Sotomayor is a mortal lock
In the past couple days I've managed to bet about $2000 that Sotomayor will be confirmed, at odds averaging 94%. I'd put another $10k on it if I could.
Here's why.
-Sotomayor needs 60 votes in the Senate. The Democrats already have 59, and may have 60 by the time the confirmation comes around. Any Dem who defects on this pick would be ostracized and face a serious primary challenge.
-A dozen Republicans voted for Sotomayor's confirmation to her position on the appeals court. They'll have a hard time explaining a switch.
-Sotomayor is popular with the public. A new Quinnipiac poll shows 54% of Americans approve of the choice, vs. 24% who disapprove.
-It's demographic suicide for the Republicans to mount a serious fight against her. Latinos are the fastest-growing racial demographic in the US. In the last election they voted about 60-40 for Obama. If the Republicans pursue an anti-Latino strategy of opposing immigration, affirmative action, and making a big deal of Sotomayor's race, they risk pushing Latinos to vote more like African-Americans, who vote Democrat 90% of the time.
-Republican insiders recognize the political realities here. There are plenty of conservative outsiders who are getting TV time smearing Sotomayor. Elected Republicans are being very quiet. A National Journal poll of Republican insiders asked, "Would it be politically smart for Republicans to try to block the confirmation of Judge Sonia Sotomayor?" 24% said yes, and 64% said no.
If Republicans don't want to see her confirmed, they better get to work on photoshopping Sotomayor snorting coke with some hookers.
Here's why.
-Sotomayor needs 60 votes in the Senate. The Democrats already have 59, and may have 60 by the time the confirmation comes around. Any Dem who defects on this pick would be ostracized and face a serious primary challenge.
-A dozen Republicans voted for Sotomayor's confirmation to her position on the appeals court. They'll have a hard time explaining a switch.
-Sotomayor is popular with the public. A new Quinnipiac poll shows 54% of Americans approve of the choice, vs. 24% who disapprove.
-It's demographic suicide for the Republicans to mount a serious fight against her. Latinos are the fastest-growing racial demographic in the US. In the last election they voted about 60-40 for Obama. If the Republicans pursue an anti-Latino strategy of opposing immigration, affirmative action, and making a big deal of Sotomayor's race, they risk pushing Latinos to vote more like African-Americans, who vote Democrat 90% of the time.
-Republican insiders recognize the political realities here. There are plenty of conservative outsiders who are getting TV time smearing Sotomayor. Elected Republicans are being very quiet. A National Journal poll of Republican insiders asked, "Would it be politically smart for Republicans to try to block the confirmation of Judge Sonia Sotomayor?" 24% said yes, and 64% said no.
If Republicans don't want to see her confirmed, they better get to work on photoshopping Sotomayor snorting coke with some hookers.
GM's going bankrupt
I assume everybody knew that already. There was a spike downward this morning in the price of the GM.CHAP11.DEC09 contract on news that the government sweetened its offer to bondholders. The original offer was that the bondholders would get 10% of the equity of the reorganized company. The new offer is that the bondholders get 10% plus warrants to purchase another 15%, in exchange for not objecting to a quick bankruptcy process. I'm not sure why people thought that reduced the chances of bankruptcy.
It's worth remembering that just a month ago GM's odds of bankruptcy were trading at 70-75%. Back then, people were concerned that GM would be the most complicated, drawn-out bankruptcy in history, and that there would be a good chance that GM would simply be liquidated, never to emerge as an intact company.
Two things have happened since then. Most importantly, the Chrysler bankruptcy has gone far quicker and far smoother than anyone anticipated. Obama's team claimed they would be out in 30-60 days, but NONE of the experts believed it was possible. The other reason GM is headed for bankruptcy is, the only alternative is for the government to continue pouring money in for the next several years.
I'm suspicious that the GM bankruptcy won't be as clean and easy as Chrysler's. There are far more creditors to deal with, including retail investors who are much more sympathetic figures than the big banks and hedge funds that lent to Chrysler. The Chrysler bankruptcy went smoothly in part because the judge agreed with the Obama Administration and permitted the Fiat merger over the objections of some creditors. What happens if the judge in the GM case decides differently? A big ol' higgledy-piggledy mess.
I fully expect there to be more last-minute negotiations like what was revealed this morning. I've pretty much closed out the short position I used to hedge my Chrysler bankrupty bet from last month. I'll be building a long position on any more late-breaking news.
It's worth remembering that just a month ago GM's odds of bankruptcy were trading at 70-75%. Back then, people were concerned that GM would be the most complicated, drawn-out bankruptcy in history, and that there would be a good chance that GM would simply be liquidated, never to emerge as an intact company.
Two things have happened since then. Most importantly, the Chrysler bankruptcy has gone far quicker and far smoother than anyone anticipated. Obama's team claimed they would be out in 30-60 days, but NONE of the experts believed it was possible. The other reason GM is headed for bankruptcy is, the only alternative is for the government to continue pouring money in for the next several years.
I'm suspicious that the GM bankruptcy won't be as clean and easy as Chrysler's. There are far more creditors to deal with, including retail investors who are much more sympathetic figures than the big banks and hedge funds that lent to Chrysler. The Chrysler bankruptcy went smoothly in part because the judge agreed with the Obama Administration and permitted the Fiat merger over the objections of some creditors. What happens if the judge in the GM case decides differently? A big ol' higgledy-piggledy mess.
I fully expect there to be more last-minute negotiations like what was revealed this morning. I've pretty much closed out the short position I used to hedge my Chrysler bankrupty bet from last month. I'll be building a long position on any more late-breaking news.
This is the easiest model ever
The CDC has gotten a real new swine flu case count up for the first time in about a week. They've gone from updating every day to updating weekdays to updating Monday-Wednesday-Friday. Monday's count was low because of the holiday weekend. Many of the state health departments were off, so they didn't communicate their new, higher counts to the CDC. Today the count is probably as accurate as it'll ever be.
The case count continues to grow at a constant, linear rate of about 290 cases per day. You can draw a straight line that fits the data very well.
Extending that line out to June 30th gives a prediction of 17787 cases when the contracts expire. Ignoring the vagaries of when the CDC updates its count, I project we will pass 10,000 cases on June 4th, and 15,000 cases on June 21st.
The CDC has said that certain areas of the country have seen a slowdown, while in others the flu continues to spread. I'm beginning to think that the actual spread of the flu will not have very much impact on the case count reported by the CDC. We already know that the lab-confirmed counts are far lower than the actual number of cases. Meanwhile, the states and the CDC have built up a testing infrastructure which seems to be able to crank out 290 cases per day. Barring new developments, I expect they will continue to perform lab tests at a constant rate. As long as the actual case count is above 290 per day, the lab-confirmed case count should continue to grow.
The case count continues to grow at a constant, linear rate of about 290 cases per day. You can draw a straight line that fits the data very well.
Extending that line out to June 30th gives a prediction of 17787 cases when the contracts expire. Ignoring the vagaries of when the CDC updates its count, I project we will pass 10,000 cases on June 4th, and 15,000 cases on June 21st.
The CDC has said that certain areas of the country have seen a slowdown, while in others the flu continues to spread. I'm beginning to think that the actual spread of the flu will not have very much impact on the case count reported by the CDC. We already know that the lab-confirmed counts are far lower than the actual number of cases. Meanwhile, the states and the CDC have built up a testing infrastructure which seems to be able to crank out 290 cases per day. Barring new developments, I expect they will continue to perform lab tests at a constant rate. As long as the actual case count is above 290 per day, the lab-confirmed case count should continue to grow.
"Trading Services Not Available"
I got this error a few times when trying to log in to Intrade this morning. I had to log out, close all my browser windows, and log back in again. YMMV.
