Nutritional Prediction Markets

The chemical resveratrol might slowdown aging.  Taking large doses of vitamin D supplements might drastically reduce your chances of contracting cancer.  It’s possible, however, that taking either supplement could worsen your health.

Some would argue that the prudent thing to do is wait for more research to be published.  But, for example, if resveratrol really does slowdown aging then if I wait five years to take it I might lose two years of life.  Similarly, if taking large quantities of vitamin D really does drastically cut the risk of cancer then waiting five years to start taking it in supplement form might cost me my life.

In ten years we will know far more about the health effects of taking these and many other supplements.  What should we do in the meantime?  What most of us are forced to do is rely on the advice of experts as presented in the popular press.  Instead someone should create prediction markets in the benefits and costs of health supplements.

If resveratrol fulfills its promise then in ten years tens of millions of Americans will be taking it.  We could therefore create a prediction market that predicts how many Americans will be taking resveratrol in ten years.

We should also create a prediction market in whether taking resveratrol proves harmful.  Many people are currently taking the drug.  A prediction market could estimate, for example, how many deaths or serious health problems will be linked to resveratrol over the next decade.

These two prediction markets would prove enormously useful to rational consumers.  For example, imagine that the first market estimated that there was a 40% chance that in ten years most Americans would regularly take resveratrol, and the second market estimated with high probability that no one would be harmed by taking resveratrol.  This would mean that with high probability taking resveratrol today would either help you or do you no harm (other than to your pocketbook.)  Based on these results many people, I suspect, would start taking the supplement.

Nutritional prediction markets would also create incentives for scientists to study the health benefits of resveratrol.  A scientist could conduct research and then play the prediction market before publishing his results.

GD Star Rating
Tagged as:
Trackback URL:
  • How would a prediction market “know” the effects of Vitamin D? If the information ain’t there, it ain’t there. Using a blind crowd to see the way doesn’t seem like a good idea.

    Oh. Ah. The last suggestion in the post, that a scientist who *does* have the information use it for insider trading, indicates that the poster knows that the prediction market doesn’t have the desired information.

    On the other hand, such a market would be a good way to:

    1) Go long on Vitamin D, don’t use it yourself, and either:

    a) VD (vitamin D, that is) keeps one alive. You die rich.

    b) VD kills. You live poor.

    2) Short Vitamin D, use it yourself and either:

    a) VD works. You live poor.

    b) VD kills. You die rich.

    3) Go long, use it yourself and either:

    a) VD works. You live rich.

    b) VD sucks. You die poor.

    4) Short it, don’t use it yourself and either:

    a) VD is the wonder drug. You die poor.

    b) You were right. You live rich.

    Ah, the wonders of markets. Able to amplify ones, what, biases? Or maybe that’s just when they are the wrong tool.

  • “Nutritional prediction markets would also create incentives for scientists to study the health benefits of resveratrol. A scientist could conduct research and then play the prediction market before publishing his results.”

    Wouldn’t that be analogous to corporate insider trading?

  • It’s a cool idea. I can easily imagine that in many research areas, the official research programs lag way behind what a smart scientist can do if they are allowed to seek knowledge on their own. Especially this is possible if there is a national regulatory agency involved such as the FDA.

    Even given such a situation, there, big-profit items face two challenges.

    First, a normal market finds efficient solutions by eliminating people who are out-competed. Losing at a long-term information market, though, takes years. Thus at least for the first several years of such a market, there will be many players that the market has not yet selected out. Lots of people may well be buying them just like lottery tickets.

    Second, for high-profit supplements, the potential profits of selling it to millions of people surely dwarf whatever money there is to be had in the prediction market. Thus if someone holds a patent on the supplement, they can dump money into skewing the information market and then, as everyone starts buying it, they can reap a bigger profit than they lost.

    These both seem fatal. Information markets seem to require either a short time scale, or large enough money involved that it’s too painful to bet and lose.

    Hmm, here’s a variation that is amusing. What if scientists got some of their pay, not in stock options, but in information-market bets? They would be allowed to choose whichever bet they liked, based on their current research. Then the money involved might be relatively small, but the incentive would be personal….

  • Felix and Brian, speculative markets know nothing until someone with info comes to trade on that info. There are many kinds of relevant info, and the whole point is to get people who have or could obtain such info to trade on it.

    Daublin, it is harder to predict the more distant future, in part because feedback is slower to tell people how good they are. But other institutions that compete with this proposal also suffer from that problem; the question is which does better. Also, it turns out that manipulation on average makes these markets more accurate on average.

  • Joseph Delaney

    What this could be really interesting for is betting on which drug should be tested in a randomized clinical trial next. There are a lot of medical questions (especially about drugs) out there and these trials are not cheap; it would be interesting to see what drug-disease combinations that people thought would be the most likely to generate important results given that they had money riding on the outcome.

  • One problem with this market (I mentioned it here) is that a market that acts as its own fundamental indicator is prone to manipulation. For example, if the market offers a 20% chance that people will take resveratrol, and I buy enough contracts to pump that probability up to 50%, and this ‘news’ convinces new users to take it up, I can use the market to manipulate the fundamentals, without the fundamentals reflecting what we’re actually measuring (the risk or reward of taking it).

    So I’d suggest either a) forcing people to identify themselves when they make a trade, so manipulators can at least pay a penalty in reputation, or b) having a futures market on the actual effect (perform a double-blind test, and have a financial instrument whose returns are based on the excess life expectancy for reservatrol users).

    This is a problem that could blindside many PM participants: in a market where changes in the market value of an instrument correlate positively with changes in the fundamental value of the instrument, there will be price trends and (possibly) manipulation. It might be nice for PM owners to warn potential buyers the way some brokers do (“This stock is close to the delisting threshold — if it drops another 10%, it’s going to be kicked off the exchange and probably drop another 20%”).

