Back On The Polls Tyler Durden Tue, 10/27/2020 - 09:56 Authored by Michael Every via Rabobank, Because 2020 isn’t finished being 2020 yet we shouldn’t really be surprised: there is water on the moon. As my wife said to me on that news breaking, this explains why Nokia was recently talking about being the first to bring the internet there too; after all, H2O and 5G are the main ingredients of modern life. Yes, the moon is only getting 4G for now, and there isn’t any atmosphere or economy: but that’s already true of lots of place in the Covid-19 world (ba doom tish). 2020 still being 2020, will the next shock be the release of uncropped pictures of the US moon landings which have Moonians in the background sitting in deckchairs and sipping water? Meanwhile, it’s now a week to US Election Day, and ACB has replaced RBG on the Supreme Court, handing Trump a major political victory. Some opinion polls are also tightening back into the kind of range they sat in back at this point in 2016. We also have the equivalent jaw-dropping alleged scandals stemming from a drip-drip set of e-mails. Despite the fact the mainstream media and social media have largely opted to ignore or censor such scandal this time round, information, like water, still finds its own level: see the Rasmussen poll yesterday suggesting a majority of all voting demographics, except Democrats, see the allegations as “likely” or “very likely” to perhaps hold some water. A different kind of pole-star (pollster?)... However, Nate Silver at fivethirtyeight.com, who still argues his 2016 election prediction for a Clinton win was accurate as he had also given Trump a 30% chance(!), still has Biden with an aggregate 9.5ppt lead and an 87% chance to win, with only a 3% chance of 2020 being a repeat of the 2016 Red Wave that took the White House, Senate, and House. Real Clear Politics has average polling projections of the same general kind. At the same time, however, a Berkeley Haas study of 1,400 polls from 11 election cycles finds that although they all claim a 95% confidence interval, their outcomes are only 60% accurate – and that is for polls a week ahead of time, with earlier polls even less accurate. Of course, there are less well-covered polls out there using a different methodology (and dubbed “quasi-partisan” by Silver) which do not sit within the mainstream election narrative. Theirs is a hypothesis, backed by most non-poll signs on the ground (including literal signs), that points to the potential for another Brexit-style class-based voting realignment, and another shock to an Establishment that once again didn’t see it coming in its models. Thus the Daily title derived from the fact that “The water in Majorca doesn’t taste quite how it ought to” should all rhyme perfectly if one is the kind of stereotypical British voter who was usually excluded in predictive likely voter modelling. One week until we know if for a Bloomberg reporter, things will turn out how they ought to. Lots is at stake. Indeed, another well-known pollster, Frank Luntz, has stated that if the major polls are completely off in 2020 again then “People like me are going to have to find a new profession.” Frank, Frank, Frank. You are too hard on yourself. There will be no need to do anything new or more socially useful a week tomorrow regardless of the outcome. We can be more than 95% sure of that. There will always be a Gramscian market for political-economy narrative dressed up as data analysis. Look at academic economists: they show one can sell demonstrably failed theories and models for decades and still keep one’s job; and if we treat pollsters the way we do central banks, we will simply increase their budgets so they can tell us --inaccurately-- what more people are thinking about more aspects of their lives. (“Today, 76.5% of men think toothpaste is made of cheese. Our model agrees.”) Or perhaps we can just fold those pollsters into our central banks? Anyway, stocks were already down on Monday because of risk-off sentiment due to the lack of pre-election fiscal stimulus; one wonders what the markets will do if they start to worry that 2020 is not ‘how it oughta’, and instead 2016 redux. Only a week at most to wait. On which note, Bloomberg market comment today says that the recent CNY rally will now pause (not linked to the above point, honest guv); that CNY is 40% undervalued; and that “regulators have already started lifting capital controls.” Now there is a water on the moon story, if so. Because if China really does lift its capital controls ahead, considered opinion is that the currency will be trading at around 10 to the US in very short order. Far, far more liquidity would flood out, not in. That is why they don’t do it. Does that, like the inaccuracy of opinion polls, really need to be made clear in 2020? Apparently so.