Yet another tool to try to assess the probability of a recession.
Betting exchange Tradesports is now offering a contract based on whether the U.S. economy goes into recession during 2007 (under their Financial – Economic Numbers category). Although trading so far has been quite thin, bettors up to this point are not putting that probability higher than 20%. The above graph should be a live link, so you can check back on this page from time to time for updates. You might also want to revisit Political Calculations to update Jonathan Wright’s yield curve calculations.
Given the concerns expressed by Stephen Kirchner about the original details of the Tradesports contract, I suggested to Tradesports a
tighter definition of what it means for the U.S. to go into a recession, which they’ve now adopted. The current contract declares that the U.S. will be said to have experienced a recession in 2007 if the Commerce Department numbers as reported on February 15, 2008 show 2 consecutive quarters of negative real GDP growth between 2006:Q4 and 2007:Q4.
Of course, here at Econbrowser we will make a formal call as to whether or not the economy has gone into a recession on the basis of my own recession probability index.
Technorati Tags: macroeconomics,
recession,
recession probability,
prediction markets
The Probability of a Recession
According to TradeSports.com, the odds of a recession occurring in 2007 are approximately 18-24% (on extremely thin volume). EconoBrowser has more….
The Tradesports contract was quite wrong about the Democrats taking over the Senate in the last election so I don’t know how much faith we should put in its predictions.
Por que não há um destes aqui?
Eu nunca entendo: se somos todos individualistas (hipnotizados por Washington, claro), globalizantes e neoliberais (malvados e feios), por que é que não existe um “Tradesports” aqui como nos EUA? Além de ser bom para apostadores, pode ser excelente …
So if the bettors are wrong, Joseph, why don’t you try to take their money?
I didn’t say they were always wrong. I was just saying that sometimes they are wrong and there’s no way to tell in advance when they are right or wrong.
Isn’t it illegal for Americans to participate in this market, now?
Pure idiocy that JDH tries to pose as economics.
What a bunch of people who bet online think may or may not have any relationship with reality.
By JDH’s logic, all those who held the NASDAQ above 5000 with their “bets” were “right”. Ridiculous.
This post is unworthy of the good work that Menzie Chinn does here, and is undermines yet further whatever small credibility JDH may have left as a useful thinker.
I agree, RN. It\’s embarrassing.
I don’t think JDH is off the wall on this. There is some evidence that the estimates of large numbers of people can be more accurate than the estimates of a small number of “experts.”
My point was that you have to accept that this is only one piece of data and that there is no guarantee that the herd is always right.
Ah, my old friend RN, on whom we can always count for some angry words, if not analysis.
So you knew the NASDAQ was going to plunge in 2000? That’s great. But can you tell us something actually useful, like where it will end up at the end of 2007? Or are you only good at predicting the stock market 7 years ago?
You can brag about what you claim to have known 7 years ago, but I have no way of verifying whether you actually profited from that call. But I do know your track record for what you were predicting 7 months ago. On May 7, 2006, when oil prices were at $70 a barrel, I wrote a post titled Have oil prices peaked?, in which I laid out some of the factors that might produce a decline in oil prices. You entered a comment on that post at that time, in which you described my reasoning as “very simplistic and not very usable analysis.”
So I apologize, but I find myself doubting your claim to be so much better than everybody else at predicting the future.
Of course, if you really could predict stock prices or oil prices or recessions, you could become a millionaire on these markets you hold in disdain. But that would be beneath your great talents, would it not? So much more fun just to brag about how much better your understanding of everything is, and use harsh words to criticize anyone whose views might differ from your own, rather than to actually place your own money on the line, as do the people whose understanding you hold in contempt.
JDH,
I think 20% is about right for the chance of recession. I believe the biggest part of the risk is the Fed holding on too long waging a war against pseudo-backward-looking-energy-influenced inflaton numbers. I don’t see it as lunacy to look at a number of forecasts (i.e. tradesports) and use it to help analyze and form and opinion. Thanks for pointing it out.
Thanks for setting the rude fellow straight, Professor.
I find Dr. Roubini’s logic compelling, and think it’s a done deal that we’ll have a recession beginning this year. I’d consider making a substantial bet on Tradesports towards that end, but have a much bigger bet — everything that I’ve got and continue to get — on gold mining stock mutual funds (VGPMX and UNWPX).
It will be an interesting 2007 and 2008, at the very least.
Follow up to last commment. Good article and charts in WSJ about inverted yield curve and predicting recession.
“Another forecasting model — developed by Federal Reserve Bank of New York economists using only the 10-year/three-month spread — puts the chances of a recession in 12 months at just under 40%.”
“A recent survey of economists by The Wall Street Journal pegged the chances of a recession within the next 12 months at 27%.”
Thanks to JDH for picking up on this contract. I don’t understand the critical comments posted about it. My understanding is that contracts like these do quite well in predicting election outcomes relative to surveys of voter intentions, so it will be interesting to see how this contract performs given the abysmal performance of most alternatives (eg, my recollection is that the Fed’s Suvey of Professional Forecasters shows most economists struggle to identify a recession in real-time, let alone predict a downturn).
