Buy Insurance—Bet Against Yourself
My friend and housemate, User:Kevin, makes a very pleasant living selling opioids on the internet, a living he expects to continue for some time, unless something awful happens like Obama losing the next election. The Intrade contract for Obama’s loss is currently trading at 42% -- what can User:Kevin do about this?
My suggestion was that he bet heavily on Obama’s loss. Say he spends $4200 buying not-Obama futures. If Obama wins, that money becomes worthless, but he gets four years selling kratom regulation-free. On the other hand, say Palin takes a surprise victory and institutes draconian regulation on various substances—User:Kevin’s $4,200 has just become $10,000, leaving a $5800 windfall to help him while he finds his next muse.
This is nothing more than what we normally call buying insurance, just extended to whatever outcomes you may want to insure against. Let’s talk about some of the effects of this action.
Leaving a Line of Retreat
Let’s say that User:Kevin now finds Obama’s loss more or less unthinkable (I don’t mean to impugn his rationality—the article’s more or less fictional from here on). Well, that’s not really very good—he needs to think about it. I suppose the proper answer is something like “If Obama will win the election, I want to believe...”—but it may also help to think that, yes, a Republican win would suck, but it would also come with this huge financial windfall up front. That is, by flattening your outcome curve a bit, you can reduce your attachment to individual outcomes, and help yourself to make more rational judgments.
Of course, if the market gives Obama 60%, User:Kevin should probably just believe that. That’s what markets are for. Which brings us to:
Helping the Market
You might think that User:Kevin’s (hypothetical) actions are antisocial. That by buying not-Obama futures, he’s sending a false signal to the market decreasing Obama’s chances. If so, Robin Hanson might like to have a word with you. Here’s why:
Let’s say that Clippy is a rational speculator in this market. He doesn’t care whether Obama wins or not, he just wants to maximize his monetary outcome so that he can purchase materials with which to create paperclips. Let’s further say that the market is filled only with rational speculators like Clippy.
Well, based on the family of no-bet theorems, they will all expect that the best-informed actors in the market will make their money from the least-informed actors. Half of them will conclude that they are the least informed, and choose not to play. Eventually those speculators best equipped to find good information and make good speculations will be alone in the market, with no one to bet with, and no incentive to produce good information. Market volume drops, put/call spreads widen, everyone goes home, no one gets informed.
Now let’s bring lots of noise to the market: irrational actors convinced they know better than the going rates because their star sign told them so; agents of one political party or another intentionally manipulating the market to energize their base; rational insurance-buyers who (may) have good probabilities, but have skewed valuations on money and are willing to accept above-or-below optimal prices. Now there is money in the market. Now there are buy and sell orders from all sides. Now it makes sense for Clippy to spend lots of computing cycles running regression tests on past electoral outcomes, buying and selling whenever the advantage is to him, making lots of money, and pricing the market accurately. Volume shoots up. Spread goes down. Prices move near-instantly with good information. Everyone is informed.
For all these reasons, gentle readers, I urge you to wire some money into an offshore account, log into intrade, find some outcome that would make you miserable, and bet heavily on it.
Why would Obama’s prospects have any kind of significant effect on kratom’s legality? Congress writes legislation. Obama could veto kratom prohibition, I guess, but he hasn’t shown himself to be more pro-drug than any other recent president.
As a senatorial candidate, Obama advocated marijuana decriminalization, a position he renounced when he ran for president. As a presidential candidate, he said he would end the DEA’s medical marijuana raids, a promise he broke after he was elected. So far he has waged the war on drugs, which he once called an “utter failure,” in pretty much the same manner as his predecessor, only with more money. The one substantial improvement in federal drug policy since Obama took office is crack sentencing reform, which he supported. But that change was in the works for years and had already attracted support from most Republicans, including George Bush.
I would expect that kratom would be illegalized via a random, trivial action of some government agency, not something at the executive level. Broadly, though there are many, many scenarios where kratom could be illegalized under Obama, there are more under a Republican president. For one reason, it is at least possible that national drug prohibition could end during a second term of Obama where that is more or less impossible under a Republic executive.
