Freight Shipping Growth and Recession Probability

Prob(recessiont) = -1.925 – 16.10 freightgrowtht + ut

McFadden R2 = 0.51 NObs = 235 (2000M01-2019M07). Coefficients significant at 5% msl bold.

Figure 1: Year-on-year growth in Cass Freight Index, in % (blue, left scale), recession probability from probit regression, in % (red, right scale). NBER recession dates shaded gray. Source: Cass Information Systems, Inc. via FRED, NBER and author’s calculations.

As Cass Information Systems, Inc. note themselves in their July report:

With the -5.9% drop in July, following the -5.3% drop in June, and the -6.0%
drop in May, we repeat our message from last two months: the shipments
index has gone from “warning of a potential slowdown” to “signaling an
economic contraction.”

 

20 thoughts on “Freight Shipping Growth and Recession Probability

  1. Moses Herzog

    Thanks Menzie, you know this is one of my favorite stats. And some day when I get rid of my lazy bones I may see if I can replicate your probit regression numbers–but as per usual that’s probably months in the making.

    Hey peeps, movie recommendation I just watched tonight: “High Life” directed by Claire Denis and has one of my favorite actresses in it—Agata Buzek. It’s kind of an adult science fiction film. Give it 9 stars on a 1-to10 scale. get it on disc or streaming or whatever, great film. The flavor of the film is more like a “Timecrimes” or a “Snowpiercer” than say a “Star Wars” if that gives you some idea what to expect. More of an adult sci-fi. than something aimed to do the “SJW” checklist.

    “Heist” is also a good film with Robert De Niro —but more B-movie fair. I give it strong 7, mostly because it has 3-4 of my favorite actors in it. But if B-movies leave you feeling cheated then you can probably skip it.

    Reply
  2. Moses Herzog

    Uh-oh….. looks like the girl who two commenters here said was going to “win over the immigrant vote” (currently polling 7% in Texas, 7% in Florida) stepped in it again. Oh no…… https://www.nytimes.com/2019/09/08/us/kamala-harris-trump-apologizes.html

    The little girl crying on the school bus on her way to the white school has “stepped in it” again. Don’t you hate it when the self-proclaimed “SJW” cannot meet the “SJW” standard?? I think I figured out what is causing all the confusion. When Copmala talks about her “homies in the hood” listening to rapper 2-Pac “back in 1982” she just means some white dudes she used to smoke weed with in the San Fran DA’s office, listening to Vanilla Ice as they giggled about all the blacks and hispanics they had put in prison on Marijuana charges. What do you guys think, did Copmala get 2-Pac mixed up with Vanilla Ice??
    https://youtu.be/rog8ou-ZepE?t=11

    Reply
  3. New Deal democrat

    I believe the Cass Freight Index includes international shipping and flights, and so is a better marker of global rather than US economic growth. Domestic trucking is 75% of domestic freight, and last month the ATA’s Trucking Index set a new record. The DOT’s Transportation Index, which includes other forms of domestic shipping, as of June was trending sideways but not down.

    Reply
      1. New Deal democrat

        Yardeni is a great resource for data graphs. I have made use of figure #3, comparing rail and trucking, several times this year. Notice that trucking is positive, while rail carloads have declined significantly. Since trucking carries about 3x as much as rail, the net is a sideways movement, as evidenced by the DOT’s Transportation Index.

        Part of this may be that trucking is more competitive and less oligopolistic compared with rail, and some is probably container ships bypassing west coast ports (which have had declining traffic all year; east coast port info is tough to come by, but the stories I’ve read strongly suggest their traffic has picked up) due to the widening of the Panama Canal, and offloading directly to trucks in the east or along the gulf coast.

        So I think the Cass Index is really showing that imports and exports are getting hit by the trade wars.

        Reply
        1. spencer

          Maybe the opening of a wider Panama Canal is causing a lot of imports from Asia being routed directly to East Coast ports.

          This should cause rail shipments from western ports to fall and local short haul trucking from eastern ports to rise.

          Could this structural change be changing the freight indices?

          Reply
        2. ooe

          Yardeni claimed in the Summer of 2007 that the economy would not go into recession so you need to ignore him. Also, it amazing that people are trying to rationalize bad statistics to prove there will no recession. The simple truth the economy is heading for a recession per the bond market. the ISM purchasing managers index indicated a contraction in manufacturing. Also, the Trump is acting more deranged so he knows the truth.

