Given the way short-term US interest rate futures are priced[1], investors seem pretty unanimous in their view that the soon-to-be-ex-Fed Chair Yellen will, for a third successive year, celebrate the festive season by continuing to slide the punchbowl away from the party ie. hike the target Funds rate by 25bp.
The rationale for a December hike was clearly outlined in the recently published FOMC minutes, namely the US economic growth trajectory remains solid, and as a result of the tightness of the labour market, which is viewed as “operating at or above full employment”, the undershoot of inflation relative to the 2% target will be eradicated due to cyclically-generated upward pressure on aggregate prices. All rather mundane and unexciting as things go… one could even go as far as to label this textbook assessment “boring”.
This is not completely unintentional. US monetary policy settings have been extreme[2] over the past several years and the Fed is attempting to “walk them back” to historic norms. Considering how the market reacted during the earlier taper tantrums, ending, then reversing, the US central bank’s massive balance-sheet expanding LSAPs as part of the policy normalization process could have gone badly wrong. Indeed, many so-called bond vigilantes have long argued that the cessation of Fed purchases of US government debt would usher in the reversal of the multi-decade bull market for bonds[3].
Understandable, therefore, that Yellen et al. want to make this transition appear as smooth and seamless as possible so as to give investors’ confidence that they know what they are doing. It also explains why the removal of the punch bowl has occurred at an extraordinarily slow pace, albeit with a slight acceleration during the past 12 months.
Assuming a December 13th hike, it will have taken Yellen more than 700 days to have achieved what her predecessor Greenspan achieved in 180 days on average during the three prior tightening cycles. (The fastest Greenspan hiked rates 125bp was 103 days during the 1994-95 tightening cycle – hardly a fabulous period for US bond investors – while the slowest, which began in June 1999, lasted 266 days.)
Relative to the level seen when the first Fed rate hike was announced nearly two years ago, the US 10-year nominal Treasury bond is just 5bp higher in yield. This stability is impressive and suggests the Fed was right to act in a slow and cautious manner. However, rather than breathing a sigh of relief that a bond market accident has not materialized, an increasing number of investors view the stability of longer-dated US government bond yields as a fresh source of concern. Specifically, because of what it may imply regards US macroeconomic prospects.
We noted above that the rationale for the Fed hiking in December is textbook. What is also textbook is that such a macroeconomic scenario (solid economic growth at supposed full-employment) should be associated with a rising yield environment, but this is not what has occurred.
As mentioned, the 10-year nominal Treasury yield is barely changed, and following four 25bp rate hikes, the gap with the 2-year nominal Treasury yield has fallen from 130bp to just under 60bp – a significant flattening of the yield curve. The reason this is stoking concern is that if this trend continues the curve will eventually invert, and this has been a very reliable lead indicator of US recessions over the past several decades. (As can be seen in the exhibit below, it correctly flagged the last two recessions.)
Exhibit 1: US Yield Curve vs. US Real GDP Growth
Source: https://fred.stlouisfed.org/
Jerome Powell, Trump’s candidate for replacing Yellen, in his confirmation hearing before the Senate Banking Committee this week gave no indication that he intends to radically depart from the methods of his predecessors. Hence, assuming the last dot-plot (September) remains valid then around the middle of next year a very reliable US recession indicator could very easily be flashing red.
Obviously, whether or not we arrive at this juncture critically depends upon whether US longer-term government bond yields stay around current levels. Many reasons have been put forward to explain why US government bond yields have stayed surprisingly low despite the US economy enjoying its third longest expansion in almost 150 years[4]. Among them are the gravitational pull from low bond yields in other developed economies, a result of their respective monetary policy stances, as well as the relative scarcity of supply following changes to Treasury issuance plans. These arguments are interesting, but there is a domestic aspect to this latest yield conundrum[5] that they do not readily explain but which is worth considering.
Comparing the difference between the nominal 10-year US Treasury bond yield and its inflation-protected (TIPS) equivalent it is possible to generate a market-based measure of inflation expectations. As can been seen in the exhibit below, outside the Great Recession this break-even inflation rate for the US has been near 2%, in line with the Fed’s interpretation of the price stability part of its mandate[6]. The major difference between the pre and post Great Recession periods, and the reason why nominal bond yields have not bounced back, is because real yields have fallen substantially (from an average of 2% prior to the Great Recession to an average of 50bp subsequently).
