Popular in Economics

Central Monetary Authorities

Macroeconomic modelling of CBDC: a critical review

European Central Bank Ulrich Bindseil

Central Monetary Authorities

Financing Private Credit

Federal Reserve Bank of New York Nina Boyarchenko

Economics

Failing Banks

Massachusetts Institute of Technology Emil Verner

Economic History

Golden Fetters

UC Berkeley Barry Eichengreen

Economic Inequalities

Visions of Inequality

City University of New York Branko Milanovic

Economics

School Finance Equalization Increases Intergenerational Mobility

Yale University Barbara Biasi

Economics

Olivier Blanchard on U.S. Pandemic-Era Inflation

Peterson Institute for International Economics Olivier Blanchard

The market for inflation risk

View more insights in Global and Regional Policy Authorities - Social Sciences - Economics - Central Monetary Authorities - Macroeconomics

View more figures from Bank of England

View the page of Saleem Bahaj

Next in Central Monetary Authorities

xtevent: Estimation and Visualization in the Linear Panel Event-Study Design

Simon Freyaldenhoven | Federal Reserve Bank of Philadelphia

Time-varying Persistence of House Price Growth

C.Y. Choi | Federal Reserve Bank of Dallas

Saleem Bahaj discusses transaction-level data on the universe of traded UK inflation swaps to characterise who buys and sells inflation risk, when, and with what price elasticity.

Read the Paper

Any views expressed are solely those of the author(s) and so cannot be taken to represent those of the Bank of England or to state Bank of England policy.

Image courtesy of interviewee. August 2, 2023

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

This video discusses inflation swaps, which are financial contracts where one party pays the other based on the rate of inflation over the contract's duration. The paper highlights how market participants, including banks, pension funds, and hedge funds, interact in the inflation swap market, providing insights into expected inflation and trading behavior.

Key points include:
1. **Inflation Swap Mechanics**: One party pays a floating rate tied to inflation, while the other pays a fixed rate. The fixed rate reflects expectations of future inflation, which central banks monitor to gauge credibility in meeting inflation targets.

2. **Market Structure**: The analysis reveals distinct trading behaviors among institutions. Banks primarily sell inflation protection (taking on inflation risk), while pension funds buy long-dated swaps for liability matching. Hedge funds engage in short-dated swaps and can either buy or sell inflation protection based on market conditions.

3. **Market Segmentation**: The paper notes segmentation between long-term swaps (traded largely by pension funds) and short-term swaps (traded by hedge funds). This video segmentation leads to differences in trading dynamics and responses to inflation expectations.

4. **Data Insights**: The research analyzes granular transaction-level data from the UK inflation swap market. It identifies institutional trading behaviors, market power dynamics, and how demand and supply shocks affect swap prices.

5. **Modeling and Shocks**: The authors propose a model accommodating various shocks (e.g., inflation expectations, banking supply constraints) to understand how these factors impact swap prices and trading volume.

6. **Informative Pricing**: Long-dated swap prices provide better signals for expected inflation due to a flatter supply curve, while short-dated prices are more volatile and less reliable. The analysis suggests that central banks should focus on long-term inflation swap prices for more accurate insights.

7. **Implications**: There are concerns about banks potentially overexposing themselves to inflation risks given their role in selling inflation protection. The relationship between trading behavior and inflation forecasts among banks suggests that individual trading actions can provide insights into expectations for future inflation.

Overall, the paper presents a comprehensive analysis of inflation swaps, emphasizing the importance of understanding market dynamics, the role of different participants, and potential implications for policymakers regarding inflation expectations.

Language:

Export as PDF

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:06.800 --> 00:11.460
An inflation swap is a financial contract where one party,

00:12.080 --> 00:16.740
agrees to essentially pay the other party,

00:16.840 --> 00:21.100
the rate of inflation over the lifetime of the contract. So, you know, if,

00:21.120 --> 00:24.940
if prices rise by by 10% and you have, let's say, a hundred pound swap,

00:25.680 --> 00:29.740
the the party that agreed to pay inflation pays 10 pounds to the party

00:30.520 --> 00:33.340
in exchange for this. The second party,

00:33.480 --> 00:36.980
the other counterparty receives what we call a fixed, like a fixed payment.

00:37.210 --> 00:41.140
Okay? So imagine they say that, look, you know, I'm gonna pay you 10% on,

00:41.140 --> 00:44.460
on my swap, and then, and, you know, it's a hundred pound swap,

00:44.620 --> 00:48.220
they would also pay 10 pounds. Um, that's the, the idea behind,

00:48.220 --> 00:51.060
behind how that swap contract works. And that means that, you know,

00:51.080 --> 00:55.180
if inflation was at 10% and the fixed part of the,

00:55.180 --> 01:00.060
of the swap contract was at 10%, they break even. Okay?