  • Hi interesting topic, I believe in 10 years most ‘rich folks’ will be taking the NCE that Sitris will be producing, while all the folks who are afraid of ‘pharmas’, or have lower spending dollars will take a resveratrol formulation of some kind.

    I also agree with Mr. Delaney here regarding skewing the market perception. If it is the intention of building relevant information, I don’t believe it can be done using a ‘prediction market’. If things were all on the up and up (basically a perfect model, where everyone identifies themselves, and has a fixed ‘vote’ and somehow does not publish any papers in journals (can be construed as marketing you know…) to skew perception… we would really be relying on statistics. And if it’s a fixed numbers game, it wouldnt be much of market, would it.

    Anthony Loera

  • Anthony, there are other scenarios more likely than either a completely fixed numbers game and everyone being completely honest and forthcoming about everything without needing any incentives to do so.

  • Anthony,

    You wrote that “I believe in 10 years most ‘rich folks’ will be taking the NCE that [Sirtris] will be producing.” The market doesn’t agree with you because the stock Sirtris Pharmaceuticals has a market capitalization of only $343 million, which is much to low if there is even a 10% chance it will develop a super-blockbuster drug. Of course, this means that if you know the market is wrong you can make a huge amount of money buying Sirtris stock.

  • Taleb in his book, The Black Swan, discusses that the market typically does undervalue low-probability, high-payoff events. So it is quite possible that Sirtris is a buy.

  • Jim

    Vitamin D3 from UV light exposure on skin is more effective than vitamin D2 from supplements. See study:

    Vitamin D3 from UV light protects against many life threatening diseases and many different cancers:

    Health and economic benefits from sun exposure are much greater than risks: study

    Tanning in moderation is healthy behavior. Enjoy the sun in moderation. Don’t sunburn.

    When you can’t get enough UV light exposure outside, a couple sessions in a tanning bed will process all the healthy vitamin D3 needed.

    Feel good about sporting a healthy tan.

  • Prediction Markets for Longevity Medicine

    Prediction markets, like all markets, are extremely useful in those places where they have taken root. In a manner of speaking, all markets perform the function of a prediction market: they measure the ever-changing opinions of the participants on a ra…

  • Julian Morrison

    1) Get the government’s heavy thumb off the scales in the medical market.

    2) Unban insider trading – the ban closes off a path for knowledge to enter the market, as Brian above notices.

    3) Voila! The actual real medical market also becomes an unbiased prediction market. No need to set up some elaborate game when the reality is ready to hand.

  • Hopefully Anonymous

    James D. Miller, great post and follow-up comments. I looked up some of your old columns published elsewhere hoping for more of this quality of thought and writing and was disappointed by the hackneyed liberal vs. conservatives motiffs. Was that performative? Were you hoping to accumulate some Sean Hannity money from that audience that you could then use rationally to maximize your persistence odds? Hoping for a transparent reply in this thread.

  • Dear Hopefully Anonymous,

    I am a politically active conservative Republican. I sometimes enjoy writing partisan political articles in which I point out (what I believe are) flaws of the left. These articles are often written for an audience that I assume share many of my political views and consequently would be inappropriate for a site such as Overcoming Bias.

    I would love to get some Sean Hannity money, do you have any advice about how I could accomplish this? As a college professor I always assumed that being a conservative would reduce my lifetime income, but I would be happy to be proved wrong.

    James Miller

  • Markets will be markets.. but I can tell you this, Resveratrol has helped me shed unwanted pounds.. And my guess is that the research is correct.. that there is indeed an anti-aging benefit.

  • Hopefully Anonymous

    Prof. Miller, I take it your conservative republicanism doesn’t lead you to an anti-persistence maximizing position, and I’m thankful for that. I’m curious if you’ve thought about ways to influence other conservative republican opinion-shapers into advancing agendas that benefit those of us who don’t want to die -agendas that are arguably in their own interest. If they approached perfect rationality, I imagine they have some sort of internal, non-transparent balancing going on, between populist positions that may be bad for advancement of persistence maximizing technology but which allows them greater wealth accumulation by taking those positions. For example, I imagine that there was a dollar benefit to Sean Hannity (or similarly positioned people) for being loudly against fetal stem cell research, in terms of increased audience, advertising revenue, and pay, that they could then use to maximize their personal odds of persistence. But I’d think they’d fear some personal cost because they might be so successful in voicing their oppostion to innovative medical technology that they could harm their personal odds of persistence. Sean Hannity may have plausible deniability that he even makes such calculations. But I don’t think you do. 😉

    So, how do you make these calculations? Or how will you from this point forward?

  • Dear Hopefully Anonymous,

    Your wrote “I’m curious if you’ve thought about ways to influence other conservative republican opinion-shapers into advancing agendas that benefit those of us who don’t want to die -agendas that are arguably in their own interest.”


  • Ryan W.

    I think there are too many variables here. For example, look at vitamin E. It turns out that natural mixed tocopherols combined with ascorbates are helpful, while a racemic mix of alpha tocopherol increases mortality by ‘crowding out’ other useful forms of vitamin E.

    Who wins that bet?

    Better to just bet on processes for producing resveratrol, or have a futures market in Japanese Knotweed or something. Or what if resveratrol is healthy, but they find a way to alter it so that it doesn’t interfere with the cytochrome system, so that people taking statin drugs can still take it. And then some people choose to take the old kind since some benefits actually turn out to hinge upon cytrochrome inactivation. (yes, I know that the primary effect is from p53 inactivation)

    When it comes down to it, I place little faith in the wisdom of crowds here. Have you seen the garbage people eat now? And don’t get me started on their exercise habits.