EFG Financial Products Day 2013 - Nassim Taleb

so welcome back to our final part of today's event before he comes back on stage let me give you a little more background on our next speaker Nassim Nicholas Taleb mr. Talib started out as a trader he worked as a quantitative analyst ran his own investment firm the more he studied statistics the more he became convinced that the entire financial system was a keg of dynamite ready to blow and as we all know this is pretty much what happened a little later he predicted it before it happened in the Black Swan and his book mr. Talib argued that modernity is too complex to understand and that Black Swan events ie hitherto unknown events unpredicted shocks would always occur now anti fragile his new book is basically a follow-up before he described a problem now an anti fragile he describes what to do about it one of the core questions is how shall we live in a world we do not understand let's hear some answers professor NASM Nicholas Taleb welcome so great thanks do you mind lowering the lights please say it's this is the only place where I'm glad you call me the professor in New York my friends former traders called me a professor one and want to make fun of me supposedly Germanic german-speaking countries you know being a professor is a very good thing very respectable so I'm glad you called me a professor here so I'm gonna talk about my book we had two options I can give a presentation or I can read from page 106 to 161 for an hour if you want okay so okay looks like you that's not the option for you but please please I beg you lower the light I'll say the truth let me give you an idea about why this title anti fragile and how the idea came to me I'm an option trader the many option traders here the registrate errs okay when I say long Dan my long volatility how many people sort of grasping I'm talking about okay so for option traders there's I'm gonna say is nothing new all right except that extent to other areas not particularly intelligent except for non option traders okay the concept is completely novel anti fragile because there's the opposite of robust the opposite sorry a fragile when you ask someone what's your upload fragile to give you a robust adaptable from adaptable all that nonsense it's not resilient words like that it's not the uppers of short gamma is what long gamma so the opposite of something that short volatility is something that gains from volatility you agree the robust doesn't necessarily gain from volatility when people tell me well complex system have adaptable properties I tell mom listen the word adaptable is not going to explain long volatility right if I buy five times the five times the normal insurance on my house five times what do I wish for floods fires all that disasters you agree right so that's not adaptable is that no so the category is long volatility the adaptable may be long volatile but that's not the point the problem we have the absence of word that our vocabulary make things very difficult people understand that notion that the opposite of fragile is anti fragile when they behave we have a lot of words you know in slang a lot of expressions that sort of convey the idea of long volatility the problem is learned it vocabulary doesn't have it someone you learn it to sort of forget about it you see the gentleman up here not particularly good-looking fellow that's not why I have his picture there's a gentleman up here that someone recognized him now Gladstone former UK Prime Minister hundreds some years ago well very interesting fellow he was the equivalent of finance minister and then between you know jobs he wrote a 1,700 page treatise on Homer and in the treatise he had he made a remark that the Greeks did not have a word for blue and effectively the ancient world did not have word for blue even if you take the Hebrew Bible there's no exact word for blue or were they colorblind no some modern researchers you know made some tests they realized that there are lot of native populations that only has three words for colors they don't have a word for blue usually blue comes last but they're not colorblind they're biologically okay but they're culturally colorblind see likewise we don't have a word for the opposite of fragile so when governments want you know to engage in action what do they say well let's not be fragile let's be the opposite of fragile let's be strong strong is not the opposite of fragile strong is what doesn't break you say so that's what so this is sort of like what my idea is about I realized that as an option trader I was frustrated over you know over my life as an option trader because people wouldn't understand that it's desirable to have things that gain from volatility but couldn't explain it to them the word robust is not precise enough so anyway this gets me to the end of my well trilogy are four and four volumes by the way because all trilogy should be in four volumes so I ruled by randomness the Black Swan there's something called The Ballad Procrustes aphorisms and this is the end and the first three books I explained that really don't understand tale events tell events are in computable mathematically you can show there's a limit and you can show empirically that there's a limit and let me explain you how empirically there's a limit to tell you how little we know about how fat that tails are meaning contributions of rare event there is a measure in econometrics called kurtosis you know what kurtosis is it tells you how fat the tails are you agree how much a distribution is non-gaussian you take the last four years of data any financial or economic or whatever even social variable take the last four years of data about ten thousand pieces of daily variations you will notice that one they may represent 80 90 percent of the kurtosis one day in 1987 the stock market crash represents 80% of the kurtosis last four years what does it mean it means we don't know how fat the tails are in these domains domains are non Gaussian it's very hard very hard to figure out what the distribution is for the rare events and the world unfortunately is becoming more and more non Gaussian less and less predictable more more dominated by the very rare event so I have here a graph they told me that you guys are professionals to give you very difficult stuff so we're gonna get into heavy math I start you know easing into it with graphs and then you get into the math right the start with the grass here we have four different different M you recognize it this time series returns okay the first one does this work the first one here is a fragile the fragile is literally licked look at this this alright if you have the returns of this the financial returns of this once you have it on the table okay it's not going to improve in quality nothing can improve it but an earthquake can break it you agree so it doesn't have upside so the fragile is what has small upside and a lot of downside for this the upper bound is unchanged so the fragile is upper bound is unchanged with some positive noise above it so this is the fragile and we're gonna see why it has to be so okay in the definition of fragile so this does this like that this if I put it on a table but the table is slanted so you know I don't want to mess up my presentation at the beginning and the end who cares right the the if I put it on a table all right nothing can improve it okay so this is short volatility this is precisely fragile because it doesn't like volatility if I hear of an earthquake injury it happens here now or no it is written notice quakes you have such a good political system that thanks alright so but if there's an earthquake let's say it's a thought experiment this is not gonna be helped so either unchanged or worse so this is the fragile the second graph is a very rare category of things sorry that can improve a lot you can have big jump up or big jump down okay let's ignore this category it almost doesn't exist but I have to put it there just you know because just to make things you know interesting but it's not the third category is a robust doesn't improve doesn't degrade it doesn't really care so a coffee cup is fragile the Brooklyn Bridge is robust you see it's not gonna get better if you ride on it you bring a truck it's not gonna get better it doesn't really care all right this is robust and a diamond is perfectly robust and then the fourth category is category of things that have small downside and huge massive upside that category needs a name so I gave it the name anti fragile as the opposite of fragile I tried every language including I don't know influence Swiss German okay the different varieties German tried 70 languages and of course