01:00.060 --> 01:04.020
So both parties pay to a 10 pounds and no, the money really changes hands, okay?

01:04.760 --> 01:08.820
And we see the fixed price of these swap contracts in advance. Obviously,

01:08.820 --> 01:13.260
inflation happens over the, you know, the life of the contract. So maybe five,

01:13.320 --> 01:16.900
10 years in some cases. And so we can think about this sort of,

01:16.900 --> 01:17.690
these fixed prices,

01:17.690 --> 01:22.420
what we call the breakeven price as being a measure of expected

01:22.420 --> 01:24.700
inflation. So if I, I'm a market participant,

01:25.120 --> 01:29.180
I'm change trading these swap contracts, I can think about, okay,

01:29.180 --> 01:31.980
well what do I need inflation to be over, over the next, say,

01:32.050 --> 01:35.660
five years for me to break even when say it's 5%,

01:36.060 --> 01:39.260
I want to have a 5% fixed price on, on my swap. Okay?

01:39.720 --> 01:42.020
So these fixed prices are carefully watched,

01:42.120 --> 01:45.740
by policy makers to gauge what expected,

01:46.100 --> 01:47.420
expected inflation's going to, well,

01:47.420 --> 01:50.340
what expect inflation looks like for the next few years,

01:50.340 --> 01:52.300
from the point of view of financial market participants.

01:53.120 --> 01:56.300
Notice also that if I'm buying what we call the floating leg,

01:56.440 --> 01:59.060
the part of the swap, which varies from infl with inflation,

02:00.000 --> 02:04.900
I'm insuring myself against prices in inflation,

02:05.080 --> 02:09.300
by paying out when inflation falls. There's also an insurance aspect,

02:09.300 --> 02:11.300
to these contracts, too,

02:11.300 --> 02:13.380
which is why certain market participants like to buy them.

02:14.040 --> 02:17.420
But going back to the expected inflation, point,

02:17.920 --> 02:21.660
the reason why we care about these, these contracts is that, you know,

02:21.680 --> 02:23.940
if I'm a central bank and the Bank of England, for example,

02:24.620 --> 02:26.580
I want to make sure that people in the economy,

02:26.690 --> 02:29.900
believe me when I say that I'm gonna hit my inflation target,

02:30.050 --> 02:34.020
that inflation's gonna be, you know, 2% at least over the medium term, right?

02:34.480 --> 02:38.220
And so if I see that, that people in financial markets are changing trading,

02:38.290 --> 02:43.060
swap contracts with fixed prices a long way above 2%,

02:43.560 --> 02:48.420
that's a sign they don't believe me. Okay? So, so, but policy makers really,

02:48.420 --> 02:51.780
really watch what's happening to the price of these contracts, okay?

02:51.780 --> 02:56.020
And so they actually react that they see swap price is going up. They may think,

02:56.050 --> 03:00.040
okay, we are losing credibility here. Inflation expectations are rising.

03:00.420 --> 03:02.600
We should raise interest rates. Okay? That's kind of,

03:02.600 --> 03:05.440
that's kind of a part of the reaction function. Now,

03:06.100 --> 03:10.520
in terms of how our paper sort of fits into this topic,

03:11.960 --> 03:15.280
everyone sort of has been able to see the price of inflation swaps,

03:16.820 --> 03:21.760
from say, Bloomberg or, or other sources for a long time, okay?

03:22.230 --> 03:26.040
What we have is the micro data on the transactions.

03:26.060 --> 03:29.160
So who's actually trading in the market, okay?

03:29.660 --> 03:31.520
And that means we can ask questions like, okay, well,

03:31.610 --> 03:35.280
who's actually demanding inflation insurance and who's setting inflation

03:35.280 --> 03:38.720
protection? So who wants to insure themselves against higher prices? Okay?

03:38.720 --> 03:43.400
And who's at risk if, if inflation rises from these contracts? But also,

03:43.400 --> 03:45.160
you know, we know that financial,

03:45.160 --> 03:50.160
markets are imperfect and sometimes the prices that we see don't reflect

03:50.820 --> 03:53.400
purely expectations about inflation.

03:53.430 --> 03:57.280
They could reflect the market power or inefficiencies in trading.

03:58.020 --> 04:02.240
And we can ask, you know, if we can see the prices and the quantities,

04:02.300 --> 04:05.800
how much the variation is actually due that we see is due to changes in

04:05.800 --> 04:09.560
inflation and how much is due to say, you know, friction on trading activity.

