A recent comment to yesterday’-s Citibank Panic/Euphoria discussion led to this post being inserted in the weekend linkfest:
According to Tradesports, the odds of the Republican Party retaining control of each Congressional House in the 2006 Mid Term Election is mixed.
In the Senate, only one third of the 100 seats are up for re-election. Votes are staggered, so every 2 years a different third of the 6-year term seats come up. The odds strongly favor GOP retention of the Senate. The Democrats would need to capture 6 seats, something previously viewed as rather unlikely —- not impossible, but unlikely.
Although several races have tightened, this remains a long shot for the minority party. The MSM is, as always, focused on the horse races but not the issues. See Race for Senate control tightens for example.
When we go to the most recent betting on the outcomes, the senate appears to be rather safe. Using trend or technical analysis, we see the GOP retention of Senate has been consistently above 70%- It has recently dipped as Tennessee and Virginia —- formerly “-safe seats”- —- have become competitive. But it would take a significant break of 70 to suggest this was anything but a long-shot.
Senate Trading
via Tradesports
When it comes control of the House, however, the prediction markets present a very different picture. All 435 seats are up for grabs.
Its also one that lends itself even more to TA:
via Tradesports
From last November til today, we see a fairly well maintained down trend. Recently, that was almost reversed on an increase in volume since hitting lows of 40% in September. That was most likely the result of a 30% drop ingasoline prices, a 20+% in Crude Oil prices, and a strong post summer rally in the stock markets.
However, the Foley/Page scandal has now reversed that. On even bigger volume, the downtrend has been maintained, and the House odds are once again near low 40s.
If the election were held today, the Senate would stay GOP, while the House would shift to the Democrats.
~~~
There is one additional note: I have long complained that the relative thinness of these markets —- the number of traders and the dollar amounts at issue —- can make their results somewhat suspect. For more on this, see our 2004 critique, Iowa and Prediction Markets.
Right now the Senate is at bid 67, ask 70, so if your theory is correct, we’ll see the trading price drop considerably more in the near future.
The only problem with using this sort of analysis is that it suffers from the hindsight is 20/20 syndrome. “If an election were held today, this is who would win” is nice to say in theory, but prediction markets are about predicting and not the analysis of past events. What I’m trying to get at is: is the current market odds correctly pricing in what will happen between now and November 7th?
Usually political markets just price in information concerning polling, but this congressional market appears to be pricing in more strategic factors (such as the memory retention of the Foley issue). If Foley was single-handedly enough to take out 15% of the Republican’s house bid, will there be additional developments of issues independent of this that will cause this to evaporate or expand?
I would claim that as this prediction market didn’t see Foley coming (Republicans were up to 55% a couple days before the scandal), it probably won’t correctly price in any short term events that will transpire between now and election day. This is where the opportunity for profit comes.
Just drew a trend line on the Senate chart and it looks like the GOP senate has broken through on high volume. Its just more of a race now – even breaking the 70% barrier doesn’t mean much. Additionally, I would consider this to be range bound trading until recently.
Maybe once political consultants realize that these markets are true markets and can be used as a hedge against their future earnings we may see the volume pick up to a level that would please you (and me) Barry. Additionally, these are probably the people we most want to be involved in this particular market, with the most insight and most direct finacial stake, and then even the most insider knowledge. The markets might become more predictive at that point.
Weak-Form Market efficiency states that you won’t gain any insight from looking at charts. Now maybe John Henry bought the Red Sox just by being lucky all the time for 30 years, but I don’t think this is the case.
Predictive markets are at least partly based on the idea that weak-form EMH isn’t valid with respect to the markets they address- or that the information they convey are ‘meta’ to the entire EMH. Are financial markets useless because of they don’t predict well? I suspect that prediction markets might be a more precise, accurate or immediate form of polling in many cases, and polls should have predictive power. It remains to be seen if predictive markets perform well. In any case, they convey valuable information.
I would suggest that most of (technical) trend-following profits over time come from 1) being long equities (so if there is a trend-following mystery, it’s partially reducible to the equity premium), 2) being long carry, either by being long fixed income or through fx, 3) more recently, being long depletable commodities (energies and metals).
The short equity, short carry and soft/ag commodity (long or short) components of trend-following returns are more flat over time, although they clearly have nice zero to negative correlations to typical portfolios, and outperform in times of stress.