Popular in Economics

Economics

Gender and Economics

University of California Santa Barbara Shelly Lundberg

Banks and Policy Organisations

Greening the financial sector

European Investment Bank Matteo Ferrazzi

Economic History

Gravity with Gravitas: A Solution to the Border Puzzle

Boston College James E. Anderson

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

FPAS Mark II: Avoiding Dark Corners and Eliminating the Folly in Baselines and Local Approximations

View more insights in Policy Organisations - Social Sciences - Economics - Banks and Policy Organisations - Macroeconomics

View more figures from Central Bank of Armenia

View the page of Martin Galstyan

Read the Study

The major challenge facing central banks today is the inability of existing policymaking frameworks to adequately deal with significant and increasing uncertainty. Martin Galstyan (Governor of the Central Bank of Armenia) proposes a new policymaking setup “FPAS Mark II” that combines successful parts of the institutional framework of Mark I systems with an MPRM approach to dealing with uncertainty.

Image courtesy of interviewee. March 15, 2024

Generating Workspace

Export as PDF

This video video is a transcript of a discussion focused on the Central Bank of Armenia's approach to monetary policy and risk management in the context of economic challenges. The speaker emphasizes that as a small developing economy, Armenia cannot solely rely on inaction for achieving macroeconomic stability, given the multitude of risks it faces, particularly highlighted during recent crises, such as COVID-19 and geopolitical conflicts.

Key points include:

1. **Understanding and Managing Risks**: The central bank aims to develop frameworks to effectively manage risks instead of avoiding them altogether. This video video involves a shift from relying on single baseline forecasts to a scenario-based approach that allows for more nuanced understanding and communication about economic uncertainties.

2. **Enhancing Communication**: There is a focus on transparency in decision-making processes, with initiatives to publish board deliberation minutes and individual board member submissions on the same day as policy decisions. This video video is designed to provide stakeholders a clearer view of the central bank's considerations, strengthening trust and credibility.

3. **Lessons from Global Trends**: The post-COVID inflationary period and its implications on policy credibility are discussed, underscoring the need for careful communication with the public and markets to maintain central bank integrity. The speaker cites the Reserve Bank of New Zealand as a successful example in adopting flexible inflation targeting and employing a 'least regrets' policy stance during crises.

4. **Fostering Public Understanding**: There is an emphasis on economic literacy among the general public as a necessary public good, contributing to societal capacity building, which, in turn, aids in enhancing the central bank's credibility.

5. **Investments in Human Capital and Analytical Tools**: The need for skilled personnel and sophisticated analytical frameworks to support effective monetary policy is highlighted, ensuring that economists bring both analytical and narrative capabilities to their work.

Overall, the discussion advocates for an evolved approach to monetary policy that prioritizes communication, scenario-based risk management, and operational transparency to navigate uncertainties effectively.

Language:

Export as PDF

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

00:00.235 --> 00:03.095
So let me begin by briefly providing some context first.

00:03.315 --> 00:05.575
Um, as a small developing open economy

00:06.655 --> 00:09.455
situated in a very precarious corner of the world,

00:09.835 --> 00:12.775
we at the Central Bank of Armenia are always aware

00:12.775 --> 00:16.295
of the fact that we cannot rely on sheer lack

00:16.295 --> 00:18.915
or inertia to achieve the price

00:18.975 --> 00:21.115
or macroeconomic stability in the country.

00:21.575 --> 00:25.595
So risks are always present more dangerous than others.

00:25.745 --> 00:27.875
Some of them are more, more dangerous than the others,

00:27.935 --> 00:30.435
and many of these risks do in fact materialize.

00:31.655 --> 00:33.795
For us, the essence of good governance

00:33.795 --> 00:36.835
and good policymaking is the ability to weather

00:37.355 --> 00:38.875
whatever comes on our way

00:39.415 --> 00:41.235
and avoid the dark corners

00:41.235 --> 00:43.115
as Olivia Blanchard described those.

00:43.495 --> 00:46.955
So for us, the question has always been not whether,

00:47.175 --> 00:48.995
we can avoid risks altogether,

00:49.535 --> 00:52.875
but rather whether we can have the type of framework

00:53.345 --> 00:55.755
that helps us to effectively deal with

00:56.215 --> 00:57.635
and manage those risks.

