This is a bit like asking Dracula why there’s a blood shortage but here’s Goldman with its best effort on why bond yields are plunging:
Q: The reflation theme in markets seems to be unwinding, led by bonds. What is recent price action signaling about the recovery?
A: Since the recent Mid-may highs, both 10y and 30y US Treasury yields have eclined by about 40-50bp. Over the same time, 5y yields have declined by a more modest 10-15bp. Both the 2s10s and 5s30s curve are 45bp and 35bp flatter respectively, and below levels seen before the February surge. Given that the reflation trade has been associated with higher yields and steeper curves, the bond market’s reversal appears to be signaling some concern along this front. We note that this isn’t purely a bond market story—underneath headline index levels, there appears to have been a significant equity rotation out of “value” and small caps, both categories expected to outperform in a reflationary environment. Commodities also appear to have put in a local peak . More recently, our macro PCA factor attribution model suggests that the price movements across asset classes have been associated with shifts in growth assessment, and with only a marginal contribution from a hawkish Fed pivot (Exhibit 1). However, until this week, the repricing was somewhat unusual in that front-end hike pricing was somewhat sticky even as longer maturity yields witnessed a sharp decline. This suggested that markets were pricing a weaker recovery in a few years (perhaps once the fiscal stimulus faded), but a near term inflation path that was worrisome enough to cause the Fed to tighten, at least for a while. That particular scenario, while a possibility, does not appear an appropriate central case to us, at least based on current data, our economists’ forecasts for future years, and the Fed’s apparent willingness in looking through near term price pressures. We would note that price action this week is onverging to a more traditional weak growth narrative, with some front-end hike pricing being reversed.
Q: Concerns about the COVID delta variant appear to be weighing on markets. Is the magnitude of repricing reasonable given the facts? Are there other factors driving the growth worries?
A: While the variant has led to a surge in new cases in many regions, hospitalizations and fatalities remain relatively low in areas with high vaccination rates. As a result, we think the economic impact for the US, Europe (and China) will be limited, although some parts of the world could see a greater hit to growth. Indeed, our US economists find that high frequency data suggest only a modest impact, and globally we see only about a 0.3pp hit to global growth this year, and none next year.
Q: The flattening of the yield curve accelerated following the June FOMC meeting. How much of recent yield curve behavior can be explained by a “hawkish” Fed?
A: The trouble with the curve isn’t that it is flattening on strong data—indeed, our post-June FOMC analysis suggested as much. After all, bear flattening is typical behavior when a central bank is expected to tighten policy. However, typical bear flattening of the curve isn’t accompanied by a decline in medium and longer maturities yields; rather, the flattening occurs because front-end yields sell off by more than those maturities. One explanation for the difference is the evolving interpretation of AIT. Prior to the June FOMC, investors likely assumed a higher inflation threshold for Fed tightening, resulting in pricing of a later-but-steeper path for hikes. It is then natural to reason that a lower inflation threshold—and by extension earlier hikes—could mean theFed wouldn’t have to hike quite as much, and therefore the terminal rate ought to be revised lower. However, this line of reasoning ignores a few key facts. While the dot plot was hawkish relative to expectations going into the meeting, once the accompanying labor market and inflation projections are taken into account, the median fed funds path is still clearly dovish when compared to past normalization cycles, including the last one. Taken in the context of the economic projections, information in the dot plot isn’t a reason to price in a lower terminal rate, in our view. Indeed, current pricing is substantially below the Fed’sprojection for “long run” rates, suggesting that markets either have a different view of this rate, are assuming a truncated hiking cycle, or are placing relatively high odds ofdownside risks to the economy in the future.
Q: A return to the “secular stagnation” narrative could be a reason for the drop in the terminal policy rate implied by the yield curve. Are current levels reasonable, and how much lower can they go?
A: Using current economic data (and forecasts), and reversing our curve models suggests that a very low natural rate (r*) assumption would be needed to justify the shape of the curve. Given that it is unobserved, and acknowledging the problems with properly identifying such a rate, we use the 5y5y nominal rate as a crude proxy of “nominal r*”, and as observed earlier in Exhibit 2, noted that current levels are close to the lows from the last cycle. In a recent report, our economists argue that the average short term real rate in the last cycle was unusually low when compared to prior cycles, and might not be an appropriate benchmark for the current recovery. Indeed, several drivers that may have kept these levels low the last cycle, such as much lower levels of fiscal (and at least initially, monetary) support, look quite different this time around.
Q: Could a shortage of safe assets be causing the decline in intermediate yields? Are excess savings being recycled into US Treasuries?
A: Our debt supply projections do indeed show that central bank purchases will keepnet supply negative for most European sovereigns, and somewhat flat in Japan. This leaves the US by far the largest supplier of safe assets to the world, though EU net supply (to fund SURE and NGEU disbursements) remains substantially positive, and is likely to be the biggest source of EUR duration. Still, these supply dynamics were known prior to the recent decline in yields, i.e., there were no surprises on this front in the last two months, and as we noted in a recent report, markets tend to price any known supply in advance of the actual issuance.
Goldman sees US 10 year yields roaring higher into year-end, up to 1.9%.
Meh. This is all much ado about nothing in my book. In the everything bubble, asset prices are unusually sensitive to the rate of change in credit and liquidity. Both monetary and fiscal impulses are falling fast:
For mine, falling yields are therefore easy to explain then:
- China slammed the brakes on credit all of this year. Growth is going to slow more not less ahead, following the Chinese credit impulse lower.
- In turn, this is going to end the global inventory supercycle and prices start to fall sharply. Next to the chopping block is commodity prices.
- The Fed has thus turned (only modestly) hawkish as precisely the wrong moment, putting a bid under DXY which exacerbates all of the above.
Bond yields are signaling that ahead is a reflationary bust. That’s fine with central banks for now. They called it, after all.
The question now is when do the Fed and PBoC panic into the growth scare that comes with all of the above and switch markets back to reflation…