Nov 30, 2016

Post-election summary
The largest takeaway from this year's election is that we are not in a golden age of forecasting and, no matter our grasp on statistics and sampling, polling can and will fail us. We still have to actually have an election to see what will happen. We have also been reminded that a widely accepted consensus of experts can be wrong.

The forecasting was so badly lopsided this year that we have to consider the possibility that Democrat turnout was suppressed due to extreme overconfidence. Nate Silver, a respected analyst who has been praised by both the right and left, was actually attacked for saying Donald Trump had a 30% chance to win; his detractors insisting the number was closer to 1%. A serious narrative developed, at one point, that Texas could flip blue this year -- Trump would carry the state by almost 10 points. The New York Times ran a hit piece on the L.A. Times for daring to say that Trump was winning in their "Daybreak" poll. The media and the forecasting community gave Democrats good reason to avoid the hassle of going out and voting.

Credit where it's due
The L.A. Times actually had to defend itself when it came under attack by the New York Times for the sample it was using in its "Daybreak" tracking poll. Shown below is the data from said poll, which bucked the consensus and showed Trump with a lead for much of the cycle. This poll was one of very few which picked Trump to win, and the only one to be pushed by one of the country's largest newspapers. Detractors are still insisting, even now, that the poll is wrong because Clinton carried the popular vote. Given the state of the rest of forecasting in this election, I don't find that argument convincing. They picked the winner using a nontraditional methodology and sample, and they deserve credit for getting it right.
They aren't forecasters or analysts, but an honorable mention in the "credit where it's due" category goes to Ann Coulter on the right and Michael Moore on the left. Michael Moore gets credit for being a lone voice on the left predicting a Trump victory, and demonstrating that he really does understand the working people of the rust belt. Ann Coulter gets credit for calling a Trump victory before he even secured the nomination, and enduring ridicule and condescending laughter on the Bill Maher show for doing so. Like Moore, Coulter called the rust belt with 100% accuracy and was one of very few conservative pundits saying that Florida wasn't a requirement for a Trump win - which ended up being true.

Battleground polling bias
Call it error, or call it bias, what's indisputable is that the battleground polling this year had it in spades. Not only was the polling bad, but it was bad where it mattered the most. In Pennsylvania, Wisconsin, and Michigan the average bias in Clinton's favor was 4.6 points. Wisconsin claims the honor this year of the largest difference between the RCP poll average and the actual outcome of the election at a massive 7.3 points. In ten out of fourteen battleground states the polling was biased toward Clinton. Most of the error seems to have come from using turnout models from 2008 and 2012 and vastly oversampling Democrats when polling.

State RCP average Actual Error/Bias
Wisconsin Clinton +6.5 Trump +0.8 Clinton +7.3
Missouri Trump +11.0 Trump +19.1 Clinton +7.1
Iowa Trump +3.0 Trump +9.5 Clinton +6.5
Ohio Trump +3.5 Trump +8.0 Clinton +4.5
Michigan Clinton +3.4 Trump +0.2 Clinton +3.6
Pennsylvania Clinton +1.9 Trump +1.1 Clinton +3.0
North Carolina Trump +1.0 Trump +3.7 Clinton +2.7
Florida Trump +0.2 Trump +1.1 Clinton +0.9
Georgia Trump +4.8 Trump +5.2 Clinton +0.4
New Hampshire Clinton +0.6 Clinton +0.3 Clinton +0.3
Arizona Trump +4.0 Trump +3.5 Trump +0.5
Virginia Clinton +5.0 Clinton +5.5 Trump +0.5
Colorado Clinton +2.9 Clinton +4.9 Trump +2.0
Nevada Trump +0.8 Clinton +2.4 Trump +3.2

The true outlier this year is Nevada
The rust belt and industrial midwest surprised everyone this year, but not much has been said about the surprise out west. Using the Real Clear Politics average of polls, there were 19 states that were forecast to go to Hillary Clinton that went to Hillary Clinton, and 27 states forecast to go to Donald Trump that went to Donald Trump. There were three states forecast to go to Clinton that went to Trump. But there was just one state that was forecast to go to Trump that ended up going Clinton, and that state was Nevada.

