Finance Research Letters 55 (2023) 104004

Contents lists available at ScienceDirect

Finance Research Letters

journal homepage: www.elsevier.com/locate/frl

Is the fed failing to re-anchor expectations? An analysis of jumps

in inflation swaps
Messaoud Chibane a, *, Ano Kuhanathan b
a
NEOMA Business School, Finance Department, France
b
Allianz Trade, Allianz Research, France

A R T I C L E I N F O A B S T R A C T

JEL: We investigate US inflation expectations de-anchoring by analyzing tail behavior of inflation

E31 swaps. We extract probabilities of jumps in inflation expectations along with their potential jump
E58 severity and look through an event study approach and a regression analysis on how Fed policy
G12
impacted these. Our findings suggest that recent increases in interest rates and FOMC an­
G13
nouncements of hikes did not manage to re-anchor short term inflation expectations as both
Keywords:
probabilities and intensity of potential inflation boom/crash anticipations remain broadly inde­
Inflation expectations
pendent of announcements and short-term interest rates dynamics.
De-anchoring
Monetary policy
Inflation swaps
Federal reserve

1. Introduction

Since 2021, inflation has been rising/high in many economies and especially in the US. Post-pandemic boom in demand, disruption
in global supply chains, tight labor markets, asynchronous Covid-19 outbreaks, large policy responses from governments to support
households and firms and the consequences of the war in Ukraine (i.e. high energy and commodity prices) are among the many factors
behind this phenomenon (see for instance Ball et al. (2022)).
Against this backdrop, the Federal Reserve (Fed) and many other central banks have recently been under growing pressure of
reining-in inflation. Market commentators and economists have been increasingly questioning the credibility of the Fed (see for
instance Editorial Board of the Washington Post (2022)). In this context, it is relevant to understand how monetary policy an­
nouncements are impacting inflation expectations. In fact, beyond actual inflation, as central banks’ action does not yield immediate
results, they also closely monitor inflation expectations of economic agents (firms, consumers, market participants) and we can argue
that inflation expectations are a key metric to assess a central bank’s credibility. Indeed, inflation expectations are key variables in two
major economic relationships: (i) the Fisher effect (Fisher (1930)) that states that market interest rates are a combination of real
interest rate and the expected rate of inflation, and (ii) the Phillips curve (Phillips (1958)) that describes an inverse relationship
between inflation and unemployment. From these seminal works emerged the rational expectations views (Lucas (1972); Sargent et al.,
(1973)) that were highly influential in Fed monetary policy during the 1970s and 1980s. Indeed, Fed Chairman Paul Volcker
recognized the importance of managing and anchoring the private sector’s inflation expectations and the ”Volcker disinflation” was

* Corresponding author.
E-mail address: [email protected] (M. Chibane).

https://doi.org/10.1016/j.frl.2023.104004
Received 24 January 2023; Received in revised form 26 April 2023; Accepted 15 May 2023
Available online 18 May 2023
1544-6123/© 2023 Elsevier Inc. All rights reserved.
M. Chibane and A. Kuhanathan Finance Research Letters 55 (2023) 104004

achieved thanks to that understanding (Goodfriend and King (2005)). Although there are legitimate debates about the parallels be­
tween post Covid-19 inflation and the 1970s inflation (see for instance Schmitt-Grohé and Uribe (2022)), anchoring inflation ex­
pectations remains a critical component and focus of policy action.
Before the development of financial instruments related to inflation, the monitoring of expectations was broadly based on surveys.
Some recent work investigated how survey-based inflation expectations are related to various factors ranging from exchange rates
(Nasir et al., (2020a) to oil price shocks Nasir et al., (2020b,c)) to past inflation and other cyclical variables (Mehra and Herrington
(2008)). Nowadays, on top of surveys, central bankers can also watch inflation-linked bond prices, derivatives, and swaps. Different
approaches have been used to estimate and extract inflation expectations, see Binder (2016) for a historical perspective. Some previous
research investigated inflation derivatives to analyze inflation expectations and their anchoring. Natoli and Sigalotti (2018) and
Gimeno and Ibañez (2018) for instance investigated euro area inflation expectations based on inflation derivatives. They both used
data on inflation options to derive probability distributions of future inflation to estimate inflation expectations. Existing research
investigating the matter of central bank credibility can mostly be split into two categories, a first one examining central bank’s policies
of explicit inflation targeting (see for instance Pham et al., (2023)) and a second one focusing on policy announcements, forward
guidance and communication (see for instance Baker and Lam (2022)). Our paper is related to the latter strand and we contribute to the
literature on inflation expectations by assessing potential extreme movements in inflation swap by analyzing tail behaviors in inflation
swaps rather than levels of future inflation. We also include the most recent period and assess central bank impact on tail behaviors in
order to examine if they contribute to reducing jump probabilities thus re-anchoring inflation expectations through an event-study
approach around monetary policy announcements. In the next section, we present an approach to estimate jumps in inflation swap
prices. After extracting these tail behaviors, Section 3 analyzes how Fed policy announcements impact them. Section 4 provides a
conclusion and offers some potential avenues for future research.