not a single language has you know an expression that captures the point of reverse fragility so is you know this is the only one a graph that has words all right I hate words so I'm not gonna leave it there for a long time so it's fragile what benefits from volatility but mathematically what benefits were involved to is benefits from cluster of things variability time because time mathematically is same almost the same as volatility incomplete knowledge Transkei us entropy randomness stressors and error you see if errors don't hurt you is gonna be neutral or positive you're anti fragile so now in probability is very simple your payoff either has a big left tail when your aunt if right when your refresher or a big right tail when your auntie fragile that's the idea I'm going through grass stress to just you know because after lunch supposedly the wakes up makes people up grass but but I'm gonna focus on you know the the the the sequence are followed in a book later on this book incidentally was written in a way to ban anyone from skimming it why there's no correspondence between chapter titles and contents okay alright for example chapter 14 when two things are not the same thing right in Brooklyn accent chapter twelve Tallis's Crete grapes chapter eleven the most interesting never marries the rock star so there's no correspondence and even go subsections same thing I did it to the Black Swan my editor said Oh we'll never sell any copies three million copies later he's sort of convinced all right so it works because you know it it prevents the journalist from skimming it so I'm gonna follow nevertheless the organization of the book because readers seem to you know follow it it's sort of like you're surprised I want to surprise the reader at every page that's my idea so we continue now this if I die after presenting the slides I'd be done this is my central idea but it took me 23 years to be able to figure it out twenty-three years to realize how universal it was okay twenty-three years this explains how not only what fragility is but how we can detect it this isn't functional space what we saw earlier is in time series space in probability that you have a big left tail if your fragile big right tail if you're anti fragile no tail if you're robust this is this explains the story sort of the convexity argument and I got the idea in there's something and Midrash Tehillim the lights are can you lower the lights and Midrash Tehillim a king has a son who committed the crime so he had to Pelt them oh I was getting confused with the lies you're supposed to Pelton who was big sorry it's supposed to hit him with a big slab of stone that was a punishment the problem is the king didn't want to give his son he had to find a solution what was the solution what do you think it was you cut the stone in small pebbles and hit him on the head with small pebbles okay very simple we have a definition of fragility you're in finance no you're fragile if a 10% move in the market hurts you more than twice a 5% move in the market okay facility cannot be in linear exposures let's take this other example here if I jump 10 meters even in Zurich you tend to be to die no that's a limit but you never know you'd never absolutely know where the Swiss you know there are always some secret somewhere that make them survive so if you jump 10 meters you die if you jump 10 times 1 meter do you die no I'm going try it if if that was so if the harm were linear you see far more linear then just walking to the office will kill you you say because you have so many more one centimeter jumps you realize probabilistically and if it were harm or linear then it would kill you let's go back to this curve there are very few large stones you say if large stones hurt me are you linearly you'd be harmed by just contact with the wall touching objects so harm has to be something nonlinear in other words very simple every additional drop in the market should hurt you more than a previous one if you're fragile that's the very definition of fragility can someone I mean let's take any example to show why the fragile results are nonlinear if I smash my car against a wall at 100 km/h it's not good for the car and not good for the wall certainly not good for me you agree if instead I Drive it against the wall ten times at 10 km/h I'm okay it means every additional kilometer should hurt you more the previous one if it weren't that way you would that you would you would die it you know from the you know contact with the wind so this is the argument for non linearity in fragility why is this central this is central because it allows us to figure out what is fragile very quickly you can figure out for example if the nonlinear effects of traffic in the city take for example New York City if you have a hundred thousand cars in Manhattan I can go to the airport in half an hour if you have 110 thousand cars it takes me thirty five minutes you add ten thousand more cars there now it's an extra hour to go to the airport so we see nonlinearities another way to illustrate the idea of the too-big-to-fail I can show here why too big is fragile just from this graph just like I can show why an elephant is vastly more fragile than a rat you don't have rats and no no but you have a yellow Hannibal's didn't leave an elephant behind or anything all right so but an elephant breaks a leg very quickly you see a rat you can toss them out the window and laugh at you all right sort of not exactly robust why let me show you the exact story here from the story associative in Arad you've heard what happened was a city in Iran they had a trader called Kol and he was hiding 50 billion dollars worth of stock in a drawer you agree I said 50 were billion worth of euros I'm sure someone here distict Lee someone might have been working at South China at a time no okay so the liquidation cost was how much do you think sorry five billion euros exactly they dropped the market on average or whatever at the end the last sale was 12% below where they opened incidentally the New York Times had headlined the next day Marcus drop from Mumbai to London on fears of recession right this is the previous day within half years of recession alright anyway so now let's think of this explains it as a thought experiment instead of having Societe Generale you have ten small banks we call a micro associative general micro all right and we'd have 10 rogue traders at different periods and the penalty distributed we call a micro care of Yale and each would be hiding five billion euros in a drawer how much does the quotation cost four five billion euros how much close to zero okay maybe five ten million I mean it's very very cheap to liquidate five billion euros increasingly so this is concavity every incremental five billion euros cost you more per unit than a previous one that's fragility and this is how we can measure fertility so if you get this idea and you're an option trader you see the short volatility yes we just have to detect where we're short volatile in life you see we can figure out from here if a company is fragile you lower sales 10% see our P&L then lower sales 10% more if it increases aha you're in trouble if it stays the same loss then it's linear they say it's okay this is sort of what we did for the IMF and we're gonna see how I'll show you later but now let me show you whatever is convex is long volatility concave a short volatility and the best way to figure it out is with this all right convex concave and I did the paper for the IMF where is the paper here we go the IMF because I blow up at the gentlemen they hired me actually it's a funny story they ask are you remember the MINIX Ross gone okay he was talking about black swans and sort of said okay let's get this guy to come over so I said okay I'm here so by the time I met them there was a pregnant lady and he was gone you know I don't read the papers and then and the first thing she said the first thing she said hi mrs. so-and-so and this is not his and then they said okay can we work together on a measure of tailor risk I say I can't measure tail risk I can't predict future event but I can tell you if Bank a is more fragile to a tale event and blind Bank B and this is how we came up with this paper and it may become policy but you know the IMF they worked very slowly very very slowly so maybe in 2071 maybe he's a third-generation survivor be applied so let me come back now to this graph because the central graph look at it again if there's an extremely near exposure up here and you have nonlinear exposure here okay look if you have a Black Swan event the difference between the two the large of the deviation the more the difference and hit between