Concerns about a global growth slowdown and recession in parts of the world are rising. The current focus of monetary policy in most advanced economies (AEs) is price stability and thus reigning in high inflation by aligning demand (which was elevated in AEs by a very generous fiscal response to the pandemic) with supply (which has been clogged due to lockdowns, shift in labour preferences, etc.). Financial markets still reflect aggressive policy rate hike expectations, although estimated terminal rates in the US are now off its recent peak and also below that implied by the Fed’s dot-plot itself.
However, consumer sentiment has crashed both in the US and EU. Will accumulated personal savings and a tight labour market continue to support spending and thus inflation? Maybe not, given falling sentiment amidst currently negative real personal disposable incomes and initial signs of lower nominal wage growth. Will early signs of an inventory buildup (based on recent commentary from retailers like Walmart and Target and also if rotation of consumption from goods to services continues) metamorphose into a broader and deeper economic phenomenon? This alongside evolution of supply decongestion, depending on stringency of Covid response in various parts of the world, would define the inflation trajectory, central bank response and the growth outlook.
While this is a more fundamental way of analysis, an often-cited harbinger of economic recession is yield curve inversions. The US 10y-2y treasury security yield spread inverted early July and remains so. The 10y-3m yield spread is still positive but is well below its recent peak in early May (Figure 1). Without admittedly getting into the probability of a US recession in the next one year, whether the Fed will be able to pull off a soft landing or whether spread inversions are the best predictors of recessions, we did a detailed empirical study of how yield spreads behaved before the three US recessions since the 1990s (excluding the short pandemic induced one in 2020). We looked at magnitude and duration of various spread inversions, the period when all spreads remain inverted, the lead it could provide on the nature and timing of a slowdown, which spread mattered most, etc. Based on this, if 10y-2y inversion persists/deepens, alongside an inversion particularly in the 10y-3m for a reasonable period, this could be a valuable empirical lead indicator of a recession.
The 1990 recession
US recessions are defined by NBER, not just based on Gross Domestic Product (GDP), but also based on indicators of employment, personal income, industrial production, etc. It is a tag with a lag.
Officially, the 1990 recession was for eight months from Aug90 to Mar91. Our study is focused on the five spreads 10y-3m, 10y-6m, 10y-1y, 10y-2y and 10y-3y. Key observations during this phase, particularly for 10y-2y and 10y-3m are:
1) 10y-2y inverted from Dec88 to Mar90 while 10y-3m inverted much later from May89 to Dec89 for a shorter period.
2) However, of the 274 days between Dec88 and May90 during which at least one of the five spreads inverted, 10y-2y inverted 70% of the days while 10y-3m inverted only for a shorter span of 36% of the days.
3) Of 274 days, both 10y-2y & 10y-3m inverted only 14% of the days and all 5 spreads inverted only 12% of the days.
4) Core phase - period of maximum duration of continuous inversion in all 5 spreads - was in Jun89 just after 10y-3m inverted, the last among all five spreads. This phase lasted for almost a month. All five spreads hit their minimum value just before/around the core period.
5) After the core phase, 10y-3m was the last to turn positive.
6) Mini-core phase - for about two weeks later in Aug89 when 4-5 spreads inverted but, unlike core phase, 10y-3m was directionally mostly moving upward and magnitude of inversions were lower than that during the core phase.
7) Then, spreads remained weak and 10y-3m inverted intermittently until early 1990 when it started moving up.
8) The eight-month official recession started one year after the end of the core-phase. However, remember recession is retrospectively defined by NBER. In real-time, one has to infer using economic data, etc.
10y-3m is thus crucial as its inversion, alongside other spread inversions, is a decisive indicator. Its direction, even when inverted, also matters. Duration of the core phase and magnitude of inversions during this is also important to decipher the nature of slowdown ahead.
The 2001 recession
The 2001 recession was for eight months from April 2001 to November 2001.
1) 10y-2y inverted in Feb00. Later in Apr00, 10y-3m inverted very mildly, when the other four spreads had already inverted, but this was only for two days after which it turned positive and moved further up. A few days later, 10y-6m and 10y-1y also turned positive. This phase was thus not a strong real-time indicator of recession.
2) In early Jun00, 10y-3m started falling and inverted in early Jul00. This was the onset of the core phase when all spreads stayed inverted for six months till end Dec00.
3) As in 1989, 10y-3m was the last to turn positive, in mid Jan01, after the core phase. After this, spreads moved sideways for two months until Mar01 when it started moving up.
4) The eight-month official recession started three months after the end of the core-phase.
The 2008 recession
The 2008 recession officially lasted for 18 months from January 2008 to June 2009.
