Comments on: Hayekian markets reconsidered

By: Tom Maguire

Tom Maguire — Thu, 21 Aug 2003 17:55:26 +0000

Love the post, but I wonder about this: …in order to be an efficient information-creating entity, a market has to have both hedgers or speculators. Although speculators are vital to the functioning of the market, you can’t have a market with nothing but speculators. How does this square with the very active sports betting markets? Other than Pete Rose, players, managers, and owners (the natural longs and shorts) are barred from gambling. Beyond that, the prices (betting lines) really are just steam on the window, to borrow your metaphor – very little economic activity, outside of the manufacture of a few bobble-head dolls, actually is affected by the news that the Giants are a 4 point underdog this weekend. That said, I believe studies have shown that the final market based odds tend to be accurate predictors of final outcomes. On a second point, I agree with the critics who are range from skeptical to dismissive of the “non-actionable” argument. Just about any one piece of info is “non-actionable” in the sense you are using it. If, for example, the President, or PM, had received a first hand, eyewitness account of Saddam personally meeting with Osama and handing him a nuclear weapon, we would not have gone to war – more steps would have been taken to verify it. (Or, if the process is hopelessly broken, then no information is actionable, but that does not count against PAM). And, as folks have noted in other contexts, war is not the only alternative. The President might, based on PAM, order more studies, or increase civil preparedness (change the color scheme!), or move a few troops, or send a diplomat to pose for photos – the menu is endless. I actually think you have a stronger objection elsewhere in the post – how do we verify the predictions of this market? With orange juice, we have lots of prices and results. Same with football bets. Same with markets predicting Hollywood box office receipts. With “Coup in Saudi Arabia”, we are going to have essentially one trial. If “the market” says there is a 20% probability of a coup by year-end 2003, we don’t learn a lot come January 2004 whether there was a coup or not. If there is no coup, it may still be the case that 20% was wildy optimistic, or pessimistic as of August 2003 – we have very little against which to compare it. The idea for PAM seems to be, markets work where there is lots of data and feedback, so they should work with very little data or feedback. Well, they might, but how would we know, either ex post or ex ante?

By: Jack

Jack — Mon, 18 Aug 2003 02:57:20 +0000

Daniel, I don’t think you are right about what Hayek needs for his critique of planning. People can have knowledge that takes propositional form, is relevant and even explicit and still need a market to put it to use. For example I know that I would like to avoid circling Heathrow waiting for a lot for my plane to land. Equally I’m sure that there is a Mr. Bloggs who knows that his life would be unbearable if Heathrow were to get a new runway. Bewtween me, him and a few million others we have knowledge of whether there should be a new airport at Heathrow that we do not individually. I also don’t believe that a culture of honesty is a full substitute for all that a market could provide. Humans are very bad at holding and updating competing views on an individual basis and markets aren’t so bad.

By: Bill

Bill — Tue, 12 Aug 2003 01:08:25 +0000

Dsquared, I see; I misunderstood your argument. I think I still do. I think that experts have both tacit and explicit knowledge. Most experts I deal with cannot put into sentences everything they know about their subject. However, they can still act on it, and could make bets about it. A futures market could aggregate all this information, and it might do better than other ways of aggregating information. So do you think a futures market populated by experts (with both tacit and explicit knowledge) and speculators (to help the market function), would provide good information (in the form of the prices of futures) to policymakers?

By: dsquared

dsquared — Mon, 11 Aug 2003 21:07:08 +0000

Bill: the key to our misunderstanding each other is Hayek’s concept of “tacit knowledge”. Your “experts” differ from my “hedgers” because they get their information in the form of sentences, to put it crudely, while my hedgers get their information in the form of empty flour bins and growing fields. Hayek’s point (and I’m not sure I believe it, but I am sure it’s the point he originally made when introducing the concept of tacit knowledge) is that the majority of human information is information which cannot be written down and put into propositional form. You note yourself that the information generated by this “market” is information which could be made available to a planner; it’s essential for Hayek’s critique of planning to go through that this isn’t the case.

