Prediction Markets on Recession by January 2021

PredictIt, accessed 4pm Pacific today:

Why the recent movement? One piece of information, the Manufacturing PMI:


Below 50 indicates contraction.

In addition, despite the assurances of the administration, markets do not seem to believe in an incipient turnaround:

Figure 1: 10 year-3 month Treasury spread, (blue), 10 year-2 year (red) and 5 year-3 month (green), all in %. Federal Reserve via FRED, Treasury, author’s calculations.

All this against a backdrop of elevated (to say the least) policy uncertainty.

Figure 2: Economic Policy Uncertainty index (blue). Source: accessed 8/22/2019.

Note that big data in the form of Google Trends suggests heightened interest in the possibility of recession.

Figure 3: Google Trends: search word “recession”. Accessed 8/22/2019, 5:10pm Pacific.

32 thoughts on “Prediction Markets on Recession by January 2021

  1. invertedintrovert

    Some great points here. You mentioned ISM Manufacturing, which certainly shaved GDP estimates from Nowcast. And yet, the model is currently looking for Q3 GDP of 1.76%, which would be higher than Q1 and Q2. The chorus of recession seems to be much greater than the data or earnings suggest. But, the yield curve has a different idea I suppose.

      1. Barkley Rosser

        Curiously enough, while he has all kinds of exaggerated worries about a future Dem presidency and a ridiculous claim that the New Deal caused a “capital strike” (language of Ayn Rand from Atlas Shrugged), he does not seem to be a fan of Trump’s and recognizes at least that Trump’s trade wars are generating a lot of uncertainty.

  2. Steven Kopits

    You should include the PredictIt question, Menzie: “Will there be a recession in Trump’s 1st term?”

    That this is 42% is actually quite remarkable, given that the bet requires two full quarters of recession before the November election. Therefore, we are speaking of, really, 2019 Q1, and 2020 Q1, Q2, Q3. Two of those have to be in recession, so PredictIt refers to a recession starting no later than April 1, 2020, that is, a mere seven months from now. At the same time, trucking tonnage surged in July, home sales look good, and initial unemployment claims are still signalling a recession a year out. Bill McBride is not even on recession watch.

    So that’s a pretty tight window, with no oil shock, no material overhang of housing, no surge in interest rates…really, none of the standard indicators beyond yield inversions, which are not to be discounted, but again, would not typically signal two quarter of recession in such short time.

    Therefore, a recession would presumably come from some other source, and again, we are circling back to China. Can China do this? Tariffs? Pure uncertainty?

    1. Menzie Chinn Post author

      Steven Kopits: Gee, it’s in the title of the post (and by the way, it’s a probability for not before the election, but before January 2021…

      1. Steven Kopits

        The Rules
        The U.S. economy shall experience at least two consecutive quarters with a negative annual growth rate in real gross domestic product, as rounded to the first decimal, according to the third estimate of such quarters’ GDP growth rates released by the Bureau of Economic Analysis. Should Mr. Trump not complete his first term in office, only those quarters during part or all of which he served as president may qualify.

        So, in two respects, the bet is different than your title.

        First, Trump must remain president during that time. So it has to do with Trump’s first term, not the date.
        Second, there must be an entire period qualifying as a recession during that time, not just the start of it. I interpreted your title as suggesting that a recession would have to start by then, which I think is a pretty fair bet at 42%. A full recession by year end 2020, that’s a pretty aggressive call at this point, I think. Not impossible, though, by any means.

        Thus the period in question is realistically Q4 2019 – Q4 2019, five quarters, but only if Trump remains president. It’s not that big a difference, but it would have been more accurate to reproduce the question as stated in the source site, I think.

        1. Steven Kopits

          Sorry, let me amend, because Trump will serve some of January 2020, so Q1 2021 still counts…

          I keep thinking his term ends when he’s voted out, but of course that isn’t true.

        2. Steven Kopits

          So if I have this right, the bet is, assuming Trump serves out his term,

          “Prediction Markets on Recession by March 2021”, if end or term qualifying as a recession, or
          “Prediction Markets on Recession by October 2020”, if the start

    2. Willie

      Bill McBride is typically not on recession watch until the recession is on us or nearly so. He has, however, made some discouraged sounding remarks lately. I read his blog on a near daily basis, and value what he has to day. That said, he’s one of the earlier commentators to recognize a recession, even if he calls it being on recession watch. My take is that the general consensus is that we are not in a recession until the recession is over anyway. At least McBride looks for it about when a recession hits, or did in 2007.

      I fully expect a recession in 2020 now, what with the escalation of trade wars with China. The Chinese know what they are doing. Trump clearly does not. He will mismanage and thrash around until he’s kicked over all the furniture and broken everything there is to break. He’s breaking the farmers who supported him, tariffs have now broken the steel workers who supported him. He is in the process of breaking Wall Street and his plans to allow the banking casino to reopen will break Wall Street in a way that breaks the rest of the country when Wall Street comes whining to the door for a bailout. I’m not sure what’s left to break, unless it’s consumer spending, and that will break on the rocks that Trump has thrown around.

