδενὸς ἐναντιευμένου, λέγω τὴν χώρην πλεῦνα ἐν πλέονι χρόνῳ γινομένην λιμὸν τέξεσθαι. ἀνὴρ δὲ οὕτω ἂν εἴη ἄριστος, εἰ βουλευόμενος μὲν ἀρρωδέοι, πᾶν ἐπιλεγόμενος πείσεσθαι χρῆμα, ἐν δὲ τῷ ἔργῳ θρασὺς εἴη.»

50. [1] ἀμείβεται Ξέρξης τοῖσιδε. «Ἀρτάβανε, οἰκότως μὲν σύ γε τούτων ἕκαστα διαιρέαι· ἀτὰρ μήτε πάντα φοβέο μήτε πᾶν ὁμοίως ἐπιλέγεο. εἰ γὰρ δὴ βούλοιο ἐπὶ τῷ αἰεὶ ἐπεσφερομένῳ πρήγματι τὸ πᾶν ὁμοίως ἐπιλέγεσθαι, ποιήσειας ἂν οὐδαμὰ οὐδέν· κρέσσον δὲ πάντα θαρσέοντα ἥμισυ τῶν δεινῶν πάσχειν μᾶλλον ἢ πᾶν χρῆμα προδειμαίνοντα μηδαμὰ μηδὲν παθεῖν. [2] εἰ δὲ ἐρίξων πρὸς πᾶν τὸ λεγόμενον μὴ τὸ βέβαιον ἀποδέξεις, σφάλλεσθαι ὀφείλεις ἐν αὐτοῖσι ὁμοίως καὶ ὁ ὑπεναντία τούτοισι λέξας. τοῦτο μέν νυν ἐπ᾽ ἴσης ἔχει· εἰδέναι δὲ ἄνθρωπον ἐόντα κῶς χρὴ τὸ βέβαιον; δοκέω μὲν οὐδαμῶς. τοῖσι τοίνυν βουλομένοισι ποιέειν ὡς τὸ ἐπίπαν φιλέει γίνεσθαι τὰ κέρδεα, τοῖσι δὲ ἐπιλεγομένοισί τε πάντα καὶ ὀκνέουσι οὐ μάλα ἐθέλει. [3] ὁρᾷς τὰ Περσέων πρήγματα ἐς ὃ δυνάμιος προκεχώρηκε. εἰ τοίνυν ἐκεῖνοι οἱ πρὸ ἐμεῦ γενόμενοι βασιλέες γνώμῃσι ἐχρέωντο ὁμοίῃσι καὶ σύ, ἢ μὴ χρεώμενοι γνώμῃσι τοιαύτῃσι ἄλλους συμβούλους εἶχον τοιούτους, οὐκ ἄν κοτε εἶδες αὐτὰ ἐς τοῦτο προελθόντα· νῦν δὲ κινδύνους ἀναρριπτέοντες ἐς τοῦτο σφέα προηγάγοντο. μεγάλα γὰρ πρήγματα μεγάλοισι κινδύνοισι ἐθέλει καταιρέεσθαι. [4] ἡμεῖς τοίνυν ὁμοιεύμενοι ἐκείνοισι ὥρην τε τοῦ ἔτεος καλλίστην πορευόμεθα, καὶ καταστρεψάμενοι πᾶσαν τὴν Εὐρώπην νοστήσομεν ὀπίσω, οὔτε λιμῷ ἐντυχόντες οὐδαμόθι οὔτε ἄλλο ἄχαρι οὐδὲν παθόντες. τοῦτο μὲν γὰρ αὐτοὶ πολλὴν φορβὴν φερόμενοι πορευόμεθα, τοῦτο δέ, τῶν ἄν κου ἐπιβέωμεν γῆν καὶ ἔθνος, τούτων τὸν σῖτον ἕξομεν· ἐπ᾽ ἀροτῆρας δὲ καὶ οὐ νομάδας στρατευόμεθα ἄνδρας.»

-Herodotus

The Histories VII 49-50

Links for Later

1. Bike messengers
2. Economics of being a model
3. Calvin Trillin (Links above via Matt Yglesias)
4. The Black Swan of Cairo

Links for Later

Felix Salmon on tail risk strategies, difficulties thereof.

Garry Tan's geographical guide for SF startups. (via @venturehacks)

Brad DeLong on the current and long term deficit outlooks. Remember, PAYGO is your friend.

