Download this post to watch the video if your feed reader does not show it to you.
Download this post to watch the video if your feed reader does not show it to you.
Don’-t Short Obama
Why political futures markets got the health care bill so wrong.
By Daniel Gross
Posted Monday, March 22, 2010, at 6:05 PM ET
It would be very difficult to tote up all the times pundits pronounced the health care bill dead, and the prospects for the Obama administration dire—especially after the election of Scott Brown in January. Intrade, the political futures market, which functions as a conventional-wisdom-processing machine, also got health care wrong. Check out this chart for the contract on health care reform being passed by June 2010. The contract is worth 100 if it is passed, zero if it is not. After Brown’-s election, it slumped to as low as 20. As recently as March 17, it was below 40. Even as late as Friday, it was trading in the mid-80s. These trading data show that “-investors”- in this market were skeptical of the Obama administration’-s ability to pass significant health care legislation, right up until the end.
Is there a larger lesson here? (Aside from the obvious one, which is political futures markets usually aren’-t very good at predicting what actually will happen in the future?) I think so. And it’-s this: Don’-t short Obama. In fact, that’-s been the lesson of Obama’-s entire career so far.
[Stock market stuff inserted here.]
On some level, it’-s tough to blame the Intrade crowd for getting Obama and health care wrong. The type of people who trade there, folks who think they’-re quite savvy about money, the market, and politics, are the same conventional wisdom hawkers who were so monumentally wrong before the financial crisis. If you’-ve tuned into CNBC or Fox Business Channel, or read the Wall Street Journal since January 2009, you would have been subject to a constant stream of money managers, pundits, talking heads, and policy wonks declaring that the U.S. economy is becoming a socialist hellhole that is hostile to business and investors. (If there were a way to short Fox Business Channel, I’-d do it in a hurry.)
The conventional wisdom market has not yet internalized the message that it’-s dangerous to your financial and professional health to short Obama. Judging by the debate in the House last night, by the talk on cable news shows this morning (full of talk about how this is going to kill Democrats in November), and by the chatter on the business networks this morning (full of talk about how the tax increases in the health care bill will destroy the markets and the economy), the shorts haven’-t learned anything.
I agree with Dan Gross that prediction markets are a “-conventional-wisdom-processing machine”-. Prediction markets incorporate expectations (informed by facts and expertise) just like the mass media do.
Prediction markets can’-t look into the far away future.
In the ObamaCare case, prediction markets have just been summarizing objectively, dynamically and quantitatively (day in, day out) what the political media were reporting about the health care reform, and about the prospect of its passing in Congress and of its signing by the President.
It would be easy for a scientist to verify that —-by comparing archived media articles with the historical InTrade prices.
ADDENDUM: To answer Hutch’-s question, the only trouble I saw in the history of this contract is the brief manipulation that happened on March 16, 2010.
UPDATE: Funny video:
Joe Weisenthal has a small opinion piece on why the CFTC allows real-money prediction markets on movie business, and bans those on politics or sports. The problem in the piece is that Joe is 100% wrong.
The CFTC is a weak institution, in the DC sphere of power. In the recent past, the CFTC lost one important battle against other parts of the US government —-even though it was the CFTC that was on the right side of the issue at the time. With politics and sports betting, the CFTC does not want to lose another battle. It is a question of survival.
Most published papers on prediction markets (there aren’-t many) paint a wildly rosy picture of their accuracy. Perhaps it is because many of these papers are written by researchers having affiliations with prediction market vendors.
Robin Hanson is Chief Scientist at Consensus Point. I like his ideas about combinatorial markets and market scoring rules, but I think he over-sells the accuracy and usefulness of prediction markets. His concept of Futarchy is an extreme example of this. Robin loves to cite HP’-s prediction markets in his presentations. Emile Servan-Schreiber (Newsfutures) is mostly level-headed but still a big fan of prediction markets. Crowdcast’-s Chief Scientist is Leslie Fine- their Board of Advisors includes Justin Wolfers and Andrew McAfee. Leslie seems to have a more practical understanding than most, as evidenced by this response to the types of questions that Crowdcast’-s prediction markets can answer well: “-Questions whose outcomes will be knowable in three months to a year and where there is very dispersed knowledge in your organization tend to do well.”- She gets it that prediction markets aren’-t all things to all people.
