Why have US interest rates been falling since April 2021? A fact check.

Various analysts reported that investors had a very negative viewof the US interest rate market at the beginning of Q2 and were accordingly short positioned to bet on or hedge against further falling bond prices. When this expectation did not materialise in Q2, parts of these short positions were closed, which supported bond prices.

If this argument is the main reason for the decline in interest rates since the end of the first quarter of 2021, then the move is primarily technical in nature. Once positioning has returned to a normal level, growth and inflation developments should drive interest rates up again, subject to a continuation of the current trend. In this case, the market movement would only be an interim correction due to extreme positioning by market participants, but it would not change the long-term upward trend in interest rates.

It is well known that extreme positioning can often be a harbinger of violent market moves in the opposite direction. Market sentiment changes can trigger a sharp move due to covering and repositioning active positions. However, this argument does not explain why such a change in expectations has occurred.

Since data on the positioning of market participants is not readily available, the repositioning of market participants can be indirectly measured by the fact that open futures positions are closed and the outstanding volume on the futures market (“open interest”) is reduced accordingly.
The chart below shows that while the volume of outstanding ten-year US Treasury contracts has declined, it has done so with an eight-week lag to interest rates. It looks like positioning has adjusted to the change in market conditions rather than the other way around. While repositioning may certainly have amplified the move, no causal effect on the reversal of the interest rate trend can be read from the data.

Open interest in ten-year treasury futures

How do the positioning of market participants and the movement in the market relate to each other?
Is the connection simultaneous or is there a time difference of a few weeks?
And how does the dynamic change once market participants have shed extreme positioning?

This argument states that raising interest rates would send the US economy back into recession. Pessimistic investors believe that over the past few years the growth potential of the US economy has been reduced and further damaged by the Corona crisis and the policy response to it. This potential is reflected in r*, the neutral interest rate at which the economy reaches full employment without overheating or cooling off.

Some market participants believe that, given major structural changes in the economy, the massive increase in debt and growth driven primarily by government stimulus, this rate is now so low that no significant interest rate cycle by the Fed can be expected. Figure 3 shows the longer-term trend of US key interest rates pointing downwards with lower interest rate peaks in each new cycle and growing intervals between cycles.

Figure 3: US key interest rates since 1990

If this argument is correct, a spillover of Japanese conditions to the US can be expected, with significantly lower interest rates, flatter yield curves and less momentum. Rate hikes occur less frequently and later in the cycle, or are met by flatter curves and falling long-term rates. Hints of tighter policy measures trigger falling interest rates at the long end. However, this argument can probably only be confirmed in the long term.

Critics of this explanation point out that such a mistake by the Fed does not fit with the still elevated inflation expectations and record low real interest rates. They also point to the highest inflation and growth rates in many years. Proponents counter that we are talking here about longer-term developments that are not contradicted by the realisation of high inflation and growth rates in the short term after the end of the recession.

Another argument is increased demand for the safest global government bonds due to an increasing scarcity of high-quality collateral in the market. On the one hand, central banks have been buying up government bonds on a large scale as part of the pandemic programmes, and on the other hand, the US
government has taken on little new debt this year due to high tax revenues and the reduction of holdings in its accounts at the central bank at the same time as massive spending programmes are flooding the economy with money.

In addition, the government has again reached the debt limit as of 1 August 2021, and currently cannot take on any new debt until that limit is formally raised. This makes the existing collateral more valuable in the market and drives down its yield.

At the same time, relief for banks with regard to collateral requirements expired in the USA in March, which the US government had temporarily lifted a year earlier as part of the Corona crisis. This has significantly increased demand for collateral since March. Since high-quality collateral is needed, among other things, in the repo market, which many market participants use for short-term financing, a hunt for collateral can trigger large reactions in interest rates.

The repo market is very large and at the same time non-transparent. It is therefore difficult to make an accurate assessment of how high the risks from the scarcity of high-quality collateral are. If this assumes large dimensions, it can lead to sudden and sharp declines in interest rates and trigger forced liquidations of risk assets. However, it is unclear how high the risk is for such a reaction.

The scarcity of quality collateral is widely acknowledged. Fed Chairman Jerome Powell said so in a hearing before Congress, and the Financial Times wrote an article on the subject in July 1. But a direct link is likely to be difficult to prove empirically. At the very least, however, the argument explains why investors have to park ever larger volumes of excess cash in the Federal Reserve’s reverse repo facility.

Volume in Fed reverse repo operations

The situation is unlikely to ease before the expected increase in the US debt ceiling in October or November, when short-term debt securities are expected to be issued again on a larger scale.
Until then, the situation can be gauged by numbers regarding the use of the reverse repo facility.

A completely different argument has been put forward by Morgan Stanley research. According to this, the rise in interest rates in Q1/2021 is said to have been caused primarily by sales by large Japanese funds at the end of the Japanese fiscal year (as of March 31st each year). Analysts at the investment bank pointed out that most of the Q1 rate hike occurred during Tokyo’s opening hours.
If Japanese institutions found the now much higher yields, steeper curves and low hedge costs attractive in the new fiscal year starting April 1st, the decline in rates may also have been driven primarily by Japan.

If the movement in interest rates were to be driven primarily by large investors from one country, Japan, it would mean that all explanations put forward would have to be reassessed and explanations derived so far might have to be reconsidered. It would also mean that one of the previously most liquid markets of all has become so much more illiquid as a result of central bank intervention that individual large investors can have a lasting influence on the direction of the market over a longer period of time.

While it is undisputed that large investors have an influence on price movements in markets, it is unclear whether a single investor can have such a lasting impact on treasuries. In recent years, there have been repeated quarters in which Russia or China have significantly reduced their holdings of US treasuries without this being reflected in significant interest rate increases.
It remains to be seen whether liquidity has declined so much since then that individual investors are now able to drive the price noticeably.

To determine whether the decline in interest rates since early April has been driven by Japanese investors, one can look at the intraday composition of interest rate movements from 4 January 2021 to 31 August 2021. Overall, US 10-year rates fell 0.44 percentage point. In the six hours that the Tokyo market is open, the market contributed 0.14 percentage point to lower interest rates, while the rest of the time Treasury yields fell 0.3 percentage point. This is roughly in line with the relationship between the trading hours, so it is not necessarily possible to discern a dominant Japanese influence in the price movements.

Is there a clear divergence in yield movements overnight and during the day? This would suggest that Asian investors are driving the market in a particular direction.