The U.S. economy is not on the verge of a recession

U.S. economic data has deteriorated a little over the past 2 weeks. Recently there have been a lot of media headlines saying “Economic indicator XYZ shows that a recession is dead ahead!”
We predict recessions because they either lead to 20% or 50% drops in the S&P 500 (depending on what our models say). Since we invest in UPRO (3x ETF for the S&P), that’s akin to a 60-95% loss in our portfolio! We do not blindly buy and hold.
You can see that a recession is not “dead ahead” when you take a good look at these indicators.

Industrial production

Zerohedge stated that Industrial Production’s rate of change over the past 30 months is negative. “This has never happened outside of a recession”.

Here’s why this indicator is useless.

  1. Using a rate of change over 30 months is a terrible idea. Economic indicators should be timely and not slow-moving. 30 months – 2.5 years – is half the lifespan of many economic cyles!
  2. The Rate of Change for Industrial Production often turns negative in the middle of a recession. I.e. this indicator lags a lot, so it is not very useful for making predictions about the future of the U.S. economy.
  3. Industrial Production has 3 components: manufacturing, mining, and utilities (i.e. energy). Mining and utilities got killed in 2014-2016 due to low oil prices. Oil prices do not drive the economy in the long term because the U.S. economy does not rely on commodity prices. So if you only look at the manufacturing component, Industrial Production has grown nicely over the past 30 months.

Loan growth is collapsing

U.S. commercial loan growth has collapsed to the point where there is no growth at all. This is important because a collapse in loan growth is ultimately what made the 2008-2009 recession so severe. U.S. commercial loan growth collapsed by 8-9% EVERY MONTH by the time July 2008 came around.
Once again, the media has been quick to point out that “this has never happened outside a recession since 1990”.
Herein is the lie. It is true that such a collapse has always resulted in a recession since 1990. But if you expand your data, you realize that this indicator is not a good predictor of recessions.
Loan growth collapsed in 1985-1986 but did not lead to a recession.
That historical case and the present case have something in common: falling oil prices.
Oil prices crashed from November 1985 – March 1986. Loan growth crashed as well in the next few months, but no recession ensued (although economic growth did slow down).
Oil prices got crushed from August 2014 – January 2016. Keep in mind that there is a time lag between falling oil prices and declining loan growth. Only after prices have fallen for a while will loan growth slow down, which is why we are only now (in 2017) seeing declining loan growth.
Loan growth isn’t just slowing because borrowers want to borrow less. Loan growth is also slowing because lenders want to lend less! With rates still compressed near all time lows, lending money isn’t very profitable for commercial banks. So although these banks aren’t raising rates, they’relending less money with stricter requirements.

European economies will not contract

With the French election this Sunday and a sudden slowdown in Eurozone economic data, people fear that the Eurozone will dip back into a recession. They say that this will happen if France votes for Le Pen, who wants to leave the EU.
The fear is that a EU recession will cause a U.S. recession. Nothing can be farther from the truth.
History shows that the U.S. is mostly immune to foreign recessions because the U.S. is a relatively closed economy. Foreign recessions will AT MOST cause a slowdown in U.S. economic growth. Growth will not turn negative. For example, Japan experienced its first recession in 20 years in 1998. Everyday the media screamed “Japan is going to drag the U.S. into a recession and corporate earnings will fall!” Nothing came of it. In 2015, every media outlet proclaimed “China’s economic downturn will hurt the U.S. economy significantly!” Once again, nothing came of it.
This time isn’t different. An EU recession will cause a small downturn in U.S. economic growth. However, an EU recession is unlikely.
In the unlikely scenario that France decides to leave the EU, it will be a long and orderly process. As a result, European economies will not suddenly contract.

12 comments add yours

  1. Troy,
    Nice blogs! I like your no-nonsense style.
    One question: when you say your model predicts a major recession, how long does the prediction lead the actual fall? What is the false positive rate?

    • It works 100% of the time. I kid you not. It sometimes leads the recession by a few weeks, but normally it’s 3-6 months. The worst case was in 1999 when we lead the recession by 1.5 years! (2nd worst case was 9 months in 2007). However, keep in mind that 1999-2000 was an extraordinary example. No rational investor could have called that top because every valuation indicator was insane. All the major hedge fund guys – George Soros, Jeremy Grantham, Jim Rogers – were calling the top in July 1998 for that bull market.
      In addition to this model, we added a backup indicator for predicting S&P 500 bear markets. We recognize that although the model’s success is 100%, perhaps the next time our model will fail. This backup indicator allows us to recognize that an economic recession just started and a bear market has started ASAP. This indicator also has a 100% success rate.
      The model is really simple. It took a long time to build because simplicity is the ultimate sophistication. Using just a few indicators, we were able to combine a wide range of economic concepts.

      • Excellent, I agree that given that we only have had a few major recessions in the past, overfitting can be a concern. If the backup is a market timing indicator, you have a perfect match.
        When we built the models, we purposely left some slack to avoid the issue of overfitting.

        • The backup is a market timing indicator as well as an economic indicator, because all bear markets with recessions (and we don’t define bear markets as “20% declines”).

  2. I agree the scenario won’t pull the US down into a recession, but I do foresee a slight downturn similar to that of Brexit. The dip just allows investors a few more chances to buy before the market gets a chance to rebound.

    • Great point. Although it’s important to not that many dips in the data don’t coincide with dips in the S&P. For example, the data suddenly diped in March-April 2014 but the S&P was resilient.

  3. Unlike the US, Canada is very susceptible to the economic state of its trading partners – especially the US. So I’m glad you’re not going into a recession!

    • Yea I think something like 22% of Canada’s exports go to the States. Like Trudeau said, “when the U.S. sneezes, Canada catches a cold”

  4. Hi Troy,
    I have to say I have been really impressed with the material you publish here.
    And I can see you appear to surf at my hometown beach of Manly! 😉
    When you say,
    “…Keep in mind that there is a time lag between falling oil prices and declining loan growth. Only after prices have fallen for a while will loan growth slow down… ”
    That’s a connection I haven’t yet grasped. How do you explain the link between loan growth and oil?
    Kind Regards,

    • Oil producers rely on commercial loans to fund new oil & gas exploration. (exploration is capital intensive). when oil/gas prices fall a little, these companies will continue their capital investment because they think that the price decline is transitory. only when prices fall for a while do these companies realize that the price weakness ISN’T transitory. then they cut down on capital investment, which causes loan growth to slow down.

Leave a Comment