2. Analyzing jumps in inflation swaps

We choose to focus our analysis on inflation swaps as they are available at a much higher frequency than survey data. We also favor
inflation swaps as opposed to Treasury Inflation-Protected Securities (TIPS) as the latter include in their prices and yields components
that are not only related to inflation expectations (i.e. future path of monetary policy, term premium, sovereign risk premium) while
inflation swaps are solely proxies for inflation expectations. To estimate tail behavior of inflation swaps, we assume that their break-
even rates from one period to another verify the following dynamics:
rt+1 = μ + ut+1 + vt+1
( )
ut+1 ∼ N 0, σ 2
⎧
⎪
⎪ 0 with probability 1 − pd − pu (1)
⎨
vt+1 = vd,t+1 with probability pd
⎪
⎪
⎩
vu,t+1 with probability pu

where both processes vd and vu are respectively crash and boom processes that are distributed exponentially:
vd,t+1 ∼ exp(αd )
vu,t+1 ∼ exp(αu )

Effectively, the inflation break-even rt+1over the period [t, t + 1]is a combination of a constant growth rate µ, a Gaussian innovation
ut+1 and a non-Gaussian jump process vt + 1 that can be positive with probability pu or negative with probabilitypd. This jump process is
idle most of the time with a probability 1 − pd − pu. When this process becomes active, the magnitude of the non-Gaussian process is
exponentially distributed with tail exponent αdif a crash occurs and αuotherwise. When both these exponents are +∞ or when the
probability of crash/boom occurrence is 0 the jump process be-comes idle, and the expected inflation rate turns out to be Gaussian.
We extract US 1Y1Y inflation swap prices from Eikon Refinitiv. Our dataset ranges from 1 October 2013 to 30 September 2022
which corresponds to the full set of data provided by Eikon Refinitiv. We perform an estimation of inflation tail parameters by moment
matching using a rolling window of 1000 days. Indeed, since fat tail events happen on a low frequency, accurate estimation requires a
substantial amount of data points for estimation. Hence, the first estimation is obtained for 31 July 2017. We then move the estimation
period by one day and repeat the same procedure until the end of the full sample. Our estimations reveal thatpu, the probability of
inflation swap prices increasing sharply has been steadily rising since Q3 2020 (see Fig. 11). Although, this probably remains very
moderate, slightly below 3%, it is double the historical average between [2017,2019]. Furthermore, we also notice that αd and αu have
consistently decreased over time by about 30% each, this means that potential shocks in inflation swaps, and especially upside shocks
could be of much greater magnitude (see Fig. 22). These observations are consistent with actual inflation data pointing towards
increased price pressures in the aftermath of the first pandemic wave and the general concern in markets that inflation might overshoot
in 2023–2024. Indeed, our estimation suggests that there is a larger probability for inflation expectations to increase and that the
potential increase could be quite large. However, we observe (i) some volatility in the probabilities and (ii) pd, the probability of crash

1
For clarity, we do not graph confidence bands. All estimation results are available upon request.
2
idem

2
M. Chibane and A. Kuhanathan Finance Research Letters 55 (2023) 104004

Fig. 1. Estimated probabilities of jumps in US 1Y1Y inflation swaps (30-day smoothing).

Fig. 2. Estimated α for US 1Y1Y inflation swaps (30-day smoothing).

has been increasing (although with a lag and less sharply thanpu) and (iii) αddecreased also over time. This tends to suggest that there is
uncertainty surrounding inflation expectations and they could also fall severely. As a consequence, we could analyze whether mon­
etary policy has a significant impact on those metrics. Especially, we check if, as one would expect, higher interest rate decrease pu
(increase) and increase pd (decrease αd). In the following section, we investigate this through two different methods: an event-study
approach around monetary policy announcements and a regression analysis on the link between short-term interest rates and jump
probabilities.