the two so if you have extreme events the nonlinear will be hurt a lot more than a linear now that's the definition of fragility we can get antifragility at the exact opposite ten percent is a lot better for me than ten times one percent I'm anti fragile so this is a standard option theory that we can apply to things and generalize so the the you know look now what we have a smile and what has a frown that's the best way to figure out the fragile or anti fragile now I was talking to Helen but I can't pronounce her last name without you know her laughing so I say to Ellen all right and I was talking to her and she's you know she likes Seneca who figured out the idea of antifragility Seneca how very simple you've heard the Stoics supposedly Stoics are robust now in theory they're not supposed to they're supposed to mentally put themselves in a situation where they do not feel any pain from losses you agree that's what people thought that they were like vegetables no that's not what they were these guys were like option traders they want the upside don't want the downside when you read Seneca or eat something you don't find in other Stoics and definitely don't find in academic papers written about Seneca because these academic papers are not written by option traders you say well Seneca what did he do every day he did something we're gonna call the barbell every day or every morning supposedly he wrote off all his possessions he was the wealthiest man in the world okay he was writing about poverty and he had 500 desks with ivory likes so you can imagine writing about poverty right so what was his idea his idea that you should write off your possessions so you're not harmed if someone takes it away from you take them away from you because you're concave you see you have wealth you get used to it it makes no difference someone takes it from you it's a lot of pain he understood that so he said no I'm gonna be one up on destiny every morning he would wake up and claim that he's poor and effectively to simulate you know being poor he would he would fake that he's in a shipwreck and then go for you know two villagers with with he said hardly any possession and hardly any slaves no more than two or three slaves incidentally by the way so he realized it was nothing his idea of poverty is very different from our idea of poverty anyway so he mitigated his downside mentally he said you can take away whatever you want from me you not gonna hurt me that's a stoic attitude but he kept the upside he said he kept the pleasure of the upside and then he would find his wealth again and feel okay so this is Seneca who discovered it and effectively in this book now I embed my idea of antifragility with Senecas idea the idea of Seneca about antifragility so I'm gonna continue now following what we can do with this notion we can do a lot of things you know this idea of what's convex and concave to reiterate sorry if I can learn to use these it'd be obsolete by the time I figure out not to use this but anyway so you see if you have more pain sorry more more pain than gain at the bottom you are concave if the market goes up one percent alright you lose more than you make if the market goes down one percent you're concave alright so you're fragile for an equivalent move alright in the opposite direction you make more than you lose your convex so in one case you're anti fragile so anti fragile means make more than you lose and fragile means lose more than you make very simple Seneca figured it out he wanted to lose less than he makes from possessions so with but with this we can really have an organizing principle to look at a lot of things around and effectively everything I've been discussing this morning everything I've been discussing anyways for last 22 years it looks like falls into this what is fragile what is robust what is anti fragile we can have this categorization based on convexity two random events this is why that makes you fragile when you have debt you don't like volatility because if the events don't play as planned you're in trouble so you have been very good at forecasting the future and you cannot make mistakes if you have equity a lot more robust if you have as we're gonna say out of the money namely venture capital and innovation and stuff then you have all the upside so can someone suggest so we can start already like a conversation what other areas this could be applied to earlier they were more aggressive okay so I continue we're gonna have a Q&A but we wanted to sort of break the ice so there's something mathematically I wanted a little break before the math but we're not gonna get it called Jensen's inequality and I'm gonna show you the graph and explain to you if this brings us to medicine okay I do a lot better having two doses of protein today and nothing tomorrow then one dose every day I am anti fragile because I like very I'd like to have variation I don't like steadiness so and we people in medicine don't know about this there's a property called Jensen's inequality which tells you if you're fresh or anti fragile locally to something if you rather have okay it's just to show something on convexity I'm putting this to show numbers for example malaria a ventilator you know long ventilators when sick people if you give them the same dose they don't survive if you give them half the dose and then one and a half time half one and a half down or randomly randomized they do a lot better than if you give them steady dose dosage same thing it looks like our buddy you need stressors so instead of getting you know 70 degrees temperature all the time you have a 70 degrees Fahrenheit you know 20 degrees you get sometimes zero for a few minutes you know for an average of 20 degrees Celsius you do a lot better so this is identifying what's anti fragile and daily life is what likes variation it's the same formula as our long volatile formula with convexity exactly the same formula and let's explore it a little more I'm gonna here explain it in two languages in English don't cross a river that's on average four feet deep so in other words the average is don't really care and then I had translated into French okay a convex function of an average is not the average of a convex function all right so in other words say your long volatility all right what would you rather have 5% moved today 5% moved tomorrow in the market or it'd rather have 0% today and 10% tomorrow 0 and 10 okay if you're long volatility you'd rather have the whole year zero and then one day 400% you agree so the average doesn't matter much well this is called Jensen's inequality and that applies to so many domains and again it has you know the meaning of average no longer the more the more the notion of average okay the less significance it has the more is irrelevant okay the more yer is a fragile or auntie fragile simple okay this is why sometimes I say forecasting doesn't matter much if you forecast the river is four feet deep you need more than the average and have a slide here to show you something it's written in Fahrenheit but you can translate I have a piece of information a grandmother spent the last two days 48 hours at 70 degrees Fahrenheit okay good news now 70 degrees Fahrenheit is excellent for grandmothers you don't say perfect like 20 degrees everybody's happy no that's the average but now you have second piece of information the grandmother spent the first 24 hours at zero degrees Fahrenheit it's like minus 16 and the second 24 hours at 140 degrees Fahrenheit I want that it's too obscene the high number all right okay for an average of 20 degrees for a night all right what do you think happened to the grandmother there you go so the more the more higher order effects matter the more fragile you are or anti fragile you are and now there's something people don't understand it's a denial of antifragility if something is anti fragile it's okay it likes variation you agree you like stressors variation and I think so by not giving something stress when it wants stress and variation what are you doing you're harming it I call that mistake mistaking the cat for a washing machine we have that idea we don't understand that anything organic alive likes variations and stressors that's how it communicates with the environment it likes some randomness in the environment you see the lion hunts all right with one part lion one part cow the cow eats what salad without dressing all right all the times very boring to be a cow you got so many of them look how bored you know they look you know it's very boring to be a cow they don't commit suicide you know but it's very boring so the cow doesn't have randomness in its feeding you agree so there's not