Anosognostic, Part 2

Victor Niederhoffer is a very bright, very nice guy who blew up his hedge fund, twice. Here, he talks with Kathryn Schulz about error, and why one doesn't see one's own mistakes:

Niederhoffer's e-mails suggested a man already obsessed with wrongness. In them, he referenced the statistical concept of path dependence; shared a series of proverbs about the game of checkers (of 5,000 such proverbs, he hazarded, about 250 concerned error); meditated on the difference between Type One mistakes (excessive credulity) and Type Two mistakes (excessive skepticism) (he himself is much more prone to Type One, he says: "I'm tremendously gullible"); observed that "one should be careful of multitasking or multiromancing"; sent me the citations for hoodoo in the Oxford English Dictionary (a hoodoo is something or someone that brings bad luck); and noted that the harpooner in Moby Dick would have made a great interview subject for this series. Finally, he pointed out that the word error has no antonym. "In retrospect," he wrote, "I know much too much about errors and much too little about the opposite, whatever it is."

The best bits of the interview include the insight that he played squash too perfectly (and safely), while playing the market without enough safety, advice on avoiding "hoodoos" (people who will bring you bad luck sooner or later), and "taking out the canes":

When the public is most frightened, only the strong are left, and that's when the market is in the best possible hands. I call it taking out the canes. Whenever disaster strikes, the very sagacious wealthy people take their canes, and they hobble down from their stately mansions on Fifth Avenue, and they buy stocks to the extent of their bank balances, and then a week or two later, the market rises, they deposit the overplus in their accounts, invest it in blue-chip real estate, and retire back to their stately mansions. That's probably the best way of making money, to be a specialist in panics. Whenever there's panic hanging in the air, that's a great time to invest.

Anosognosic

This five part series of interviews by Errol Morris deals with the fact that we believe we know more than we do, are self-deceptive, confabulate, and generally get things wrong. This set of articles could make a fine book all by themselves. Neuroscientists, biographers, Donald Rumsfeld and Woodrow Wilson make appearances, alongside a very confused bank robber:

Wheeler had walked into two Pittsburgh banks and attempted to rob them in broad daylight. What made the case peculiar is that he made no visible attempt at disguise. The surveillance tapes were key to his arrest. There he is with a gun, standing in front of a teller demanding money. Yet, when arrested, Wheeler was completely disbelieving. “But I wore the juice,” he said. Apparently, he was under the deeply misguided impression that rubbing one’s face with lemon juice rendered it invisible to video cameras.

It's a phenomenon that affects both the fool and the genius. We don't notice when we're screwing up. The best answer I've heard for why this is the case is that if we were realistic about our knowledge and capabilities, we'd never do anything.

Links for Later

Rajiv Sethi on Minsky's evolutionary perspective on market instability. Additional information.

Paul Krugman: "the facts have a well-known Keynesian bias".

Steve Waldman: "Austerity is stupid..."

Bryan Caplan on trembling hand perfectionism.

Douglas Adams on Tail Risk

The major difference between a thing that might go wrong and a thing that cannot possibly go wrong is that when a thing that cannot possibly go wrong goes wrong it usually turns out to be impossible to get at or repair.

(via Mark Thoma, via Paul Kedrosky, via Ezra Klein)

Why CDO's Can't be Priced

Paper of the day:

Computational Complexity and Information Asymmetry in Financial Products Arora, et al.

via Felix. Interesting passage from his commentary:

...the solution to model risk isn’t more complex models, its less reliance on models altogether. And anybody who applied a simple smell test to the mortgages underlying the CDOs in question — rather than deciding instead to trust various quants both in-house and at the ratings agencies — would have come to the right conclusion without any computing power at all.

In other words, strategic thinking beats statistical thinking here too.

Power Laws

When are power laws in effect? A statistical note. It's interesting that wealth is not one of the distributions where power laws appear to apply.

Overconfidence

Dominic Johnson and James Fowler have created a model to explain the evolutionary underpinnings of overconfidence:

By creating a mathematical model of the way overconfident individuals compete against ordinary individuals, they show that there is a clear advantage in overconfidence.

In fact, if the potential reward is at least twice as great as the cost of competing, then overconfidence is the best strategy. In fact, overconfidence is actually advantageous on average, because it boosts ambition, resolve, morale, and persistence. In other words, overconfidence is the best way to maximize benefits over costs when risks are uncertain.

...