1) There was a short period of about a month (Jan06 to Feb06) when spreads inverted very mildly but 10y-3m inverted only for a few days. In real-time, this is not a strong indicator.
2) 10y-2y inverted in Jun06. Later in mid Jul06, 10y-3m inverted last (as in the previous two recessions). Although 10y-2y and 10y-3y intermittently turned mildly positive in Jul06 and Aug06, 10y-3m continued to stay negative.
3) Core phase (but 10y-3y was not always inverted) lasted long for 11 months from Jul06, during which spreads were directionally moving down for the first 8 months. The subsequent 2008 recession was longer than in 1990 & 2001.
4) 10y-3m and 10y-6m hit their minimum values at 4.5 months, and again at 7.5 months, into the core phase. These were much deeper than that in 1989. The subsequent 2008 recession was deeper than in 1990 and 2001.
5) 10y-3m was thus a clear indicator, inverting sharply and consistently while 10y-2y inverted only mildly throughout.
6) After the core phase, 10y-3m turned strongly positive in two weeks but had a 3-week inversion period again. However, this inversion was much milder and shorter. 10y-2y and 10y-3y never inverted and were moving marginally upward. Thereafter, 10y-3m turned positive and moved up strongly alongside other spreads.
7) The eighteen-month official recession started seven months after the end of the core-phase.
Inferences on spread behaviour before the 3 ’R’s
- 10y-3m inversion matters. Among all spreads, 10y-3m is a clearer indicator although it inverts for a shorter period of time vs. other spreads. 10y-3m inversion alongside 10y-2y inversion for a reasonable period is a strong indicator. Also, 10y-3m is typically the last to turn positive and breakout of the core phase.
- 10y-3m typically inverts last among all the five spreads. Thus, to call a recession in real-time, simultaneous inversion in other spreads (particularly in 10y-2y), direction of 10y-3m (falling or rising), etc. also matter.
- Duration, magnitude of inversions and direction of spread movement during the core phase matters. Even volatility, i.e. frequency and magnitude of up and down fluctuations in spreads, during this phase is important. These clearly indicate the nature of recession that could follow.
- Core phase before the 2008 recession was defined more by the continuous inversion of three spreads, while 10y-2y and 10y-3y inverted only mildly but stayed flat and mildly positive. However, 2008 recession was the deepest among the three recessions. This was despite that lowest magnitude of spreads touched in 2008 was only next to that before the 2001 recession. This is because, duration of the core phase before 2008 was almost double of that before 2001. Thus, details about magnitude and duration of spread inversions have to be looked at closely.
- There could be a short period of spread inversions before the core phase. This is likely not a strong indicator.
- There could also be a ‘mini-core’ phase or a period of weakness in spreads, after the core phase, but one needs to again look at the magnitude, duration, diffusion and direction of spreads.
- Onset of recession as officially defined by the NBER occurred in 3-12 months from the end of the core phase, based on the three episodes.
So, to gauge the possibility and the nature of a potential recession in the US, one needs to look at both 10y-2y and 10y-3m spreads with all the above nuances and considerations. Currently, 10y-3m is still positive, albeit well off its recent peak and falling while other spreads have inverted.
Disclaimer:
MUTUAL FUND INVESTMENTS ARE SUBJECT TO MARKET RISKS, READ ALL SCHEME RELATED DOCUMENTS CAREFULLY.
The Disclosures of opinions/in house views/strategy incorporated herein is provided solely to enhance the transparency about the investment strategy / theme of the Scheme and should not be treated as endorsement of the views / opinions or as an investment advice. This document should not be construed as a research report or a recommendation to buy or sell any security. This document has been prepared on the basis of information, which is already available in publicly accessible media or developed through analysis of IDFC Mutual Fund. The information/ views / opinions provided is for informative purpose only and may have ceased to be current by the time it may reach the recipient, which should be taken into account before interpreting this document. The recipient should note and understand that the information provided above may not contain all the material aspects relevant for making an investment decision and the security may or may not continue to form part of the scheme’s portfolio in future. Investors are advised to consult their own investment advisor before making any investment decision in light of their risk appetite, investment goals and horizon. The decision of the Investment Manager may not always be profitable; as such decisions are based on the prevailing market conditions and the understanding of the Investment Manager. Actual market movements may vary from the anticipated trends. This information is subject to change without any prior notice. The Company reserves the right to make modifications and alterations to this statement as may be required from time to time. Neither IDFC Mutual Fund / IDFC AMC Trustee Co. Ltd./ IDFC Asset Management Co. Ltd nor IDFC, its Directors or representatives shall be liable for any damages whether direct or indirect, incidental, punitive special or consequential including lost revenue or lost profits that may arise from or in connection with the use of the information.