00:58.775 --> 01:02.155
In other words, when we make forward looking policies,

01:02.505 --> 01:04.715
does our framework help us mitigate

01:05.295 --> 01:07.555
rather than abbreviate the risks

01:07.615 --> 01:09.155
and uncertainties that we face?

01:09.815 --> 01:11.435
So this way of thinking about the risks

01:11.435 --> 01:15.915
and uncertainty became, probably more crystallized

01:16.055 --> 01:17.355
during the covid crisis

01:18.095 --> 01:21.035
and the inflationary period that followed the covid crisis.

01:21.585 --> 01:25.235
Looking globally under particularly advanced economies,

01:25.375 --> 01:28.755
we were clearly in the first real inflationary period

01:29.055 --> 01:32.675
for decades, and there existed enormous amounts

01:32.675 --> 01:36.715
of uncertainty about what was driving it

01:36.775 --> 01:39.235
and how policy owed to respond.

01:40.715 --> 01:42.325
This created new communication

01:42.345 --> 01:45.085
and credibility challenges for advanced central banks.

01:45.565 --> 01:48.605
Problems that many thought have been put in bed since

01:49.085 --> 01:51.885
probably 1980s in advanced economies,

01:51.905 --> 01:54.805
the post covid period served as a difficult reminder

01:55.275 --> 01:58.365
that policy credibility cannot be taken for granted.

01:59.345 --> 02:00.635
That policy errors

02:00.635 --> 02:04.395
and uncertainty can create real welfare implication

02:04.905 --> 02:06.795
that ripple throughout the globe.

02:07.335 --> 02:11.275
So, in Armenia, this period made

02:12.535 --> 02:13.755
all the more challenging

02:13.755 --> 02:17.115
because of the fact that we were facing a complex

02:17.315 --> 02:18.955
layering of crises.

02:19.705 --> 02:22.195
Many of those that included not just covid,

02:22.335 --> 02:25.515
but also we had war and regional geopolitical shifts.

02:26.175 --> 02:30.395
So while we, appropriately dealt with these challenges

02:30.515 --> 02:34.835
and avoided worst case outcomes in dark corners,

02:35.755 --> 02:38.395
I think we start thinking at the same time

02:38.465 --> 02:39.955
what the best policy framework

02:39.955 --> 02:40.995
would look like in the future.

02:41.735 --> 02:44.955
So in that sense, we asked several questions to ourselves,

02:45.415 --> 02:48.395
and first would be like in the face of such uncertainty,

02:49.175 --> 02:52.395
how could one develop a single interpretation?

02:52.415 --> 02:53.955
And I want to stress the word single

02:54.145 --> 02:57.035
because most of the central banks, including ourselves,

02:57.265 --> 02:58.475
were relying on this

02:59.245 --> 03:01.985
Single forecast line, which we call baseline.

03:02.725 --> 03:04.625
So, and the forecast that future.

03:05.125 --> 03:09.585
So can we forecast the future with any degree of legitimacy?

03:09.975 --> 03:11.105
That was the first question.

03:11.565 --> 03:12.985
The second one was, what type

03:12.985 --> 03:16.985
of communication would effectively convey the richness

03:17.105 --> 03:21.145
and complexity of our policy discussions within the board

03:21.765 --> 03:25.345
and without minimizing uncertainty that we were facing?

03:25.525 --> 03:28.625
So we didn't want to give our general public

03:28.685 --> 03:31.105
or professional audience this false assurance

03:31.495 --> 03:33.905
that we are the ones who know the best.

03:34.775 --> 03:37.075
So there is, there were rich discussions

03:37.305 --> 03:39.235
with multiple scenarios in the board,

03:39.455 --> 03:42.755
and what we thought is a good idea is to replicate that

03:42.775 --> 03:44.755
and to provide this richness to the public

03:45.175 --> 03:46.915
so they can make their own decisions.

03:47.375 --> 03:50.755
And the third thing that we tried to have, have our heads

03:50.755 --> 03:54.235
around was whether we should serve.

03:54.295 --> 03:58.265
And that would be probably the, the kind of extension

03:58.285 --> 04:02.545
to problem number two, whether we should act as a

04:03.485 --> 04:05.745
expert forecaster in this case,

04:06.325 --> 04:09.865
or we should effectively manage the risks and uncertainty

04:10.125 --> 04:11.305
and be very honest.

04:11.965 --> 04:14.185
And humility would be part of that.

04:14.375 --> 04:17.665
Also telling that we are not the ones who know all the best,

04:18.285 --> 04:20.705
but these are the options that we have discussed.