In the battleground states the average polling bias was +2.2 points in Clinton's favor. In Nevada the apparent bias is +3.2 points in Trump's favor, which makes for a total deviation of 5.5 points from the average. Furthermore, the battleground average is actually skewed by 0.3 points just from Nevada being present in the calculation. Without Nevada the average bias is even greater in Clinton's favor. The worst of the bad polling in the battleground states was at least consistently bad in that it universally favored Clinton. Nevada's polling was not only bad, but inconsistent with the general trend of favoring Clinton.

Prediction markets failed as well
Prediction markets failed to shrug off the forecasting consensus this year, and the "wisdom of the crowds" proved no better than the wisdom of the experts. In 2008 and 2012 prediction markets did very well, but so did the general forecasting community. This year would appear to vindicate those who say that prediction markets reinforce the expert consensus rather than coming to an independent conclusion. PredictWise had Clinton at a 89% favorite to win, and PredictIt 82%. BitCoin prediction market Predictious was down at 75%, five points higher than Nate Silver who had Clinton at 70% going into election day. Pivit was hovering around the 80% mark for most of October.

Because of outages and problems getting data from the site, did not use the Pivit prediction market for the final forecast -- and ended up with a more accurate prediction after changing data sources. Staying with Pivit would have meant getting North Carolina and Florida wrong as well.

Pivit had a good early season
Pivit was at peak accuracy in late September and early October, between roughly 9/20 and 10/5. It was at this time that Pivit had 47/50 states correct, missing only Pennsylvania, Michigan, and Wisconsin. Missing those states means missing the election, of course, but nevertheless this was the most accurate period for Pivit. Shown below is the forecast from October 5, which reflects the relatively stable betting of late September and early October.

During that timeframe, I noted that was diverging from Pivit somewhat rapidly. Pivit had Trump maintaining a lead in Florida, Ohio, North Carolina, and Iowa while the 538 forecast shifted toward a Clinton victory in all of those states. The divergence was the most extreme in Florida and North Carolina, two states that ended up blue in the Nate Silver forecast but went to Trump. I believed this was significant because it showed a prediction market going against the grain and shrugging off polls and other forecasts. Ultimately, that changed.

Something happened to Pivit around October 23
As I noted in a previous news post Pivit experienced enormous shifts in betting over a roughly 100 hour period. These shifts were larger than any previous jumps in the betting data on Pivit, and moved Florida and North Carolina out of Trump's column and into Clinton's column. We now know that Trump did indeed carry those two states, and that Pivit was more accurate before the large and sudden shifts. So what happened? We can only speculate. It's not unreasonable, since Pivit is a play-money prediction market, to consider the possibility that it was gamed. There is no way to know without access to Pivit's internals. But the following stands:
  1. Pivit was accurate in early October when it held Florida, North Carolina, and other swing states for Trump in spite of most other forecasts shifting toward Clinton.
  2. Pivit lost accuracy in late October, suddenly and erratically, when there were jolting shifts (as large as 24% in the case of Florida ) toward Clinton.
  3. After the shifts the Pivit forecast ended up identical to the 538 forecast.
  4. After these events, made a more accurate forecast by removing Pivit as a data source.
Going forward will probably move to a more custom forecast, rather than rely on a single prediction market for data.

The Recounts
What are the odds that the proposed recounts change anything? Let's put it this way: the prediction markets haven't even created events for recounts in Michigan, Wisconsin, and Pennsylvania. Michigan, the tightest of the three races, has already been counted twice and Trump won both times with a margin of 10,000 votes. For reference, the most enthusiastic Al Gore supporters claim a statewide recount would have netted Gore a swing of about six hundred votes (out of six million) in the Florida recount in 2000.