I doubt there is much difference at all. Prop 19 was a lot more modest than ending national drug prohibition, yet the White House declared it would not respect state law, and would continue raiding pot dispensaries (if 19 passed). Admittedly, it is possible that Obama is secretly an anti-prohibitionist, and that he has just been acting like the exact opposite for political reasons. But that applies equally to the Republicans. And, of course, since we don’t yet know who the republican candidate will be, there is some small chance that a libertarian-leaning republican will win, and actually be somewhat anti-prohibition.
Of course, this is all overwhelmed by the fact that the president’s effect on kratom legality is trivial (compared to Congress, FDA, other bureaus, state legislatures, local governments, etc.) Basically, worrying about who holds the white house is kinda silly when you consider how widely distributed regulating power really is.
I knew a girl a while back who was planning a trip to Iran but was anxious about it for obvious reasons. I told her to hedge by buying lots of “AIRSTRIKE.IRAN.DEC10” and “IRAN.NUKE.TEST.DEC10″ contracts. She didn’t get it, which is kind of sad because she worked for an investment bank.
I came very close to making a large bet against Prop 19 passing in CA, because the universe where marijuana was legalized would have been A very desirable place for me and my business to find ourselves. If the US Department of Justice was busy suing California they wouldn’t have the resources to spend on scheduling kratom.
Alas, I didn’t (for the same reason I have never bet on In Trade, hyperbolic discounting, basically… I want to make my bet right now, not wire it and wait a week), so I am several thousand dollars poorer than I would have been and marijuana isn’t legal in California.
I’m also reluctant to buy not-Obama futures right now because I think Obama’s chance of winning is closer to 75% and that the market will correct itself eventually and I’ll be able to buy right around the inflection point.
This would seem to suggest that it would be rational for you to be selling not-Obama futures now. Until the market reaches your prediction modified by your risk-aversion incentives and overhead.
Or they would look for politically easy drugs nobody has heard of, in order to convince conservative voters that they’re doing something. Few people have heard of kratom, so it’s a great candidate for this. The more sensationalist news sources will report that kratom is a dangerous, highly addictive drug that is used by some people during sex. Eventually the debate will be over whether it should be Schedule 1, or whether there are acceptable medical uses for it.
On the plus side, though, throwing kratom under a bus might help distract from California’s sadly-hypothetical legal marijuana issue.
This is actually one of the best long term ways for my business to wind down in 10-15 years.
Kratom has a very obvious medical use. It’s an opioid and there is evidence that it has fewer side effects than morphine in mice (less constipation, and less “rewarding action” (euphoria) per unit of analgesia).
There is a patent pending for all possible derivatives of kratom: http://appft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=/netahtml/PTO/srchnum.html&r=1&f=G&l=50&s1=%2220090221623%22.PGNR.&OS=DN/20090221623&RS=DN/20090221623
I think the debate will likely be about whether kratom derivatives should be schedule 2 or schedule 4.
Terrible advice, BTW—too much could go wrong. (When I tried to make a bank transfer over SWIFT, my institution laughed at me and told me Intrade’s supplied information was wrong in some inscrutable way.)
Just snail-mail them a personal check.
(Checks are easy, if slow; this is why I was able to profit $25 off Prop 19′s failure.)
Aren’t checks a US thing? In some countries it can be impractical to even cash out a check.
It’s a pity that their supplied information is wrong, is there other way?
They have a somewhat involved way of accepting credit/debit card (involves you mailing or faxing bills or something, IIRC), and I think they also take money orders & cashier’s checks. But personal check seemed the easiest method for me, all in all.
Cheques aren’t only in the U.S., but they do seem to be rare in most of Europe any more. However, as long as your bank is willing to make a cashier’s cheque for you, then it looks like Intrade can use it.
These are the only bets I’m prepared to make on betting markets—I’m too wary of optimism bias to make a bet where I independently want the outcome that wins me money.