04:09.980 --> 04:14.240
So data we have is, is essentially transaction level data.

04:14.380 --> 04:18.880
So every single transaction of an inflation swap for the UK market,

04:19.060 --> 04:23.760
so say a, a bank, a dealer bank as we call them,

04:24.090 --> 04:26.880
sells a swap contract to a hedge fund,

04:27.140 --> 04:30.520
on a particular day at a particular time, we, we see exactly that transaction,

04:30.830 --> 04:34.920
okay? And that, and that's reported as part of a regulatory dataset,

04:35.110 --> 04:39.600
that the Bank of England has access to, and we have access via, via them.

04:40.100 --> 04:44.200
And what's interesting is that we can use this data to sort of describe the

04:44.400 --> 04:48.760
features of the market, okay? And we have basically three different facts from,

04:49.070 --> 04:53.880
from these, from this transac transactional information. The first is that,

04:55.020 --> 04:58.240
essentially banks are

04:59.770 --> 05:02.950
net long in selling inflation protection, if that makes sense.

05:02.950 --> 05:06.070
So they're the ones who actually end up bearing inflation risk.

05:06.290 --> 05:09.270
So if you actually do function numbers,

05:10.700 --> 05:15.030
they end up having about a hundred billion dollars worth which dollars they say

05:15.030 --> 05:17.830
we pounds anymore. Uh, we dollars.

05:18.020 --> 05:21.950
They have a hundred billion dollars worth of fixation, exposure,

05:22.130 --> 05:26.070
at the beginning of 2021 when we saw the recent increase, in inflation.

05:26.380 --> 05:29.750
Okay? What's that mean? How do, how do we make sense of that? Well,

05:29.760 --> 05:34.350
we've seen about, a 50% overshooting, in the uk,

05:34.600 --> 05:39.070
price index since that, that time being 2021, okay? Compared to, say,

05:39.070 --> 05:43.390
the bee's target of 2% inflation, we've seen a 15% increase in, in,

05:43.490 --> 05:45.550
prices beyond that, okay?

05:45.890 --> 05:50.870
And that means that banks have sent about $15 billion just in cash

05:50.880 --> 05:54.910
terms, extra in meeting their commitments,

05:55.480 --> 05:56.750
under their inflation swap,

05:57.220 --> 06:01.640
Uh, contract. So it has to be 15 million, dollar outflow,

06:01.670 --> 06:06.080
from the banking system. Uh, and that's about 3% of bank capital. So it's not,

06:06.110 --> 06:07.800
it's not a tiny figure. Um,

06:08.230 --> 06:11.760
obviously banks have benefited from rising rates at the same time. So yeah,

06:11.760 --> 06:14.240
they have sort of hedged themselves for the means,

06:14.260 --> 06:18.000
but still it's a relatively substantial, cost for them.

06:19.090 --> 06:22.920
Who's on the other side of this transactions are fact Two is that we show that

06:22.920 --> 06:24.280
facing,

06:24.290 --> 06:27.320
banks who is buying protection from the banking system are pension funds.

06:27.580 --> 06:32.360
So pension funds in the UK have, have large, real,

06:32.880 --> 06:36.120
liabilities are pensions are to an extent index linked.

06:36.780 --> 06:40.040
And pension funds want to ensure that when inflation rises,

06:40.550 --> 06:44.600
they are protected from that and don't have to end up, you know,

06:44.600 --> 06:48.240
paying us all very large pensions when they haven't got necessarily assets in

06:48.240 --> 06:49.440
the balance sheet to meet those payments.

06:49.500 --> 06:53.350
So they buy inflation tax from the banking system. And what's interesting is,

06:53.470 --> 06:53.890
'cause our,

06:53.890 --> 06:58.310
our pensions are very long-term liabilities to the pension fund sector.

06:59.260 --> 07:01.590
They buy very long dated swaps, okay?

07:01.590 --> 07:06.510
So we see a lot of trading in sort of 10, 15, 20 year inflation swaps

07:06.510 --> 07:10.550
between pension funds and, and, and the banking system, okay?

07:10.850 --> 07:13.870
And a lot of volume that is some very, very long maturity swaps,

07:14.410 --> 07:18.110
but there's also a lot of trading at the short maturity of the market, okay,

07:18.170 --> 07:21.830
too. And here we see a very different participant.

07:22.090 --> 07:25.190
So our fact three is that in shorter maturities,

07:25.250 --> 07:29.270
we have hedge funds essentially trading against the banking system. Uh,

07:29.270 --> 07:33.430
they act as like informed traders. So when inflation is falling, they tend to,

07:33.690 --> 07:38.370
end up selling inflation protection. So they benefit when inflation falls,

07:38.880 --> 07:40.290
when inflation starts to rise,

07:40.300 --> 07:44.010
hedge funds switch their position and they end up buying protection.