Nov 25, 2016

Michigan officially goes to Trump
The state of Michigan is set to certify the official election results on the 28th. All of the state's county clerks have certified their counts and Trump has maintained a margin of around 10,000 votes.

The Electoral College will convene on December 19th and, barring the exceedingly rare faithless elector, will make Donald Trump the President Elect at 306 electoral votes.

Nov 9, 2016

Election Forecasting Dealt Devastating Blow
The forecast missed Nevada, Wisconsin, and Pennsylvania and will more than likely miss Michigan as well. These states represent a substantial value in terms of electoral votes, and were more than enough to change the forecasted outcome of the election.

Other forecasting outfits, in fact most of them, did substantially worse. This includes Nate Silver at FiveThirtyEight who missed all of the above states, except Nevada, and also missed another 44 electoral votes in Florida and North Carolina. The biggest takeaway from this forecasting blowout is the principle of "garbage in, garbage out." Nobody had a model that saw past the bad polls in the rust belt. With bad polling comes bad forecasting.

Stay tuned for much more post-election analysis here in the coming days.

Nov 7, 2016

Final forecast locked in!
The final forecast combines weighted data from the PredictIt prediction market with the Real Clear Politics poll average. The prediction market data is favored unless some specific conditions are met: State-by-state, the prediction market data is compared against the poll average and the poll trajectory. Where the prediction market data disagrees with the latest polls, more weight is given to the polls but only in states where the trajectory of the poll indicates a large swing in previous five days.

There are 3 states which are essentially too close to call, and Trump retains a razor-thin advantage in each as shown here:

Nevada 51%
Florida 51%
North Carolina 56%

Clinton can miss in any of these states and still pull off a fairly easy victory, whereas all of them are must-win for Trump under most scenarios. Without a large upset like Pennsylvania or Michigan they are critical to his path to 270. This is why Clinton's overall odds of winning stand at ~80% to Trump's 20%. There are lots of places where Clinton can miss. Trump can't miss once.

Some final overall probabilities for a Clinton victory as reported from different sources:

PredictWise 89%
NYT Upshot 84%
PredictIt 82% 81%
Predictious 75%
Nate Silver 69%

Oct 25, 2016

Trump's uphill battle
If nothing changes, Donald Trump faces three realistic scenarios on election day. The first is that the polls are accurate and he is defeated handily by Hillary Clinton, even with some of the in-play states breaking in his favor. The second scenario would see the polls slightly biased towards Clinton, in which case Trump picks up most of the swing states but still loses. The third scenario, and the only one in which Trump can win, is that the polls are repeating their Dem+6 bias of 2014. If that were the case, Trump would probably grab all of the in-play states and maybe even an upset in a blue state.

Even in the third scenario Trump would need to pick off a high-value blue state like Michigan or Virginia, or two smaller states like Nevada and New Hampshire. If nothing changes, even calling this an "uphill battle" is being generous. The forecasters echo this sentiment with their probabilities of a Clinton win:

Pivit 95%
PredictWise 91%
Nate Silver 87%
PredictIt 85%
Predictious 77%

Oct 23, 2016

Unusually large swings on Pivit
Over the last 100 hours, there have been some extreme changes in some of the swing state forecasts on Pivit. These movements are several times larger than the largest swings seen previously this season. Even more strangely, the bulk of the movement happened overnight on a single night, Weds. Oct 19th. I also encountered errors accessing Pivit on that day with error messages saying the server was at capacity. The swings are as follows:

Florida Clinton +24%
North Carolina Clinton +14%
Arizona Clinton +12%
Ohio Clinton +12%
Iowa Clinton +10%

Pivit is not a real-money prediction market, and its internals are not transparent. This makes it susceptible to both internal and external manipulation. The site's odds, according to Pivit itself, are derived from "historical data, real-time info, and the wisdom of the crowds" which suggests that their algorithm isn't using 100% user betting data. So, this extreme swing could have been the result of an internal adjustment made by the site's operators. The massive swing could have also come from a large influx of new users betting heavily on Clinton in the swing states. Or, less likely I believe, the swing could have been completely organic with a large portion of the user base making a sudden change in their betting.