If a lot of other people feel the other way, that might even be a good strategy!
Aren’t there other ways of reducing attachment to individual outcomes that are not quite as expensive as making irrational bets? Can’t you just become not quite as attached?
Surely piling crazy betting behaviour on top of crazy beliefs is not to be recommended.
The point isn’t to make irrational bets, but rather to make rational ones. Ones where the odds you’re getting are reasonable, and you’re ensuring a certain wealth/happiness minimum for yourself.
Just because you consider it GOOD if Obama wins doesn’t mean you can rationally conclude “Obama will win”. In fact any way that you can conclude an Obama victory from your liking Obama is obviously irrational.
However concluding that Obama losing will cause you to have financial difficulties means that insuring yourself against such a loss is very useful.
According to the post: “the best-informed actors in the market will make their money from the least-informed actors”. If you bet without paying any attention to the most likely outcome, you are more likely to be among the least-informed actors—and so will lose money on top of the transaction charges for playing.
Insurance doesn’t appeal to everyone. Many of my countrymen regularly play the National Lottery—increasing their risk-taking, not minimising it. Basically, insurance people always take their cut—which makes insurers poorer, on average. Minimising your risks is an especially poor strategy for males.
This post is trying to sell readers on insurance. However, I don’t want insurance, thank you!
Insurers do take their cut, and it may be enough that insurance is not worth it. However, that’s not certain; while customers will be poorer on average, they may also be better off on average, since money is worth less the more you have (diminishing margin of returns, concave utility function, etc).
Minimizing risk can be a good idea whenever you have a threshold of risk past which you’d have a very hard time returning (i.e. severe illness or poverty).
Of course there are circumstances when insurance makes sense. What the post says, though is:
There is no mention of considering whether you might want to take more or fewer risks. This idea that risks are bad and that people should insure against them is not a biologically realistic one. The council given in this post needs to be taken with multiple pinches of salt.
Insurance is as far as I can tell a hedge with particularly high overhead, but it is a hedge. You’re paying to reduce your exposure to unwanted risk—a move that as you correctly deduce averages out to putting less money in your pocket, but which can nonetheless be quite rational when you take into account the Gambler’s Ruin and similar effects.
There’s also the psychological effects of risk to consider, as well as the diminishing-returns effect that TobyBartels mentioned.
Unless of course you are the CEO of a company with a large contract to manufacture computerized voting machines. In that circumstance, you could also conclude a not-Obama victory if you don’t like Obama.
I’ve heard about people doing this for sports events they care about—e.g. betting against their national team qualifying for the soccer World Cup.
Just a small link for those reading this in 2011 or later: “2011 Intrade fee changes, or, Intrade considered no longer useful for LessWrongers”.
Yes! I told people to do this after the immigration bill in Arizona made them much more indignant than usual.
Choosing so that all paths lead to victory...is recommended.
Is the getting an offshore account important part? How do you recommend getting it? When I needed to transfer money from my bank account, I had to go to the central office of my bank and fill up the forms.
Intrade: “is safely hosted in Ireland, from which it targets primarily U.S. customers, its development is nonetheless hampered by U.S. laws requiring banks and credit card companies not to process payments to offshore betting websites. Nor is Intrade allowed to promote its services with advertising in the U.S. and the bigger European countries.”—source
This strategy would also, if developed further, make legislative bodies impotent. If everyone could smooth out their risk profiles by betting on negative outcomes, then the only effect of any proposed legislation on people’s behavior would be a transfer of wealth reflecting individuals’ indifference curves (i.e. the outcome sets they are indifferent between). New laws would lose most of their ability to change behavior.
(Btw, why “User:Kevin”?)
How so? If someone recieves money when their current business becomes illegal, this doesn’t mean that they’ll carry on in the business, ignoring the change in law. In fact, it makes it easier for them to cease their current business immediately.
I may be missing some major point, so if so clarify, but it seems to me that this would in no way decrease the effectiveness of legislation at discouraging actions after it was enforced: all it would do is decrease the effect of THREATENED legislation (that has not yet been passed).