An Honest Paper
To some extent, all of the researchers over-sell the accuracy and the range of useful questions that may be answered by prediction markets. So, it is refreshing to find an honest article written about the accuracy of prediction markets. Not too long ago, Sharad Goel, Daniel M. Reeves, Duncan J. Watts, David M. Pennock published Prediction Without Markets. They compared prediction markets with alternative forecasting methods for three types of public prediction markets: Football and baseball games and movie box office receipts.
They found that prediction markets were just slightly more accurate than alternative methods of forecasting. As an added bonus, these researchers considered the issue that prediction market accuracy should be judged by its effect on decision-making. So few researchers have done this! A very small improvement in accuracy is not considered material (significant), if it doesn’-t change the decision that is made with the forecast. It’-s a well-established concept in public auditing, when deciding whether an error is significant and requires correction. I have discussed this concept before.
While they acknowledge that prediction markets may have a distinct advantage over other forecasting methods, in that they can be updated much more quickly and at little additional cost, they rightly suggest that most business applications have little need for instantaneously updated forecasts. Overall, they conclude that “-simple methods of aggregating individual forecasts often work reasonably well relative to more complex combinations (of methods).”-
For Extra Credit
When we compare things, it is usually so that we can select the best option. In the case of prediction markets it is not a safe assumption that the choices are mutually exclusive. Especially in enterprise applications, prediction markets are heavily dependent on the alternative information aggregation methods as a primary source of market information. Of course, there are other sources of information and the markets are expected to minimize bias to generate more accurate predictions.
In the infamous HP prediction markets, the forecasts were eerily close to the company’-s internal forecasts. It wasn’-t difficult to see why. The same people were involved with both predictions! The General Mills prediction markets showed similar correlations, even when only some of the participants were common to both methods. The implication of these cases is that you cannot replace the existing forecasting system with a prediction market and expect the results to be as accurate. The two (or more) methods work together.
Not only do most researchers (Pennock et al, excepted) recommend adoption of prediction markets, based on insignificant improvements in accuracy, they fail to consider the effect (or lack thereof) on decision-making in their cost/benefit analysis. Even if some do the cost/benefit math, they don’-t do it right.
Where a prediction market is dependent on other forecasting methods, the marginal cost is the total cost of running the market. There is no credit for eliminating the cost of alternative forecasting methods. The marginal benefit is that expected by choosing a different course of action than the one that would have been taken based on a less accurate prediction. That is, a slight improvement in prediction accuracy that does not change the course of action has no marginal benefit.
Using this approach, a prediction market that is only “-slightly”- more accurate, than those from alternative forecasting approaches, is just not good enough. So far, there is little, if any, evidence that prediction markets are anything more than “-slightly”- better than existing methods. Still, most of our respected researchers continue to tout prediction markets. Even a technology guru like Andrew McAfee doesn’-t get it , in this little PR piece he wrote, shortly after joining Crowdcast’-s Board of Advisors.
Is it a big snow job or just wishful thinking?
[Cross-posted from Toronto Prediction Market Blog]
In part #2, he speaks about the books he is writing:
Business leaders rely on metrics and data to inform decisions around new products and opportunities, but traditional forecasting methods suffer from bias and lack of first-hand information. That’s why business forecasting is an ideal target for the application of crowd wisdom. While bets are made anonymously, some prediction market software applications have built-in reward systems for accurate forecasters. And the accuracy of prediction markets over traditional forecasting methods is proven again and again. […] Prediction markets will then aggregate this knowledge to produce actionable, people-powered forecasts. The result is an ultra-rich information source that will lay the foundation for smarter, better-informed company decisions. […]
Robin Hanson debates a Mencius Moldbug on prediction markets, decision markets, and…- futarchy:
Download this post to watch the video —-if your feed reader does not show it to you.