too much Jensen's inequality effect but how about the lion the carnivores how often does a lion get hits a prey very rarely once twice a week all right and and typically you miss a lot of hunts and and all right so we're one part cow one part lion so therefore we should have some randomness in a way we get our protein you agree well effectively doctors are now discovering that when you deprive someone of protein for a few days they have Auto Faudree they start eating your cancer cells if you give someone if you feed someone steadily they tend to develop the prototype eta is the time to develop diabetes but if you feed them randomly who feast and famine replicate the structure of the environment then they get they don't have diabetes if you do intermittent fasting seems to be all that comes from the same equation that we can test its convexity to a source of nutrient or not so all of these will fall together so mistake in the cat for a washing machine the economy is organic you agree the economy the problem is economics textbooks make it look like a clock or a watch Adam Smith's a watch unless misunderstood this sort of organic but he thought it was like a watch he called it the watch this watch is actually a Swiss watch it doesn't self heal believe me it doesn't sell feel it wants you have to pay to hunt the stock at home at home this you write to for it so there's no self feeling all right you go you got a bill when you take it for okay repair your cat self heals no the problem is the washing machine the clock all these things need constant monitoring the economy if you give it constant monitoring you kill it if I bring the human being if I put him in a room with absolutely zero stressors zero stressors okay and or send someone spend six months reading war and peace you know different translations and stuff like that in a space shuttle alright or a space station and then bring back to what happens to their bone density they can break a leg just by walking alright and of course they're not not a lack of exposure to viruses and and and bacteria weakens their immune system so everything organic that we deprive them of stressors weakens and this is what I call the Soviet Harvard illusioned is it the economy need stressors so the restaurant business works beautifully because it's able to convert failure into a huge amount of benefits for the collective fragility of components confers robustness alright and gains for the overall because you know if you know every time a restaurant fails food gets better in the city so otherwise would be eating soviet-style cafeteria food likewise when it takes there to different differences between businesses businesses that are supported by the state tend to become fragile businesses that are not like California or the restaurant business they use randomness as as fuel and benefit from it and it's quite this is what I was saying this morning about Greenspan he didn't understand that if the forest if you stifle every fire in the forest what you do you let the flammable material accumulate on the side and then then the big one would be uncontrollable if you let the forest clean itself it'd be much more manageable so here up top is a process with volatility constant volatility scares people it's like the political system in Italy you know yeah but it works it's a lot of volatile you know newspapers but not in real life right the bottom is political citizen Khanna me after Greenspan or political system in Syria nothing happens until something happens and guess what puffs a huge jump so the second one has fatter tails nothing happens and big jumps the one up top always makes mistakes and converts them and to you know try to use it use its variation as fuel so this is the thing that libertarians tend to like that having noise in a system constant noise is good for it having no noise you know is bad for it and there are a lot of systems that actually benefit from randomness this in physical in the physical world like you inject randomness in some systems they stabilize something called stochastic resonance or something called and annealing where randomness and noise protect the system now again the government's come in and they over stabilize they cause this weakness this is what's gonna happen to Saudi Arabia but don't tell anyone when in the Black Swan as I mentioned Syria and Saudi Arabia one went already one was Syria we have no political volatility you stifle variation you weaken the system and then look what happened and now Saudi Arabians who's responsible for this the United States foreign policy is there to repress any kind of volatility but you can't keep doing that you see because you're weakening the system now also there's wait another thing I discuss is there's two different businesses as there's a business that uses error as an engine for improvement so every time a bridge collapses whether in Idaho or in Geneva okay every single bridge on a planet is likely to improve you see it gets safer every time it claimed crashes the next plane ride is safer for everyone so the probability of a crash of planes drops after every plane crash well that's the system that benefits from every mistake let's compare it to the bank when a bank crashes it doesn't make the next Bank less likely to crash okay which means the system is not built properly which is what you get here systems like these okay or the the errors are monstrous so and this comes from disease of naive interventionism that also happens to your health if you you know we're supposed to have exposure to you know just like the economy to stressors and stuff like that if you deprive yourself of stressor you become weaker no well this is why we go to a health club but they have health clubs here no means there's an implicit acceptance of the anti fragility of the human body that needs to be stressed and the stressor comes in a factor Jensen's inequality the same thing applies you see it's a lot better to lift a hundred-pound once and then nothing for the rest 24 hours then lift a tenth of a pound you know every hour for there you get the idea so it's the same with Jensen's inequality you have gains that are convex up to a point so this is sort of the idea and then with this we can make a table tableau and finance what makes you fragile as a system was dead robust anti fragile and anything biological system efficiency makes you fragile redundancy makes you robust there's something called functional redundancy degeneracy it's called but it's not really some misnomer so we can use this but here the nation-state centralized is fragile anti fragile city-state decentralized anti fragile you guys are even what step below the city-state its compound state no and then whatever compound can't do you flows up so you're flowing things up rather than making things come down so this is we can do a lot with this so I have an extra what ten minutes Ellen before ten minutes of presentation I will take Q&A unless you're in a hurry to contradict me in the raise your hand accommodate now one discovery I made was trial and error okay that every history book of technology every book in the history of technology makes you believe that technology comes from science okay and that wealth comes from education same process like universities generate science and then technologists use that science to do technology no you agree what's wrong with the model what's wrong with it I mean we live here in a place that contradicts our model yes they're very good so what happened is that it looks like science comes from technology and the mechanism I call it lecturing birds how to fly is that University's claim credit when in fact it's entrepreneurs and thinkers and people who are doers you know with their hands when faculty discover things now let's backtrack a little bit every single story out of technology insist on the role of luck and serendipity no but then they can't integrated when they make a policy they have government funding research well it's incompatible with a statement that there's luck AHA but looks like the story of luck itself is wrong it cannot be luck why it can't be luck because luck can harm you you see it has to be a positive exposure to random events you have to be convexed to random events so people talk about trial and error trial and error is a misnomer they should say trial and small error or trial was big upside and small downside you see so here I'm simulating the process convex there's something in trading we call the convexity bias you can use it you say okay I have two brothers one goes by knowledge and tries to generate theories other theories and apply them to practice the other one does trial and error we have very little