Prior to this sudden shift, Pivit was one of the few forecasts that had Trump ahead in Florida, Iowa, and Arizona and tied with Clinton in Ohio. Now, Pivit is very much aligned with Nate Silver's FiveThirtyEight forecast. Although less interesting, maybe it's more accurate now. In the end, accuracy is what counts! I look forward to revisiting what happened here on Pivit after the election.

Oct 18, 2016

Clinton continuing to gain ground in swing states
Hillary Clinton continues to chip away at Donald Trump's narrow lead in Florida, Iowa, Arizona and North Carolina. All four of those states are under a 60% probability of win for Trump on the Pivit prediction market, which makes them almost too close to call for either candidate. Ohio falls in that same category, with Clinton holding a 53% chance of victory.

Trump needs every state mentioned above, plus an upset in another state somewhere else, to secure a win. His fall in these vital swing states make it virtually impossible for him to win, and the Monte Carlo simulation using these probabilities confirms that: Clinton's chances of a win are now over 90%.

At this point, Trump's only real chance of winning lies in a major media event bringing Clinton down, or the polls being wildly inaccurate. FiveThirtyEight's analysis of the 2014 Senate race polls revealed an average bias of Democrat +6. If that were the case again this year, Trump might stand a reasonable chance of taking some key swing states. But polls tend to be more frequent and more robust during Presidential Election years versus off-year Senate races.

Oct 14, 2016

Nevada, Ohio back in Clinton territory
Nevada is now solidly back in Clinton's corner, registering a 63% chance of victory. Ohio is trending in her direction and she currently has a very small lead there. Other key swing states, Florida, North Carolina, and Iowa are also trending in the direction of Hillary Clinton and have all dipped below the 60% threshold for Trump -- somewhere they haven't been in weeks.

The aggregate probability of Clinton winning the election now stands at 88%.

Oct 5, 2016

FiveThirtyEight shows much larger Clinton uptick than Pivit

In both 2008 and 2012 prediction markets closely tracked with Nate Silver's FiveThirtyEight forecast. The prediction markets and the analysts all ended up with very accurate forecasts, and they all had basically the same forecast. Indeed, that has been the case with the Pivit prediction market this season as well. Even when the curves had some separation, their trendlines still followed the same pattern - until the first Presidential debate on September 26th. As of today, the Pivit electoral map is still registering 259 electoral votes for Donald Trump, whereas the FiveThirtyEight map is almost 40 electoral votes lower.

There seems to be a significant deviation developing between the Nate Silver probabilities and those on Pivit and, oddly enough, it's the most pronounced in the volatile swing states. The charts below illustrate the gap developing between the two forecasts, and it's clear that major deviations began after the 9/26 debate. It is hard to overstate how odd this is. Maybe Pivit just hasn't caught up to the polling cycle as fast, or maybe Pivit traders weren't as swayed by the debate. Only one thing is certain: if the trend continues they can't both be right.

Oct 4, 2016

Recent history of prediction market accuracy
In 2012 the Intrade prediction market had a perfect forecast for 24 consecutive days between mid September and mid October. This also marked the longest stretch of consecutive days during the entire season with the same forecast in place. This year seems to be following the same pattern of a relatively stable forecast going into October. With the exception of Nevada fluctuating back and forth, the forecast has remained relatively unchanged since September 14th. In 2008, the Intrade prediction market did not stabilize at its most accurate forecast until October 10th.

