Wednesday I gave a brief talk (audio, slides) at the annual meeting of the Society for Risk Analysis. It seems many risk analysts are like futurists in disliking numerical/probability estimates, preferring to qualitatively discuss “scenarios.” They note one can’t think of all possible relevant events, and point to past numerical estimates that now seem way off.
Inferring the probability of the world ending from your option price is not nearly as simple as you make it out to be. The underlying asset of your option is *not* the world ending, but rather the world ending plus a lot of other stuff. Maybe you can account for some of the other stuff, but if you don't have a complete asset structure (and you almost surely do not), then you're going to have a biased estimate of the probability of the world ending.
What I described was simply an option. Do you not think option prices give information about underlying assets?
The interest rate on that loan is affected by dozens of other factors and so will not give you reliable information about when the world will end.
Jess R., you and I couldn't agree more on this topic. It's why I favor high levels of redistributive/pigovian taxes, for things like subsidized prediction markets investigating "high impact events".
say I want to bet against the end of the world. Nobody will take that bet because nobody will be around to collect.
You can loan out money now on the condition that you are paid back with interest after some ostensible end-of-the-world date. This is equivalent to betting that the world will not end.
I think this is implausible.
Prediction markets for "disaster" events don't work very well. As an extreme example, say I want to bet against the end of the world. Nobody will take that bet because nobody will be around to collect. So, prices in that prediction market would tell us nothing about the probability of the world ending.
In this case, if I buy a refuge future, the probability that I'll actually get to use it in the event that I need to is quite small. Prices in this market would be a poor predictor of the probability of a major disaster.
This is worth investigating, of course. But the key benefit of prediction markets, as I understand it, is that you can subsidize them. The quality of the info being fed into any kind of market (whether natural, like that for bomb-shelters, or artificial, like subsidized refuges) will be limited by the resources that the participants are willing to invest into researching/calculating/theorizing. It's likely that investigating low-probablity/high-impact events is not usually worth it for self-interested individuals, so society (through governments or charities) needs to subsidize the prediction markets to provide the proper incentives.
Sounds worthy of trying out, to me.I wonder if we can use "natural" data to get similar approximations (# of bomb shelters/panic rooms being built, population outflows, etc.)For example, what can we see from Korean data right now, wher existential risk seems to be increasing. Was there anything discernable in the period of time before the political change was obvious?
What if there was a chance that the refuge would be sealed when no catastrophe?It would at least test them, and their inhabitants.
Hmmm... Did you happen to have this particular organization in mind? See http://www.gizmag.com/vivos... and http://www.terravivos.com/s... .Price is $50k per person, apparently. I wonder what they would think of your (quasi airline-style) overbooking idea?
I did not propose that "large organizations and/or governments" run these refuges or markets.
Robert, Robert, and Adam, the company I linked to has sold a thousand refuges so far. What fraction of those refuge slots do you think are for people who personally paid for such refuges, or their close friends and family? I'd guess at least 25% of those slots are for others. Which suggests a non-negligible chance a ticket owner could use it, which is all we need to get meaningful relative prices.
I agree with Robert. The incremental risk that one's "right" to entrance in the refuge would not be recognized in the event of a disaster would destroy the futures spread (and the predictive power of the scenario you propose).
What are the odds that four years will be enough time, even assuming that a refuge is the right strategy?
Also, The Digwell Carol.
I'm going to go out on a limb here and assert bluntly that this simply won't work. Consider that the kind of people who are truly interested in refuges also tend not to be the kind of people who would trust large organizations and/or governments to manage or run them. And frankly, I sympathize with their viewpoint. After all, your apparent assumption that law and order, including contract law (!), would prevail in times of *extreme* societal distress/disorder is, in my view, flawed, to say the least. Consider, for example, that contract law was already blatantly overruled in the process of the GM bailout, which was not by any remote reasoning a TEOTWAWKI ("the end of the world as we know it.") situation. And really, who is so naive to believe that his or her "ticket" to a lifesaving refuge would actually be honored, while the (presumably-government administrators/policemen) at the doors to the refuge would stalwartly block their own families and friends from entering instead, all to preserve your contracted space for you? And who would trust these same people (and their likely-unionized employees) from occasionally (or more often) looting these refuges of their valuable foods, medicines, flashlights, etc, in advance? OK? A better solution: Private, small-group, out-of-the-way refuges, built, operated, guarded, and maintained entirely by the people who pay for them, who know/trust each other well, and who intend to use the facilities personally if/when they need them. This may seem boring and non-grandiose, but it is also workable, since some of it is already happening. And if you want to see more of it, then maybe you could offer people a tax break or some free land or something. (But then, don't count on them actually taking the tax break if it requires them to reveal where their refuge is, or let the government inspect it, etc. See what I mean?)
Seems like there's too much political risk for these futures to give you a reliable price. You're really paying for the opportunity to assert that the future you bought 5 years ago entitles you to a spot in a refuge ahead of some pregnant woman. Good luck with that one...