07:44.010 --> 07:46.090
So they benefit from when inflation rises.

07:46.350 --> 07:50.410
So they sort of acting to get ahead of the cyclones, were benefiting from,

07:50.630 --> 07:54.290
rising and falling prices depending on, on their, on their,

07:54.620 --> 07:55.690
their net positions.

07:56.310 --> 08:00.250
And what's interesting about this is you have this very extreme segmentation,

08:00.910 --> 08:03.810
between these different markets. So in, in long maturity swaps,

08:04.230 --> 08:06.410
so very long-term inflation swaps,

08:06.920 --> 08:09.570
it's really pension funds trading against banks, and that's it.

08:09.990 --> 08:14.290
In the short maturity market, you have a hedge funds trading against banks.

08:14.750 --> 08:19.290
And so segmentation is very stark in the data. So you a segmented market, okay?

08:20.350 --> 08:23.010
We have this segmentation. So, so how, how I think about this thinking about,

08:23.010 --> 08:25.570
okay, we want to model the trading behavior of,

08:26.270 --> 08:30.930
of banks vis-a-vis the pension funds in the long inflation swap market

08:31.390 --> 08:35.650
and banks vis-a-vis hedge funds in the short inflation swap, market.

08:36.430 --> 08:38.210
And our goal here is to say, look, okay,

08:38.210 --> 08:41.970
imagine that these institutions are all,

08:42.730 --> 08:46.010
choosing what portfolio they want to invest in. They have some,

08:46.120 --> 08:47.530
some assets and liabilities,

08:47.530 --> 08:50.850
and they wanna optimally choose their portfolio knowing that they're trading

08:50.850 --> 08:54.010
against each other. So banks know they're trading against pension funds and,

08:54.350 --> 08:55.360
and, and,

08:55.450 --> 08:57.760
hedge funds military against banks and banks are trading against both

08:57.760 --> 09:00.840
counterparties actually at the same time. Okay? That's a segmentation view.

09:01.700 --> 09:04.800
But what we do is we also add frictions to this model and say, look, you know,

09:04.910 --> 09:08.880
it's not the case that we can all perfectly buy and sell insurance,

09:08.880 --> 09:11.080
particularly because we know pension funds, for instance,

09:11.080 --> 09:12.840
have constraints from their mandates.

09:13.510 --> 09:17.680
There's margining requirements attached to, buying a tenant. Inflation swap.

09:18.170 --> 09:21.080
Banks have capital requirements and legacy constraints on, on,

09:21.140 --> 09:24.880
on the transactions that they, that they engage in. And these,

09:25.050 --> 09:28.960
these constraints generate sort of deviations from the optimal choice, okay?

09:29.660 --> 09:31.920
And so you think about how this plays out,

09:32.500 --> 09:35.040
you can imagine a world where there's no sort of deviations,

09:35.040 --> 09:39.640
everything's perfect, okay? And we observe, you know,

09:39.640 --> 09:44.440
imagine about long, long dated inflation swaps. We observe some supply curve,

09:44.940 --> 09:45.580

09:45.580 --> 09:49.880
for selling inflation protection from the banking system towards pension funds

09:50.500 --> 09:53.400
and some demand curve that arise from the pension fund sector.

09:53.420 --> 09:57.680
That's what our model predicts in a frictionless case, okay?

09:58.220 --> 10:03.160
We reached equilibrium where pension funds by protection potentially because

10:03.160 --> 10:07.600
they're more risk averse than the banking sector or because,

10:07.600 --> 10:10.000
they had different expected inflation than the banking sector.

10:10.140 --> 10:13.800
That's two explanations why I have this positive position on the part of pension

10:13.810 --> 10:17.080
funds, okay? But that's a frictionless case.

10:17.780 --> 10:20.960
We can have other trading constraints that emerge. Okay?

10:20.960 --> 10:24.560
So it could be the case that pension funds just can't buy things of mandates

10:24.630 --> 10:25.800
that we talked about previously.

10:26.740 --> 10:31.560
Or it could be the case that because pension funds have real liabilities on

10:31.560 --> 10:32.393
their balance sheets,

10:33.070 --> 10:36.080
they're unable to properly ensure those of their other assets.

10:36.140 --> 10:38.400
And so they have to go to the swap market to get protection.

10:39.020 --> 10:40.000
That's kind of like a,

10:40.120 --> 10:44.200
a non fundamental or non expected inflation driven demand for inflation

10:44.210 --> 10:45.720
protection. Okay?