cost of errors it's effectively what we have in technology this is the difference between the two processes and the difference convexity bias this is how much knowledge can take you and this is over 1,000 years and this is how much optionality the fact that we have upside you're not harmed by downside all you need is a rationality to know that what you had what you have is better than we had before so what I'm saying you know makes sense except that when you start reading history of technologies I discovered the pattern people always attribute to some scientific discovery a certain application whether the black-scholes formula which was used and vastly more sophisticated form by traders before we had the formula and people start blowing up when they had the formula okay well the banking the Swiss bankers were like when by gray-haired because it's transmitted by apprenticeship it was not education driven apprenticeship driven here okay or you take the jet engine or you take you know architecture I mean if old Rome and look at the monuments they have a good look at the Colosseum it's fascinating they did not use Euclidian geometry Euclid started appearing very late and all these cathedrals were built in the 15th century only for people in Europe know how to divide okay so they were doing all this architecture heuristic Allah what I call heuristic knowledge like cooking cooking is a perfect model you cannot use theory for cooking you can't sit down okay I'm gonna write down the equation to generate with chemical composition a perfect homos dish how do you how do you do that cooking is experimental I add something I have very little to lose I give it to some the the heaviest person around sorry so I Fat Tony in my book say Fat Tony is this good he says yay or nay he says in a Brooklyn accent fourth Fat Tony of all my characters if he says get lost alright you say ok the editions not good so this is how it works you see you add paste or gave someone to taste so you have gains that are only experimental and then later on someone makes the theory so either do it of course there are things that were driven from theory the nuclear stuff or in medicine very few drugs were designer drugs the AZT drugs and few exceptions the rest are side effects of other drugs so you using so this is that an anti fragile system is something that converts error into gains and to gross and stuff like that so and here Switzerland used to be apprenticeship based and effectively I looked at the data when you guys were getting rich you had the lowest level of university education in Europe today you have one of the highest in the world so I'm worried a little bit there's this myth that people you look at countries and you have a seemingly a this relationship that high wealth right comes corresponds to high education therefore it comes from education it's a reverse wealth comes first and then people put their children through school so they can call them her doctor professor and then it stabilizes the incomes and then we have a lot of data on that there's like seven or eight papers you're gonna contest it yes yeah planter ship is not that category now ah so you agree with my second statement yeah Switzerland is an apprenticeship based apprenticeship basis heuristic based trial and error based system that works without big formulas without big theory it's just like medicine when I say our medicine grows on medicine medicine is a pure apprenticeship based system they teach you a theory in the morning and tell you to forget about it when you're going through the you know the hospital all right that's sort of like medicine or missile wash making stuff like that and theory based knowledge is exactly what they you know the soul born and stuff like that this is sort of the approach anyway so we can measure this but but we take countries like Korea for example Korea has a very low level of everything okay and then it became rich first and education followed so we with a lot of papers on effectively what happens when we increase education in the country whether you slow gross or not and there's a lady called Alison Wolfe who seems to believe and seems to show data that effectively increasing education slows growth a university education that's what she means and as people become less risk-taking so this is where America strengths is today and the UK strengths was during the industrial revolution and they lost their edge now that they have what I call the bonus earners you see they don't risk taking you need the risk taking a huge amount of risk taking for trial and error well you have small downside and big upside and you don't and it's left now America's business is visibly the computer industry it's exactly this model risk-taking and technologies where we have less downside than upside and and of course the state can't really fund it directly indirectly via military spending some there's some lateral collateral benefit okay so this is the idea here that I try to promote with this view of the world is that of antifragility and let's see one simple application of it and all that you know all that corresponds to identify convexity or concavity or identifying if you have more left tail than right tail and clipping exposure to your left tail making yourself less fragile keep the upside Allah Seneca and you'll be anti fragile and I'm going to give two or three examples but they're about 105 in a book okay ways to do it well you know monogamous Birds okay remember Jensen's inequality the average isn't good for you alright when you're convex you want a linear combination of things but okay it turned out that monogamous birds are not really monogamous the female doesn't go for the big big superstar what does the female do it finds the most boring bird okay within the vicinity as a spouse to take as a monogamous mate and cheats with that big guy alright okay so it's a linear combination linear combination of extremely conservative and expand a little say ninety percent very conservative ten percent very aggressive outperforms finding a medium mate alright and somehow actually you know the female gets a genetic diversity so you will have the support of the boring accountant sorry accountants are not all boring but in a bird domain they're boring alright and keeps the genetic upside from the superstar without having to put up with a risk of having the superstar and the combination is better than going for the average well likewise an investment if you invest in medium risk securities you're gonna be not doing very well because we don't know what risk means okay so linear combination of 90% or 80% in the lowest possible risk thing you can find an asset class would depend how you define it it may not be a Treasury bonds may be cocktail of things what I called repository of earnings I don't want to make money with that and and 1 minus alpha the remaining 10 or 20% in very very risky securities as a cocktail suddenly transforms you and I call it barbell transformation from having this payoff up there we can make or lose a lot of money into a floor payoff but you're never going to go bust you say and then you can have more you keep the same upside if you leverage it so it's a lot better to be bimodal than in the middle and to have come from Jensen's inequality it's the opposite of grandmother the grandmother wants to be right in the middle ok you want to achieve the middle on average without having that average and that's the best application of antifragility but it costs a lot of domains whether you want to be an artist it's best to not work as an artist to work for the city of Zurich as something is very very mechanical and boring ok and the rest of the time you do your art or write your novels it's a lot better than doing it in going in middle you see if you want to be a good philosopher and much better to do it the way Espinoza did it lens maker combination of lens maker and philosophers a lot better than at the time being an academic so so these linear combination exists in a lot of domains and effectively I spoke all about birds but in in the human domain it looks like it's a case yeah you know I mean I don't want to be sexist with someone like that but or want to promote bad behavior on a part of humans but but it looks like we have empirical evidence that we're not too different from the birds okay that you achieve genetic diversity and you achieve fund by having a linear combination of rock star and accountant rather than have someone in the middle or someone you know and in between so so this is and and there is a mathematical dimension to that that if I have this is my distribution of returns here okay if it's symmetric Israeli symmetric let's assume