          Reply
  4. AS

    Professor Chinn,
    I finally got my output to agree with yours using log approximation to the actual twelve month percent change. Using the actual percentage change, the probability of recession now is about 18% and the log probability is about 17%, so both very close.

    I used FRED data series USRECM as the dependent variable. Did you use USRECM?

    I am not sure, however, about the meaning of “MSL”. Is this short for “mean significance level”? I searched my old statistics texts and did not find MSL, but of course did find definitions of “significance level”.

    Interesting, I found a citation for “median significance level” for small samples.
    Joiner, Brian L. “The Median Significance Level and Other Small Sample Measures of Test Efficacy.” Journal of the American Statistical Association, vol. 64, no. 327, 1969, pp. 971–985. JSTOR, http://www.jstor.org/stable/2283479.

    Reply
  5. 2slugbaits

    Menzie Just to be clear, this is not predicting the probability of a recession in 12 months, but is telling us the likelihood that we are currently in a recession, right?

    Reply
      1. Willie

        Thank you. That is a detail that was not clear to me right off. Looking at the plot of the data from 2014 to 2016 makes it clear that some parts of the country may have felt like it was a recession then. Some parts of the country may be more depressed now than they were last year, whether we are in a recession or not. Calculating localized economic performance would probably require more data and more time than is available. It would make for some interesting analysis if it could be done.

        Reply
  6. AS

    2Slugs or other experts, any thoughts on the following model? Maybe the trucker will begin to show-up in the following ratio of icsa/unemploy. Icsa represents new claims for uninsurance benefits and unemploy is the level of unemployment.

    I have not seen a probit model done with a trend.

    Using FRED series data from 1975m07 to 2019m8
    1.new claims for unemployment insurance (icsa)
    2.level of unemployment (unemploy)
    3. and also using a trend factor to account for the declining trend in icsa/unemploy,
    results are shown below.

    Probability of recession within one month 2.4%
    Prob(usrecm(t+1)) = -11.203 + 1.653* icsa/unemploy + 0.00529*@trend u t+1
    McFadden R2 = 0.35
    NObs= 529. Coefficients significant better than 1%
    While the current probability is low, what seems interesting is that once the ratio of icsa/unemploy starts to spike, a recession follows even though some spikes are followed by what seems to be an improvement in the ratio.

    Reply
    1. Moses Herzog

      I don’t know how this plays out economically, just “thinking out loud” here. “Intermodal” might be an interesting number to track. As I understand it, it’s contracted to the trucking companies, and then they turn right around and ship it by rail. Rail is always cheaper than truck (with the possible exception of short distances), that’s why you see tons of “BJ Hunt” trailers (little yellow logo on a white trailer) and Schneider trailers (solid orange trailers, Menzie probably sees them all the time when he’s on the interstate, in trucker slang called “pumpkins”) on trains, often you’ll even see the trailers doubled up on top of each other. You’ll even see “Werner” trailers from time to time, (like a solid powder blue). Others of course. But BJ Hunt and Schneider are the biggies. Schneider used to brag they were the largest single U.S. employer of ex-military guys. I don’t know if that’s true or not but they certainly were floating that rumor around the 2000 year timeframe.

      Reply
      1. Moses Herzog

        CRST is another huge one, but honestly I don’t remember seeing much of their stuff done by rail, but I’m certain they must do some “intermodal” as it’s just cheaper. When CRST first started (in the mid-1950s) the owners noticed livestock would be shipped from Iowa to Chicago and then the trailers would come back to Iowa empty. They figured out the simple idea (and “market need”) that if those empty trailers on the run back to Cedar Rapids were loaded with cheaply shipped steel they’d have themselves an efficient operation. And CRST took off like rocket boosters.

        Reply
      2. Moses Herzog

        Oops, JB Hunt. Don’t ask me how I muddled that one up. There’s enough vulgar jokes in the trucking industry, I don’t dare guess how that name would go with the typical semi-driver. Last of this summer heat must be getting to me.

        Reply
  7. Moses Herzog

    This one’s for all the HK “mofos” (and I mean “m****” [edited by MDC] in the most affectionate way possible), and if Menzie doesn’t let this one past the filter for the girl who got her eye taken out by HK police (read Beijing officials) , he’ll feel guilt forever and his ancestors will take revenge.
    https://www.youtube.com/watch?v=eilzQSthLCY

    Reply

Leave a Reply

Your email address will not be published.