Sotomayor in, Intrade down
Well, it looks like Intrade's initial impression was correct: Sotomayor was the early frontrunner, who was later outpaced by Kagan and Wood. Personally I figured it wasn't Sotomayor, simply because all the right-wing fire was concentrated on her, and there didn't seem to be any counter-fire from the left. Some commentators were concerned that the right would define Sotomayor before Obama could. Evidently Obama made a calculation that most of the public didn't pay much attention to the SCOTUS handicapping.
I had a modest short position on just about everybody, including Sotomayor. I would have done very well with a pick out of left field, lost a ton if it had been Granholm or Clinton, and made a small amount with anybody else.
So....
WTF, Intrade? No explanation, no apology, no mention of the down time? I understand Intrade doesn't have the resources to ensure 99.99% uptime like the NYSE or Comex, but c'mon...
Here's what you do: say you're sorry, take steps to fix whatever it was, and most importantly, have a backup website which can display basic information like how long it will be until you're back up. Otherwise you get members like my commenters whose first thought is that you've disappeared with all their money.
___
Update: they have apologized, and it looks like it was a screw-up by their ISP. I still think having a backup status site is an easy thing to put together. www.intradestatus.com is available.
I had a modest short position on just about everybody, including Sotomayor. I would have done very well with a pick out of left field, lost a ton if it had been Granholm or Clinton, and made a small amount with anybody else.
So....
WTF, Intrade? No explanation, no apology, no mention of the down time? I understand Intrade doesn't have the resources to ensure 99.99% uptime like the NYSE or Comex, but c'mon...
Here's what you do: say you're sorry, take steps to fix whatever it was, and most importantly, have a backup website which can display basic information like how long it will be until you're back up. Otherwise you get members like my commenters whose first thought is that you've disappeared with all their money.
___
Update: they have apologized, and it looks like it was a screw-up by their ISP. I still think having a backup status site is an easy thing to put together. www.intradestatus.com is available.
Virginia Governor -- The Weasel You Know
This just came on my radar recently. Turns out there is an honest-to-god election going on in Virginia on June 9th. Specifically, it's a 3-way contest for the Democratic gubernatorial nomination. The Washington Post gives a good summary of the state of the race:
Polling has consistently shown McAuliffe in the lead.
Research 2000 (Daily Kos) 5/18-5/20:
Moran 22
McAuliffe 36
Deeds 13
Survey USA 5/17-5/19:
Moran 22
McAuliffe 37
Deeds 14
Public Policy Polling 5/19-5/21:
Moran 20
McAuliffe 29
Deeds 20
I don't know that much about Moran or Deeds. I mainly associate McAuliffe with the transparently cynical triangulation strategy used by Bill Clinton in the 90s. Like, say, Arlen Specter, his core political belief is that he should be in office. In a centrist state like Virginia, his money advantage and poll-approved positions could be enough to get him elected.
McAuliffe has the highest negative ratings in the race, but thus far he's been quite successful in splitting the "anybody but McAuliffe vote." There are also quite a few undecideds, and nearly 50% of voters indicate they could change their minds. Deeds just got the highest-profile endorsement in the race, from the Washington Post. That could be quite significant, as metro DC was his weakest area in the state. If he can consolidate the anti-McAuliffe voters, he might have a shot at this. Check this somewhat hyperbolic blog post for more.
At this point, I don't have any position, I'm just trying to raise interest in the Intrade community about the race. I don't think it's a lock for anyone, and we're going to see some interesting twists and turns in the next couple weeks.
The leader is Terry McAuliffe, a celebrity wild card who chaired the Democratic National Committee for four years and last year chaired Hillary Clinton's presidential campaign. Second-place is another legislative veteran -- Brian Moran, former Democratic leader of the House of Delegates and brother of congressman Jim Moran... [Third place is] Creigh Deeds, a veteran Charlottesville-area state senator.
Polling has consistently shown McAuliffe in the lead.
Research 2000 (Daily Kos) 5/18-5/20:
Moran 22
McAuliffe 36
Deeds 13
Survey USA 5/17-5/19:
Moran 22
McAuliffe 37
Deeds 14
Public Policy Polling 5/19-5/21:
Moran 20
McAuliffe 29
Deeds 20
I don't know that much about Moran or Deeds. I mainly associate McAuliffe with the transparently cynical triangulation strategy used by Bill Clinton in the 90s. Like, say, Arlen Specter, his core political belief is that he should be in office. In a centrist state like Virginia, his money advantage and poll-approved positions could be enough to get him elected.
McAuliffe has the highest negative ratings in the race, but thus far he's been quite successful in splitting the "anybody but McAuliffe vote." There are also quite a few undecideds, and nearly 50% of voters indicate they could change their minds. Deeds just got the highest-profile endorsement in the race, from the Washington Post. That could be quite significant, as metro DC was his weakest area in the state. If he can consolidate the anti-McAuliffe voters, he might have a shot at this. Check this somewhat hyperbolic blog post for more.
At this point, I don't have any position, I'm just trying to raise interest in the Intrade community about the race. I don't think it's a lock for anyone, and we're going to see some interesting twists and turns in the next couple weeks.
Pop culture
Oh, I heard that a television show called "American Idol" had its finale last night, and someone unexpected won. Huh.
Swine flu slowing?
All the swine flu contracts are plummeting today based on the new update from the CDC. The CDC's daily updates had been showing 200-400 new confirmed cases each day. Today's update had only 54 new cases. Looking at the state-by-state numbers, we see this was not because of a one-time downward revision -- no states decreased their reported numbers, and only a handful made small increases.
Why such a big slowdown? It's not clear. The urgency of monitoring has certainly diminished over the past couple weeks. The CDC doesn't even update over the weekends anymore. State health departments are probably also reducing their vigilance. The spread of the flu is also slowing as the weather gets warmer. Most new cases over the past couple weeks came from northern states. Still, either of these factors should cause a gradual slowdown, not a sudden drop of 80% fewer new cases.
The extreme slowdown today is probably a statistical blip, and I'd guess we see >100 new cases tomorrow. If we don't, though, all the contracts should go much, much lower.
Disclosure: This is very good news for me. I'm short in all of these contracts, mainly in the >10,000 and >25,000. I had pretty much expected the >10,000 to expire at 100, but this is very hopeful for me.
Why such a big slowdown? It's not clear. The urgency of monitoring has certainly diminished over the past couple weeks. The CDC doesn't even update over the weekends anymore. State health departments are probably also reducing their vigilance. The spread of the flu is also slowing as the weather gets warmer. Most new cases over the past couple weeks came from northern states. Still, either of these factors should cause a gradual slowdown, not a sudden drop of 80% fewer new cases.
The extreme slowdown today is probably a statistical blip, and I'd guess we see >100 new cases tomorrow. If we don't, though, all the contracts should go much, much lower.
Disclosure: This is very good news for me. I'm short in all of these contracts, mainly in the >10,000 and >25,000. I had pretty much expected the >10,000 to expire at 100, but this is very hopeful for me.
Thanks, Math!
Initial unemployment claims come in at 631,000, right in the middle of my mathematical model's 600,000-650,000 range.
SCOTUS: Picking losers, not winners
As with the vice presidential selection process, it's often much easier to pick losers than winners. My major positions right now are short Hillary Clinton and short Jennifer Granholm. Clinton for obvious reasons, Granholm because of a host of scandals in her past. Granholm's stock rose recently because she traveled to DC for a public appearance with Obama, which turned out to be about raising gas mileage standards.