By: Bill

Bill — Sun, 10 Aug 2003 19:36:07 +0000

Daniel, “The market exists because of the hedgers … It follows from this [and the other arguments] that in order to be an efficient information-creating entity, a market has to have both hedgers [and] speculators.” It seems to me that if you replace hedgers with experts, the market still works (according to your theory). In other words, a market will be an information-gathering entity if it has experts and speculators, with no hedgers. If true, this invalidates most of your article, since you use ideas that only apply to hedgers, not experts. You seem to know more about these ideas than I do, so perhaps you can tell me where I have misunderstood your argument. Hedgers trade, not to profit from their information, but to insure themselves against uncertainty. For example, if I am a baker, I want to focus on baking, not tracking the flour market to see when is the best time to buy my flour supply. So I buy futures, thus removing that uncertainty. Similarly with wheat growers. Speculators might find someone who is willing to pay $5 a contract, someone who is willing to sell for $4 a contract, and complete it, making a $1 per contract profit. However, this can work not just with hedgers, but experts who want to make money off their information. So, if I am an expert in the middle east, and you are an expert who disagrees with me, a speculator can make the same kinds of deals between us as between hedgers, thus setting prices. For example, if you think there is a 95% chance of rain, and I think there is only a 5% chance of rain, we don’t have to argue about it, we can just make a bet. If it rains, I pay you $100; if it doesn’t, you pay me $100. We both like this deal, and we both are trying to profit off our information. With lots of people doing this, speculators can “help” us set a price (more complicated than the above) that in some way describes all of our information together. If all of us were weather experts, we would expect this price to be a good indicator or the weather, despite the fact that none of us are hedgers, just experts. Similarly, if we were all experts in the Middle East situation, you would expect the price to give information about the Middle East situation. So your argument that “It follows from this that in order to be an efficient information-creating entity, a market has to have both hedgers [and] speculators.” doesn’t seem correct. You are confusing (literally, fusing together) the concept of a hedger with an expert, and basing your conclusions on the idea of hedgers without thinking about experts. For example of this error, you say that no one has an interest in more terrorist attacks, so the market wouldn’t work. You are right, no one would “hedge” this position, but experts might still have information about it. Therefore, the market might still work. You are also, it seems, confusing the idea of a speculator with an expert. For example, you say that betting markets are “entirely speculative”. This is false, given your definition of speculator as someone with knowledge only of the market, not the underlying situation. Actually, a betting market is about experts going to speculators (the bookmakers). The problem is not that these markets are “entirely speculative” its just that the experts aren’t particularly expert.

By: Bill

Bill — Sun, 10 Aug 2003 18:56:18 +0000

Daniel, “So, what would actually happen if the Saudi contract roofed it?” Presumably it would be one input into the decision-making process. It would be information that you combine with other expert’s information to decide what to do. You have to assume the CIA is reasonably competent and reasonably interested in making good decisions, but if you don’t assume that, you are arguing against the CIA, not against the futures market.

By: Bill

Bill — Sun, 10 Aug 2003 18:52:56 +0000

Daniel, “I actually think it’s something approaching a category-mistake to suppose that anyone could be in a position to have tacit knowledge relevant to the question “Is the chance greater than 22% that the government of Saudi Arabia face a coup attempt this year?”. It’s a question which demands an answer in the form of a proposition, not like the question “Is orange juice for September delivery too dear at $5 8/16 a contract?”, which demands an answer in the form of an action.” This is quite confusing. Can you explain the difference between the following, showing me how one “demands an action” and one “demands a proposition”? – Should I accept a deal where I get $100 if the government of Saudi Arabia faces a coup attempt by September, and pay $100 otherwise? – Should I accept a deal where I get $100 if September’s OJ price is less than $5 8/16 a contract, and pay $100 otherwise? There are many differences between the two, but I don’t understand which differences make one “demand an answer in the form of an action” and the other “demand an answer in the form of a proposition.”