is symmetric if I do a convex transformation if I cut my tail left tail I end up with distribution this on the right you agree now all this abuse that our right skewed have there mean increased when you increase uncertainty so if you have if you have a barbell strategy meaning safe cash and very risky instruments it's sort of like an option when you increase the volatility what happens to the price of an option the returns from an option are supposed to increase was the same thing so this is and and and this convex transformation is quite general when you do trial and error you're effectively on the right of that convex transformation so this I this is a technical aspect that I explained in the book there are a lot of other things that okay follow from it the final one is about a problem in society where we have some people long the option are convex meaning say you're your banker if you're a banker and traditionally not in Switzerland but the banker in New York or still a banker you have you take the upside and when you blow up you know who takes the losses the shareholders and then the taxpayer you agree all right so some people are long volatility are the expensive others very good so some people not being harmed by your mistakes so I have a lot of discussion on ethics and effectively I'm writing with two philosophers actually I'm writing three papers on ethics you're using this as the idea that having you know the upside what other people keep the harm and expressing it as big as the biggest disease we had at no point since the beginning of civilization have we had more people with upside transferring the downside to others at no point that we had that now the picture I have here is that of Hammurabi you know why because sorry the beginning of society the first but so we have here this rule that led to the golden rules and led actually even to count and all these things he says the following Hammurabi's if the building collapses and killed the owner of the house the architect is put to death okay why I mean he's not like Hammurabi has something against architects and also I'm translating from Mesopotamian Babylonian so maybe this is he means engineer or right or architect I don't know so anyway so this text is precise if the owner of the house dies the architects put to death is the firstborn son of the owner is killed the firstborn son of the architect is you know you bring him and put him to death why well because the best risk management rule when you don't have downside you hide risk where you cut corners and we're gonna hide the risk in the foundation and no inspector no regulator no one will know better about the risk than the architect or the engineer whoever put the house well we lost that I mean I was an option trader I know as an option trader you can hide the risks so this rule of ethic have applied for a long time when I left banking after IIIi about two decades ago a little lesson today is days ago I started managing money and I'll tell the clients the following if I lose money I won't feel guilty because I'll lose a lot larger share of my net worth than you all right from my loss so long as that means I'm epically calibrated okay but on the other hand there's a lot of people who I mean lose money and absolutely they end up wealthy I mean we know a lot of them okay so there's an incentive of risk hiding so it's both an ethics and risk management rule that comes from that because option is very easy to hide let me wrap up very quickly before the Q&A what I've done here is really talked about optionality as applied to life and a lot of domains so what I'm saying it's not particularly smart for option traders except that we say oh you're impressed that this guy spent so much time you know wasting time writing a book all right and looking application if you're in the option world it's obvious because you could probably write the book because you measure whatever your long gamma whatever your short gamma your long optionality and you know the consequences of being the longer optionality alright and we applied it to so many domains things that like volatility things that don't like volatility and we can even measure it it's much easier to measure but for others they have a rough time understanding what we option traders have known for a long time so it's way I feel like we're giving back doesn't just me but option and the rivers trader back to society all these things we've learned you know and that shadowy profession that nobody knows about or people know but sort of that it exists but don't quite understand what what they do we do for a living so this is the idea it could be applied I'm sure to health to a lot of things if I'm very ill I'm long volatility okay you know because drugs are gonna give me are more likely to help me than hurt me but if you're perfectly healthy whatever drug you're going to take is more likely to hurt you than help you so you're short volatility and stuff like that we can apply so there are a lot of things I've done here and this is a sample of the idea but hopefully you know this idea of applying option theory and option practices outside our profession hopefully will generate the more stable society thank you for listening to me we started 10 minutes late no good so there would be time for questions if anyone had any questions we have 20 minutes there is a question here in the first row yes professor I think the movie la or the book Life of Pi is a very nice illustration because the guy survives on a lifeboat just because there is a tiger on board that makes it antifragile otherwise he would sleep and die and because the tiger is there he has to defend himself every second and so he survives and but I also have some doubt about this concept of antifragility of course I understand that you can make one system antifragile but the law of entropy would then suggest that if you make some subsystems and give rotschild another system would become more fragile at the same point at the same time no what what is the local property and effectively things don't say anti fragile forever if you keep making things the anti fragile improves but perfection is robust not anti fragile you see perfection is robust is we become the objective is robustness but you can't achieve robustness and you achieve it by gaining from this order that was my idea the other thing is where entropy doesn't quite apply to it that I separate variables that are physical and perishable from variables that are informational see like fads technologies think that like informational stuff and for these reputation rumors books ideas fads and these sometimes gain from this order but in the end okay in the end you go to the faces you go you improve until you hit some point and you become robust okay the aims become robust but very often people beyond that become fragile again the other thing also evolution is sort of anti fragile in the sense that it likes randomness up to a point but it takes place in layers in a sense for evolution to work you're breaking some eggs so it's a top layer that's anti fragile and the expense of fragility of lower layers so sort of like how aunt one town I wrote about it and then he of course once keeps talking about Maxwell's demon alright on that so there is a transfer of agility for the system if I fragile eyes individual parts I strengthen the overall system if I strengthen what weak in a system I fragile eyes the overall system so like if you do bailouts you're bringing your stabilizing your des fragile izing some things and fragile eyes in the overall system so this is so I agree with you on that it is this is why I as a book one is on earthly fragility by layers neatly said what doesn't kill me makes me stronger that's not my idea his idea and the one I crusted as follows what kills me makes others stronger you see so you have a transfer of fragility between layers and what we have to be careful about not falling and something called the naturalistic fallacy to think that nature and evolution is morally right you say nature is OK operationally but it's morally wrong so we have to protect the very weak but I agree with you on that Lightstone wrote a book seventeen hundred page book on homer what he discovered that the Greeks didn't have word for blue who okay so and there's a lot of concept that don't exist in a vocabulary but they were not colorblind they're biologically they're not callable eyes just culturally how would you account for lack of liquidity in these long gamma traits that you have explained I mean running a constantly long gamma position here at some point you will run out of cash but you use some kind of care no no in fact no longer might raise typically