3. Does monetary policy have an impact on jump probabilities?

3.1. Analyzing monetary policy announcements

As we seek to isolate the effect of the Fed’s monetary policy announcements, we use an event study methodology. Event study
methodologies are highly popular in many research topics, from asset price dynamics (recent work include Ho et al. (2022), Demi­
rgüç-Kunt et al., (2021), Marmora (2022)), to macro and policy related questions (see for instance Adams et al., (2022)). Especially,
some previous work investigated the impact of monetary policy announcements, among others we can cite: Wongswan (2009) who
analyzed the impact of Fed announcements on equity prices or more recently Sun (2020) who found that interest rates respond to
changes in the regulated retail interest rate and the required reserve ratio in China. Since our variable of interest is the estimated
boom/crash probability component in inflation swap returns, we cannot estimate a “normal” probability like we would be able to for
equity prices. Instead, we have to opt for a more straightforward approach, taking time into account. Hence, we estimate regressions
where boom/crash intensity and probabilities are a set of time-period fixed effects as follows:
pu = β0,pu + βt, pu + ∈pu (2)

3
M. Chibane and A. Kuhanathan Finance Research Letters 55 (2023) 104004

pd = β0,p + βt,p + ∈p d (3)

d d

αu = β0,αu + βt,αu + ∈αu (4)

αd = β0,α d + βt,α d + ∈α d (5)

Due to the quasi-monthly frequency of Federal Open Market Committee (FOMC) two-day meetings, we choose to consider a
window of 15 days, centered on the final day of each meeting. We estimate our equations around FOMC meetings that raised its target
for the federal funds rate (FFR) corresponding to 9 meetings within our sample’s period. In Tables 1 and 2 we display the estimates (β)
and standard errors (σ) for the time-varying coefficients of our regressions. We observe that for βpu there is no clear positive or negative
coefficient either before (from t = 0 to t = 6) or after the meetings (after t = 7). And more importantly, all estimated coefficients are not
statistically significant, which suggests that there is no significant impact of FOMC announcements on βpu . Regardingβp , estimates of
d

our fixed effects after the FOMC meetings are, as expected, mostly positive (i.e. the FOMC announcements in times of high inflation
should indeed increase the probability of a sharp decrease in inflation expectations). However, here again, they are not significant.
Likewise, for both βαd and βαu , we observe that our estimated coefficients after FOMC meetings’ announcements are not significantly
different from 0. Overall, these results tend to suggest that FOMC meetings’ announcements have no significant impact on jump
probabilities of inflation swaps (βp and βpu ) and their potential magnitude (βαd andβαu ). The probability and the potential magnitude of
d

boom and crash of future inflation seem to evolve independently of policy announcements.

3.2. The nexus between short-term interest rates and jump probabilities

As one could argue that market participants can anticipate monetary policy decisions and that speeches or other communications
from central bankers can lead to adjustments in interest rates and inflation expectations prior to official Fed announcements (see for
instance Szyszko et al., (2022)), we investigate the relation between jump probabilities and short-term interest rates - which are
considered as proxy for monetary policy rates and are available at a higher frequency. To do so, we run the following regressions on
1000-day rolling windows:
pu = cpu + γpu × Δr + ∈pu (6)

pd = cp d + γp × Δr + ∈pd (7)
d

αu = cαu + γαu × Δr + ∈αu (8)

αd = cαd + γαd × Δr + ∈αd

With c being a constant, Δr the change in short-term interest rate. Here, we use the inter-bank overnight rate as the short-term
interest rate. The regressions are fitted with White’s heteroskedacity robust estimator. We find a positive and increasing relation
between short-term interest rates and both jump probabilities (puandpd) - see Fig. 3. In the meantime, interest rates have been rising.
Theoretically, rising rates should have decreased upside jump probabilities and the coefficient should have been negative or at least
decreasing. Furthermore, we note that the coefficient for pd increased. This indicates that with rates increasing the probability of a
crash in inflation expectations increased. Although this is in line with economic intuition, a crash in inflation expectations reveals that
the rise in rates may be too fast/hard (i.e. suggesting a potential policy mistake). Also, we note that the coefficient forpuis higher and
increased more than forpd. Likewise, we note that the coefficients for αpu and αpd decreased over time (see Fig. 4). Overall, this tends to