10:45.980 --> 10:50.400
And that means that that tho those two things can shift around demand for

10:50.400 --> 10:55.280
inflation swaps from the pension fund sector separately from

10:55.940 --> 10:58.240
shocks to expected inflation, as it were. Okay?

10:58.500 --> 11:00.240
And a similar thing applies to the making sector too.

11:00.240 --> 11:04.080
They also have constraints if say bank capital regulation gets tightened,

11:04.230 --> 11:08.200
that can shift around the supply curve of inflation swaps from the banking

11:08.200 --> 11:10.680
sector. And of course then we have two markets, right?

11:10.680 --> 11:11.600
We have a segmented market,

11:11.600 --> 11:15.360
we have a market for long data swaps and a market for short data swaps between

11:15.360 --> 11:18.840
the banking system and hedge funds. And again, you can tell a similar story.

11:18.840 --> 11:21.440
You have a demand shifter from the hedge funds. You know, they may,

11:21.470 --> 11:24.360
they may face some margin requirements they can't meet, for example,

11:24.860 --> 11:27.240
and banks face supply constraints, okay?

11:27.580 --> 11:30.680
So what we end up with in the model is basically four different shocks.

11:30.940 --> 11:32.320
The first is what we care about really,

11:32.320 --> 11:36.760
which is what we call a fundamental shock when expected inflation moves. Okay?

11:36.940 --> 11:39.640
That's the one to actually get at if I the bank of even care about when expected

11:39.640 --> 11:43.400
inflation move. But we have also three other shocks, okay?

11:43.460 --> 11:48.160
One is it's a shock to the ability of the banking system to supply

11:48.630 --> 11:52.000
inflation protection, supply supply inflation derivatives, inflation swaps,

11:52.310 --> 11:54.440
okay? And then a swap,

11:54.480 --> 11:58.280
a demand shock from the pension fund sector and a demand shock from the hedge

11:58.280 --> 12:00.360
fund sector. So we have these three different,

12:00.360 --> 12:02.640
four different shocks that emerge from the model.

12:03.260 --> 12:07.400
And the question is how do we take the two prices that we see the price of a

12:07.400 --> 12:11.880
long-term swap and the price of a short-term swap and the quantities that we see

12:11.880 --> 12:12.180
as well.

12:12.180 --> 12:16.560
So the quantity that that pension funds buy from banks and the quantities that

12:16.630 --> 12:18.120
that hedge funds buy from banks.

12:18.780 --> 12:23.760
And take those four time series and figure out how much of the

12:23.760 --> 12:26.160
variation in the prices that we're seeing is actually driven by expected

12:26.160 --> 12:30.640
inflation as opposed to these other shifters that can explain how the market is

12:30.640 --> 12:33.200
moving. What we do next then is say, look, we,

12:33.200 --> 12:37.560
we have these four different shocks. How do we actually disentangle,

12:38.140 --> 12:39.720
the effects in the data we have?

12:40.260 --> 12:43.480
And what we do is propose three different strategies for doing that.

12:44.020 --> 12:45.720
The first is that, you know,

12:45.980 --> 12:50.680
on certain days we know that inflation news comes out. So for example,

12:50.840 --> 12:55.240
D O N S comes out and publishes, estimates of inflation,

12:55.670 --> 12:58.720
okay? So sorry, estimates the release of inflation for that month.

12:58.720 --> 13:01.160
Basically that's, that's the data, that's what we're thinking about.

13:01.620 --> 13:02.600
And so on that day,

13:03.430 --> 13:08.250
there's much more news about expected inflation than there are

13:08.350 --> 13:12.610
on other dates, okay? So we have what we call ity,

13:13.270 --> 13:15.970
identification strategy that emerges from this.

13:15.990 --> 13:20.130
We can use the fact that the relative variants of these two shocks changes

13:20.870 --> 13:22.930
to separate out the fundamental shock,

13:22.980 --> 13:26.970
which is very volatile on the inflation release dates from the other shocks that

13:26.970 --> 13:30.530
hit, hit the, hit the system of the, the data we have. Okay?

13:30.530 --> 13:35.410
So that's one strategy that we look at. The second strategy that we look at is,

13:35.670 --> 13:38.050
what we call granularity. Okay?

13:38.110 --> 13:42.370
So we look exploit the fact that you know each individual institutional later.

13:42.390 --> 13:46.610
So we see bank by bank pension, fund by pension fund, hedge fund by hedge fund,

13:46.610 --> 13:49.250
their trading behavior. And we say that look,

13:49.250 --> 13:52.730
because there are some very large funds or banks in the data,

13:54.150 --> 13:56.130
you know, their idiosyncratic behavior,

13:56.780 --> 13:59.250
which is not affected by effective inflation, for example,

13:59.870 --> 14:04.650
can actually generate movements in demand or movements in supply independent of,

14:04.710 --> 14:07.130
of fundamental needs about inflation. Okay?