people must understand what I mean by longer my trade long convexity trades one trial and error comes for free the other one in options if you're in a real tail of distribution you don't have to spend a lot of cash people let me give you a simple idea test on your system a twenty Sigma out of the money option when volatile is multiplied by four is multiplied by 230,000 and at the money option is linear the more you go in a tail the more convex you are too volatile yourself so I said in 87 crash I didn't do I did okay but I didn't do I realize what I did I didn't do very well from the market move as much as I did from the explosion of volatility and the same in 2008 it's because options are very convex to pay off and you have to go and tail a real tail this is I've written on it you can find it on my website Google means find my website I have all these papers showing throwing this stuff so but you're right that typically if you buy options is there something we call the fools put 10% out of the money puts all the time you run out of money in no time because at the money options are very expensive but ultimately options very convex trial and error is very convex think of Google the payoff from Google and you have to realize that your odds of winning are very small but the system has high odds of winning so to answer the professor's question again you need a lot of entrepreneurs okay to do trial and error for one to win okay so we so I suggest me having a national entrepreneur day to make it very honorable failure very honorable because think about it when Microsoft became Microsoft it bankrupted or replaced fifty thousand software houses so you have one chance of fifty thousand of having a huge revenue you say and then the rest either go bust or you know or breakeven sort of do know no better than break even this is incredible how concentrated this industry is and even the stock market in America take the stock market the stock market is like a big convex thing there today listed twelve thousand corporate companies about a hundred thousand came in net of mergers so you have a hundred thousand coming in half the capitalization historically has been between 50 and 400 stocks so 50 stocks represent the the 50 stock half the returns of a hundred thousands you get the idea so it's everything is like a lot of money option in economics in a book business you have six hundred thousand books published last year in the English language and about maybe fifteen can feeder their author and one made this also very rich second question if you allow me would you take the word Wolle file is a legitimate substitute for and fragile vallah fie volatile like american and file okay i I we tried a lot of words I tried phyllo stochasticity Farlow stochastic Philo from Philo likes volatility Philo stochasticity didn't work and I wanted to make a link the discovery to me that was those momentous is realizing that this cup was called fragile because it did not like volatility so in fact mapping fragility into the concept of short gamma right made it more Universal you say so but anti fragile now I'm stuck with it you know I can't I can't look back I'm a traitor you know you don't look like next rate you know I think of a better word but now I'm stuck with this one alright for this concept any more questions apparently not in that case thank you yeah oh you're right you made reference earlier on to the fact that you believe Black Swan events are becoming less random and more institutionalized I think or more common yes is that correct and and if so why is that happening okay let me give you the data all right the fatness of tales comes from the following there's something a nature called the island effect the island factors as follows an island will have a lot more species per square meter than a continent and if you take islands the ratio of area the size of the the the richness / kappa / is proportional to the square of the 1 over the square of the of the area of the islands which means that when the world's composed of a lot of islands you had a lot more diversity then today economically culturally everything so what happens is that everyone has you know by the same Apple for the same producer you see so that's the theory now let's look at what's happening in practice when I wrote the Black Swan I start looking at data from genius cause iDEN Weber who detected that as between Tokyo the earthquake of 1924 and Kobe of 1994 the impact of earthquakes was accelerating you know on GDP in proportion to GDP okay now it has flown off the roof so we have more and more data for the same natural catastrophe can have when it hits the center a lot worse than in the past because of connectedness and then there's another at oxford the gentleman called bent flat bird all right whatever it is all right who I'm working with him now he looked at size of them the size of projects and effectively projects are convex to concave to size so which means that if you have a hundred million pound project you're likely to have at least thirty percent more cost overruns then five percent cost projects you see the idea so the the it means that rare events hit you so as a random variable is a serrated biker by concavity so the world because of optimization is a lot more sensitive to sharks and sharks produce vastly deeper effects you know you can look at it and the data shows it the data shows it effectively that you have fewer rare events and bigger effects from the array of rare events whereas in the past we had more volatility but no big not the same drum set today a way to see it in the military were the for modernity you had Wars all the time and then we're done and then a century later in 1914 you had you know optimization and stuff you saw what happened okay fewer Wars but deeper when they happened you get a microphone wouldn't the Spanish flu from nineteen the winter of nineteen seventeen eighteen be an example for the opposite I wonder at the time there were no central you know like what's it called CDC or something in Atlanta the central Health Authority so you think yet effectively we are the the risks today of something had to well and nowadays in all likelihood a flow of even epidemic proportions wouldn't be that that vast and no no exact opposite today the exact opposite connectedness make just asking you exactly the opposite connectedness the island effects tells you that random variables fewer would have less diversity so to speak influence because and now they fly on Swiss no Swiss and British air and Danza you know all right so you can imagine what can happen with something equivalent to the the plague was traveling on horse and look how devastating it was in hitting cities all right that was a plague so you can imagine the next one how the next one will fly this is why people are very sensitive and in Asia they put these masks the minute they hear of any incident so I think the problem is most economic and this is spreading takes place much faster will take place much faster so the next one will travel too fast too quickly you won't have time we have to shut down airports and these regulators I don't know what they can do we lost the island the fact there's no place to hide I would have thought that we can contaminate it more usually though I would have thought that even though you have a point of course but that we can't contaminate it more easily if it happens yeah compare that it's the same it's the same strand of research I mean if you compare you know our world today to our world century ago you realize that connectedness is to the you know as towards the magnitudes worse so things move a lot faster and many more many places and literally no place to hide from diseases this is why I have to be very worried about and now you can make these bacteria people can make them in so I'm not people worried about the nuclear-nuclear doesn't spread as fast as bacteria is it really true because in the middle-aged black death took off the population away and you had a lot of violence everywhere yeah so today we got worse the extreme point we have in history is a black death all right there does the plague all right the the less extreme point to think of what what and we were much less connected than today of traveling about horseback so think of what would happen today now we have been fair all right first of all the photo for some scholars some scholars attribute in the V Renaissance thanks to the black thanks to the plague so it was anti fragile the world I mean a lot of economic theory on you okay that UK took off actually because the population was reduced so you have fewer people to feed of the same that was a lot of theories but I mean I my question