04:21.125 --> 04:22.305
And this what we think

04:22.325 --> 04:24.025
how the risk management should look like.

04:24.535 --> 04:28.665
This is an analytical work which is based on,

04:30.065 --> 04:34.865
comments or approaches made by Alan Greenspan in 2004,

04:35.105 --> 04:36.425
I think, if I'm not mistaken.

04:36.765 --> 04:39.705
And he called it monetary policy as a risk management.

04:40.045 --> 04:41.385
So we take that approach

04:41.765 --> 04:44.345
and try to create an operational framework,

04:44.595 --> 04:47.225
which is based on models and everything underneath.

04:47.645 --> 04:51.385
But the basic idea was to apply to this monetary policy

04:51.485 --> 04:53.145
as a risk management framework

04:53.205 --> 04:56.345
and to enrich that the term monetary policy

04:56.365 --> 05:00.565
as risk management originated from Alan Greenspan,

05:00.865 --> 05:02.605
who implemented versions

05:02.605 --> 05:04.605
of this approach at the Federal Reserve.

05:05.075 --> 05:06.965
However, the fed's use of

05:07.685 --> 05:11.685
MPRM largely happens on the internal level,

05:12.495 --> 05:14.115
and a lot of the richness

05:14.115 --> 05:17.875
and benefits of this approach is lost when it comes

05:17.895 --> 05:19.275
to external communication.

05:19.975 --> 05:21.075
So for example,

05:21.185 --> 05:23.955
they do not immediately publish certain documents

05:24.055 --> 05:27.275
or analysis that shed light on the different risks

05:27.335 --> 05:30.915
and uncertainties being considered at the board of FYMC.

05:31.575 --> 05:35.435
Our approach under FPAs mark two, is to take a lot

05:35.435 --> 05:36.635
of the foundational

05:36.695 --> 05:40.155
and conceptual elements of NPRM approach,

05:40.775 --> 05:44.355
but augmented with significantly more transparency

05:45.645 --> 05:48.195
about how we arrive to our decision.

05:48.535 --> 05:49.555
So that is critical.

05:49.815 --> 05:53.075
So this includes not only following the very highest

05:53.435 --> 05:56.995
standards for policy communications, but also publishing

05:57.105 --> 06:00.845
Our key communication materials in virtually real time.

06:01.545 --> 06:03.965
So for example, our monetary policy report on the

06:03.965 --> 06:05.285
attributed minutes of the board.

06:06.305 --> 06:09.245
So how board deliberates on on the, the, those issues

06:09.755 --> 06:12.245
will be published on the same day as the decision is made.

06:12.545 --> 06:14.445
So we are trying hard to do that.

06:14.985 --> 06:17.565
We had just first example couple of days ago.

06:17.955 --> 06:18.965
It's extremely hard

06:18.965 --> 06:21.165
because it puts a lot of pressure on the team

06:22.075 --> 06:25.135
who does the stuff plus on the board and including myself

06:25.135 --> 06:26.975
because I need to prepare myself also

06:26.975 --> 06:28.015
for the press conference.

06:28.315 --> 06:29.815
So that takes hell of a time,

06:30.155 --> 06:33.775
but we are committed to be completely transparent.

06:34.315 --> 06:36.735
So in that sense I think that's, that's a good approach.

06:37.395 --> 06:40.415
In terms of application, I think the Federal Reserve Bank

06:40.415 --> 06:44.015
of New Zealand is probably the most relevant for us,

06:44.535 --> 06:45.615
considering their success

06:45.615 --> 06:47.935
with implementing flexible inflation targeting

06:48.275 --> 06:49.735
and least regrets policymaking.

06:50.495 --> 06:53.255
Additionally, like us, like Armenia,

06:53.885 --> 06:55.695
they are also a small open economy

06:56.155 --> 06:58.415
and despite being a small country, they managed

06:58.415 --> 07:02.055
to set the global standard for central banking practice,

07:02.545 --> 07:05.375
which is also something that we aspire to do

07:05.375 --> 07:06.735
with this new framework.

07:07.275 --> 07:09.575
One good example of Reserve Bank

07:09.575 --> 07:13.975
of New Zealand has been their use of policy of least regrets

07:14.235 --> 07:17.575
during the covid, which is also an important component

07:17.635 --> 07:19.895
of the decision making process in our framework.