This is a fun little tid-bit: Colorado judge Christine Arguello says she's been approached by the Obama team for vetting. (Hat tip to The Page) Interesting because it shows that Obama is throwing a wide net that includes individuals that have never appeared on media short lists.
Those short lists seem to have been narrowed to Diane Wood (who recently canceled a law-school class to travel to DC), Elaine Kagan, and Sonia Sotomayor. I have small long positions on Wood and Kagan, but I don't think any of them are good buys at their current prices.
This is a fun little tid-bit: Colorado judge Christine Arguello says she's been approached by the Obama team for vetting. (Hat tip to The Page) Interesting because it shows that Obama is throwing a wide net that includes individuals that have never appeared on media short lists.
Those short lists seem to have been narrowed to Diane Wood (who recently canceled a law-school class to travel to DC), Elaine Kagan, and Sonia Sotomayor. I have small long positions on Wood and Kagan, but I don't think any of them are good buys at their current prices.
I make a lot of graphs
I think the new markets on the initial unemployment claims is a great idea. Someone has been nice enough to make a market with spreads only 10 wide with 20 contracts on each side.
This is a market that is especially well suited to mathematical modeling. No one on Intrade has much of an intuitive sense of what the next unemployment number is most likely to be, much less a sense of how much variability there is in that estimate. I ran some quick numbers on this. Here are the reports for the past couple months.
I modeled this as a gaussian distribution, taking the mean and stdev from the past 5 weeks. I get the following probabilities:
>550k >99.9%
>575k 99.3%
>600k 91.5%
>625k 47.2%
>650k 6.6%
I made bets on all the offers that were substantially outside of this range -- long >575 at 95, long >600 at 85, and short >650 at 25.
Is this model appropriate? Any model has flaws. This one assumes that nothing particularly different happened this week compared to the last 5. I think that's pretty accurate. I also tested the model using all the data shown in the graph. The precise probabilities were different, but they all told me to make the same bets. Is the distribution truly gaussian? Probably not. There's always fat tails, black swans, etc. But for each of the bets I made, the model told me that the market maker's odds were off by a factor of 4 or more. That's a very big margin of safety.
This is a market that is especially well suited to mathematical modeling. No one on Intrade has much of an intuitive sense of what the next unemployment number is most likely to be, much less a sense of how much variability there is in that estimate. I ran some quick numbers on this. Here are the reports for the past couple months.
I modeled this as a gaussian distribution, taking the mean and stdev from the past 5 weeks. I get the following probabilities:
>550k >99.9%
>575k 99.3%
>600k 91.5%
>625k 47.2%
>650k 6.6%
I made bets on all the offers that were substantially outside of this range -- long >575 at 95, long >600 at 85, and short >650 at 25.
Is this model appropriate? Any model has flaws. This one assumes that nothing particularly different happened this week compared to the last 5. I think that's pretty accurate. I also tested the model using all the data shown in the graph. The precise probabilities were different, but they all told me to make the same bets. Is the distribution truly gaussian? Probably not. There's always fat tails, black swans, etc. But for each of the bets I made, the model told me that the market maker's odds were off by a factor of 4 or more. That's a very big margin of safety.
Knitting is boring
I know I said I would stay away from the Entertainment/Box Office lines, but I just shorted some NIGHT.MUSEUM.+$100.0M at 2.3. I don't think it's going too far out on a limb to say that Night at the Museum 2 will not be one of the top 11 weekend openers of all time. I just need to do this 450 more times to make up what I lost on Watchmen.
Swine flu rule changes
The swine flu contracts were originally going to be expired based on the number of confirmed cases, as published by the CDC at http://www.cdc.gov/h1n1flu/update.htm. Now the CDC has stopped publishing the number of "Confirmed" cases and is only publishing the "Confirmed + Probable" case numbers.
It's not obvious to me how Intrade should handle this type of situation. I think they did a good job when they designed the contract. They chose a metric which was quantitative, authoritative, and freely available on the internet. Of course, to properly evaluate these contracts, traders had to recognize that the number of lab-confirmed cases would be only a small fraction of the actual number of cases. But after making this adjustment, the contracts offer quite a good vehicle for predicting whether the flu will peter out or continue to spread.
So what to do when the CDC replaces the number of interest with a different number? Intrade have already decided to use the Confirmed + Probable number to evaluate the contracts. In practice, this probably won't make much difference. As you can see from the graph below, the Confirmed tracks the Confirmed + Probable pretty well, with a lag of about 2 days.
This change will only make a difference if we are right at the margin of one of the contracts on June 30th (e.g. 25001 Confirmed + Probable cases). Suppose the number does come in at 25001 on June 30th, and it's clear that at least 2 of those cases are Probable, rather than Confirmed. I think there's a strong case to be made that the >25000 contracts, bought or sold before the rules change, should be expired at zero. [Full disclosure: I probably have more riding on this than anyone else on Intrade. Of the 1043 SWINE.FLU.US.JUN09.25000+ contracts that have traded hands, 605 of them were sold by me.]
That is a highly unlikely scenario. The bigger question is, what will they do if the CDC stops publishing any numbers? I asked the exchange manager, who replied, "We do not expect the CDC to stop publishing numbers. If they do however, then a new source of expiry information will be found and everyone advised." This is not particularly informative, but they don't have a lot of good options. Here's what they could do, in decreasing order of terribleness:
-Use the estimated number of cases in the US. This would be absolutely wrong. Everyone knows that the actual number of cases is far higher than the number of lab-confirmed cases. If we went by estimated cases, all the contracts would have expired at 100 already. All the trading and all the prices are based on predicting lab confirmations, not estimated case numbers.
-Use the last-published numbers. This is what is suggested by letter of the contract, which states that "The figures as published at the time of expiry will be used to expire this market." Presumably, the most recent (non-updated) numbers would still be available on the CDC's site. It's rather unfair as it does not really capture the spirit of the contract.
-Cobble together numbers from state health departments and CDC seasonal monitoring. This is difficult to do, but it would closely track the intent of the original contract.
-Find an alternative authority such as the WHO. This would be ideal, but may not be available.
Really only the latter two are fair ways of dealing with the situation. Anyone have other suggestions?
It's not obvious to me how Intrade should handle this type of situation. I think they did a good job when they designed the contract. They chose a metric which was quantitative, authoritative, and freely available on the internet. Of course, to properly evaluate these contracts, traders had to recognize that the number of lab-confirmed cases would be only a small fraction of the actual number of cases. But after making this adjustment, the contracts offer quite a good vehicle for predicting whether the flu will peter out or continue to spread.
So what to do when the CDC replaces the number of interest with a different number? Intrade have already decided to use the Confirmed + Probable number to evaluate the contracts. In practice, this probably won't make much difference. As you can see from the graph below, the Confirmed tracks the Confirmed + Probable pretty well, with a lag of about 2 days.
This change will only make a difference if we are right at the margin of one of the contracts on June 30th (e.g. 25001 Confirmed + Probable cases). Suppose the number does come in at 25001 on June 30th, and it's clear that at least 2 of those cases are Probable, rather than Confirmed. I think there's a strong case to be made that the >25000 contracts, bought or sold before the rules change, should be expired at zero. [Full disclosure: I probably have more riding on this than anyone else on Intrade. Of the 1043 SWINE.FLU.US.JUN09.25000+ contracts that have traded hands, 605 of them were sold by me.]
That is a highly unlikely scenario. The bigger question is, what will they do if the CDC stops publishing any numbers? I asked the exchange manager, who replied, "We do not expect the CDC to stop publishing numbers. If they do however, then a new source of expiry information will be found and everyone advised." This is not particularly informative, but they don't have a lot of good options. Here's what they could do, in decreasing order of terribleness:
-Use the estimated number of cases in the US. This would be absolutely wrong. Everyone knows that the actual number of cases is far higher than the number of lab-confirmed cases. If we went by estimated cases, all the contracts would have expired at 100 already. All the trading and all the prices are based on predicting lab confirmations, not estimated case numbers.