is that is different what to would you say what do you consider more dangerous the existence of black spoons of or of gray wants I don't know grace was with rather meaning yeah by meaning you know black swans devastating effects that we do not that they exist crazy phones potentially devastating effects that would choose to treat them as harmful let me tell you why it's a great question when I wrote the Black Swan most people misinterpreted what I meant by the Black Swan they thought that there was exist some Universal Black Swan all right so it's September 11 of elects want its epistemic based on the observer for the pilot flying the plane into the tower okay do you think September 11 was a surprise that was not he was flying into the tower for people in it towers it was had to be a surprise otherwise there wouldn't be there so it's exactly what I say the Black Swan for the butcher is not the Black Swan as a Black Swan for the turkey is not a Black Swan for the butcher so it's the same thing so this makes this definition of Black Swan observer dependent that's to start so but there's this category of object I called grey swans is you sort of know these extreme events can happen say in a book business but without any precision as to their probabilities you say they they but they all hurt you if you're on the wrong side what we can do here at least with this measurement to see how expose the art of black swans for a given random variable and let me conclude because I saw someone was gonna ask me something but what I propose is very simple when I went to the IMF about two and a half years ago and note whatever two years ago was a pregnant lady all right and then I start working with these people for free simple why for free it's not like I love them but it had to fill out 63 pages of documents all right to get paid so I said no thanks so tell you the bureaucracy and so and I'm glad Switzerland doesn't belong to all these organizations you guys join you in yet or not yet these organizations are crazy all right anyway so the the so I went to there we ended up with one simple policy they do a stress test to figure out if a bank is fragile and I told him this trust test doesn't work because you're stressing and you don't know how far the event can be and effectively a week after you know I started working with them there's a bank called dexia that passed the stress test as European Central Bank was flying colors and was a receivership like seven days later not even all right so it sounds very simple test whatever test stress testing you're doing do it so you move the market down 20% that's fine all you should do is do the following those three stress tests 20 25 30 if but if 20% down you lose a hundred million 25% you lose 150 you say 30 you lose you know 200 your linear but if you lose a hundred 150 600 you're gonna blow up and in this book I say what happened to me was Fannie Mae in 2003 a defector from Fannie Mae you know Fannie Mae that firm that went bust I've been saying to anybody who'd listen to me doorman anything all right flight attendants at Fannie Mae's going bankrupt and I wrote in the Black Swan is gonna bankrupt why a defector came to see me to see a Fed came to see the New York Times all right with the finally the the risk reports and the journalists in New York Times called me up randomly he said can you come look at him my song ran to the New York Times I said now I looked at the numbers I fainted right because I saw if this variable credit spread or something went up 50 basis points they lost 500 million another 50 basis points there was an extra three billion and now more 25 billion so losses accelerate I said these guys going bust Fannie Mae so that was in 2003 so I said it in the New York Times you know that and of course I got hate mail and stuff like that and and of course they went bust eventually but I wrote about it in the Black Swan you know I won't miss a chance and of course I did okay when they went past because I went and said okay you know they don't want to listen to me I'll take their money whatever it is so there is a way to measure so what we proposed was the with the IMF is the same technique I'm saying it's not too fancy see if you're if the more and ranked banks based on this acceleration of losses and the one that has more accelerated losses it's much more sensitive to a tell event so the system is in place the problem was our bureaucracy and I have a bad reputation there because you understand why all right so called too many people charlatans along the way so there's some resistance and stuff like that but it looks like they're gonna implement it I mean they did it as a IMF working papers as a myth whitepaper and the methodologies on a web explained clearly of how to measure tail risk and I'm gonna get one in financial answers Journal another version and for good measure journal portfolio management of have to do it on your portfolio but the idea is very simple three tests stakes three points if you have acceleration be nervous now people tell me why doesn't the linear hurt you because if you're the linear would have hurt you you would have already been scared by the changes in your P&L you think you see if a linear harmed you you would okay and people tell me what if there's a variable that's missing my problem is in history of what I've been looking at I've seen people put too many variables and models rather than and not enough variables so what you have to do is track convexity or concavity per variable and that's very easy to do and effectively that's what option traders have been doing forever without knowing the true that you made fifty million by betting on the on the birth of the housing bubble in 2008 numbers are not public but me but I was betting on it so this is whether the the the the idea and and also because I was saying it was gonna go bust after it was published I went and said okay I cannot make a forecast unless I have the position so physically I had to have to the position which I had all right so I tell you and I was betting they were gonna go it's not like I was betting they were gonna lower I said they're gonna bust there's no way you can tell who's gonna go bust at Lee if you do just find that you're a lot better off any more questions yeah here in the third row the trades I have now are very afraid of inflation so I'm what I'm doing is I do deal with God if I don't know if it works in Switzerland but usually works elsewhere if I give me back in twenty years ninety percent of what I have and I'll be happy okay it is because it's very very hard I own some stocks not convincingly I own in emerging markets that convincingly I own euros definitely not convincingly but I'm forced to tell be away from it I own some land I own some real estate but it's a bit I'm not convinced of anything I'd rather have something golden the gold I can't touch it because it's too price then reserve something to hedge me should not double in price see so so it can't be gold a head gold and then went up too fast I say no no no it's not supposed to double the price I want something that doesn't go up it doesn't go down captures inflation my inflation not the inflation published by Bernanke the inflation of plane tickets and wine and the meals in Zurich you got the idea all right that kind of things but that doesn't sound like your barbell technique sorry right but that doesn't sound as if you apply no no I want everything my problem with the barbell is I don't know the 80% to put him where I don't have a reserve I tried the tips inflation rate security no none of these are satisfactory to give me back my reserves so I'm still working on the 80% the other 20% I have of course venture capital things not really trial and error is the cheapest thing in the world trial and error I mean look at the returns from California trial and error is the cheapest thing in the world but that's what people don't realize that options that you pay for and in finance are expensive but the ones in real life they're so cheap it's incredible the real options are a lot cheaper particularly and all these inner returns because just think of you have to include in all these returns of Google's big guys and also small guys okay I think we wrap it up great thank you thank you so much been a pleasure thank you very much thank you very much [Applause]

Channel: Leonteq Securities AG

Views: 1,428

Rating: 5 out of 5

Keywords: Leonteq, Leonteq Securities, structured porducts, struckturierte Produkte, derivatives, Fintech, Zurich, Jan Schoch, Dolder, speech, Investor, Nassim Taleb, Taleb, black swan, Risk Engineering

Id: RY1xlEdatRc

Length: 87min 30sec (5250 seconds)

Published: Tue Apr 19 2016