14:07.670 --> 14:11.610
And so we use that idiosyncratic movements of large institutions as instruments

14:12.190 --> 14:16.970
for shocks of demand and supply. Okay? The third thing we do, okay,

14:16.970 --> 14:19.130
so we do three different strategies to try and answer this question,

14:19.150 --> 14:20.570
but they actually will lining up.

14:20.790 --> 14:25.010
But the third thing we do is something called a sign restrictions approach.

14:25.470 --> 14:27.490
And just so that look, okay,

14:27.490 --> 14:32.410
imagine that there's no spillover from a demand shock between,

14:32.590 --> 14:36.010
the long and the short-term market at a very high frequency. So we say look,

14:36.010 --> 14:36.890
within a trading day,

14:37.550 --> 14:41.490
the bank's desk that's trading long-term inflation swaps and the bank's debt

14:41.490 --> 14:44.370
trading shortterm inflation swaps don't talk to each other too much.

14:44.370 --> 14:47.570
They're busy trading and not really doing anything in that situation.

14:47.720 --> 14:50.770
When pension funds try to buy more swaps, it doesn't affect

14:50.770 --> 14:55.240
The price of the short dated, market. Basically that,

14:55.240 --> 14:58.440
that segmented market we talked about, okay? And vice versa,

14:58.470 --> 15:00.520
when hedge funds try to buy more swabs, okay?

15:01.140 --> 15:04.120
We also make the assumption that basically hedge funds have the most

15:04.120 --> 15:07.960
information. Uh, so when, when the inflation rises,

15:07.960 --> 15:10.160
they're the ones that react the most. Okay?

15:10.620 --> 15:15.480
And those two assumptions combined means we can actually attach signs to each of

15:15.480 --> 15:16.260
these shocks. So,

15:16.260 --> 15:20.120
so a demand shock raises prices and quantities in the loan market if it's from

15:20.120 --> 15:22.760
pension funds, but has no effect in the short market, for example.

15:22.760 --> 15:25.680
That's one thing we do. We can use those different signs, which you say, okay,

15:25.680 --> 15:29.520
let's identify the model that way. Okay? So we have these,

15:29.530 --> 15:33.720
these three different strategies and what they do is quite interesting.

15:33.720 --> 15:38.560
They allow us to identify because we've, we've got supply and demand shocks,

15:38.620 --> 15:39.453
in both markets.

15:39.980 --> 15:44.360
We identify the slope of supply and demand curves for long data inflation,

15:44.390 --> 15:48.440
swap and pension in funds and banks and short data inflation swaps in pension

15:48.440 --> 15:52.000
funds and banks, okay? And this is kind of like, you know,

15:52.000 --> 15:54.440
in economics you care a lot about slope supply and demand. It's one of,

15:54.440 --> 15:56.960
one of the most interesting things, and we can measure it in the,

15:56.960 --> 15:57.840
in these two different markets.

15:58.940 --> 16:01.880
And what's fascinating about this is that the demand curves are very similar.

16:02.100 --> 16:05.960
So slopes are the same, for both pension funds and, and hedge funds.

16:05.960 --> 16:08.560
They're very, very similar demand curves in terms of their slope.

16:09.220 --> 16:13.720
But banks have a much flatter supply curve when they're supplying

16:13.930 --> 16:18.920
swaps to pension funds compared to when they're supplying swaps to

16:18.920 --> 16:22.030
hedge funds. And that has a very,

16:22.180 --> 16:25.350
important economic consequence. It means that, you know,

16:25.350 --> 16:28.750
imagine you have a very flat supply curve, a demand shock.

16:28.750 --> 16:32.510
There's very little to the price because supply is able to adjust very easily,

16:33.380 --> 16:34.170
okay?

16:34.170 --> 16:38.730
And that means most of the variation in a long dated

16:39.320 --> 16:43.170
swap contract. So one more than 10 years in, in,

16:44.230 --> 16:45.030
in maturity.

16:45.030 --> 16:48.890
So most of the variation in the price of those sorts of contracts is not

16:48.960 --> 16:53.610
explained by demand shifters because you're able just to move along a very flat

16:53.610 --> 16:56.090
supply curve very easily, okay?

16:57.110 --> 17:00.410
And all the variations and explained either by shocks the bank supply,

17:00.710 --> 17:05.490
so the supply curves of up or down or expect inflation moving, okay?

