Good morning. It is Fed day. We expect a message from the Fed very similar to what we heard in Jackson Hole last month: the tightening will continue until the data improves, neutral rates and market weakness be damned. Email us: [email protected] and [email protected]
Long bonds in a tightening cycle
The short end of the yield curve looks attractive, as Unhedged has mentioned several times recently. Here is the Treasury curve, as of yesterday (light blue) and a year ago (dark blue):
A year ago, the short end was about a teensy bit of yield being better than no yield at all. Now, short yields are higher than long ones and, in a very volatile rates environment, don’t bring much rate risk with them. Owning some seems like a no-brainer.
But spare a thought for the long end, too. Yes, there is one scenario when 10-year bonds (sovereign or corporate) will be big losers. If the labour market stays tight, wages keep rising and the economy stays hot, the Fed will push the policy rate well past the market’s current expectation of 4.5 per cent or so. This will drag the long end higher; the yield curve can only invert so far. We might get a little taste of this today, if the Fed’s message is particularly tough.
Believers in this scenario include BlackRock’s Gargi Chaudhuri, Americas strategist at iShares. He argues that the Fed’s balance sheet runoff policy is a key consideration:
We believe we will see more attractive levels to enter longer-duration positions in the next few months. We also believe that longer rates may rise given that the Fed’s quantitative tightening (QT) is just getting started. In September, the Fed doubled the pace of QT to $95bn/month, meaning that over the next year, $1tn in Treasuries and MBS will have to find a new home. This supply dynamic should support somewhat higher long rates. In addition, we believe the volatility premium associated with QT has yet to be fully priced into longer-dated maturities.
This view makes sense, if you take the intuitive position that the when the Fed is not buying bonds all the time, the price of bonds will fall. Unhedged does not happen to agree: our best guess is that by draining liquidity from the financial system, QT will depress risk appetites, increasing demand for long Treasuries.
However you think QT works (see below), there are two other scenarios where long bonds might work pretty well. Optimistically, if the Fed sees inflation starting to cool, and remembers that policy works with (all together now!) “long and variable lags”, it might pivot towards smaller hikes, with an eye to ending rate increases altogether. The market believed this was happening in June and July, and the 10-year rallied nicely. The market was wrong then, but the pivot may happen yet.
There are pivot believers out there, despite the horrific August CPI report. One such is Robin Brooks, chief economist at the Institute of International Finance. He thinks that financial conditions have tightened fast, and the effects are starting to show. We are in a housing recession, thanks to a huge move in real mortgage rates. The strong dollar is already pinching US and global manufacturing. Global economic activity indicators are decelerating fast. While the August report showed prices that had been rising fast rose even faster, there was a ray of hope. The number of price categories that are seeing high inflation is falling. Here is his chart:
Given all this, Brooks sees the door to a Fed pivot as wide open:
We think we get a 75 [basis point] increase this week, and a data-dependent press conference that will keep the option open to slow hikes if the data turns benign — which we think they will . . . if the chair signals a more mixed outlook, we will get a rally in long bonds, as the policy outlook shifts and the inflation risk premium declines.
The final scenario is the most straightforward one: the Fed overtightens badly and drives the US right into a proper recession. If that suddenly looks like the most likely outcome, 3.5 per cent on a 10-year Treasury is going to look pretty good, because the Fed will be in loosening mode soon enough. Monica Erikson, who manages investment grade credit at DoubleLine, told Unhedged last week that her firm’s view is that there is “a pretty high probability” of a Fed overshoot. Given that, and the beating that long corporates have taken recently, driving their prices into the 70s and 80s, she is adding duration to her portfolios — even though credit spreads are not particularly wide.
Unhedged agrees that the overtightening and recession scenario is the most likely, for the dumb reason that this is what has most often happened at the end of tightening cycles in the past. Long bonds are looking interesting.
A note on QT and non-linearity
While Rob was on vacation, I wrote a news story on quantitative tightening and the Treasury market with our markets colleague Kate Duguid. Here’s a summary (here is the whole thing):
The Treasury market is very important. If it doesn’t work, global finance breaks.
Trading conditions in the Treasury market are a mess. Liquidity is lousy and double-digit swings in (basis point) yields are commonplace.
The Treasury market is structurally unsound. It relies on private market makers, called primary dealers, to make the wheels turn, but those primary dealers are struggling, thanks in part to bigger deficits (ie, more Treasuries) and post-crisis regulation. This Bank of America chart shows their diminishing market-making power nicely:
The biggest buyer of Treasuries, the Fed, has pulled back. It is up to private actors to absorb the bonds that the Fed is no longer sucking up. Maybe they can do this, maybe they can’t. If they can’t, things will get bad.
Unhedged has spent a barrel of ink (including in the above segment) on QT’s many possible channels. The most standard idea is that QT pushes up long yields by altering the supply-demand balance for Treasuries. We think liquidity effects will predominate, but we are not enormously confident.
After speaking to a bunch of analysts for our story, I’ve found another lens on QT helpful. Try separating its impacts into two buckets: linear and non-linear. The distinction is between effects that mount steadily, and those that play out slowly and then suddenly. When Fed researchers ask, “How many rate hikes does quantitative tightening equal?” they are asking about linear effects, from liquidity or supply-demand or some other channel. When analysts worry QT raises the risk of abruptly breaking the Treasury market, that is a non-linear effect.
I put this framing to Joseph Wang, the Fed Guy, who said he has started worrying more about such “non-linear tail risks” since the repo market blow-up of 2019:
Repo is a stable, sleepy market, and much more liquid than cash Treasuries . . . If [rates in] the repo market can suddenly quadruple, then [rates in] cash Treasuries can as well.
Markets for safe assets, we should remember, can be fragile, with big consequences for every other financial market. Something to watch as QT chugs on. (Ethan Wu)
One good read
New York nightlife’s new normal looks like a whole lot of places closing early.
Read the full article here