      1. Steven Kopits

        Here’s the last thing Bill McBride wrote about a recession, August 14th:

        Don’t Freak Out about the Yield Curve

        There are reasons to be concerned. The global economy is slowing. The US economy has slowed. Current policy (especially on trade) is a drag on growth.

        But I wouldn’t freak out about the yield curve.

        In mid-1998 the spread between the 10 year and the 2 year went slightly negative, and a recession didn’t start until 2001 – over 2 1/2 years later. Of course the Fed cut rates in 1998 – just like the current situation.

        When the spread turned solidly negative in 2000, the Fed was raising rates. That would be a more concerning scenario.

        Also, with overall yields so low, I’m not sure this indicator is as useful as it has been. The yield curve is indicating economic weakness, but I’m not currently on recession watch.

        Bill reports a lot of economic news. Some of those indicators have been soft. But by no means all.

        June home sales were a cycle high, and July was up 4.1% yoy

        Black Knight: “Mortgage Delinquencies See Strong Recovery from June Spike”

        Weekly Initial Unemployment Claims decreased to 209,000 (at or near cycle lows)

        “ATA Truck Tonnage Index Surged 6.6% in July”

        By no means all grim.

        1. Willie

          No not all grim. Grim enough that the overall tone of the message was gloomier than McBride has been for quite some time. Remember that he went on recession watch almost the same time the Great Recession started. He was gloomy but not on recession watch before it hit. I am reading tea leaves, because that is how I make predictions. He is a pretty good tea leaf to read, in my opinion.

    1. pgl

      Are we also Hereby Ordered to start reading those worthless blog posts by Steven Kopits? Yep – let’s go full blown Command and Control!

  3. pgl

    Trump must think he is Stalin:
    Trump orders U.S. companies to “start looking for alternatives” to China

    President Donald Trump is demanding that U.S. businesses “immediately start looking for an alternative to China,” including bringing their manufacturing back to the U.S. The president issued the directive in a series of tweets on Friday morning after China hiked tariffs on $75 billion of U.S. products.

    Trump calls Fed Chairman Jay Powell ‘enemy,’ compares him to Chinese President Xi

  4. pgl

    George Conway on Donald Trump:
    George Conway: ‘I hereby order White House staff’ to admit Trump to ‘Walter Reed’

    Is Conway being unfair to Trump. Conway was alarmed at a few of the recent comments from Trump and Trump’s comment on the dollar was so off the charts that Walter Reed may not even be able to help.
    “Is the U.S. dollar stronger against other currencies than it’s ever been? President Donald Trump said so — at least four times in one day. First, Trump tweeted on Aug. 21, 2019, that the United States has the “Strongest Dollar in History.”
    USA! USA! Oh wait – this hurts net exports. But Trump is celebrating the strongest dollar in history. Then again – he is wrong.
    ‘The most common way to track the overall strength of the dollar is the U.S. Dollar Index, a statistic that has been calculated for decades. It’s based on a weighted average of U.S. exchange rates with six currencies: primarily the Euro, but also the Japanese yen, the British pound, the Canadian dollar, the Swedish krona, and the Swiss franc. According to that index, the dollar was stronger as recently as many of the days between December 2016 to March 2017, and it was stronger still between late 1999 and early 2003. The dollar was at its peak historical strength during portions of the early to mid 1980s.’

    I know, I know – FAKE NEWS! So let’s check with FRED:

    Trade Weighted U.S. Dollar Index: Major Currencies, Goods

    This index is near 92 right now. It was 100 in 1973 when the series started and reached 145 in early 1985.

    Yes – our President is truly bonkers.

  5. Sebastian

    Little known/little interesting fact: Just prior to The Great Recession, the FOMC eased at the September, 2007 meeting, (the first in the series for that cycle) again in the October, 2007 meeting, and then again in the December, 2007 meeting…by which time we were in recession.

    Not a prediction, just an example of FOMC inadequacy when the economy is going to go where it’s going to go.


    1. Barkley Rosser

      Oh, this is hilarious. The article Moses cites refers to this poll in NH that so much is being made of as consisting of “a few hundred” participants, almost certainly well below the 600 he found insufficient in the Iowa poll post first debate that had Warren and Harris ahead of Biden (and even further ahead of Bernie). For Moses that poll was to be denounced for its small sample size, but here this one must be taken seriously. Offhand it looks like the DNC is doing what he previously claimed is what should be done, only count polls with larger sample sizes.

      Which is it, Moses? And why are you so enamored of this homophobe who visits a destroyed city with the man who destroyed it?

      Oh, and of course, this stuff from you is just more worthless off-topic dreck.

      1. pgl

        A sample of 600 isn’t as bad as that Trump tweet over how Google swayed millions of voters to Hillary based on nothing more than some dubious claim about how searches are done and a survey of 21 voters. As Hillary noted – that is half of the number of Trump toadies who have been convicted so far.

        1. Barkley Rosser

          There are maybe five or six people on that list I more or less respect: Acz, Butos, Cebula, Rubin, Salerno, Walker, although in some of there cases, such as Salerno, an uber-Misesian Austrian, I am not too surprised, although most of the Austrians I know do not like Trump. Then there are two I do not respect at all: Laffer and Lott. There are several that are more or less non-economist jokes like Luskin, and then a whole bunch I have never heard of from pretty obscure places.

Comments are closed.