17:05.990 --> 17:08.290
And that means that long dated swap prices,

17:08.430 --> 17:12.330
say more than 10 years are actually quite informative about face expectations.

17:12.350 --> 17:12.770
So you should,

17:12.770 --> 17:16.130
you should look at those long dated prices to get a sense of like what are

17:16.130 --> 17:20.850
long-term face expectations in contrast in the short market,

17:21.240 --> 17:23.490
because you have an upward sloping supply curve.

17:23.490 --> 17:27.090
Now it's not very flat demand stocks are up mattering a lot and generate a lot

17:27.090 --> 17:31.410
of volatility, and they move the price around a lot, okay? And so short dates,

17:31.410 --> 17:35.410
inflation swap prices say less than three years and much less informative but

17:35.570 --> 17:38.930
expected inflation, which is kind of frustrating if, for the Bank of England.

17:39.130 --> 17:43.490
'cause they care about like what's the immediate expectations of, of,

17:44.070 --> 17:45.610
exp expect of expectation as long as,

17:45.710 --> 17:48.490
as well as a long-term thing they care about both. But in terms of getting a,

17:48.650 --> 17:49.560
a short-term gauge,

17:50.400 --> 17:54.440
actually these short dated swap prices aren't super informative. Okay?

17:54.440 --> 17:58.680
So that's, that's what we do. Um, and we can also do decomposition too.

17:58.700 --> 18:03.570
We can ask, okay, if I take a long-term inflation swap price, how much of the,

18:03.590 --> 18:06.170
of how much of the, the sort of the time series that we see,

18:07.310 --> 18:08.650
how much of that is the,

18:08.710 --> 18:12.450
the variation is due to fundamental factors expected inflation and how much is

18:12.450 --> 18:15.970
due to these demand and supply shocks? And what we see in the data we have,

18:15.970 --> 18:16.890
we have quite a short time.

18:16.910 --> 18:21.250
So these five years is that the events where there was a negative shift in

18:21.250 --> 18:25.050
inflation expectations? So covid for instance, when I face inflation,

18:25.160 --> 18:27.330
fell quite a lot in that time period,

18:27.950 --> 18:32.040
the 10 year swap price overestimated the fall, okay?

18:32.040 --> 18:35.840
So the swap prices were looked like inflation expect were falling a lot more

18:36.150 --> 18:39.080
than they actually were within decomposition, okay?

18:39.900 --> 18:43.440
And the opposite is true most recently. So, so when we had, for example,

18:43.440 --> 18:47.320
in the autum this year around the mini budget crisis, a rise in,

18:47.500 --> 18:50.560
in long-term face expectations, there was also overshooting.

18:50.900 --> 18:55.160
So actually the market prices were saying that face expectations have risen more

18:55.160 --> 18:57.640
than they actually had done. Once you take out the effects of,

18:57.740 --> 19:00.640
of supply the demand shocks and just focus on fundamentals,

19:01.190 --> 19:05.600
so's overshooting aspect too, that's there in, in the, in the data.

19:06.420 --> 19:10.820
The third thing we do in the paper is look at individual

19:11.410 --> 19:13.820
trading behavior. And just to say, you know,

19:14.460 --> 19:18.820
I said that the short term prices are aren't a useful guide, to,

19:19.640 --> 19:20.820
excess inflation.

19:21.520 --> 19:24.220
The other issue is that when you go into institution level beha,

19:24.220 --> 19:25.660
pricing behavior, trading behavior,

19:26.900 --> 19:30.180
actually it turns out that really the short term market for inflation swaps in

19:30.180 --> 19:31.580
the uk, so less than three years,

19:32.680 --> 19:37.220
all the pricing behavior was explained by about four banks and another four or

19:37.220 --> 19:39.020
five hedge funds. So it's a very,

19:39.090 --> 19:43.180
very small number of institutions that are actually doing most of the trading

19:43.360 --> 19:46.020
and reacting to the news that we're seeing. Okay?

19:46.020 --> 19:49.820
So we're aggregating the information about accepted inflation for a high number

19:49.820 --> 19:53.300
of institutions in the short market in a long-term market. However,

19:53.480 --> 19:56.860
you have hundreds of pension funds trading and many, many more banks. So again,

19:56.860 --> 20:01.420
it's a much, much, you know, richer information set that's emerging from, from,

20:01.480 --> 20:05.980
from long dated swap prices. Okay? So again, you know,

20:05.980 --> 20:09.780
it's another reason to perhaps pay less attention to the short dated swap.