Okay, my thinking is basically that the insurance mechanism’s predicted effects feed back to the legislature-level, not necessarily the outlaw-level (though it can do that too—see the end of this comment).
In other words, in anticipation of this insurance-like reallocation, a legislature will find most of its potential laws less appealing. For example, if they want to ban something, they are doing it to reward a politically-powerful group and punish one that is less so. They want, e.g. “lots of wodget-sellers jailed so that honorable wodget-avoiders don’t have to deal with the wodget menace”. With the insurance mechanism, however, they can’t get that by itself; at best, they can get that plus a large transfer of wealth from the “good guys” to the “bad guys”—which is much less politically appealing.
Even if it doesn’t have that influence on the legislature, and they persist in such bans, the general effect of the insurance (transfer of financial resources from those supporting [in this case] the ban to those opposing it) may not mean that the sellers of the banned product “take the money and run”. It could just as well mean they can raise better barriers against law enforcement (like what happened with alcohol dealers under prohibition, who could make enough to disrupt the laws against them without them being overturned), allowing them to continue in their work.
So perhaps “impotent” is an exaggeration, but it definitely weakens the power of the legislature through several effects (legislators playing the prediction markets, political unattractiveness of enriching the target of the laws, and the shift in financial power each law brings).
If the ban on wodgets is expected to be binding, then a proposed ban on wodgets with probability p of being enacted does decrease the expected value of wodget-selling capital. Wodget-sellers can choose whether to cut their losses by taking out insurance, keep their risk exposure unchanged or even reverse their exposure by betting more than their existing interest—but they cannot get a free lunch unless they have magical foreknowledge of future events.
Yes, to the extent that wodget-sellers trade their risk with wodget-avoiders, the political positions of either will shift towards neutrality. But this will hardly make legislatures useless.
As for your suggestion that having more money could help against enforcement, well, let’s put it in decision-theoretical terms: do you honestly think that transferring money from the possible world in which selling wodgets is legal to the possible world in which doing so is illegal is actually a sound investment? Because this is what the transaction boils down to.
A genuine problem with political prediction markets is that niche speculators can rent-seek by manipulating the legislation process. I don’t have a real answer to this, but it seems to be a self-limiting problem: e.g. a speculator who tries to manipulate the legislature into enacting some law and fails has wasted effort in addition to losing his bet.
I appreciate your explanation. It certainly adds more dimensions, and while the effect would be even more complicated, and I don’t want to even try to model it (with my limited economic and political knowledge) I now understand your position.
Referring to Keven as “User:Keven” has become somewhat of a convention. Initially used by Clippy but adopted by the professed humans too. The idiosyncracy appealed.
I think Clippy refers to everyone as “User:[username]”, but since he refers to Kevin especially often, that’s what’s most remembered.
This.
Why was this voted down? It seems to be the most efficient way to answer the question, and necessary because it was MBlume’s use that was questioned.
There’s more than one Kevin on the site… it seems slightly presumptuous to just call me “Kevin” in posts.
That with a link on first mention would do the trick.
Um, this is wrong. The effect of insurance on behavior is called “moral hazard”, and it is a market imperfection. People who contract for insurance will be best off if they minimize resulting changes in behavior, and insurance contracts with severe moral hazard issues are simply not traded. (For instance, unemployment insurance is not traded on the private market.)
That’s only true for cases where the insured has significant influence over whether the insured event happens. That’s not the case for legislation, unless you’re an influential legislator or lobbyist; and, to the extent that moral hazards cause a problem there, it still works to make the legislature impotent (since they can no longer trust each other not to use their votes to game the prediction market).
And I think you’ve significantly overstated the impact of moral hazard problems: there are numerous instances of insurance contracts where the insured has significant control over whether the insurer must pay out—the solution is to make sure the payout doesn’t fully return the insured to their indifference curve (e.g. through deductibles), not for the market to fail to exist. Otherwise, how could there even be liability insurance?