07:20.035 --> 07:23.495
So in the immediate, onset of the Covid pandemic,

07:24.155 --> 07:27.935
the RBNZ communicated policy of least regrets,

07:28.895 --> 07:33.495
recognizing early on that they were knowing knowingly taking

07:34.375 --> 07:38.615
a very easy monetary policy stance in order

07:38.715 --> 07:40.335
to support demand in the economy.

07:40.335 --> 07:43.255
During the covid lockdowns, they recognized

07:43.275 --> 07:47.215
and communicated that they simply did not know the extent

07:47.555 --> 07:49.055
of the decline of the demand

07:49.995 --> 07:52.925
that would occur when the economy shut down.

07:53.435 --> 07:57.805
Therefore, early on they communicated that a clear risk

07:57.905 --> 07:59.005
of this policy is

07:59.005 --> 08:02.645
that it would potentially mean higher inflation in the

08:02.645 --> 08:04.365
future during the recovery phase,

08:05.025 --> 08:08.365
but that they were willing to take this type of risk

08:08.365 --> 08:12.445
because the risk of not doing enough to support the economy

08:12.705 --> 08:15.245
during the pandemic was unacceptable.

08:15.345 --> 08:17.205
So potential might be lost.

08:17.545 --> 08:20.445
So they were kind of contemplating between two risks

08:20.825 --> 08:23.565
and they came up and they were extremely honest in their

08:23.965 --> 08:26.325
communication, telling their general public

08:26.675 --> 08:30.805
that they clearly understand that the outcome might not be

08:30.905 --> 08:33.085
as what they wish to be,

08:33.835 --> 08:36.935
but there are more dangerous risks on the downside

08:37.225 --> 08:39.375
where the entire economy could collapse.

08:39.555 --> 08:42.095
So those are the few examples that I just want to mention

08:42.115 --> 08:44.215
to you, if I may call it innovation.

08:44.635 --> 08:47.575
The key innovation of our framework from an operational

08:47.635 --> 08:48.735
and procedural point

08:48.735 --> 08:52.575
of view is shifting the focus away from a single

08:52.955 --> 08:54.695
narrow baseline forecast

08:55.185 --> 08:57.525
To a scenario based approach as I mentioned earlier.

08:57.945 --> 09:02.005
So during each forecasting ground, our economists,

09:02.385 --> 09:04.485
who work with a board very closely,

09:04.955 --> 09:09.205
they built the best arguments, the best cases they can make

09:09.345 --> 09:12.725
for why the future path of the policy rate should be higher,

09:13.295 --> 09:17.205
which we call case a higher than what market expects

09:17.545 --> 09:19.845
or lower as we call the case B.

09:20.195 --> 09:22.085
Then what is currently priced in.

09:22.185 --> 09:25.245
So the financial markets have their own forecast,

09:25.825 --> 09:29.775
the one which is higher, we call it case A, which

09:30.475 --> 09:32.895
it means basically that we do expect

09:32.895 --> 09:34.695
that inflation will be higher for longer

09:34.755 --> 09:38.175
or something like that, or downside risks on the demand

09:38.175 --> 09:40.455
side, which would be incorporated in case B.

09:41.225 --> 09:44.485
So in practice this happens through intense discussions

09:44.485 --> 09:47.445
between the team and the board in terms of coming up

09:47.445 --> 09:50.165
with a very broad set of relevance scenarios.

09:51.145 --> 09:54.845
Um, and then from this broad set of scenarios,

09:54.905 --> 09:58.685
we choose two illustrative ones for case A

09:58.685 --> 10:02.085
and case B, which are basically used

10:02.105 --> 10:03.765
for our external communication.

10:04.505 --> 10:05.525
But discussions

10:05.545 --> 10:08.805
and decision making itself continue to be motivated

10:09.425 --> 10:11.245
by the broader set of scenarios.

10:11.625 --> 10:15.485
So though we illustrate and communicate the exemplary ones,

10:15.905 --> 10:19.365
but at the same time within the board, within the boardroom,

10:19.705 --> 10:23.325
we discuss much broader scenario base that we have at our,

10:23.625 --> 10:24.885
under our control.

10:25.305 --> 10:27.525
So this achieves a few goals for us.

10:27.695 --> 10:30.755
First, through this process, we hope to embrace

10:31.105 --> 10:34.355
what is known as adversarial collaboration.

10:34.355 --> 10:37.115
Daniel Kahneman has got a paper on that about him

10:37.115 --> 10:39.475
and his wife being working on that field.