But it is Johnson and Fowler's predictions that are most worrying. Their model implies that optimal overconfidence increases with the magnitude of uncertainty. So the greater the risk, the more overconfident individuals should become.

Yes, Yes, "Correlation Does Not Equal Causation"...

...so what does?

Causal inference in statistics, a commentary on Judea Pearl's review article on the topic.

Magical Thinking

Roman Frydman and Michael D. Goldberg on the Rational Expectations Hypothesis and its discontents.

the assumptions that underpin these theories are largely inscrutable to those without a Ph.D. in economics. Indeed, the debate is full of terms that mean one thing to the uninitiated and quite another to economists.

Consider “rationality.” Webster’s Dictionary defines it as “reasonableness.” By contrast, for economists, a “rational individual” is not merely reasonable; he or she is someone who behaves in accordance with a mathematical model of individual decision-making that economists have agreed to call “rational.”

The centrepiece of this standard of rationality, the so-called “Rational Expectations Hypothesis”, presumes that economists can model exactly how rational individuals comprehend the future. In a bit of magical thinking, it supposes that each of the many models devised by economists provides the “true” account of how market outcomes, such as asset prices, will unfold over time.

The economics literature is full of different models, each one assuming that it adequately captures how all rational market participants make decisions. Although the free-market Chicago school, neo-Keynesianism, and behavioural finance are quite different in other respects, each assumes the same REH-based standard of rationality.

In other words, REH-based models ignore markets’ very raison d’etre: no one, as Friedrich Hayek pointed out, can have access to the “totality” of knowledge and information dispersed throughout the economy. Similarly, as John Maynard Keynes and Karl Popper showed, we cannot rationally predict the future course of our knowledge. Today’s models of rational decision-making ignore these well-known arguments.

Risk Perception

Summary of Slovic's 18 factor risk perception model at Lone Gunman.

Science News Roundup

Scientists create rogue microwave model useful for simulating rare events that emerge from random processes.

Eric Heller at Harvard University and various pals have studied rogue waves for some time. Today they show how microwaves propagating through a forest of scatterers which the team call a "quasi-two-dimensional resonator with randomly distributed scatterers, each mimicking an r^−2 repulsive potential".

The results are fascinating because they clearly show the rogue waves (or hot spots in microwave terms) appearing more often than conventional thinking (Rayleigh's law for the wave height distribution) allows. In fact the team says the probability in their set up of a rogue wave appearing is 15 orders of magnitude greater than Rayleigh statistics predict. They attribute the difference to ray refraction rather than to resonance effects as conventional thinking might suppose.

Mouse retinas contain cells that detect approaching objects, possibly providing an advance warning system.

While investigating mouse eye cells, Botond Roska at the Friedrich Miescher Institute for Biomedical Research in Basel, Switzerland, and colleagues noticed that one type behaved unusually in response to movement. Further analysis of this one kind of retinal cell revealed that it fired only when an object approached.

The researchers suspect that people have similar cells, which alert us to approaching objects faster than our brain cells can. "It's an alarm system that's as close to the front end of the organism as possible," says Roska. "If you left it to the brain to respond, it might be too late."

It could also be that these cells form part of the signal processing apparatus that enhances visual acuity.

Sweeping Out the Hump

This post on Lone Gunman got me thinking about Joel Spolsky's use of Taiichi Ohno's Five Whys coupled with the procedure of Fixing Things Two Ways as a resiliency mechanism for non-Black Swan problems. When something (e.g., a server) fails, it has to be repaired, but the underlying causes of the problem (the five levels of the Five Whys) should also be fixed. This strategy differs from the standard metrics-based approach in that it generates a system where costs due to previously known and realized causes go down over time.

Spolsky gives the excellent example of server downtime in comparison to airplane crashes:

Measuring the number of minutes of downtime per year does not predict the number of minutes of downtime you'll have the next year. It reminds me of commercial aviation today: the NTSB has done such a great job of eliminating all the common causes of crashes that nowadays, each commercial crash they investigate seems to be a crazy, one-off, black-swan outlier.

As time goes on, an increasing proportion of problems derive from rare events. The high-frequency events at the hump of the frequency distribution get "swept out", along with any early occurring rare problems from the tails. This frees up attention for the truly bizarre and unforseeable events. Note that this doesn't work well if your Black Swans are catastrophic. If you get hit by a civilization ending asteroid or a spontaneously business-ending event, you have little opportunity (or benefit) to learning from experience, but this kind of problem is thankfully extremely rare in our collective prior history.