As we step into the second month of '22, markets and analysts get more and more confident in multiple Fed rate hikes this year. 90 bps increase in 2-year US Treasuries since 1st September last year suggests three to four rate hikes. In the last five months, the price of fed funds futures contract expiring in January 2023 also went down from 99.72 to 98.71. That gives a 100 bps increase in implied rate - four 0.25 bps rate hikes. Analyst forecasts seem to be mostly aligned with the market, maybe only except BofA, which expects the Fed to hike every meeting, i.e. seven times, this year. While four rate hikes this year are completely plausible, seven are not.
High growth rebound and low unemployment forces the Fed to reverse the fed funds target to a higher, preferably the pre-pandemic 1.25%, level. This is good news. Pushing up interest rates when the decision to tighten is based on strong fundamentals is always a positive sign of strengthening economy. Problems come when a central bank either changes the policy rate suddenly and for no particular good reason (monetary policy shock), or when it changes the rate on the basis of information that is not complete (wrong decision). Since 2008, modern central banks, and particularly the Fed, don't like unnecessary volatility in the markets and to avoid this, CBs simply follow the economy and "condition" the markets. We could see this type of conditioning on the last FOMC meeting where J. Powell's overall message was [not a citation] "we will tighten, and we may do it very hard". This is forward guidance.
Just after the meeting, BofA and other three major banks updated their forecast by increasing the number of times, they think, the Fed will hike this year. However, what these analysts know but don't put much weight to, is that the long-end of the curve has been signaling a completely different outcome. The yield curve has almost inverted. First, in the eurodollar futures market two months ago the curve slightly inverted, and now in the Treasury market it has flattered.
Flat or inverted dollar curve is a very strong indicator. In fact, it is the best forward-looking business cycle measure, the best known forecast of recession. The consensus compared to the yield curves look like a complete statistical noise, yet once again it is being disregarded. Reason? The Fed. According to Ethan Harris, the head of research at BofA, "the yield curve is heavily distorted by huge central bank balance sheets and US bond yields are being held down by remarkably low yields overseas". In theory, the rationales seem reasonable but in fact, empirically, are completely wrong.
Some QE history
Treasuries, and so the yield curve, do react to central bank purchases, but the effect that these purchases have on the market is vastly overstated. In theory, QE can affect the bond market in two different ways. The first one is the "flow effect" in which Treasury prices react in times of purchase announcements or actual purchases. The second one is the "stock effect", that increases yields by permanently reducing the total amount of Treasuries. Flow effects can be very clearly seen in the first LSAP program (QE1), when a day after the announcement of purchases, Treasury securities of longer maturity fell by roughly 50 bps. Subsequent rounds of QE were less impressive in terms of flow effects. In fact, there were no significant flow effects in the following purchase programs.
As for the stock effects, these in practice are also statistically nonexistent. Some argued that the original LSAP had some stock effects, however, a later BIS publication disproved claims of practically any effects that central bank asset purchases may have on prices of US Treasuries and UK gilts.
Anyway, analyzing literature and running statistical models is not necessary to figure out that there is a lack of relationship between QEs and yields. Analyzing yields themselves suffices.
The story is that the Fed's purchases should lower yields, and then close the spreads, or in Harris' words "heavily distort the yield curve". Surprisingly, however, historically, yields increased during quantitative easing phases. During QE1, Treasury 10s went from around 3.0% to 3.88% in three months, and at the end of the program closed at around 3.4%. Treasury 5s and 30s followed the same direction. From September 2009 to April 2010, in the time of QE2, 5s yield increased by 38 bps, and 10s by 60 bps. The QE3 period saw an increase in 10s and 30s from 2.13% and 3.27% to 2.65% and 3.69% respectively. Every quantitative easing program ended with higher or roughly equal Treasury yields compared to the time when these programs were announced.
As yields rise, spreads widen. Thus, asset purchases periods saw yield spreads, mostly increasing as it is shown in plots below.
Changes in the maturity of the Fed's Treasury portfolio also doesn't seem to cause any particular movement in Treasury spreads, i.e. the yield curve. The proportion of bonds (notes) held by the Fed to the total marketable supply of bonds (notes) shows this very clearly. To see how the yield curve changes in the maturity structure of the Fed's Treasury holdings, we can analyze the spread between these two ratios.