-Use the last-published numbers. This is what is suggested by letter of the contract, which states that "The figures as published at the time of expiry will be used to expire this market." Presumably, the most recent (non-updated) numbers would still be available on the CDC's site. It's rather unfair as it does not really capture the spirit of the contract.
-Cobble together numbers from state health departments and CDC seasonal monitoring. This is difficult to do, but it would closely track the intent of the original contract.
-Find an alternative authority such as the WHO. This would be ideal, but may not be available.
Really only the latter two are fair ways of dealing with the situation. Anyone have other suggestions?
Only Huntsman can go to China
Nobody saw this coming. Obama has named John Huntsman Jr., erstwhile Governor of Utah, to be ambassador to China.
I understand 100% why Obama would do this. Huntsman has great political talent, and might have challenged Obama in 2012 (though I predicted Huntsman would be the Republican nominee in 2016). Following thousands of years of historical precedent, Obama removed a potential rival by sending him to a post in a faraway land. Not only that, Huntsman is actually quite well qualified: he did his Mormon mission in Taiwan, he speaks fluent Mandarin, and he's been ambassador to Singapore.
But why on earth would Huntsman accept the job? Despite his lefty-moderate social positions, he was quite popular as governor of Utah. He probably saw that 2012 wasn't his year to run for President, but he could have stayed on as governor until 2016. Governing Utah is fairly provincial, and being ambassador to China certainly allows Huntsman to check the "foreign policy experience" box on his Presidential resume. But doesn't working for Obama get Huntsman labeled as a traitor to the Republicans? Unless there is a seismic shift in the Republican party, I think this could poison the well for him in 2016. I suppose it depends on how long he serves and what he does afterwards. Maybe he recognizes that his religion makes it difficult for him to win a Republican primary, and he's either given up on the Presidency or plans to switch parties? I'm sure he's thought out a plan, but it's hard to guess what it is.
I understand 100% why Obama would do this. Huntsman has great political talent, and might have challenged Obama in 2012 (though I predicted Huntsman would be the Republican nominee in 2016). Following thousands of years of historical precedent, Obama removed a potential rival by sending him to a post in a faraway land. Not only that, Huntsman is actually quite well qualified: he did his Mormon mission in Taiwan, he speaks fluent Mandarin, and he's been ambassador to Singapore.
But why on earth would Huntsman accept the job? Despite his lefty-moderate social positions, he was quite popular as governor of Utah. He probably saw that 2012 wasn't his year to run for President, but he could have stayed on as governor until 2016. Governing Utah is fairly provincial, and being ambassador to China certainly allows Huntsman to check the "foreign policy experience" box on his Presidential resume. But doesn't working for Obama get Huntsman labeled as a traitor to the Republicans? Unless there is a seismic shift in the Republican party, I think this could poison the well for him in 2016. I suppose it depends on how long he serves and what he does afterwards. Maybe he recognizes that his religion makes it difficult for him to win a Republican primary, and he's either given up on the Presidency or plans to switch parties? I'm sure he's thought out a plan, but it's hard to guess what it is.
SCOTUS
I haven't taken any big positions on the next Supreme Court justice. I like Wood and Kagan, but haven't bought of lot of either of them. My main move has been arbitraging by betting against everyone. At certain points over the last couple weeks, the sum of the bids on all the potential nominees totaled more than 100. It was more than 120 at one point. So I simply sold them all, locking in a small guaranteed profit, with the chance of a much larger profit if Obama picks someone who isn't on the list.
Election Law Blog has a very smart piece about why it will be Wood. No surprise her stock has risen over the past week.
Election Law Blog has a very smart piece about why it will be Wood. No surprise her stock has risen over the past week.
Terrible new market
Intrade listed a new market on climate change. This market was sponsored by TheClimateBet.com via Intrade's new "list a market for $100" program. This is a poorly conceived excuse to promote the "bet" between Marketing Professor Scott Armstrong and Al Gore. It reflects a total lack of understanding of climate science and probability.
The contract goes through some contortions to create "projections" representing Armstrong and Gore's views. The bet is on the "global average temperature anomaly" in 2010, in other words the difference between the 2010 temperature and the average temperature between 1961 and 1990. Theclimatebet.com claims Armstrong's prediction is 0.284 degrees C, which is the anomaly in 2007. This is a strange prediction for someone who thinks global warming is a myth. Shouldn't his prediction be... zero? Or does he think that the climate changed between 1990 and 2007, but that it will stop from now on?
More importantly, there is tremendous year-to-year variability in global temperature. This variability completely swamps the difference in Armstrong's and Gore's predictions. As the contract interprets it, the predictions for global average temperature anomaly differ by a value of 0.09. The standard deviation of the GATA since 1979 is 0.18.
Look at this picture: the black dots are the yearly averages, and the red and blue dots are Armstrong and Gore's alleged predictions. Does any serious person think that the winner of this bet has anything to do with whether global warming is real?
The contract goes through some contortions to create "projections" representing Armstrong and Gore's views. The bet is on the "global average temperature anomaly" in 2010, in other words the difference between the 2010 temperature and the average temperature between 1961 and 1990. Theclimatebet.com claims Armstrong's prediction is 0.284 degrees C, which is the anomaly in 2007. This is a strange prediction for someone who thinks global warming is a myth. Shouldn't his prediction be... zero? Or does he think that the climate changed between 1990 and 2007, but that it will stop from now on?
More importantly, there is tremendous year-to-year variability in global temperature. This variability completely swamps the difference in Armstrong's and Gore's predictions. As the contract interprets it, the predictions for global average temperature anomaly differ by a value of 0.09. The standard deviation of the GATA since 1979 is 0.18.
Look at this picture: the black dots are the yearly averages, and the red and blue dots are Armstrong and Gore's alleged predictions. Does any serious person think that the winner of this bet has anything to do with whether global warming is real?
Quick Hits
I like to keep track of how my posts move the Intrade markets. After yesterday's Olympic Host ramblings, I see there are about 200 new bids on North America between 52 and 55.
____________
Daniel Gross does the math on the GM bankruptcy. Looks someone will make $2.4 billion from credit default swaps if the company goes bankrupt. As a caveat, we don't know how much overlap there is between people who hold the CDS and people who hold the bonds. CDS can be used to speculate as well as to insure. My guess? We see a flurry of last-minute negotiations, which bring in a large majority of the bondholders but not the necessary 90% threshold. GM gets sent off to a more complicated version of Chrysler's pre-packaged bankruptcy. But I'm not going long now at 92.5.
____________
Why no volume in the FRANKEN.SEATED contracts? After all, this is very important politically, with plenty of Senate business on hold until Franken gets seated.
1. There hasn't been any news.
2. There's too much ambiguity. To estimate a date, you need to know a) when the MC SC will rule, b) whether they will remand it to lower courts, c) whether the lower courts will order further recounts, d) when the final results are in, will Pawlenty sign the election certificate? I can guess at D (yes). People who watch the MN SC closely estimate a ruling in mid-to-late June, but they don't really know. NO ONE knows about B and C. Nobody has a precise estimate of fair value, so nobody wants to take long-ish odds on a position.
____________
Daniel Gross does the math on the GM bankruptcy. Looks someone will make $2.4 billion from credit default swaps if the company goes bankrupt. As a caveat, we don't know how much overlap there is between people who hold the CDS and people who hold the bonds. CDS can be used to speculate as well as to insure. My guess? We see a flurry of last-minute negotiations, which bring in a large majority of the bondholders but not the necessary 90% threshold. GM gets sent off to a more complicated version of Chrysler's pre-packaged bankruptcy. But I'm not going long now at 92.5.