Sep 28, 2016

Clinton gets small bump on Pivit post-debate
The first Presidential debate between Hillary Clinton and Donald Trump, which took place two nights ago, has resulted in a slight uptick for Clinton on the Pivit prediction market. Her chances of winning the election rose roughly four percentage points to 72%, using the state-by-state betting data from Pivit. Nevada moved slightly into Clinton territory, now posting a 53% chance of breaking in her favor. Trump remains eleven hard-to-get Electoral College votes away from a victory in the latest map.

Sep 21, 2016

September forecasts increasingly favorable for Trump
The chart below tracks the overall chances of a Clinton victory during the month of September using both Pivit and Nate Silver state-by-state probabilities. Note that this is not the output of FiveThirtyEight's proprietary computer model, but rather a simple monte-carlo simulation using the individual state by state probabilities reported on Likewise, this chart uses the state-by-state betting data from Pivit, rather than the single contract for a Clinton win. In both cases this method results in slightly more favorable odds for Clinton, and is probably a good measure of the upper bound on her probability of victory.

It remains to be seen if this is a trend that will continue into the final 47 days before the election, or if this spurt of Trump momentum lacks staying power. We are heading into the period of time that forecasts were at their highest accuracy during 2008 and 2012 -- I plan to do a news post about that when October gets here.

Shown below is the Real Clear Politics poll average for the same period of time in the chart above. We can see that Pivit and FiveThirtyEight track the poll average closely, indicating the poll-driven nature of election forecasting, even for prediction markets. It is notable that even though the poll average has narrowed to a single point, the forecasters are still giving Clinton roughly a 70% chance of winning the election. This is because of the difficult math Trump faces in the Electoral College, and how hard it will be for him to pick up any additional states for a win -- such as Colorado, Pennsylvania, Virginia, or Michigan.

Sep 15, 2016

Nevada goes red on Pivit
For the first time this election season Nevada has gone red on the Pivit map, giving Donald Trump a total of 265 electoral votes to Clinton's 273. This is by far the best the map has looked for Trump, putting him within five electors of a win. But it's going to be very hard for Trump to pick up those last five votes; he needs a difficult win in a blue state. If he pulled off a huge upset in Pennsylvania, Michigan, or Virginia that would do it. So would a victory in Colorado, or a combination of New Hampshire and a single elector from Maine (Maine can split its votes.)

The difficulty of Trump sealing the deal on one of those remaining states is reflected in the overall probability of a Clinton win, which still stands at 67%. If the election were held today, Hillary would still have a 2:1 edge over Trump according to the Pivit map. And that isn't likely to budge much, even with Trump's momentum, unless we see a big movement in the polls in one of these key states.

Sep 14, 2016

Can Trump win Pennsylvania?
Pivit traders currently give Hillary Clinton a 68% chance of winning Pennsylvania, which is a few points lower than PredictIt and So, can Trump win? The general consensus seems to be that Trump has roughly a one-in-four chance of winning the state, which hasn't gone red since 1988. Trump is ahead by large margins with white voters who have less than a college degree, while Clinton remains strong with the urban poor. Pennsylvania has no shortage of both demographics, and both campaigns seem to understand that it's going to come down to turnout. Trump supporters in PA point to the nearly 2:1 edge they have in party conversions during early 2016. Going into the last few months of the election cycle, they seem to have the momentum and enthusiasm to win -- but that doesn't necessarily translate to a victory, just ask Bernie Sanders.

There are currently no Pennsylvania polls recorded by Real Clear Politics since Thursday, September 8. Those polls, taken before Clinton's health problems were revealed, had her up five points in the Keystone State. That margin, if it follows the trend of the rest of the polls, has undoubtedly tightened since the events at the 9/11 memorial and the announcement that the candidate has pneumonia.

Sep 8, 2016

Clinton dives 16 points in ten days, still sits at 80% probability of winning
Hillary Clinton's chances of winning, according to Pivit, peaked around 96% and have since dropped 16 percentage points to today's 80% figure. Once solidly ahead in North Carolina and Iowa, those two states are now in Trump territory, with Trump sitting at 68% in North Carolina - larger than Hillary's lead in Pennsylvania.