EDIT: Easter egg: see if you can find an internet article by me arguing the opposite of what I’ve said here. Don’t embarrass me further by posting it when you find it.
What’s “the insured event” here? If it’s the enactment of legislation, then yes, it is possible in principle that insurance will affect the market for laws, by shifting risk to those most able to bear it. If it is the behavior of people who would be affected by legislation (say, a breach of some kind which gives rise to liability), then my above point stands: the “impact of moral hazard problems [is] reduced” iff insurance does not change behavior much.
Liability insurance only covers negligence and strict liability, not intentional torts; and insurers seek to minimize their payouts by (a) imposing deductibles and (b) monitoring their clients’ observable precautions. In some cases, insurance will actually enhance the incentives inherent in tort law, by shifting part of the risk of negligence to a third party which can specialize in monitoring against it; but this is a matter of technical efficiency.
Degree of caring isn’t actually the thing that makes insurance worthwhile.
Find an outcome that would make you financially miserable, and bet on it such that if it comes to pass you won’t be financially miserable. In a sufficiently non-rational market (ie people betting on things because they want them to happen) the size of the bet will not make you financially miserable if the outcome doesn’t happen and you lose the bet.
I think this is the core idea of the post, and the stuff under the headings is some of the positive results of this course of action, not necessarily a justification for it.
This is reminding me of The Quants, and not in a good way.
It’s a book about some very smart people who thought they could beat the stock market by making large bets on what they thought would go up and smaller bets on what they thought would happen if the first bet was wrong.
They didn’t have enough experience to realize that both their bets could go wrong.
I think the second bets were apt to involve margins.
So it is extremely important that you critically evaluate whether the market is sufficiently non-rational, and that you attempt to find out the possibility of a third outcome that doesn’t satisfy the bet but still makes you financially miserable (eg User:Kevin bets on not-Obama, Obama wins, Obama cracks down)
Yeah.
I strongly recommend The Quants. It’s very interesting about how the guys who invented the systems (notably Edward Thorp) understood the theory underlying the math better, and had clear ideas about how much it was safe to bet. The later mathematicians and (iirc) physicists just looked for the big win.
Financially wasn’t the missing nuance. In fact this works in principle even for causes of non-financial misery. But the feature that is needed is risk aversion. Or a non-linear utility to money relationship. Anything that makes you care more about a winning bet than a losing bet of the same amount. Caring (potential misery) is actually not directly relevant.
Don’t forget about liquidity constraints.
For most people in the West, $10,000 is a negligible fraction of their lifetime wealth, so it is not rational to be anything other than risk-neutral about it. But User:Kevin needs his $5,800 as soon as the election ends, not over his lifetime. He may or may not be risk-averse, but he is definitely liquidity-constrained.
To the downvoter: If you don’t believe that risk-neutrality is rational for relatively small sums of money, I’ve got some extended warranties to sell you. And if it was legal, I’d throw in some tickets for the local numbers bank as an alternative.
I’ll admit I can’t make much sense of what you’re saying, but
this is already in the post—you care about a winning bet because it saved you from hard times, you don’t care about a losing bet because you profit in other ways from this outcome.
I didn’t disagree with the post, nor suggest the post was lacking. I pointed out that the concluding exhortation misses the mark.
It absolutely does. I sacrificed some precision for clarity so that I could end with a ringing exhortation. When I have a moment I’ll probably footnote this.
Honestly, part of me is still a little confused about what I’m supposed to do at the ends of essays other than stop talking when I’ve said all the stuff I have to say.
ETA: On further reflection, the exhortation is almost right. The target you want to optimize for is “outcome in which money is worth more” but “outcome I’d really hate” is a cheaper target to compute—it’s emotionally salient, and can be quickly processed, probably in parallel—while still being a decent pointer to the true target—you can use a deliberative, serial process afterwards to pick the outcomes you actually should bet on.
Exactly!
I like this article, but the formatting is really wonky.