“-Not inflation”-, the gold critics will shout, in one of their go-to arguments. This is what we hear from CNBC’-s Mark Haines at every possible chance: since 1980, gold has not kept up with the CPI and so shouldn’-t be used as an inflation hedge. One would point out to Mark that this is analogous to arguing for global cooling based on that one 2005 start date. If you pick basically any other start date but the one corresponding to gold’-s 1980 peak, you see something different, even giving CPI a long head start over floating gold prices:
|Cumulative Increase Through December 2009|| -|
| -||CPI||Gold||Gold/CPI Increase Ratio|
|From:|| -|| -|| -|| -|| -|
|Jan-55||808.3%||3129.5%||3.87|| -|| -|
|Jan-70||476.9%||3113.9%||6.53|| -|| -|
|Jan-75||320.1%||514.8%||1.61|| -|| -|
|Jan-80||185.0%||143.8%||0.78|| -|| -|
|Jan-85||105.4%||254.2%||2.41|| -|| -|
|Jan-90||71.8%||176.3%||2.45|| -|| -|
|Jan-95||44.5%||197.8%||4.44|| -|| -|
|Jan-00||28.5%||298.5%||10.46|| -|| -|
|Jan-05||13.3%||155.6%||11.74|| -|| -|
But in shorter time-frames gold critics do have half a point. Since 2003, on a daily basis, gold returns have only been 12.5% correlated to changes in the inflation rate implied by 10-year TIPs. On a monthly basis, gold returns are 9% correlated to those of the TIPs spread.
We can look back further if we examine the the monthly performance of gold versus year-over-year changes in the CPI index. The CPI index for a given month is released in the subsequent month, so CPI monthly values are shifted forward in this study to correspond to the month of their release. The YoY change in CPI is further assumed to be the market’-s expectation of future inflation. All gold prices here are daily averages based on the London PM fix through December 1974, and Comex/CME spot thereafter.
Ignoring the fact that gold generally rose in this period, it doesn’-t do particularly well when inflation is elevated by this definition. A cut-off of 4% was used because it was the round number that most nearly bisected the 501 months in question, but the pattern holds-up when this parameter and other assumptions are varied:
|Monthly Gold Price Changes By Inflation Rate, Apr 1968 –- Dec
| -||Sum||Number of Months||Average|| -|| -|
| -|| -|| -|| -|
|>- 4%||180.4%||225|| -||0.80%|| -|| -|
|<-= 4%||232.3%||276|| -||0.84%|| -|| -|
So what does gold hedge against? Gold does well when real returns are low. You can’-t consider inflation without looking at prevailing rates and growth. The rates used below are the average of daily 10yr constant maturity rates (GS10) within a given month. As Larry David would say, “-pretty …- pretty good”-:
|Monthly Gold Price Changes By Real Rate, Apr 1968 –- Dec 2009|
| -||Sum||Number of Months||Average|| -|| -|
| -|| -|| -|| -|
|<- 3%||414.0%||268|| -||1.54%|| -|| -|
|>-= 3%||-1.3%||233|| -||-0.01%|| -|| -|
3% was used because it is again the round number that most nearly bisects the observations, but it can be varied without changing the essential result. There are also simple ways to define low real returns without a fixed parameter that show similar performance breakdowns with very different distributions of months. Now, these are retrospective studies, not trading systems, but obviously there is little chance that those returns were drawn from populations with the same mean.
It’-s surprising that thoughtful types like Nouriel Roubini and Martin Feldstein have questioned gold’-s inflation hedging, but didn’-t mention this point —- it seems glaring: people hold the relatively useless metal when real rates and opportunity cost are low. This simple point somehow never comes through in the noise surrounding gold: the glib Spam-sagacity vs. the Fall of The Republic, all the go-to arguments.
Clearly there are other factors that may throw the model off for long stretches of time. These may be false positives (e.g. non-dollar weakness) or false negatives (e.g. if gold is monetized to the point that it rises in deflation).
Putting aside the current weakness related to the Euro and elevating risk aversion, since I’-m expecting real rates to be on the low end compared to the late 20th century, my bias is still long gold. If yields should rise, especially if they are driven by vigilance, gold might make less sense.
[Cross-posted with minor changes from Seeking Alpha]
So far as new applications of derivatives markets I think one possibility is we may see more people making, creating derivatives markets, betting markets on policy, public policy outcomes. We’-ve already seen that with regard to the Federal Reserve. There is a market now in which people are able to make, take positions on the likelihood of a change in the Federal Reserve Bank policy at their next meeting of the Federal Open Market Committee, so and these markets are concerned with the question of what the Federal Reserve Bank rate will be set at. So I think we may very well see more of these kinds of markets and this could very well provide some indication of how the participants in these markets evaluate some of the policy proposals that governments are making.
Watch the video (download this post if your feed reader does not show it to you):