Clinton has plummeted in Ohio as well with the state now sitting almost at 50/50 with Trump trailing by less than half a percent. Florida is also a dead-heat at this point with Hillary only carrying a 2% greater probability-of-win over Trump. However even with the recent gains by Trump, he still has an uphill batte to 270 electoral votes. Even if he picks up three key states that Mitt Romney lost, Florida, Ohio, and Iowa, he'd still need another 10 electoral votes to get the win. That means some combination of New Hampshire, Colorado, and Nevada. Either that, or pull of something that no Republican has done since 1988: a win in Pennsylvania. And that is why Mrs. Clinton still sits at 80% to win.

Aug 24, 2016

Polls swing toward Clinton, then Trump
Hillary Clinton had an amazing week in the polls following the conclusion of both party conventions, but that seems to be fading in recent days as states like North Carolina and Iowa trend back towards Trump. But Pivit traders continue to give Hillary Clinton an easy win with 347 electoral votes, more than Barack Obama claimed in 2012.

Pivit traders also continue to create arbitrage with the state-by-state betting which is in disagreement with their own overall election contract, by ten percentage points. The electoral map proposed by Pivit traders aggregates to a Clinton probability-of-win topping 92%, where the straight Clinton-to-win contract sits at 82%. There are no real-money prediction markets that have Clinton above 80% currently. The Pivit Electoral Map is sitting twenty points higher than Bitcoin prediction market Predictious.

Forecast Clinton Win
Pivit Electoral Map     92%
Nate Silver 85%
Pivit 82%
PredictWise 79%
PredictIt 73%
Predictious 72%

Aug 2, 2016

This Year's Most Likely Electoral College Ties
Although the likelihood of an exact 269-269 tie in the Electoral College is fairly low, 1-2% probability, there are a few maps this year that are quite plausible as ties. In the event of a tie in the Electoral College, the newly elected Congress breaks the tie, with the House voting (one vote per state) on the President and the Senate voting on the Vice President. All indications currently are that the Republicans would win both votes, as they are expected to retain dominance of both houses.

The map depicted below is the tied-vote most consistent with current polling that has Trump ahead in Nevada and New Hampshire. Clinton would have to hold on to her lead in Virginia, and Trump would have to lock down highly contested Florida and Iowa. The rest of the map would surprise no one, and is in line with current polling.

In the map below Trump wins Florida and pulls a surprise, rust-belt upset in Michigan. Hillary holds on to new, blue, western strongholds in Colorado and Nevada.

The map below has the west mimicking the 2000 election, but looking like 2012 in the southeast. The surprise here would be the northeast, with Trump picking up Pennsylvania and New Hampshire.

Jul 25, 2016

Internally inconsistent betting on Pivit
The forecast on Pivit for the winning party for the 2016 election stands at 61% for the Democrats to win, as of July 25. This is in line with other prediction markets such as Predictious (55%), PredictWise (68%), IEM (64%), and Predictit (65%). Nate Silver of gives Hillary Clinton a 53% chance of winning the election. However, Pivit traders seem to have thrown all that out the window when it comes to building an actual electoral map.

The state-by-state betting on Pivit aggregates to an overall probability of Democrat victory of 87% - over 25 percentage points higher than Pivit's actual contract for the aggregate victory. The bets could both be wrong, but they cannot both be right. If the state-by-state bets are right, then the 2016 winning party contract is wrong. If the winning party contract is right, then there is at least one state on the map that is very wrong, or several that are somewhat wrong. It's probably the latter.

This sort of thing usually doesn't happen in prediction markets, because two ways to make the same trade at vastly different prices represents an arbitrage opportunity. However, on Pivit a number of things allow for an inconsistent market. For one, it's a point system and there is no real money at stake. The incentive to make a small but guaranteed profit is not there in the Pivit market where it would be in a real-money market. Secondly, Pivit is not 100% crowd-driven. The internals are not transparent, as the site hints at "historical" and "real-time" information being infused into the odds. Thirdly, it's not a trivial thing to look at an electoral map and derive the aggregate odds of victory, you pretty much need a computer to do it and Pivit traders probably are not going to the trouble to do the calculations.