____________
Why no volume in the FRANKEN.SEATED contracts? After all, this is very important politically, with plenty of Senate business on hold until Franken gets seated.
1. There hasn't been any news.
2. There's too much ambiguity. To estimate a date, you need to know a) when the MC SC will rule, b) whether they will remand it to lower courts, c) whether the lower courts will order further recounts, d) when the final results are in, will Pawlenty sign the election certificate? I can guess at D (yes). People who watch the MN SC closely estimate a ruling in mid-to-late June, but they don't really know. NO ONE knows about B and C. Nobody has a precise estimate of fair value, so nobody wants to take long-ish odds on a position.
Olympics Host
The host of the 2016 Olympics is a surprisingly widely traded contract. We've got a total volume of nearly 20,000, with big offers and tight bid/ask spreads. I took a closer look when Politico put out an article this morning about Obama's support for bringing the Olympics to Chicago.
The candidate cities are Chicago, Madrid, Tokyo, and Rio de Janeiro. From my brief study of the history of the games, there are 4 components that go into the selection of a host:
1) Technical preparedness. Will the host have all the necessary stadiums, hotel rooms, public transportation, and other infrastructure? Is there sufficient funding, and do the local citizens support hosting the games?
2) Continental rotation. Since 1952, the Summer Games have never been hosted on the same continent twice in a row.
3) Geopolitics. Every 20 years or so, the Olympics goes to a developing nation to let it highlight its improvement on the international stage. I'm thinking of Mexico City in 1968, Seoul in 1988, and Beijing in 2008.
4) Bribery. After scandals related to the 1996 Atlanta games and 2002 Salt Lake City games, there has been a supposed crackdown. We'll see... I'm sure clever hosts can still find ways to funnel money to important Olympics officials.
My impression is that technical preparedness is more of a threshold issue. The International Olympic Committee (IOC) doesn't always pick the country with the best infrastructure, but it must ensure that the infrastructure is up to a certain standard. All four cities meet this standard. The one knock against Chicago is that games in the US are privately financed -- unlike other countries, the government will not guarantee funding.
Continental rotation would seem to rule out Madrid (the Games will be in London in 2012), and to favor Chicago or Rio. The Olympics have not been held in North America since 1996, and they've never been held in South America. Supporters of Madrid and Rio argue that the London games are "Anglo-Saxon" games, and that Asian and Anglo-Saxon games should be followed by Latin games.
Geopolitically, there is certainly momentum to hold the games in Rio as a nod to the development of Brazil and Latin America. However, China's 2008 games may count as the "developing nation" games this time around, working against Rio's chances. Obviously, Obama's international popularity and his close ties to Chicago are a huge factor in that city's chances.
As far as bribery, I'll go ahead and indulge in stereotypes. While the Spanish and Brazilians may be known for petty corruption in their police and civil service, Chicagoans take the cake for sophisticated organized racketeering.
I bought some North America contracts at 55. Just as Rome's games in 1960 and Japan's in 1964 showed they had rehabilitated themselves following World War II, Chicago 2016 games will display an America rehabilitated from the Bush era.
______
As a side note, the Politico story has a link to www.gamesbids.com, a clearinghouse for news about the selection process. Gamesbids also has a BidIndex(TM) number, which is a "secret formula" which incorporates technical evaluations along with geopolitics. It's also hilariously inaccurate, having predicted only 1 of the past 5 bids correctly, the no-brainer of Beijing in 2008. I mean, seriously, the guy created this formula retrospectively. It's based on a few hard numbers like the IOC's bids, but there are also gooey inputs like "geopolitics." He couldn't tweak the design a little bit to get the right answers? To be fair, the BidIndex(TM) does seem to identify the top contenders. The actual winner was usually #2 on the BidIndex(TM) rankings published just before the vote. Still, this is a great example of people putting way too much confidence in models.
The candidate cities are Chicago, Madrid, Tokyo, and Rio de Janeiro. From my brief study of the history of the games, there are 4 components that go into the selection of a host:
1) Technical preparedness. Will the host have all the necessary stadiums, hotel rooms, public transportation, and other infrastructure? Is there sufficient funding, and do the local citizens support hosting the games?
2) Continental rotation. Since 1952, the Summer Games have never been hosted on the same continent twice in a row.
3) Geopolitics. Every 20 years or so, the Olympics goes to a developing nation to let it highlight its improvement on the international stage. I'm thinking of Mexico City in 1968, Seoul in 1988, and Beijing in 2008.
4) Bribery. After scandals related to the 1996 Atlanta games and 2002 Salt Lake City games, there has been a supposed crackdown. We'll see... I'm sure clever hosts can still find ways to funnel money to important Olympics officials.
My impression is that technical preparedness is more of a threshold issue. The International Olympic Committee (IOC) doesn't always pick the country with the best infrastructure, but it must ensure that the infrastructure is up to a certain standard. All four cities meet this standard. The one knock against Chicago is that games in the US are privately financed -- unlike other countries, the government will not guarantee funding.
Continental rotation would seem to rule out Madrid (the Games will be in London in 2012), and to favor Chicago or Rio. The Olympics have not been held in North America since 1996, and they've never been held in South America. Supporters of Madrid and Rio argue that the London games are "Anglo-Saxon" games, and that Asian and Anglo-Saxon games should be followed by Latin games.
Geopolitically, there is certainly momentum to hold the games in Rio as a nod to the development of Brazil and Latin America. However, China's 2008 games may count as the "developing nation" games this time around, working against Rio's chances. Obviously, Obama's international popularity and his close ties to Chicago are a huge factor in that city's chances.
As far as bribery, I'll go ahead and indulge in stereotypes. While the Spanish and Brazilians may be known for petty corruption in their police and civil service, Chicagoans take the cake for sophisticated organized racketeering.
I bought some North America contracts at 55. Just as Rome's games in 1960 and Japan's in 1964 showed they had rehabilitated themselves following World War II, Chicago 2016 games will display an America rehabilitated from the Bush era.
______
As a side note, the Politico story has a link to www.gamesbids.com, a clearinghouse for news about the selection process. Gamesbids also has a BidIndex(TM) number, which is a "secret formula" which incorporates technical evaluations along with geopolitics. It's also hilariously inaccurate, having predicted only 1 of the past 5 bids correctly, the no-brainer of Beijing in 2008. I mean, seriously, the guy created this formula retrospectively. It's based on a few hard numbers like the IOC's bids, but there are also gooey inputs like "geopolitics." He couldn't tweak the design a little bit to get the right answers? To be fair, the BidIndex(TM) does seem to identify the top contenders. The actual winner was usually #2 on the BidIndex(TM) rankings published just before the vote. Still, this is a great example of people putting way too much confidence in models.
I am not James Surowiecki
This made me laugh this morning:
Thanks for the link. Though I'm flattered that my writing is good enough to confuse someone into thinking that I work for the New Yorker, I really am a neuroscientist doing this in my spare time. Obviously the name of my blog is a reference to Surowiecki's book The Wisdom of Crowds. Beyond that, we've got no connection. And my only tie to Intrade is that I bug their exchange manager with enough questions that we're on a first-name basis.
See also Wiser Than The Crowd, who is probably the pseudonym of a very smart financial expert, financial entrepreneur and “prolific” book author —with ties to InTrade (which he should disclose, if it were the case).