By: Bill

Bill — Sun, 10 Aug 2003 18:43:38 +0000

Daniel (Dan?, Dsquared?), “The point being that if the overall intelligence process is corrupt, even the addition of an honest indicator isn’t going to help matters.” I don’t think this is necessarily true; an overall bad decision process may have a bad information processing system. Improving the information processing system will probably improve, or at least be a step in improving, the decision-making process. If by “corrupt” you mean “No matter what information they get, they’ll do something stupid” then you are assuming what you want to prove. I’m not saying you’re wrong :-) but it isn’t a good argument.

By: Bill

Bill — Sun, 10 Aug 2003 18:30:24 +0000

Nabakov, “When, say, a suicide hijacking seems imminent, what’s to stop traders from cashing in, therefore causing the price to dip and communicating a sudden drop in the threat level unrelated to the actual danger?” How exactly does that work? I have a future that is, imminently, going to be worth $1. How is it that I start selling it for less than $1? Lots of people would like to buy it for less, but no one would be willing to sell it for, say, $0.90. So I guess I don’t understand why the price would drop as the attack became imminent. I might “cash in”, but only for approximately $1, and never much lower. “what would have stopped [the terrorists] from using the TFM as advance warning of much was known about their plans?” Not sure I parsed this correctly, but I’m not sure that this is a bad thing. If I am a terrorist, and suddenly I hear that my enemies know what I am about to do, then I don’t do it. Is this not a good thing? Apart from actually catching the terrorist, preventing their attacks would be the next best thing, right? Or are you worried that I, the terrorist, might have a portfolio of attack plans, and I would use the futures market to determine which is the best one to carry out? That is, to determine the one that is least suspected? Or are you worried that I, the terrorist, would never get caught, since I would be forewarned of any attack that had a large probability of failure?

By: Nabakov

Nabakov — Fri, 08 Aug 2003 08:39:47 +0000

Couple of things I don’t understand about that terrorist futures market (TFM) proposal. Firstly markets work on self-interest. The higher the price on a given predictive event, the higher its likelihood of occurence, right? But what about “profit-taking”? When, say, a suicide hijacking seems imminent, what’s to stop traders from cashing in, therefore causing the price to dip and communicating a sudden drop in the threat level unrelated to the actual danger? And secondly, terrorists can read the market too (the airline futures thang just before 911 has not been properly explored or explained to my knowledge) so what would have stopped them from using the TFM as advance warning of much was known about their plans?

By: kokmo

kokmo — Thu, 07 Aug 2003 23:32:25 +0000

I fail to see how the information PAM would provide is any less actionable than information provided by a human analyst.

By: dsquared

dsquared — Thu, 07 Aug 2003 06:37:50 +0000

Basically because most “weather” derivatives in which there is any trading are derivatives on hurricanes, tornadoes and other extreme weather events.

By: Anurag

Anurag — Thu, 07 Aug 2003 01:54:50 +0000

Why is trading on weather derivatives one-sided? Eg if it is an exceptionally hot summer, crops and farmers do worse, but more AC is used and energy companies benefit. An opportunity to reduce risk exists for both. Also, I don’t understand why the only people with tacit knowledge about terrorism are the terrorists themselves. I am told that if I were to wake up 100 people waiting for a delayed plane and ask them when it will depart, the mean will be a very accurate estimate. Presumably, this is not because they are all airplane mechanics, but because they live in a world where planes are delayed.

By: dsquared

dsquared — Wed, 06 Aug 2003 17:38:29 +0000

Oh go on then …

By: Dan Hardie

Dan Hardie — Wed, 06 Aug 2003 16:07:07 +0000

Oh, be gracious, amend the article to give credit where due… The simple model of what a futures market needs to work seems excellent to me and ought to be published as a short working paper asap.