The shape of the Yield Curve under QT

I don’t follow many market related discussions on Twitter, primarily because I am not a “theory” guy and secondly because I have total contempt for most Twatter-based furus, given the undeniable fact that they never get anything right, but talk incessantly about their crap theories about how the world will end by 4pm today, or why the sun will rise in the West tomorrow.  At best they are broken clocks and get something right completely by accident once a decade and then scream “I told you so” for the next 9 years. Which makes them intellectually dishonest and just garbage people.

Joseph Wang (@FedGuy12) is NOT one of those.  From the few interaction I have had with him, he seems a very modest, knowledgeable, and thoughtful person who thinks deeply and honestly about the markets and draws suitable, logical conclusions.  His knowledge of the plumbing of the bond and repo markets is undeniably excellent and correct.  Although I follow him, I had missed this piece he published on Jan 19th on his blog https://fedguy.com/the-qt-timebomb/, which everyone should read, and my attention was drawn to it today by @INArteCarloDoss, who, like me (I guess) is more of a practical market person than a theorist.

I thought it useful for everyone if I published my commentary on this piece.  I will not comment of the “plumbing” or the transmission mechanisms he describes in his piece, apart to say that is all spot on and excellently explained.

What I want to argue with is the eventual market response and my reasoning.  I want to make it clear from the onset: if I thought this piece had no merit, I would simply toss it in the mental bin, as I have with so many others, and not comment.  It is precisely because I think it has much merit that I want to argue only with the market response. In no way is this intended to be a hit piece. It is two market professionals arguing about the outcome. What markets are all about. It is always useful for everyone (me included!) to look at things from a different perspective and engage in argument.

So, here goes:

1. QE steepens the curve. QT should flatten the curve. Evidence in chart of 30yr yields below. The 2 biggest periods of QE introduction drove long yields UP, not down (as indicated by blue arrows).  That is because Risk On (or Off) flows are much larger than funding shifts or issuance pressure and completely swamp them.  And also because the market anticipates the success of QE on the economy and therefore normalization of future conditions, which includes a normal, positively sloped yield curve. In every episode of QE the market reacted well BEFORE it started and then REVERSED that flow, as participants switch from Risk Off to Risk On. With QT the opposite should happen: if long yields are to rise, they will rise NOW and reverse LATER, as soon as QT is announced.

2. If the above is correct, then consider the following: banks are basically machines that transform short duration into longer duration and capture that spread. They borrow short (overnight, 1 month, 3 months, combinations of these, whatever) and lend long to their customers (5yr corporate loans, 30yr mortgages, etc.). Most of the really long stuff, like mortgages, are packaged and sold, so on their books they retain mostly an average of 5yr corporate loans, for argument’s sake, and capture the funding spread to them.  You can see from the below chart that there is an excellent correlation between the speed of steepening of the spread between 3-month money and 5yr notes (blue line) and the value of XLF. As the blue line steepens (or rather, the acceleration of the steepness increases), so XLF goes up, anticipating better future profits as the funding spread increases.

As the blue line falls, which I argue it will (and dramatically so, when the Fed first hikes rates and then does QT), XLF will fall in outright terms and also in relation to SPX. You can see how XLF outperformed SPX during QE (blue arrows) and now logic would dictate the opposite should happen during QT.  I know that this is also a correlation with expected defaults in a recession (among other factors like impairments) and that QE helps alleviate fears of recession and that therefore it helps XLF out of a hole and makes it rise relative to SPX (ceteris paribus does not exist in real life, unfortunately), but I would argue that a negative yield curve will be enough to smack both Risk On and XLF relative to it.

Ergo, is the trade to go short $UB_F when Fed does QT?  I would argue that if long bonds are to go up in yield, it will be well before QT is announced.  And the moment it is, the trade is LONG $UB_F and short XLF.  In any case, I am a buyer of a dip down to 127-116 area in TLT and a seller of XLF (on spread, wherever it is when TLT hits those values), whenever that might be. It will probably take till the summer.

My conclusion is that rate hikes and QT will reflect themselves in future bank profitability much more so that in long bond yields.

P/L, as always, will be the judge. It will be interesting to follow this in the months ahead.

Currently: XLF: 40.42, TLT: 138.60

I could easily be wrong and @FedGuy12 right.  Would not be the first time and certainly not the last.  In which case: well done, my congratulations in advance. But basically this whole argument boils down to the shape of the curve under QT conditions.  He thinks steep, I think flat to negative. Rien ne va plus. The roulette wheel starts spinning sometime this spring/summer.