Exhibit 2: Contributions To US 10-Year Nominal Treasury Yields
Source: https://fred.stlouisfed.org/
Such a composition is consistent with Wall Street taking the view that the Fed is a highly credible central bank operating in a “new normal” world defined by a lower (falling?) US potential economic growth rate. (The long-run economic growth projections of FOMC members published every quarter alongside the dot-plot charts clearly show the Fed is of a like mind regards the direction of travel for US potential GDP growth). This “new normal” narrative is clearly persuasive and has been reinforced by the relatively tepid pace of the latest US economic expansion: real US GDP growth, averaged over a 10-year rolling period, has fallen to 1.4% – its lowest level in the entire post-war period.
The problem with potential economic growth rates is that they are unobservable – it is an intellectual construct of economists that has to be estimated. When it comes to setting policy this represents a challenge – it is akin to driving a car by looking in a rear view mirror (a metaphor we have employed previously[7]) – because estimates can be wrong. Only with the passage of time do these errors become more visible.
Together with the relatively tepid nature of real GDP growth, another facet of the US economic expansion that has surprised investors, and Fed officials alike, is the subdued nature of inflation. As can be seen in the exhibit below, on the targeted inflation measure – core PCE deflator – the Fed has undershot its 2% goal almost every single month since it announced a numeric target in January 2012[8].
Exhibit 3: US Core PCE Deflator vs. Fed Target
Source: https://fred.stlouisfed.org/
Think about that for a moment.
Monetary policy impacts feed through into the economy with a lag, but almost all macro models have the full effect coming through in two years. Hence, a fair conclusion to draw from the Fed having missed its inflation target for almost five consecutive years is that US monetary policy has been overly restrictive (assuming it planned to achieve its numeric target). At the same time, economic growth has been more subdued than expected. Perhaps there is no puzzle at all – the Fed (and investors) lulled by the narrative appeal of “new normal” have simply been overly pessimistic about US potential economic growth rates.
At this stage we should issue a mea culpa, we too have found the “new normal” narrative seductive. What got us thinking about this alternative explanation was the evolution of crowd-sourced sentiment data in the US over the past five years. When we plot US economic growth and inflation outlook sentiment we see something markedly different than what we observe looking at the components of US nominal bond yields (see exhibit 2 above).
Since 2012, crowd perceptions of US inflation have been persistently negative – the Trumpflation surge being a notable outlier, this is significantly different to the period prior to the Great Recession. By contrast, economic growth perceptions have tended to be positive and more in line with the pre-Great Recession trend – see exhibit below.
Exhibit 4: Crowd-Sourced Sentiment – US Growth and Future Inflation
Source: www.amareos.com
The crowd, it appears, has been much better at comprehending the underlying soft inflation dynamics evident in the US economy than either Fed officials, their macro models or many investors have been.
Boots on the ground 1, DSGE Models 0.
Turning back to Fed policy, what this alternative scenario implies is that by continuing with its interest rate hiking cycle, the Fed may be on the cusp of committing a policy error that results in recession. By sticking to its analysis of the situation, and attributing the inflation undershoot to transitory or unexplained/unknown factors, the logic of continuing to tighten is internally consistent. CPI inflation should rise necessitating more monetary tightening.
However, if the alternative scenario we have just outlined – one that is corroborated by the evolution of crowd sourced sentiments over the past several years – is correct, this rise in CPI inflation will remain elusive. Then the appropriate course of action would be to pause. Allowing the US economy to run a little hotter (less cold may be a more apt description), economic growth would pick-up and the undershoot of inflation relative to target would eventually be eradicated. Combined this macro-mix should exert upward pressure on longer-term US Treasury yields, allowing the Fed to resume normalizing its policy stance without risking inverting the yield curve and potentially triggering a recession – something it is not well equipped to deal with due to the dearth of monetary headroom.
Such a change in direction, would require a significant mental flexibility by the Fed. This may be too much to hope for, so to conclude this Market Insight, it is worth considering the implications of the Fed continuing to hike interest rates and inverting the curve.