Thanks for the link. Though I'm flattered that my writing is good enough to confuse someone into thinking that I work for the New Yorker, I really am a neuroscientist doing this in my spare time. Obviously the name of my blog is a reference to Surowiecki's book The Wisdom of Crowds. Beyond that, we've got no connection. And my only tie to Intrade is that I bug their exchange manager with enough questions that we're on a first-name basis.
Get in there, Mortimer, and SELL!!
There's been a sharp decline in the number of new swine flu cases. Here's the updated graph.
It's still hard to tell exactly where this is going, but I've become bearish on all the remaining contracts. Not because swine flu isn't going to spread, but because the CDC will stop testing every suspected case. News reports are saying most states are going to a more relaxed testing stance, where they test people who have been hospitalized, and in areas where there is no virus yet, but don't test the vast majority of people presenting with flu symptoms.
I closed out my long position on the >10000 at 85, and I've gone short. Trimmed my long position in >5000, and put in small shorts in all the others.
Hat tip to Delphi from the Intrade forums, who got me thinking seriously about the change in CDC testing policies. He asked me to include these caveats, because he doesn't want to be accused of manipulating the markets:
From listening to the CDC conference call this afternoon, I can say that the answers are 1) many states did not report over the weekend, and 2) correcting errors. The CDC said they had finished out most of the testing backlog, but that they were continuing to receive 300-400 new samples per day. No word on how those samples are expected to turn out. Meanwhile many state health departments have begun doing their own testing.
I expect tomorrow we'll see a big surge in case load, as the states that took the weekend off will report 3 days worth of cases in 1 day. The following days will be instructive in how aggressive the CDC will be in testing over the rest of the summer.
It's still hard to tell exactly where this is going, but I've become bearish on all the remaining contracts. Not because swine flu isn't going to spread, but because the CDC will stop testing every suspected case. News reports are saying most states are going to a more relaxed testing stance, where they test people who have been hospitalized, and in areas where there is no virus yet, but don't test the vast majority of people presenting with flu symptoms.
I closed out my long position on the >10000 at 85, and I've gone short. Trimmed my long position in >5000, and put in small shorts in all the others.
Hat tip to Delphi from the Intrade forums, who got me thinking seriously about the change in CDC testing policies. He asked me to include these caveats, because he doesn't want to be accused of manipulating the markets:
1. Did state health departments give their (perhaps stressed-out) staff part or all of the weekend off? It's hard to say. I do note that only 9 of 51 states reported new numbers at all, according to the table, today. I tried to look for a dip in the numbers last weekend, adjusted for the background surge as the illness spread, but it's hard to tease out. And besides, the same "due for time off" rule probably didn't apply 7 days ago, since we were still in such a high alert mode. I can't find any mention in news reports whether state agencies were working on this over the weekend.
2. California. What's up with their numbers? Sudden drop in confirmed cases from 282 to 191 in one day. Could be a glitch in today's numbers, or correcting errors in previous (redundant?) reports from the state.
From listening to the CDC conference call this afternoon, I can say that the answers are 1) many states did not report over the weekend, and 2) correcting errors. The CDC said they had finished out most of the testing backlog, but that they were continuing to receive 300-400 new samples per day. No word on how those samples are expected to turn out. Meanwhile many state health departments have begun doing their own testing.
I expect tomorrow we'll see a big surge in case load, as the states that took the weekend off will report 3 days worth of cases in 1 day. The following days will be instructive in how aggressive the CDC will be in testing over the rest of the summer.
Tinkerbell Stress Tests
Anyone who's paying attention knows that the bank stress tests that were announced yesterday are bullplop. If you're just tuning in, Barry Ritholtz has some posts that go into convincing detail about the overly optimistic economic projections and the extremely lax 25-1 leverage targets. In short, this is a charade.
Geithner's plan is what I call the Tinkerbell strategy: if we all clap our hands and believe in fairies, the banks will be solvent and the recession will go away. To some extent, this is really true. The root word of credit is credere, Latin for "to believe." No banking system can work if no one trusts anyone else. Trust is a necessary but not sufficient component of the system. But how does the promotion of these transparently phony stress tests promote trust?
A bubble occurs when individual speculators stop caring about the fundamental value of an asset, and start concerning themselves only with what other speculators will pay for it. In the early stages of a bubble, it makes rational sense to overpay for the asset, with the strategy of re-selling to a "greater fool" at an even more inflated price. When it becomes apparent to all involved that the trading value of the asset is utterly unsupported by the fundamentals, the supply of fools runs out, and the trading price collapses back to the fundamental price.
Geithner is attempting something which has never been done before. He is trying to re-inflate the banking bubble. With the full faith and credit of the US government behind him, he's trying to convince the fools to dive back into the market. Armed with a sufficient number of fools, even the smart, rational players will get involved again. In theory, the process of the initial bubble will repeat itself. Even the smart money will buy in, intending to re-sell to the fools.
It won't work that way. Bubbles are based on a confluence of psychological factors. In particular, they rely on participants genuinely believing that the assets involved are special. During the run-up to the real estate crisis, we heard over and over how house prices had never fallen nationwide. How "they aren't making any more land." How financial innovation and derivatives had distributed risk to those best able to bear it, making the system robust and more stable than ever before. No one is going to fall for these lines again. Everyone will be watching everyone else, waiting for the first sign of panic, knowing that they need to be the first ones out the door when it starts. It's a recipe for a quick boom and an even quicker bust.
Geithner's plan is what I call the Tinkerbell strategy: if we all clap our hands and believe in fairies, the banks will be solvent and the recession will go away. To some extent, this is really true. The root word of credit is credere, Latin for "to believe." No banking system can work if no one trusts anyone else. Trust is a necessary but not sufficient component of the system. But how does the promotion of these transparently phony stress tests promote trust?
A bubble occurs when individual speculators stop caring about the fundamental value of an asset, and start concerning themselves only with what other speculators will pay for it. In the early stages of a bubble, it makes rational sense to overpay for the asset, with the strategy of re-selling to a "greater fool" at an even more inflated price. When it becomes apparent to all involved that the trading value of the asset is utterly unsupported by the fundamentals, the supply of fools runs out, and the trading price collapses back to the fundamental price.
Geithner is attempting something which has never been done before. He is trying to re-inflate the banking bubble. With the full faith and credit of the US government behind him, he's trying to convince the fools to dive back into the market. Armed with a sufficient number of fools, even the smart, rational players will get involved again. In theory, the process of the initial bubble will repeat itself. Even the smart money will buy in, intending to re-sell to the fools.
It won't work that way. Bubbles are based on a confluence of psychological factors. In particular, they rely on participants genuinely believing that the assets involved are special. During the run-up to the real estate crisis, we heard over and over how house prices had never fallen nationwide. How "they aren't making any more land." How financial innovation and derivatives had distributed risk to those best able to bear it, making the system robust and more stable than ever before. No one is going to fall for these lines again. Everyone will be watching everyone else, waiting for the first sign of panic, knowing that they need to be the first ones out the door when it starts. It's a recipe for a quick boom and an even quicker bust.
Swine Flu: Movin' on Up
All of the swine flu contracts are substantially higher over the past few days. The US case count has continued its exponential growth unabated. If the case count continues to grow at the 35% daily rate it's been averaging over the past week, we'll have 10,000 cases by May 15th. If we continue seeing 250 new cases per day (as we have the past 2 days), we'll have 10,000 cases by June 12th.
I'm pretty certain we will have more than 10,000 cases in the US by the end of June. In fact, there's a decent chance we'll have 100,000. The question is, how many of those will be laboratory-confirmed by the CDC? The CDC and WHO will continue to take this very seriously. They're concerned about a much more widespread, possibly mutated version of the virus coming back next fall. Even if it doesn't, they see this as a training run for future outbreaks.