10:39.975 --> 10:41.915
So second, with this new approach,

10:42.015 --> 10:44.075
we are much more focused on, on understanding

10:44.295 --> 10:47.795
and debating what the policy relevant risks are

10:48.145 --> 10:49.875
that could threaten price stability

10:50.215 --> 10:53.515
and how we ought to respond to them rather than

10:54.195 --> 10:56.035
structuring our policy discussion

10:56.035 --> 10:58.035
around the baseline scenario.

10:58.495 --> 11:02.275
So this is, probably two major important ones

11:02.295 --> 11:03.835
and probably the third is

11:03.835 --> 11:07.835
that the framework requires engagement within development

11:07.835 --> 11:08.915
of financial markets.

11:09.485 --> 11:12.155
Since our scenarios are prepared relative

11:12.155 --> 11:14.795
to the current market pricing in the financial markets.

11:15.215 --> 11:17.035
So we should be very clear with them

11:17.325 --> 11:20.435
where we see the future would like look like the summer

11:20.455 --> 11:25.035
of 2021 is probably almost a textbook example of,

11:25.255 --> 11:27.275
there being two possible

11:27.295 --> 11:28.755
or plausible scenarios

11:29.345 --> 11:31.355
that were very different implications

11:31.355 --> 11:32.595
from the policy perspective.

11:33.475 --> 11:35.055
Our firm perspective is

11:35.055 --> 11:38.335
that relying on such a scenario based approach during

11:38.365 --> 11:40.455
that time would have been very helpful

11:40.595 --> 11:42.415
to prepare the central bank

11:42.595 --> 11:43.655
and financial markets

11:43.795 --> 11:47.175
for higher interest rates when inflationary risks began

11:47.175 --> 11:50.255
to materialize and to have avoided some

11:50.255 --> 11:52.055
of the policy errors at that time.

11:52.715 --> 11:55.525
So in our framework, is,

11:56.595 --> 12:00.095
so we had this scenario with that the case say with a more,

12:00.835 --> 12:04.735
sticky kind of inflation where interest rates would need

12:04.735 --> 12:06.815
to be higher than market expectations.

12:07.115 --> 12:11.375
So this echoed views expressed by many in the field,

12:11.375 --> 12:15.935
including our, one of our most famous ones, Larry Summers,

12:16.395 --> 12:18.535
who argued that the combined fiscal

12:18.595 --> 12:20.615
and monetary stimulus was too big

12:21.235 --> 12:23.615
and represented a grave economic error.

12:24.235 --> 12:27.895
In particular, the primary concern that this camp voice was

12:27.895 --> 12:31.695
that aggregate demand was already pushing up against

12:31.695 --> 12:32.695
aggregate supply

12:33.275 --> 12:35.575
and thus translating into higher inflation,

12:36.385 --> 12:39.775
which would be exacerbated by such high stimulus

12:40.365 --> 12:42.575
once the economy fully reopened.

12:43.035 --> 12:44.335
And we somehow agreed,

12:44.475 --> 12:46.295
but there is also the second part,

12:46.715 --> 12:49.655
so the other scenario which is inflation is

12:49.655 --> 12:50.815
transitory in us.

12:51.685 --> 12:54.335
That scenario, this is similar

12:54.595 --> 12:58.295
to probably other arguments voiced at that time that

12:58.435 --> 13:00.255
supply side factors connected

13:00.275 --> 13:03.015
to the covid lockdowns were the main contributor

13:03.015 --> 13:04.135
to the rise of inflation.

13:04.805 --> 13:08.055
Therefore, given that was assumed to be a temporary nature

13:08.515 --> 13:10.775
of the supply disruptions, inflation expected

13:10.955 --> 13:14.735
to peter out over coming quarters, so it will come back.

13:15.205 --> 13:19.895
Thus the PR policy rate could be normalized in line

13:19.895 --> 13:21.495
with still recovering labor market.

13:22.405 --> 13:27.265
And then what we were telling through our framework is

13:27.265 --> 13:28.545
that if you go out

13:28.565 --> 13:31.185
and communicate with the markets both of these scenarios

13:31.945 --> 13:34.925
and trying to tell them the richness of the story,

13:35.755 --> 13:39.925
that would've preserved your credibility much more than

13:39.945 --> 13:43.805
trying to compress everything this diversity into one

13:44.245 --> 13:47.325
baseline, which was believing in transitory

13:47.785 --> 13:49.125
and somehow being wrong

13:49.465 --> 13:52.045
and then somehow risking your own credibility

13:52.265 --> 13:55.085
and tying your hands in terms of policy action.