Table 1
Estimated β for pu and pd.
σ σ
βpu βpu βpd βpd

t1 0.112 0.137 -0.034 0.066

t2 -0.025 0.042 -0.06 0.035
t3 0.017 0.042 0.036 0.05
t4 -0.014 0.033 0.031 0.036
t5 0.013 0.042 0.059 0.032
t6 0.004 0.025 0.04 0.03
t8 -0.034 0.024 0.031 0.04
t9 0.027 0.034 0.06 0.031
t
10 -0.012 0.034 0.05 0.035
t
11 -0.022 0.025 0.011 0.021
t
12 -0.006 0.028 0.032 0.044
t
13 -0.03 0.025 0.007 0.031
t
14 0.001 0.029 0.054 0.047
t
15 0.018 0.041 0.04 0.043

4
M. Chibane and A. Kuhanathan Finance Research Letters 55 (2023) 104004

Table 2
Estimated β for αu and αd.
σ β σ
βαu βα u αd βαd

t1 -0.001 0.001 -0.879 0.734

t2 0.001* 0 0.264 0.509
t3 -0.001 0.001 -1.805 1.285
t4 0 0.001 -1.288 0.8
t5 -0.001 0.001 -1.691 1.006
t6 0 0.001 -1.555 1.031
t8 0 0.001 -0.718 0.745
t9 -0.001 0.001 -1.575 1.034
t
10 0 0.001 -1.435 0.777
t
11 0 0 -0.376 0.221
t
12 0 0.001 -0.592 0.726
t
13 0 0 -0.53 0.601
t
14 -0.001 0.001 -1.2 1.153
t
15 -0.001 0.001 -1.397 1.111
*
Denotes 5% significance.

Fig. 3. Estimated γ for pu and pd with overnight rate.

reveal that despite rising rates, a growing share of market participants feel that either policy makers are reacting too abruptly or too
softly in their objective to tame inflation.
To confirm our findings, we also perform the same regressions replacing inter-bank overnight rate by the 3-month Libor and find
very similar results. As a matter of fact, differences in coefficients on average are lower than 1.e− 5 for all γ (see Figs. 5 and 6).

4. Conclusion and extensions

The results presented in this paper are three-fold. First, by looking beyond headline figures of inflation-related assets prices and
estimating tail components of inflation swap prices, we find that market expectations of an inflation overshoot in the US have been
steadily increasing since 2020Q2. Second, we find that uncertainty around inflation expectations has increased because expectations of
a strong disinflation have also increased (although to a lesser extent). And third, more interestingly our event-study and regression
analysis suggest that neither Fed policy announcements of hikes nor increasing short-term interest rates did weigh on tail risks of
inflation swap prices. Both did not reduce overshoot probability and they did not reduce probabilities of strong disinflation, thus
suggesting that the Fed is not managing to re-anchor market expectations of inflation through short-term interest rate hikes. Such
findings pave the way for further research on underlying factors supporting higher inflation expectations (e.g. excessive liquidity, job
market issues, disruption in global supply chains) and also on potential policy options to tame them (e.g. would more aggressive
monetary policy action / communication be more effective? Should authorities look into a different policy mix including structural
reforms?). Moreover, this question is likely to take another turn in the context of financial stress in the US. Indeed, with the recent
collapse of some small banks in the US, the Fed will also need to tackle financial stability concerns while pursuing its efforts to tame
inflation. Thus, it is facing a difficult balancing act which raises many potential research questions on the role, objectives, and priorities
of central banks in the current environment as well as its articulation with financial regulation.

5
M. Chibane and A. Kuhanathan Finance Research Letters 55 (2023) 104004

Fig. 4. Estimated γ for αu (left-hand scale) and αd (right-hand scale) with overnight rate.

Fig. 5. Estimated γ for pu and pd with 3-month Libor.

Fig. 6. Estimated γ for αu (left-hand scale) and αd (right-hand scale) with 3-month Libor.

6
M. Chibane and A. Kuhanathan Finance Research Letters 55 (2023) 104004

Declaration of Competing Interest

We wish to confirm that there are no known conflicts of interest associated with this publication and there has been no significant
financial support for this work that could have influenced its outcome. We confirm that the manuscript has been read and approved by
all named authors and that there are no other people who satisfied the criteria for authorship but are not listed. We further confirm that
the order of authors listed in the manuscript has been approved by all of us. We confirm that we have given due consideration to the
protection of intellectual property associated with this work and that there are no impediments to publication, including the timing of
publication, with respect to intellectual property. In so doing we confirm that we have followed the regulations of our institutions
concerning intellectual property. We understand that the Corresponding Author is the sole contact for the Editorial process (including
Editorial Manager and direct communications with the office). He is responsible for communicating with the other authors about
progress, submissions of revisions and final approval of proofs. We confirm that we have provided a current, correct email address
which is accessible by the Corresponding Author.

