Can inflation miraculously melt?

January 22, 2023 – Weekly comment

Just a few thoughts here that don’t necessarily lead to any strong conclusions.

Liz Ann Sonders tweeted this on Friday: Used Car Value index from Manheim…”has been less bad MTD…but -13.7% yoy drop is still one of the worst readings on record.

Fed’s Waller on Friday: “The market has a very optimistic view that inflation is just going to melt away.  We have a different view, that inflation is not going to just miraculously melt away.  It’s going to be a slower, harder slog to get inflation down, and therefore we have to keep rates higher for longer and not start cutting rates by the end of the year.”

It’s obvious from the Manheim chart that car prices were a huge contributor to inflation.  I’m not saying that Waller is wrong about the decline of inflation being slower, but the drop in used car prices does appear to be a miraculous meltdown.  Could the Fed be just as wrong about the speed of inflation’s decline as they were about ‘transitory’?

Of course, used car price data naturally leads to questions about rising delinquencies.  A quick search found that Auto Loans delinquent by 90+ days for Q3 2022 is 3.89% and rising.  While I was on this rabbit-hole search I saw a BBG note from November that the delinquency rate for subprime auto loan bonds was 5.13% as of October, based on borrowers late 61 days or more, as compared with 3.76% from the previous year.  I also found this delinquency rate on credit card loans from smaller banks (and note that smaller banks are the ones increasingly tapping the Discount Window, according to the NY Fed [link at bottom]). 

At 6.88% for Q3 2022, it’s the highest delinquency since 1991, outside of the covid spike in Q1 2020.  It way exceeds the GFC.  When loans go bad, the price of collateral falls. We’ve seen that recently with crypto.  What’s the collateral on credit card loans?  Certainly not all the consumer crap that was purchased; it’s credit scores and borrower income.  The fraying starts at the edges.  Of course, we know all that.  We know that calls for recession have been non-stop ever since the definition had to change after the two consecutive quarters of GDP contraction in the first half of last year.  The only question is, of course, how fast and how deep?

One other quick haphazard research item: I had seen a note that the decline in the TGA (Treasury Gen’l Acct) at the Fed was partially offsetting QT. In May 2022 TGA was $945B and last week was $339B.  It’s not a linear decline, call it $80b per month.  That dynamic doesn’t seem likely to continue, and will probably hasten the end of QT.  But here’s another random data point: US currency in circulation was $1.799T in December 2019.  As of the end of last year it’s $2.297T.  An increase of nearly $500 billion in three years.  Where’s the demand for that? (Besides suitcases for Hunter).  I never see anyone paying in cash besides me. Is it for those sad ubiquitous tip jars that usually have 1 or 2 crumpled singles and a pile of nickels and pennies?  Or is the demand for $100 bills related to a store of value in Russia and elsewhere; a unit of anonymous transaction for those that don’t wish to be tracked?  I’d bet, well, I’d bet $100 that there are a LOT of $100 bills circulating around in Davos.  The treasury wants to track transactions of $600 in banking accounts.  Just seems odd that US ccy in circulation has increased by 27% in three years.  By the way, that three year increase of $500 billion is more than the market cap of bitcoin ($440b). Another fun coincidence, the dollar price of gold is up a similar 32% since the end of 2019. 

There are many technical aspects of monetary policy affecting markets which are way beyond my grasp.  For example, the circus related to the debt ceiling.  Here’s what I DO know.  The current limit of $31.38 trillion is gargantuan.  According to the Fed’s Z.1 report, Fedl debt outstanding at the end of 2019 was $19.04T.  In the GFC, private debts were shifted to the government.  By comparison the latest shift was much larger. How does the market accept a 3.5% yield for US ten-year treasuries?   


As I mentioned last week, a sale of FFG3 is an easy hedge for those concerned about the possibility of a 50 bp hike on Feb 1.  Fed Effective (EFFR) has been 433 since the last hike.  Feb has 28 days.  One day at 433 and twenty-seven at 458 would lead to a settlement price of 457.1 or 9542.9.  FFG3 settled 9542 on Friday.

So the market is expecting 25 in Feb.  The next meeting is March 22; there is no meeting in April.  FFJ3 settled 9522 or 4.78%, or 20 bps higher than what will likely be the new EFFR of 458.  So we’re looking at high odds of 25 at the March meeting. 

SOFR Rate is around 3 bps lower than EFFR. FF futures settle to the daily average of the EFFR while SOFR futures settle to the daily compounded rate of SOFR, so after compounding, the rates are currently very close. 

This past week there was huge buying of call condors and butterflies that are a play for one or two more 25 bp hikes and then a slight bias toward ease.  These plays have been well covered on twitter and elsewhere so I will just note strikes and settlements from Friday. 
SFRH3 9512.5/9518.75/9525/9537.5 broken call condor settled 2.75 ref 9516; max value 6.25 between 9518.75 and 9525.
SFRM3 9500/9512.5/9525/9537.5 c condor settled 3.75 ref 9511; max value of 12.5 between 9512.5 and 9525.
SFRU3 9500/9550/9600 c fly settled 14.25 ref 9528.5; max value of 50 is if SFRU3 expires exactly at 9550.

SFRU3/SFRU4 is the most inverted one-year calendar on the strip.  It settled -169.5 but traded as low as -181 on Thursday.  Over this period, the market is forcefully expressing the need for easing.

UST 2Y423.4418.0-5.4 wi 412.0
UST 5Y361.1356.5-4.6 wi 354.2
UST 10Y350.9348.2-2.7
UST 30Y361.9365.33.4
GERM 2Y259.2257.7-1.5
GERM 10Y216.8217.70.9
JPN 30Y161.0152.3-8.7
CHINA 10Y289.8292.83.0
SOFR H3/H4-97.5-104.0-6.5
SOFR H4/H5-101.5-100.01.5
SOFR H5/H6-8.5-3.55.0
CRUDE (CLH3)80.1181.641.53
Posted on January 22, 2023 at 7:37 pm by alexmanzara · Permalink
In: Eurodollar Options

Leave a Reply