Tuesday, January 11, 2022

Bonds: Yields on the Long End Will Come Down

 As noted in the WSJ over the weekend, the US 10yr Treasury yield finished last week at its highest yield since January 2020. This was apparently due to a more hawkish than expected tone from the latest FOMC meeting minutes, with three interest rate increases expected and an earlier-than expected wind-down of the Fed's asset purchases.

The framing in the media that the Fed will start "tapering" its asset purchases is misguided, because the Fed has already been de facto tapering since March 2021, when it launched its Reverse Repo program (RRP), which has eclipsed $1.5 trillion. The Fed holds assets such as Treasuries and agency MBS on its balance sheet, and has been lending those securities out as part of the RRP. At $1.5 trillion, that is over a year's worth of asset purchases, assuming $120 billion/mo. So it's hard to see an asset purchase "taper" having any meaningful impact, given the securities purchased are offset by them being lent out (technically a sale with promise to repurchase at a later date) by 12x over.

What's interesting is, contrary to popular beliefs, the Fed's QE programs have historically led to an increase in the yield on the 10yr (or perhaps more accurately, a steepening in the curve). It's unclear exactly why this is, because "easing" implies lower rates, and people also blame high financial asset valuations with the Fed's supposed rate "suppression." How can it be then that QE causes rates (at least the long end) to rise?

The most plausible explanation in my view is what I'll call the Collateral Scarcity Thesis (CST). CST proposes that when the Fed does QE, it is effectively removing high quality, good collateral (think Treasuries and agency MBS) from the market by placing it on its balance sheet in exchange for bank reserves. As argued by IMF economist Manmohan Singh, collateral is an important financial lubricant because institutions can reuse it for new credit creation, whereas reserves cannot be used in the same manner. He further argues that the Fed unwinding its balance sheet itself is a form of easing, because it creates capacity for banks to make new loans. Because QE removes collateral, it creates a shortage and therefore tightens credit. Tightening credit is usually associated with higher rates. Looking at historical yields (or in this case, the 2yr10yr spread), QE has been associated with rising yields/steepening curve, which is a sign of credit tightening. See below:

The green dots represent instances when the Fed has announced QE (green means go!); red they slowdown and/or stop.

As we can see, QE causes the long end to rise, and reversing it causes yields to fall. This is in direct contradiction to the narrative that QE "suppresses interest rates;" it clearly raises/steepens.

One thing to note - the red dot for March 2021 does not represent the Fed announcing an unwind of its balance sheet per se. Instead, it represents the date that the Fed announced the RRP. As stated earlier, RRP is a de facto unwind of QE, because the Fed takes in reserves and releases collateral. As would be predicted by CST, the RRP eases credit conditions by creating additional capacity for bank lending. Increasing the supply of collateral should lead to growth in bank credit. And it worked! Year-over-year growth in bank credit fell substantially in March 2021, which coincided with rising bond yields. When the Fed launched RRP at the end of March, the long end came crashing down and bank credit bottomed before starting to grow again:

So we know the taper doesn't matter because the Fed has already been tapering since March. Does this mean bonds are a buy here? That depends. I think yields will be lower six months from now, but that doesn't necessarily mean bonds are a screaming buy. It's possible that yields rise in the short term due to seasonal and/or technical reasons. For example, companies drawing on credit lines to purchase inventory to restock following the holiday season.

If demand for credit is robust, then I wouldn't expect the long bond to perform well. Higher demand for credit = price of credit (i.e. yields) go up, all things being equal. I think chronic underinvestment by the federal government for decades has the US economy operating far below its productive capacity. We haven't been doing enough to support broad-based growth. The best evidence of this can be seen in the weak Labor Force Participation and Capacity Utilization rates: 

In general, when more people are working and generating income, their capacity for borrowing (read: credit demand) increases. We are lapping tough comparative periods because of the massive fiscal stimulus that took place in 2020 and 2021, which means we may have hit peak credit demand. Higher inflation also depletes savings. And the Fed may raise rates, which I expect to flatten the yield curve. Slowing credit demand and diminishing supply as the Fed raises short-term rates are bullish for the long bond.

No comments:

Post a Comment

On Inflation

[Note: I originally started writing this piece on December 23, 2022, then got held up with holiday festivities. More posts for the new year ...