Data availability

Data will be made available on request.

References

Adams, C., Meer, J., Sloan, C., 2022. The minimum wage and search effort. Econ. Lett. 212, 110288.
Baker, J.D., Lam, J.-P., 2022. Assessing the credibility of central bank signals: the case of transitory inflation. Econ. Lett. 220, 110875.
Ball, L.M., Leigh, D., Mishra, P., 2022. Understanding U.S. Inflation During the Covid Era. October. National Bureau of Economic Research. Working Paper 30613.
Binder, C.C., 2016. Estimation of historical inflation expectations. Explor. Econ. Hist. 61, 1–31.
Demirgüç-Kunt, A., Pedraza, A., Ruiz-Ortega, C., 2021. Banking sector performance during the covid-19 crisis. J. Bank. Financ. 133, 106305.
Editorial Board of the Washington Post (2022). The federal reserve has a credibility problem. The Washington Post.
Fisher, I., 1930. The Theory of Interest. Mac Millan, New York.
Gimeno, R., Ibañez, A., 2018. The eurozone (expected) inflation: an option’s eyes view. J. Int. Money Financ. 86, 70–92.
Goodfriend, M. and R.G. King (2005). The incredible Volcker disinflation. J. Monet. Econ. 52 (5), 981–1015.
Ho, K.C., Gao, Y., Gu, Q., Yang, D., 2022. Covid-19 vaccine approvals and stock market returns: the case of Chinese stocks. Econ. Lett. 215, 110466.
Lucas, R.E., 1972. Expectations and the neutrality of money. J. Econ. Theory 4 (2), 103–124.
Marmora, P., 2022. Does monetary policy fuel bitcoin demand? event-study evidence from emerging markets. J. Int. Financ. Mark. Institutions and Money 77,
101489.
Mehra, Y.P., Herrington, C., 2008. On the sources of movements in inflation expectations: a few insights from a var model. FRB Richmond Econ. Q. 94 (2), 121–146.
Nasir, M.A., Balsalobre-Lorente, D., Huynh, T.L.D., 2020a. Anchoring inflation expectations in the face of oil shocks & in the proximity of ZLB: a tale of two targeters.
Energy Econ. 86, 104662.
Nasir, M.A., Duc Huynh, T.L., Vo, X.V., 2020b. Exchange rate pass-through & management of inflation expectations in a small open inflation targeting economy. Int.
Rev. Econ. Financ. 69, 178–188.
Nasir, M.A., Huynh, T.L.D., Yarovaya, L., 2020c. Inflation targeting & implications of oil shocks for inflation expectations in oil-importing and exporting economies:
evidence from three Nordic kingdoms. Int. Rev. Financ. Anal. 72, 101558.
Natoli, F., Sigalotti, L., 2018. Tail co-movement in inflation expectations as an indicator of anchoring. Int. J. Central Bank. 14 (1), 35–71.
Pham, T.A.T., Nguyen, T.T., Nasir, M.A., Duc Huynh, T.L., 2023. Exchange rate pass-through: a comparative analysis of inflation targeting & non-targeting asean-5
countries. The Q. Rev. Econ. Financ. 87, 158–167.
Phillips, A.W., 1958. The relation between unemployment and the rate of change of money wage rates in the United Kingdom, 1861-1957. Economica 25 (100),
283–299.
Sargent, T.J., Fand, D., Goldfeld, S., 1973. Rational expectations, the real rate of interest, and the natural rate of unemployment. Brookings Pap. Econ. Act. 1973 (2),
429–480.
Schmitt-Groh´e, S., Uribe, M., 2022. What Do Long Data Tell Us About the Inflation Hike Post Covid-19 Pandemic? August National Bureau of Economic Research.
Working Paper 30357.
Sun, R., 2020. Monetary policy announcements and market interest rates’ response: evidence from China. J. Bank. Financ. 113, 105766.
Szyszko, M., Rutkowska, A., Kliber, A., 2022. Do words affect expectations? the effect of central banks communication on consumer inflation expectations. The Q. Rev.
Econ. Financ. 86, 221–229.
Wongswan, J., 2009. The response of global equity indexes to U.S. monetary policy announcements. J. Int. Money Financ. 28 (2), 344–365.