13:55.155 --> 13:57.245
Because previously you over

13:57.265 --> 14:00.365
and over have repeated that this is transitory phenomenon

14:00.905 --> 14:05.405
and then when you somehow betrayed your first inception,

14:05.875 --> 14:09.005
then it's for you hard to go and explain how wrong you were.

14:09.305 --> 14:12.605
So let me begin by mentioning that FPA mark one

14:13.325 --> 14:14.645
I think was an important

14:14.785 --> 14:16.085
and necessary first step

14:16.225 --> 14:20.285
for developing the analytical capacity in our own bank,

14:20.925 --> 14:22.325
I think in other banks as well.

14:23.145 --> 14:27.335
And, there are many core aspects of FPAs mark one

14:27.335 --> 14:29.055
that have stood the test of the time

14:29.595 --> 14:31.895
and owed to be preserved and augmented.

14:31.975 --> 14:35.935
I think with that being said, the past few years have

14:35.995 --> 14:39.255
of course drawn attention to several drawbacks related

14:39.255 --> 14:40.495
to FPAs mark one,

14:40.835 --> 14:42.695
and that will touch upon just a few of them.

14:43.565 --> 14:45.745
So the current approach of monetary policy,

14:45.845 --> 14:47.785
the current approaches of the monetary policy, and

14:47.785 --> 14:50.785
therefore the communication strategies critically

14:50.795 --> 14:53.005
Hinge on the ability of the central bank

14:53.585 --> 14:56.645
to project the future and design optimal policy responses.

14:57.105 --> 14:58.125
Now, that is a problem

14:58.555 --> 15:01.285
because we can easily see that it,

15:01.465 --> 15:05.885
it is essentially impossible to make a forecast in times

15:05.905 --> 15:07.565
of critical uncertainty.

15:08.275 --> 15:12.335
So moreover, dropping such projections in the close

15:12.355 --> 15:15.415
of assurance and confidence has the potential

15:15.415 --> 15:17.255
to mislead both policy makers

15:17.595 --> 15:20.735
and economic agents by providing this false sense

15:20.735 --> 15:25.115
of security when we are planning at the same time

15:25.255 --> 15:28.715
and as much as the central bank tries to add disclaimers

15:28.775 --> 15:31.035
to it, projections, the mismatch

15:31.035 --> 15:32.555
between baseline projections

15:32.575 --> 15:36.195
and what, what actually materializes can have very real

15:36.255 --> 15:39.635
impact on credibility as we were discussing in the case

15:39.635 --> 15:40.635
of 2021.

15:41.065 --> 15:45.475
Finally relying on baseline single forecast that's,

15:45.545 --> 15:49.715
somehow a narrow vision, how risks will materialize

15:49.715 --> 15:52.995
and what will happen in the future can create huge problems

15:52.995 --> 15:54.435
in terms of policy agility

15:55.375 --> 15:58.635
and unnecessary type policymaker hands.

15:59.695 --> 16:02.315
So, another thing that we have noticed

16:02.375 --> 16:04.035
and I would like to highlight is that

16:04.815 --> 16:08.595
FPA mark one has a tendency to be a model centric approach

16:08.895 --> 16:10.475
to monetary policy analysis.

16:11.355 --> 16:15.035
Although models are very important component of,

16:15.535 --> 16:19.275
any solid framework, I have nothing against,

16:19.565 --> 16:21.645
solid approach to that.

16:21.945 --> 16:23.845
But we do have a rule at the central bank

16:23.865 --> 16:27.685
and the rule is that our economist never say model says,

16:29.035 --> 16:30.495
so it's their own judgment.

16:30.605 --> 16:32.415
They can use models as tools,

16:32.875 --> 16:35.255
but that that would be the economist own judgment.

16:36.205 --> 16:40.105
So, the greatest issue that we see here is that,

16:40.695 --> 16:42.385
over dependency on models

16:42.485 --> 16:46.145
and their outputs rather than underlying economics

16:46.245 --> 16:49.345
or quality of the e economists running the models

16:49.645 --> 16:52.665
and conducting analysis is, is a real problem.