At this point, it's not particularly useful to identify every single individual case. Indications are that authorities will scale back the intensity of their monitoring efforts, and will track the swine flu like they track seasonal flu. In all likelihood, the new cases will level off below 1000/day, mainly due to lab testing limitations. That's still plenty to hit 10,000 cases before the end of June.
I'm long on all of these contracts.
I'm pretty certain we will have more than 10,000 cases in the US by the end of June. In fact, there's a decent chance we'll have 100,000. The question is, how many of those will be laboratory-confirmed by the CDC? The CDC and WHO will continue to take this very seriously. They're concerned about a much more widespread, possibly mutated version of the virus coming back next fall. Even if it doesn't, they see this as a training run for future outbreaks.
At this point, it's not particularly useful to identify every single individual case. Indications are that authorities will scale back the intensity of their monitoring efforts, and will track the swine flu like they track seasonal flu. In all likelihood, the new cases will level off below 1000/day, mainly due to lab testing limitations. That's still plenty to hit 10,000 cases before the end of June.
I'm long on all of these contracts.
More informed SC speculation
Worthwhile Money, the guys who put me onto the Souter departure contracts, have a long post about Souter's replacement.
I agree with their overall assessment of how Obama will make his pick, but I don't think Clinton will see even a temporary bounce, and Cass Sunstein is a dude. There's about a 95% chance the nominee will be a woman.
I agree with their overall assessment of how Obama will make his pick, but I don't think Clinton will see even a temporary bounce, and Cass Sunstein is a dude. There's about a 95% chance the nominee will be a woman.
Arlen Specter: "My only moral conviction is that I should remain in the Senate."
Interview with the NY Times:
Specter soon apologized, saying, "In the swirl of moving from one caucus to another, I have to get used to my new teammates...I’m ordinarily pretty correct in what I say. I’ve made a career of being precise. I conclusively misspoke."
In other words, "I'm sorry I said what I meant, instead of what was politically expedient. I'm ordinarily a pretty good liar."
At this point, I'm not sure what reason Specter could give to Pennsylvania Democrats as to why he should represent them. He'll almost certainly face a primary challenge from Congressman Joe Sestak. Specter is way ahead of Sestak in a recent poll, leading 57 to 20, but that's mainly because no one knows who Sestak is. Sestak's favorable/unfavorable numbers are 15/3, with 82% not sure. Once a proper race gets underway, Sestak will have full-throated support of the Dem base and netroots community. In a closed primary, that gives him a very good chance.
I don't expect Specter to make any decision based on policy merits. But just out of self-preservation, he better start leaning way, way left to head off that primary challenge.
With your departure from the Republican Party, there are no more Jewish Republicans in the Senate. Do you care about that?
I sure do. There’s still time for the Minnesota courts to do justice and declare Norm Coleman the winner.
Which seems about as likely at this point as Jerry Seinfeld’s joining the Senate.
Well, it was about as likely as my becoming a Democrat.
Specter soon apologized, saying, "In the swirl of moving from one caucus to another, I have to get used to my new teammates...I’m ordinarily pretty correct in what I say. I’ve made a career of being precise. I conclusively misspoke."
In other words, "I'm sorry I said what I meant, instead of what was politically expedient. I'm ordinarily a pretty good liar."
At this point, I'm not sure what reason Specter could give to Pennsylvania Democrats as to why he should represent them. He'll almost certainly face a primary challenge from Congressman Joe Sestak. Specter is way ahead of Sestak in a recent poll, leading 57 to 20, but that's mainly because no one knows who Sestak is. Sestak's favorable/unfavorable numbers are 15/3, with 82% not sure. Once a proper race gets underway, Sestak will have full-throated support of the Dem base and netroots community. In a closed primary, that gives him a very good chance.
I don't expect Specter to make any decision based on policy merits. But just out of self-preservation, he better start leaning way, way left to head off that primary challenge.
The Supremes
My RSS feed is rife with uninformed speculation about Obama's replacement for David Souter. I think there will be very few real leaks about this, and plenty of guessing and over-reading of tea leaves. As a rule, if Drudge puts up a headline about a candidate, you can go ahead and short their bounce.
Obama was a constitutional law professor. He knows way more about the prospective replacements than any previous president. I wouldn't be surprised if he had made up his mind before he was even sworn in. He probably feels like he can find someone with a center-left reputation who will turn out to match his own legal thinking quite well.
I think it's a mistake to assume Obama wants to nominate the biggest liberal he can get confirmed. In his appointments so far, he has tended to go for individuals with strong abilities, without too much consideration for ideological purity. I think he'll look for someone with a very strong legal mind and the ability to influence the thinking of other justices.
It's also a mistake to think that anyone Obama nominates can get confirmed without a big fight from the Republicans. They've made it clear that they will do whatever they can to derail Obama's agenda. Sucking up all the political oxygen for a couple weeks over a confirmation battle is exactly what the leadership is looking for. Obama knows this. He won't play into their hands by nominating, say, Pamela Karlan, who's a lesbian and vocal gay-rights activist. But he also won't choose a moderate for the sake of trying to avoid a fight.
I haven't done too much research into the individuals yet. Here are my initial impressions on the five who are leading on Intrade right now.
Granholm: No. She's been attorney general and governor of Michigan. Her record has been adequate, but not particularly impressive. She has no history as a judge, and, other than a few years clerking for a 6th circuit appeals court judge, has little top-level legal experience.
Sears: An African-American woman who is currently chief justice of the Georgia Supreme Court. She plans to leave this position to "pursue other opportunities," whatever that means. I'm guessing she doesn't particularly relish being a judge, and wants to be the dean of Emory Law School. I'm sure she would consider a Supreme Court nomination if it were offered, but she doesn't appear to chomping at the bit. She was also found guilty of minor violations of campaign laws in her 2004 re-election campaign.
Kagan: Obama already appointed her to be the current Solicitor General. The appointment raised eyebrows at the time because she had never argued a case at trial, and has not argued before the Supreme Court. She comes from the same institutions as Obama, Harvard Law and U Chicago. She has an extensive background in both government and legal academia.
Wardlaw: Could well be her. There is a totally unsubstantiated blog report that she is being vetted. She wrote the lower-court opinion in Redding v. Safford, about whether school officials were correct to strip-search a young girl for prescription-strength Advil.
Sotomayor: Has been pegged as the front-runner, and is drawing 10 times more criticism from the right than anyone else. I think people assume that Obama has a checklist of characteristics he's looking for from a justice, and Sotomayor fits the bill. Female? Check. Minority? Check. Nominated by both Bush I and Clinton? Check. Rags-to-riches life story? Check. She's an accomplished jurist, but there are plenty of others out there. Opponents are pushing the story that she's a bully and not particularly smart (Jeffrey Rosen's shoddy article being the chief example.) Marc Ambinder points out that if she is going to be nominated, her supporters better start pushing back publicly. I suspect Obama is happy to have her take the fire, so he can roll out his actual nominee as a blank slate. I can't rule her out, but I don't see any value in betting on her at 38.
Obama was a constitutional law professor. He knows way more about the prospective replacements than any previous president. I wouldn't be surprised if he had made up his mind before he was even sworn in. He probably feels like he can find someone with a center-left reputation who will turn out to match his own legal thinking quite well.
I think it's a mistake to assume Obama wants to nominate the biggest liberal he can get confirmed. In his appointments so far, he has tended to go for individuals with strong abilities, without too much consideration for ideological purity. I think he'll look for someone with a very strong legal mind and the ability to influence the thinking of other justices.
