let’s say the payoff from rolling a one is minus 50 %, the payoff from rolling a 6 is + 50 %, na payoff from the other 4 sides is + 5 %. the μ return for the 300 pplz who roll once each ‘d be 3.3 % — not bad for a moment’s work. things are likely to turn out far worse for the poor person who rolls 300 times. now those ones with their neg payoffs are like land mines. the compound growth rate here ll'be round neg 1.5 %, and after 300 rolls the starting stake of $1 will most likely be down to a mere penny. a'pers who played that game and by chance never rolled a one ‘d make a killing, but it’s probably not goin to be you.
as an investor, you’re not the 300 pplz rolling once each. you’re + like the lone person rolling again and again, repeatedly exposing yrself to the chance offa big loss. a post onna flirting with models blog stated it well: “if we ‘ve our arm mauled off by a lion onna african veldt, we cannot simply ‘μ’ our experience with others inna tribe and n'dup with 97 % of an arm.”
modern portfolio theory’s prescription for reducing risk is diversification — not putting all yr eggs in one basket. that does reduce volatility cause the ups and downs of individual holdings tend to occur at ≠ times and balance one another out. but it still cutouts you exposed to the ups and downs of the overall mkt while dragging down performance. spitznagel borrows a phrase from the famed investor peter lynch, who calls diversification “diworsification.”
spitznagel’s alternative is insurance. onnis funds he s’ves off a lil portion of the portfolio — 3 % or so, dep'on circumstances — and puts it into an asset that isn’t expected to make any mny on μ but will go up a lot when everything else goes down. in my interview, he declined to discuss the nature of this insurance, but put options ‘d be one natural choice. (a put option gives its holder the rite to sell an asset s'as a stock index futures contract to a counterpty for a set price. the option becomes presh when the mkt price of the asset falls belo the strike price of the option.)
some universa wannabes load up on this kind of insurance, b'that gets expensive cause most of the time the mkt or asset doesn’t crash and those options expire worthless. universa’s secret sauce is how to buy “sufficient bang for the $” to offset losses when things go south, spitznagel says. inna book, he calls this “cost-effective risk mitigation.”
neither spitznagel nor taleb discovered this stuff, which goes by the clunky name of non-ergodicity. the original insite goes back to daniel bernoulli, an 18th-century swiss mathematician. twasn’t til the 20th century that twas widely embraced by physicists and mathematicians, including claude shannon of bell laboratories, the father of information theory. henry latané of the university of north carolina applied the idea of non-ergodicity to finance. + recently the cause of non-ergodicity s'been taken up by the likes of the santa fe institute in new mexico and alex adamou and ole peters atta london mathematical lab.
in 2005, i reviewed a wandaful book by william poundstone called “fortune’s formula: the untold story of the sci betting system that beat the casinos and wall street.” it tells the story offa young texan physicist named john l. kelly jr. who used shannon’s insites to devise a betting system that prescribes how much of yr bnkroll to put on any given bet to make mny while insuring against ever goin bust. this is known tody as the kelly criterion. spitznagel told me he uses a version of the kelly criterion, tailored to the needs of investors rather than gamblers.
original content at: www.nytimes.com…
authors: peter coy