20:10.600 --> 20:13.020
The last point I want to make this is sort of like as a crosscheck of what we're

20:13.020 --> 20:13.840
doing here,

20:13.840 --> 20:18.460
is that the banks in our sample all produce monthly forecasts

20:18.760 --> 20:22.340
of inflation. Like what they expect inflation to be over the next,

20:22.360 --> 20:26.220
12 months or so, let's say. And we can say, look, okay,

20:26.400 --> 20:31.100
we extract our shock to expect inflation, from our model and say,

20:31.100 --> 20:34.540
how do those forecasts change with respect to the shock? Okay?

20:35.120 --> 20:39.580
And then compare how that forecast change maps into bank's trading behavior

20:39.640 --> 20:43.420
change in response to the shop. And actually those two things run really,

20:43.420 --> 20:47.760
really well. So banks are trading on the forecast,

20:47.760 --> 20:49.920
that they're producing, which we found very interesting.

20:49.920 --> 20:52.720
So typically we sort of think that their economics department and their trading

20:52.720 --> 20:55.400
department are relatively separate. Doesn't seem to be the case.

20:55.400 --> 20:59.560
There's actually a mapping between, inflation forecast,

20:59.580 --> 21:02.920
but individual banks and their trading behavior as well. So some,

21:02.920 --> 21:06.520
there's something informative about, about individual institutions trading,

21:06.520 --> 21:09.640
that we're seeing. And if there are qualifications, I think we have,

21:09.920 --> 21:13.880
a couple. The first is, you know, there is, there is information contained in,

21:13.880 --> 21:16.320
in the inflation, stock market.

21:16.440 --> 21:17.880
I think that's kind of like what we learned from this.

21:17.990 --> 21:20.320
It's just not a perfect proxy, okay? So you should,

21:20.700 --> 21:23.800
you should take those measures of a slight pinch of salt,

21:23.800 --> 21:26.040
and not think of them as perfect measures of inflation.

21:26.040 --> 21:30.200
And it could be other things, driving the price that that po that you see.

21:30.580 --> 21:31.960
And policymakers should better that in mind.

21:31.960 --> 21:34.360
That's what the overarching message from this.

21:34.980 --> 21:38.440
Having said that the long dated swap is much more informative,

21:38.500 --> 21:42.960
it appears than short data swaps for the reasons we we sort of described, right?

21:43.030 --> 21:46.560
It's both. The flat supply curve basically eliminates the impact of,

21:46.560 --> 21:48.840
of demand shocks. And because you have many,

21:48.840 --> 21:51.160
many more actors trading in that market,

21:51.340 --> 21:52.960
so you're aggregating information across many,

21:52.960 --> 21:55.760
many more agents when you see the prices that you see. Okay?

21:55.760 --> 21:57.640
Those are the two things to bear in mind there. Short let,

21:57.640 --> 22:01.960
just swaps more problematic. Secondly, I mean, there's a slight, you know,

22:01.960 --> 22:03.800
sort of from a more macro prudential perspective,

22:04.630 --> 22:07.520
there's a slight risk that banks are, you know,

22:07.520 --> 22:11.880
selling a lot of inflation protection. Um, and you know,

22:11.880 --> 22:15.040
if there is an inflationary shock that persists for a long time,

22:15.330 --> 22:18.600
given these are very long dated swaps, 10, 20 years,

22:19.350 --> 22:21.360
that could actually have an impact on bank profitability,

22:21.360 --> 22:22.560
even knock on effect in the economy.

22:23.420 --> 22:26.000
I'm not sure how regulations should be reacted to that,

22:26.000 --> 22:27.320
whether it's a concern or not,

22:27.650 --> 22:31.640
given that banks have other hedges against inflation or more precisely against

22:31.640 --> 22:34.280
wide interest rates, their, their interest margin have to rise,

22:35.120 --> 22:39.000
interest rate rises. So they have some protection from that. However, you know,

22:39.000 --> 22:42.360
that's, there is a concern about, about all this protection that are selling,

22:42.700 --> 22:46.960
from the banking system. Uh, the third thing is that, you know,

22:47.030 --> 22:50.480
this mapping to to individual trading behavior,

22:50.620 --> 22:52.560
we found that really exciting that the fact there is,

22:52.560 --> 22:56.360
there is actually a mapping between the trading behavior of, of,

22:57.690 --> 22:59.750
of individual institutions,

22:59.750 --> 23:02.790
and their inflation forecast that just come straight from the swap,

23:02.790 --> 23:05.190
the swap market and their inflation forecast's gonna be linked together.

23:05.490 --> 23:08.310
We think that's incredibly novel. We've never seen a paper showing that before.

23:08.730 --> 23:10.350
Um, and we were very excited to see that.

Faculti AI can make mistakes

Research

Policy

Analysis