I know, it looks good in the editor, and I’m still having trouble making the spacing look right on the page. I might just poke at the HTML directly and see if that helps things
ETA: There’s a reason I try to stay away from anything at work involving that thrice-cursed markup language.
This seems like a long chain of inference—imho, Palin is likely to be chaotic. If there’s money to be made in making kratom illegal, she’s more likely to do it, but I expect that the prison-industrial complex will focus on harsher enforcement of existing laws long before they get around to making more drugs illegal.
People other than me seem to be consistently more optimistic about the future prospects of my business. I’m not sure what to make of that.
The Eye of Sauron might turn to face you, but I think it will be more a random thing than something predictable by a change in administration.
How pessimistic are you generally?
It wouldn’t surprise me if there’s a bias for people to worry about things going badly for themselves (high stakes), and for them to be reassuring about things going badly for other people (improve the emotional atmosphere). This doesn’t say anything about who’s right.
What do other kratom sellers seem to think?
Not very pessimistic compared to the average LWer. I have weird ontological views that make me extremely optimistic in the larger pessimism/optimism spectrum, but applied only to myself I’m still more optimistic than average if only because my life so far has shown that optimism would have been more correct than pessimism in retrospect. I did not expect one year ago or when I started this business almost four years ago that it would be doing as well as it is.
We don’t talk about it. I should probably ask, but let’s just say that there is a status quo of cooperation among the kratom vendors (the specifics of which I don’t want to get into here) and it seems much simpler to maintain the current state of non communication. Though there are a few vendors I could and probably should ask.
My opinion that kratom will remain legal indefinitely if marijuana is legalized was an outside opinion from a rather accomplished drug legalization lobbyist.
Inside view vs outside view, probably. You are looking at the myriad ways in which the system could absentmindedly crush your business; other people are extrapolating from their knowledge of previous similar cases.
Possibly it is just social convention. You have to be pessimistic about your business because it’s rude to gloat, and other people have to be optimistic about your business because it’s a form of well-wishing.
You also have to pessimistic about your business to make appropriate insurance decisions! (And not just literally buying insurance, or making side bets.)
I feel “but” here is misplaced. It should indicate tradeoffs or opposing arguments, but in this case you are just enumerating separate elements of one of the two possible outcomes.
His but is fine.
“Bad thing but better thing that more than offsets it.”
blushes
Thanks, that’s always nice to hear. =)
Glad you appreciated it. I put work into fitting in that pun without making it seem too contrived. :P
Decisions should be made not on absolute badness or goodness of any given outcome, but on comparative goodness of alternative outcomes following from alternative possible decisions.
On my reading, the comparisons are there, but they are implicit. Here are what I took to be the implicit assumptions. For every state S of affairs under consideration, if the money is worthless in S, then S is worse than the all-else-equal state S’ where the money is not worthless. Similarly, in any state T where kratom is unregulated, that state is better than the all-else-equal state T’ where kratom is regulated. (In particular, states are being considered at a low-enough resolution that you can just flip the “the money is worthless” or “kratom is unregulated” bits and get another state. Also, while some fanciful states might be made worse if kratom were unregulated and all else were equal, those states are not being considered because their probabilities are too low.)
So, according to the above implicit comparisons, losing money always makes things worse. It is therefore valid to say that losing money is bad. Similarly, it is valid to say that keeping kratom unregulated is good.
Given this, it is valid to make statements of the form, “If X happens, then A will happen (which is bad in the above sense), but B will happen (which is good in the above sense).”
You are considering the wrong alternatives, things that can possibly happen independently on your actions, while the alternatives that should be considered result from your taking different actions. The alternatives under the factors you don’t control are just internal structure of your model of uncertainty inside a particular possible decision.
I’m not sure what you mean by “thing that can possibly happen independently on your actions”. But “the money becomes worthless” is not independent of your action — the money’s becoming worthless requires that you choose to bet.
This is true (well, once you :s/, but not/but on/) however it does not make the original ‘but’ misplaced in any way.