It's also a mistake to think that anyone Obama nominates can get confirmed without a big fight from the Republicans. They've made it clear that they will do whatever they can to derail Obama's agenda. Sucking up all the political oxygen for a couple weeks over a confirmation battle is exactly what the leadership is looking for. Obama knows this. He won't play into their hands by nominating, say, Pamela Karlan, who's a lesbian and vocal gay-rights activist. But he also won't choose a moderate for the sake of trying to avoid a fight.
I haven't done too much research into the individuals yet. Here are my initial impressions on the five who are leading on Intrade right now.
Granholm: No. She's been attorney general and governor of Michigan. Her record has been adequate, but not particularly impressive. She has no history as a judge, and, other than a few years clerking for a 6th circuit appeals court judge, has little top-level legal experience.
Sears: An African-American woman who is currently chief justice of the Georgia Supreme Court. She plans to leave this position to "pursue other opportunities," whatever that means. I'm guessing she doesn't particularly relish being a judge, and wants to be the dean of Emory Law School. I'm sure she would consider a Supreme Court nomination if it were offered, but she doesn't appear to chomping at the bit. She was also found guilty of minor violations of campaign laws in her 2004 re-election campaign.
Kagan: Obama already appointed her to be the current Solicitor General. The appointment raised eyebrows at the time because she had never argued a case at trial, and has not argued before the Supreme Court. She comes from the same institutions as Obama, Harvard Law and U Chicago. She has an extensive background in both government and legal academia.
Wardlaw: Could well be her. There is a totally unsubstantiated blog report that she is being vetted. She wrote the lower-court opinion in Redding v. Safford, about whether school officials were correct to strip-search a young girl for prescription-strength Advil.
Sotomayor: Has been pegged as the front-runner, and is drawing 10 times more criticism from the right than anyone else. I think people assume that Obama has a checklist of characteristics he's looking for from a justice, and Sotomayor fits the bill. Female? Check. Minority? Check. Nominated by both Bush I and Clinton? Check. Rags-to-riches life story? Check. She's an accomplished jurist, but there are plenty of others out there. Opponents are pushing the story that she's a bully and not particularly smart (Jeffrey Rosen's shoddy article being the chief example.) Marc Ambinder points out that if she is going to be nominated, her supporters better start pushing back publicly. I suspect Obama is happy to have her take the fire, so he can roll out his actual nominee as a blank slate. I can't rule her out, but I don't see any value in betting on her at 38.
Really? You're betting on swine flu?
I feel a bit of a moral qualm in placing wagers on a potential disease epidemic. I feel the same way about contracts on hurricanes, whether Osama will be caught, or whether dictators will be removed from power. There's a certain distaste to profiting from someone else's misfortune. Yet I think the existence of these markets are a net positive, from a utilitarian perspective.
The objection to these markets is really a visceral one. Say you win a bet that Kim Jong-Il will remain in power -- does that make you happy? Millions of people continue to live in poverty and oppression, and you're $25 richer? It's not a good feeling. And you probably don't tell your friends about it, unless you're the kind of guy who goes to the ballgame wearing the visiting team's colors.
But this "yuck" factor is intangible, and these markets can produce material benefits. It is vital that policymakers get correct information about what will happen in the world. Prediction markets could send strong signals that swine flu was about to spread wildly, or that Mugabe's regime was about to collapse. These markets are far from omniscient, but they often represent quite a good way of pooling collective knowledge. These merged individual intuitions can provide an additional piece of information about what will happen in the future.
I'll take knowledge and a queasy feeling over sitting and hoping.
The objection to these markets is really a visceral one. Say you win a bet that Kim Jong-Il will remain in power -- does that make you happy? Millions of people continue to live in poverty and oppression, and you're $25 richer? It's not a good feeling. And you probably don't tell your friends about it, unless you're the kind of guy who goes to the ballgame wearing the visiting team's colors.
But this "yuck" factor is intangible, and these markets can produce material benefits. It is vital that policymakers get correct information about what will happen in the world. Prediction markets could send strong signals that swine flu was about to spread wildly, or that Mugabe's regime was about to collapse. These markets are far from omniscient, but they often represent quite a good way of pooling collective knowledge. These merged individual intuitions can provide an additional piece of information about what will happen in the future.
I'll take knowledge and a queasy feeling over sitting and hoping.
Swine Flu
Cliff notes: Epidemiological modeling is hard. What I present here is pretty much the equivalent of a 4-year old drawing the Mona Lisa in crayon.
There are a couple professional models out there. This model, from Northwestern University, uses information about how people move around the country to try to predict where and how many cases will develop. A team from Indiana University is using a similar technique, and both models predict that there will be about 2000 cases in the US by the end of May.
In making these models, the researchers have to guess at certain constants. Most important is the virulence of the disease: when an infected person meets an uninfected person, how often is the disease transmitted? They calibrate their models based on available data from the CDC, but this is incomplete. There are certainly people out there who have swine flu but have not yet been tested. The researchers account for this, but it involves a lot of guesswork.
So far the models are somewhat accurate. Early predictions have underestimated the case count by about 50%.
Here are graphs of the case count in the US since the outbreak began, and the rate of growth.
So far, the case count appears to be growing exponentially, at a rate of about 25% per day over the last week. If you apply that growth rate through the end of June, it predicts 150 million Americans will have swine flu. Obviously, that's not going to happen. The growth will slow at some point, but I really have no basis for predicting how long the exponential growth will continue. We're already at the end of flu season, so transmission will slow as we enter the warmer months of the year.
I'll keep an eye on it, but I don't have a strong sense of how this will play out.
There are a couple professional models out there. This model, from Northwestern University, uses information about how people move around the country to try to predict where and how many cases will develop. A team from Indiana University is using a similar technique, and both models predict that there will be about 2000 cases in the US by the end of May.
In making these models, the researchers have to guess at certain constants. Most important is the virulence of the disease: when an infected person meets an uninfected person, how often is the disease transmitted? They calibrate their models based on available data from the CDC, but this is incomplete. There are certainly people out there who have swine flu but have not yet been tested. The researchers account for this, but it involves a lot of guesswork.
So far the models are somewhat accurate. Early predictions have underestimated the case count by about 50%.
Here are graphs of the case count in the US since the outbreak began, and the rate of growth.
So far, the case count appears to be growing exponentially, at a rate of about 25% per day over the last week. If you apply that growth rate through the end of June, it predicts 150 million Americans will have swine flu. Obviously, that's not going to happen. The growth will slow at some point, but I really have no basis for predicting how long the exponential growth will continue. We're already at the end of flu season, so transmission will slow as we enter the warmer months of the year.
I'll keep an eye on it, but I don't have a strong sense of how this will play out.
April Results: Back to Braggin'
After last month's $1500 loss, I'm back in the black for April. Total account value plus withdrawals is $61,352, a gain of $8800 vs. last month. The big gainer was the NY-20 election, with additional wins on Chrysler bankruptcy and Souter leaving the Supreme Court.
Everything's Coming Up Milhouse!
Supreme Court Justice Souter to Retire
I've been accumulating this in lots of 25-50 contracts at a time at a price around 25. It was pretty obvious that he was next, imo. All the other justices had hired clerks months ago. Souter always hires late, but in past years even he has hired his clerks by the end of March. The other major candidates, Bader-Ginsberg and Stevens, had made it pretty clear they planned to stay around. Stevens will want to stick around for a couple years and become the oldest SC justice in history; his reputation is that his mind is still as clear as it ever was. Bader-Ginsberg was diagnosed with pancreatic cancer, which can in many cases be fatal. She made several public appearances to dispell any rumors that she might be going anywhere.
Hat tip to worthwhilemoney.com, who put me on to this. This won't expire until the end of the SC term in June, so I suppose it's possible another justice might die in a freak accident or something, but Souter is upper 90% to be the next to leave.