So, what is most likely wrong with the Pivit electoral map? The betting on Florida, Ohio, Iowa, New Hampshire, and Nevada is inconsistent with the latest polls which have Donald Trump ahead in those states. If those wagers reflected the polls, the Pivit map would look identical to the Nate Silver map. And one thing you can be sure of is that FiveThirtyEight's aggregate probability is in mathematical agreement with its electoral map. But maybe we're not giving Pivit traders enough credit: maybe they're looking past the convention bump that Trump got in the polls, assuming it will fade as it often does. Maybe the Pivit map is right, and the winning party contract is wrong -- in which case Donald Trump is in real trouble.

Jul 19, 2016

Source of data for this year's forecast.
For the 2008 and 2012 elections, based all its predictions on data from The Irish company was the world's highest volume prediction market for U.S. political events and it performed amazingly well, correctly forecasting 49/50 states in 2012 and missing the Electoral College tally by a single vote in 2008. Unfortunately shut down in late November of 2012; more on why below. Although the company has indicated that "Intrade 2.0" is under development, we don't yet know when it will go online. There are no other real-money prediction markets with the kind of volume and the scope of contracts that Intrade offered.

Prediction markets are unique in that there is no single person, polling outfit, or think tank modeling the election, analyzing the polls, and determining the odds of a particular outcome. Rather, a multitude of individuals do the legwork - buying and selling contracts on the outcome of an event and the market clearing price of those contracts tells us the odds. It's a proven model, and quite remarkable in its simplicity. It's the shortest and most elegant route to an aggregation of all the polling data and analysis available to the public. In keeping with the philosophy of aggregating as much data as possible, will make its 2016 forecasts using data from the Pivit prediction market, created by the founders of Intrade. Although Pivit is not a real-money prediction market, and hints at some internals that might not be entirely crowd-driven, it is one of the few prediction markets this year offering a contract for each state in the union.

So what happened to Intrade?
Intrade suffered a series of setbacks starting in 2010 when U.S. banking regulators prohibited U.S. citizens from funding their accounts with credit or debit cards. In May of 2011 CEO and founder John Delaney passed away on Mount Everest, just fifty meters from the summit. Then In 2012, just after the election, the U.S. Commodity Futures Trading Commission (CFTC) filed suit seeking an injunction against U.S. citizens trading on the site, claiming that Intrade contracts were "commodity options" and that the company was soliciting U.S. citizens to buy and sell unregulated securities. In the end, Intrade was forced to suspended the accounts of U.S. citizens which quickly decimated trading volume on the site. Shortly thereafter the site ran afoul of Irish banking regulations as some sizeable "financial irregularities" were discovered on the company's books, and the site closed down.

Although the company is far from blameless in its own downfall, the power and reach of U.S. government regulators is well illustrated in the story of Intrade, an Irish company located on Irish soil. Despite correctly forecasting two presidential elections, Intrade was treated by the U.S. government as just another illegal online casino. All the usual buzzwords (gambling, money laundering, terrorism) were bandied about, and the very thing that made it so successful (lots of American participants) is what ultimately painted the target on its back. It is unlikely we will ever see a prediction market with political contracts operate with the blessing of the U.S. government, as the official position of the CFTC remains that such contracts "involve gaming and are contrary to the public interest." Read more about the death of Intrade at

back to top

All Posts

Nov 30
Nov 25
Nov 9
Nov 7
Oct 25
Oct 23
Oct 18
Oct 14
Oct 5
Oct 4
Sep 28
Sep 21
Sep 15
Sep 14
Sep 8
Aug 24
Aug 2
Jul 25
Jul 19