As mentioned above, an inverted yield curve has proved to be “a reliable lead indicator of US recessions”. Recessions are not equity-friendly macroeconomic outcomes – see exhibit below – hence it would appear sensible to begin reducing equity allocations as soon as the yield curve inversion occurs.
Exhibit 5: S&P500 Index vs. US Real GDP Growth
Source: https://fred.stlouisfed.org/
However, one of the key words in the penultimate sentence is lead – in both of the two previous recessions it was early by at least seven months. Many analysts and investors no doubt would argue – and we recall a very lengthy discussion with a technical analyst[9] on this very subject – that stock markets are forward looking and will, just like the bond market, being to price in a recession well ahead of time.
This rationale appears to be further supported by the logic of backward induction that we outlined in the previous Market Insight (albeit in a different context). All investors will get the same piece of information – the yield curve inverting – at the same time, and all can expect market conditions to be less favourable for selling equities as the recession approaches. Hence, getting ahead of the pack by selling first appears to be the optimal strategy. But, if every other investor can follow the same logic, the outcome is that stocks sink as soon as the yield curve inverts, as all investors try to front-run each other.
As appealing as such logic is, when we look at how US large cap equities performed in between the US Treasury curve inverting and the start of the US recession (as defined by the NBER GDP-based recession indicator not the NBER’s business cycle dating committee’s subjective assessment) we find that on both prior occasions, stocks continued to rally for some considerable time – a result that is also corroborated by an interesting experimental economics study[10].
Exhibit 6: S&P500 Cumulative Return After Curve Inversion (10 year – 2 year)
Source: https://fred.stlouisfed.org/
If one squints at the left hand side of the chart it is possible to detect a period of equity market weakness as the curve inversion occurs, consistent with investors reacting to the potential recession warning. However, in both cases this weakness proved short-lived. History therefore, if it is a useful guide at all, suggests that there will be plenty of time for investors to scale back their exposure to US large cap stocks as/when/if the US yield curve inverts.
Regards how wide this window of opportunity may be, based on the above chart, stock markets peak anywhere between 150-350 days after the curve inverts. In the last instance, this coincided with US crowd-sourced sentiment towards US economic growth and equities falling into negative territory – see exhibit below (crowd-sourced sentiment data are unavailable prior to January 2005). That marks the point when the public’s faith in the recovery is finally extinguished and the bear market begins.
Exhibit 7: Crowd-Sourced Sentiments – US Growth and Equities
Source: www.amareos.com and https://fred.stlouisfed.org/
In conclusion, with the yield curve still some 60bp away from inverting, and with the possibility (hope?) that the Fed backs away from hiking rates next year, it appears too soon for investors to abandon US stocks. Moreover, stock price trends and crowd-sourced sentiments show investors do not give up on the economic recovery at the first sight of curve inversion providing investors with a potentially valuable “get out of jail free card”.
(Note:This Market Insight deals purely with cyclical arguments, it ignores geopolitical aspects which we covered in our previous Market Insight[11] and whose pertinence has been underscored by the latest North Korean missile launch.)
Amareos sentiment analytics incorporate MarketPsych indices.
FOOTNOTES:
[1] See: http://www.cmegroup.com/trading/interest-rates/countdown-to-fomc.html
[2] Although perhaps not extreme enough – see below.
[3] We never subscribed to this much anticipated call – see: https://www.amareos.com/financialresearch/caveat-venditor/
[4] See: https://www.amareos.com/financialresearch/the-market-sentimentalist-were-going-on-a-bear-hunt/
[5] Greenspan proposed the first conundrum back in 2005. For background and a review of factors the Fed sees as driving Treasury yields – see: https://www.frbsf.org/economic-research/publications/economic-letter/2017/november/new-conundrum-in-bond-market/
[6] It has three not two mandate objectives as is commonly thought. In addition to full-employment, it is also supposed to keep long-term interest rates stable – something it has been rather successful at doing.
[7] See: https://www.amareos.com/financialresearch/the-market-sentimentalist-rear-window/
[8] See: https://www.federalreserve.gov/newsevents/pressreleases/monetary20120125c.htm
[9] Hat tip: MG
[10] See: http://people.exeter.ac.uk/dgbalken/BEEM10908/Nagel.pdf
[11] See: https://www.amareos.com/financialresearch/the-market-sentimentalist-winter-is-coming/