Speaking of your recession probability index, what is it reporting now? We’d love to know monthly updates.
thanks
Actually, charts, it’s by definition a quarterly series. Next update should be Jan 31.
Having skimmed your paper, I wonder whether you couldn’t turn out a monthly toy version. If the markets weren’t opening in eight hours I might reread the thing and answer my own question. Alas, they are, so I’m off to bed.
FWIW, “true value” of that contract could be 20 and it could still be a buy at 24 if you think headline data is due for temporary weakness. Selling now is not just predicting that the probability of thus-defined “recession” is less than 20%, but that forthcoming data will improve. The time to short is when things look worst, right before they turn.
My suspicion is that that time will have arrived in just about a quarter, which is to say that I think 06Q4 GDP will be weak.
I’d consider making a substantial bet on Tradesports towards that end, but have a much bigger bet — everything that I’ve got and continue to get — on gold mining stock mutual funds (VGPMX and UNWPX).
PG,
The price of gold is strongly determined by the federal government. Invest in such instruments with extreme caution. They change the rules at the drop of a hat.
wcw,
GDP in Q4 weak? Hows come? The PCE stuff so far has been really good. Other components, such as inventories and capital spending, will be a drag, but weakness elsewhere is what it will take to stave off a GDP reading above 3%, given what we’ve seen from consumption spending.
As for Tradesport doubters, this isn’t a subject about which we need to lean on single data points – got the mid-term election wrong of whatever. There is actually a body of research on the subject. “The Wisdom of Crowds” offers a good starting place for those not interested in reading journal articles.
I frankly don’t know what the record of prediction markets is on recession forecasts, but it seems we are about to get a new data point. The problem, I’d guess, is that recessions are sufficiently rare that it’s hard to put together a record on whether prediction markets make decent recession calls. One thing I’d like to know, but am too lazy to find out for myself, is whether betting markets on recession contain any information not available in (give answers different from) other methods of recession forecasting.
Actually, we have much more than 1 data point, even just using the 2006 election. Tradesports correctly predicted the outcome in 32 out of 32 of the individual Senate races. The one thing they got wrong, to which Joseph refers, was the composite contract that the Democrats would win the Senate. There can be a 60% probability the Democrats will win Maryland, and a 60% probability they will win Missouri, and a 60% probability they will win Tennessee, and a 60% probability they will win Virginia, but a much less than 50% probability they will win all four.
In fact, we did not know for sure even the morning after the election whether the Democrats had indeed carried the Senate. It is a well-established common fallacy of the human mind that we often end up concluding that the event that was observed (in this case, the Democrats winning control of the Senate) must have had a high probability of occurring.
After reading the original post my first thought was why doesnt a contract for recession probability or GDP exist, maybe it does and I am unaware? Seems to me this would a very functional way for a company to transfer or manage risk. It could be structured much like fed-fund-futures with each contract paying a certain amount for each 0.1 of GDP or maybe its only in-the-money if we actually do reach recession. I realize that many businesses particularly those that are publicly traded manage risk in many other ways but this could become one piece to the puzzle, consumer discretionaries and/or retail for instance.
Weak GDP (by which I mean 2% or lower) I see from autos, construction, and middlin’ holiday sales. No great surprise on any of them. The big risk here is energy prices; depending on the deflator, the same nominal number I think gives us 2% could almost as easily exceed 3% in real terms. As for the PCE, that’s just what I’m looking at. So far, the numbers are “fine” — and fine numbers gave us 2% last quarter. I see just enough in Q4 to hold the lid on.
I could be wrong, but then I’m not trading this weird little contract. Know what you trade etc etc.
KHarris,
I have read “Wisdom of Crowds” and would be interested in reading further. Are there some specific journal articles on this behavior you could point me to.
TY
Technically it is not a success for the prediction market to have gotten 32 out of 32 races right, unless the implied percentages were something above 97%. If a market is predicting an event with 67% probability, as it did for the Republicans holding the Senate, then if that price is right, it *better* be “wrong” 1/3 of the time. So actually the fact that the markets got the Senate majority wrong is very consistent with their price behavior.
Hi, I am the guy who proposed this contract back on Nov 30 (http://www.midasoracle.org/2006/11/30/recession-contract-proposal/).
I also posted the same caveat that Kirchner did on my blog, a day before he did (http://caveatbettor.blogspot.com/2007/01/contract-for-us-recession-in-2007.html)
Just thought you would want to keep the attributions at the highest of academic standards, as I am a longtime reader of your fantastic blog.
Thanks for letting us know, cb, and forgive the oversight. To partially atone, we’ve added your interesting site to our blogroll.
Very kind of you, Professor H. As a 6 month blogger, I have seen that this activity often deviates from optimality.
Best wishes on your site, and the respective careers and families of its august contributors.
one should note that thinly traded contracts have less predictive value than widely traded “bets”. That would be one reason to doubt this particular prediction even if you believe in the concept per se
Bellanson