Since folks seem to be discounting at 100% the previous post on why the TEMP.2009.HIGH stock seems massively overvalued, I started worrying that traders out there know something I don't. So, I threw the data into STATA instead and ran a few prediction models with lag effects and a time trend. The best fit model predicts a 2009 temperature anomaly of 0.38821, which is higher than I previously was estimating. But, the probability of exceeding 1998 seems largely unaffected because my confidence bands are tighter. The standard deviation of the regression residuals is 0.100832. The residual tells you the difference between the actual observation and the estimate predicted by the model for any particular year. So, if the actual temperature anomaly turns out to be 0.546, the residual would be -0.15779: the prediction minus the observation gives you that as residual. How often do we observe residuals at least that large on the down side? 7 times in 154 observations: 4.5% of the time.

The simple model explains 84% of the variance in the time series using only lagged dependent variables and a time trend.

Want to replicate this? Get the dataset, add in a null 2009 observation, throw it into STATA, type:

gen year2 = year^2

reg deviation L1.deviation L4.deviation year year2

predict y_hat

predict resid, residuals

Then just take (y_hat - 0.546) for 2009, and see how often the residual is smaller than that value:

sum resid if resid < -0.15778

7 observations in 154.

I wonder what the folks buying at 0.22 know that I don't. No way anybody's using this for hedging against a hot year, not at these stakes....

* Note I kept the first and fourth lags after a general to specific reduction starting with 7 lags.

## Thursday, January 22, 2009

Subscribe to:
Post Comments (Atom)

## 13 comments:

I also tried adding in a break in the time trend at 1970 instead of having the squared term; didn't help. I also tried some ARMA models which didn't provide any substantially different results.

Unless one of our traders is sitting on preliminary data results showing that we've had a huge increase in temperatures over the last couple of months, something odd is going on here.

Maybe whoever is pushing the price up doesn't understand how unlikely the outcome is? Or just feels that the yeah will be hotter? Some people go on gut feeling for stocks I have noticed....

There's so much here that makes no sense to me.

Temp.2009 is trading at 0.55. So, there's a 55% chance that next year's temperature anomaly is > 0.324. But a 24% chance that it's also > 0.546? Yes, the standard deviation of the 150 series is high enough to get that, but certainly not the standard deviation of the year on year changes!

Maybe whoever is pushing the price up doesn't understand how unlikely the outcome is? Or just feels that the yeah will be hotter? Some people go on gut feeling for stocks I have noticed....Jeanette Fitzsimons maybe?

People are idiots, who cares why - let's make money off them! ;-)

I think you just need to be patient with these low priced stocks. I've noticed that prices can trade well away from obvious fundamentals for quite some time because it is hard for people to go short in this capital-constrained iPredict world. Eventually even those supporting the price run out of money, and the stock invariably subsequently collapses as the buying stops and the longs begin stopping out of their position. In the meantime you should just be happy to be able to sell at these levels!

I'm now short >1000 shares. How much does it take?!

People don't believe that statistics can predict the weather. Or they don't want to. Or so far they can't.

I'm not sure, but short that shit.

Now just hoping that we don't have a bunch of traders from NIWA who've preliminary numbers on the likelihood of an exceptional el nina event that would cause us to have one of the biggest year-on-year temperature increases in the last 150 years. But with the massive cold snap in the Northern hemisphere for the start of January, I'd be pretty surprised if we this year were able to manage that kind of increase.

It also takes a lot less margin for folks going long than for folks going short: someone with a long position of as many shares as me has less than a third as much capital frozen.

YEs I think that shares are always bias towards the middle. If you short a share at 90 cents you tie up (and risk) only 10 cents a share. But if you buy it you tie up and risk 90 cents.

The reverse is true for lower priced shares. Buying a share at 10 cents costs, well 10 cents. But shorting it can be risky and tie up 90 cents.

So I would say there is a small but significant bias towards 50 cents a share no matter what the price and fundementials are.

There is evidence of favourite/longshot bias in markets generally: longshots price higher than outcomes and favourites are underpriced. That also holds in sports betting markets. But that's for stuff in the 5% tails and it's a small bias. Here it looks like the price is about 18 cents too high.

I think there must be a bias towards the centre, perhaps just because people aren't willing to wait.

Take the EFA Repeal for example, that's been under 90c for ages, despite National repeatedly saying that it's in their 100 day plan, and their 100 day plan ends about 60 days before the stock close.

I'd like to see Ipredict's maximum deposit raised towards 10k and then we would see who has some conviction in view and thus get to the right answer more rapidly.

Post a Comment