More bonds at the Fed relative to notes should imply lower bond yields compared to note yields, meaning that yield spreads should fall. At least, that is the story, and the story turns out to be wrong again. From January 2011 to May 2013, bonds ratio and notes ratio went from 18% and 15% to 42% and 21% respectively. Did the yield spreads change? In the interim, yes. In the end, no. During 2015-2017, normalized bond holdings fell compared to notes, and the yield spreads also fell. Complete opposite to "the story", and "Fed's yield curve distortion".
Current term structure curves are worrisome
A downward-sloping yield curve doesn't just forecast an economic slowdown, but also contributes to one. Higher low-term yields and lower high-term yields make commercial banking business quite problematic. Why? Well, it is because banks engage in a maturity-transformation, that is, they finance themselves with short-term liabilities, mostly deposits, and invest in longer-term assets, that is mostly the loan portfolio. When the yield curve is rising across maturities and is upward sloping, then banks can earn money by actually engaging in their core business, as their assets earn much more than their financing expenses. In times of an abnormal yield curve, that is flat or maybe even inverted, the business becomes complicated. Interest margins get squeezed and lending slows as banks are trying to survive.
As of 5 February 2022, the difference between 7-year and 5-year Treasury yield is 12 bps and the gap between 10s and 7s is only 4 bps. Nine-months ago, the same gaps were at 45 bps and 34 bps, respectively. J. Powell's change in rhetoric, i.e. suddenly inflation is not transitory, caused the short-end to increase, while the long-end haven't moved that much. In the last 9 months, the yield on 2-year T-note has increased by 103 bps, and only 19 bps on the 10-year note.
If the recent surge of shorter maturity yields is really caused by the FOMC talk, then the interest rate transition mechanism is faulty. The markets don't believe in the long-term recovery from everything that has happened in the last three years.
A eurodollar curve looks a bit better now than a few weeks ago, but still inverts for one period at the March '25 contract. The short-end is extremely steep and that is most likely caused by hike expectations and the traders that want to cash out, these expectations increasing.
Contracts that mature later than December '26 are a little different, as these are significantly less liquid. However, long maturity ED futures normally would be at least flat and what we see is a downward-sloping long end that starts at Sep '28 contract.
At this moment, both UST and eurdollar curves tell us that the market prices in only hike expectations, and not necessarily the future path of the interest rate. Hawkish expectations can be exploited and traded, similar to the mechanics of a stock market bubble. UST 5s, 7s, 10s as well as eurodollar Sep '23 and onwards show that there is something wrong with the mainstream tightening cycle narrative.
Does China really follow a different cycle?
As BoE and a few other smaller central banks have increased their policy rates, The PBoC has already cut the main policy rate two times since December last year. China's 1-year loan prime rate (LPR) decreased, in total, 15 bps, to 3.70%. Similarly, the 5-year LPR has also been reduced by the Chinese central bank by 5bps to 4.60%. Lowering both rates implies that the PBoC wants to boost both, the business activity and the housing market, as the 1Y LPR is the benchmark for corporate loans, and the 5Y LPR sets the level of mortgage lending rates. A significant implication of such action is that credit conditions got tight.
A domino of Chinese developers defaults, that started with Evergrande, is now spreading further into companies that previously were regarded as relatively healthy. As insolvency problems increase, financing is more and more difficult to secure. Another effect is a slower infrastructure development that feeds into the manufacturing capacities.
In January, China NBS and Caixin Manufacturing PMI fell to 50.1 and 49.1, respectively. Caixin PMI is now on the lowest level since the outbreak of the pandemic. New orders, a PMI metric that moves in advance of all other indicators, is at 49.3. New orders have now been drifting below 50 for five consecutive months. Industrial production has somewhat recovered, though, is still on a low year-over-year growth level of 4.3%. In short, the present state and the near-future outlook for the Chinese manufacturing-side of the economy is quite dim.
Of course, the real estate sector cannot be solely blamed for the current situation. High prices of raw materials and the local omicron lockdowns are also somewhat responsible for the slack. Nevertheless, the services side, although a little stronger, also struggles. Retail sales increased only 1.7% yoy and the non-manufacturing PMI is currently at 51.1. To put these numbers into perspective, in 2015-16 retail sales growth was at about 10%, the PMI level about 53.4, and these years weren't really prosperous for the Chinese.
If the slowdown in the Middle Kingdom continues, then spillovers to the global economy will be certain. Global growth slack coming from the second-biggest economy and on top of that a flat, downward-sloping yield curve do not instill much optimism. Seven hikes this year? In this global environment, four funds rate increases that are not followed by a country-wide recession would already be an enormous success.