Carpe Diem

Financial stress in US financial markets remains low by three statistical measures, and has been trending downward

The charts above show three different statistical measures of financial stress that were released this week by the Federal Reserve district banks of Chicago (weekly), St. Louis (weekly) and Kansas City (monthly).  None of these three financial stress indexes are suggesting any conditions in the financial markets that would indicate that the U.S. economy is in a recession already, and none of them are exhibiting any upward trends that would point to a pending recession. In fact, all three stress measures have been trending downward through 2012, indicating that there is less financial stress now than at the beginning of the year. Here are descriptions and more details for each stress index:

1. The St. Louis Fed Financial Stress Index (STLFSI) is calculated using the principal components procedure based on 18 weekly data series that include seven interest rates, six yield spreads, and five other financial variables.  For the last nine weeks, the STLFSI has been below zero (lower values indicate less stress, see blue line in top chart), and has generally been trending downward for the last year. The current measure of financial stress for the week ending last Friday at -.194 is below the historical average of zero.

2. The Chicago Fed National Financial Conditions Index (NFCI) is based on 100 financial indicators consisting of 47 weekly, 29 monthly, and 24 quarterly variables, and has proven to be a highly predictive and robust indicator of financial stress at leading horizons of up to one year. Empirical analysis indicates that the NFCI is 95 percent accurate in identifying historical crises contemporaneously.  Increasing risk, tighter credit conditions and declining leverage are consistent with tightening financial conditions are associated with positive values for the NFCI, while negative values indicate the opposite.  The NFCI has been negative for the last three years, and has been trending downward for the last year (see red line in top chart).  At -0.71 for the first week of November, the NFCI is well below its historic average of -0.40.

3. The Kansas City Fed Financial Stress Index (KCFSI) is a monthly composite index of 11 financial variables (yield spreads and asset prices) and the KCFSI is calculated using the principal components analysis.  According to the today’s press release, the Kansas City Financial Stress Index (KCFSI) for the month of October continues to indicate that financial stress in the U.S. financial system remains low. The October KCFSI was -0.40 in October, a slight increase from September’s index, but below its long-run average of zero.  For the last nine months starting in February, the KCFSI has been negative (lower values indicate less financial stress), and the general trend over the last year has been downward.

Bottom Line: Based on statistical measures of financial market stress from three district Federal Reserve banks, financial stress in the U.S. financial system remains low through the end of October (for the monthly KCFSI) and into the first week of November (for the weekly indexes).  Past recessionary periods have been associated with rising index values for all three financial stress indicators, and we’re now seeing the opposite – all three indexes have been trending downward for the last year. Taken together, these three stress measures provide no evidence that the U.S. economy is in recession now, and no evidence that a recession is pending.

7 thoughts on “Financial stress in US financial markets remains low by three statistical measures, and has been trending downward

  1. As the housing prices declined through 2006, so did the stress indexes. The first indications of trouble in the subprime market appeared in February 2007 (I believe that’s the little blip in the first part of 2007 on those charts). The market then continued on its merry way right into the summer of 2007. Strangely, I remember the first day I noticed craziness in fixed income was 26 June 2007 when high yield debt sold off, then the volatility spiked in both equity and fixed income markets.

    The graphs show stress as it became apparent to all of us. I don’t see where they predicted in a previous period what a tick up in stress in the next period. Unlike the VIX, which uses implied vol (the expected standard deviation of price in the underlying), these statistical measures use historical data, so they aren’t looking forward. I’m just unclear what the takeaway is here. Markets haven’t seized up and credit spreads are tightening. This is pretty well known. These models would be more useful if they were predictive, but even the one meant to be predictive doesn’t seem to be predicting anything.

  2. a question:

    at a time when interest rates and rates spreads, which are the key components of these indexes , are being deliberately manipulated by literally unprecedented central bank manipulation, why would we expect any sort of meaningful result from these “stress indexes”?

    that seems like cargo cult thinking.

    it seems akin to claiming at a patient is health because his fever has gone down and ignoring that you have him in an ice bath.

    the issue that such indexes ignore is that leverage at banks does not look good.

    if you own us treasuries at 10 or 15 to 1 gearing, the end of twist and zirp will lead to massive capital hits.

    consider what happens to the value of a bond portfolio at 10:1 leverage if rates go back to even historically low values like 2%.

    that’s a crisis in the making and one that these sorts of indexes will completely miss.

    • Morganovich,

      I don’t think you have to worry about ZIRP going away any time soon. It’s the only thing keeping TBTF crony banks profitable.

      Plus, there’s lots of government debt to inflate away. Of all the options available to the government, inflation is the easiest to implement. All they have to do is pray that they lose control of it.

      • methinks-

        it will be interesting to see what helicopter ben does at year end when twist expires.

        that will drop purchases from about $85bn a month to about $45bn.

        will he increase the size of qe3? extend twist? actually let purchases drop?

        going to be a key issue.

        what are you hearing on his likely replacement in 2014? (ben seems to have enough sense to get out of dodge before this totally blows up. he likely learned that trick from greenspan.)

        yellen?

        • Morganovich,

          Oh, I think the Fed will be torturing the curve for the foreseeable future. I’m pretty sure they’ll find an excuse to extend the twist or some version of it.

          what are you hearing on his likely replacement in 2014?

          Gideon Gono.

          What other living banker has the steady hand and experience to get this job done?

    • that’s a very interesting chart. the divergence of this very fundamental and tangible index with the rate spread based ones is extreme.

      i’m more inclined to trust an index like this at the moment given how manipulated rates are.

      it does not look like it generally gives much warning in advance, but current levels certainly seem worrying.

      this gels with what we just saw from earnings.

      for Q3 we saw revenues for the S+P 500 drop 0.9 % yoy in nominal terms (which is close to a 3% drop in real terms using cpi) earnings dropped by 0.6% nominally as well.

      those are numbers consistent with recession (though they do not prove it). q4 guidance got slashed and a great many companies spoke of strong negative linearity in the q with September much weaker than the prior months.

      q3 GDP was reported as higher that q2, but that was all an increase of government spend. gdp-d also looked quite low again, even compared to cpi.

      use cpi and private gdp was stall speed.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>