16:53.125 --> 16:57.265
So we decided that we need to equip our own economists

16:57.775 --> 16:59.345
with tools necessary,

16:59.885 --> 17:02.105
but at the same time there should be a solid

17:03.185 --> 17:04.865
economic judgment behind everything.

17:05.165 --> 17:06.225
And probably you remember

17:06.225 --> 17:09.745
what Stanley Fisher said about Bob Solo is, do you know

17:09.745 --> 17:12.545
that phrase he used during one of his set?

17:12.565 --> 17:15.745
He said, I would rather have Bob solo than an

17:15.745 --> 17:18.825
econometric model, but I would rather have Bob solo

17:18.855 --> 17:21.545
with an econometric model than without one.

17:22.005 --> 17:23.385
So that would be my short answer.

17:24.045 --> 17:27.705
So our robust communication framework is arguably the most

17:27.705 --> 17:29.585
important aspect of this new framework.

17:29.605 --> 17:32.945
So we'll put a lot of emphasis on communication, so

17:32.975 --> 17:35.265
that is very important from external point of view,

17:35.885 --> 17:37.105
our framework is first

17:37.245 --> 17:41.505
and foremost built upon the need of transparently clearly

17:42.045 --> 17:44.505
and effectively convey to our key stakeholders

17:45.005 --> 17:47.465
how the central bank is thinking about key sources

17:47.565 --> 17:48.825
of risk and uncertainty.

17:49.575 --> 17:52.805
Using this idea as our main motivator,

17:52.905 --> 17:55.605
we built a comprehensive framework for communications,

17:55.885 --> 18:00.165
transparency, accountability, which seek to achieve,

18:00.645 --> 18:02.445
probably six fundamental principles.

18:02.445 --> 18:04.725
Principle number one would be to provide clarity

18:05.385 --> 18:07.085
on monetary policy objectives.

18:08.345 --> 18:10.845
Number two, offer transparency about decision

18:10.845 --> 18:12.205
making process itself.

18:12.995 --> 18:16.265
Three would be ensure policy agility in the face of risk,

18:16.335 --> 18:20.185
uncertainty and dynamically evolving economic conditions.

18:20.765 --> 18:22.025
Fourth, provide markets

18:22.205 --> 18:26.065
and stakeholders with a combination of qualitative

18:26.445 --> 18:28.185
and quantitative forward guidance.

18:28.845 --> 18:30.145
Number five, support

18:30.165 --> 18:32.145
and strengthen central bank credibility.

18:32.645 --> 18:35.905
And number six, provide full transparency about analytics

18:36.405 --> 18:40.225
and data and the toolkit that the central bank uses

18:40.225 --> 18:41.265
to conduct the analysis

18:41.265 --> 18:42.785
and the support of its own decisions

18:43.365 --> 18:44.865
to achieve this principles.

18:45.205 --> 18:48.825
Um, we recognize that we have numerous different groups

18:48.885 --> 18:50.945
of stakeholders that serve

18:51.085 --> 18:53.705
as our target audiences in our communication.

18:54.445 --> 18:57.025
Uh, I think those group would include the general public

18:57.025 --> 18:59.345
first and foremost, but also medium journalists,

18:59.655 --> 19:01.825
financial markets specialist, academia,

19:02.235 --> 19:04.225
government ifi and so on.

19:04.525 --> 19:06.065
So of course the target audience

19:06.065 --> 19:09.865
for monetary policy communications I di are very diverse

19:09.925 --> 19:10.945
and wide ranging.

19:11.445 --> 19:14.625
So which means that we need to work,

19:14.695 --> 19:17.225
also on economic literacy for the general public,

19:17.475 --> 19:20.505
which is from our perspective is providing the social

19:20.805 --> 19:22.785
or public good in that sense.

19:23.125 --> 19:25.985
But at the same time we have different approaches

19:26.005 --> 19:29.745
for different layers and groups how to communicate.

19:30.205 --> 19:33.105
So under FPA mark two, we place a very high degree

19:33.205 --> 19:36.145
of emphasis on being highly transparent about

19:36.765 --> 19:38.305
how we're thinking about the economy

19:38.565 --> 19:40.945
and how we arrive to our decision.

19:41.985 --> 19:44.805
In support of this, we have began publishing attributed

19:44.805 --> 19:47.005
minutes of the deliberation

19:47.005 --> 19:49.805
that led the board's decision on the same day

19:50.185 --> 19:51.845
as decision itself is published.

19:52.105 --> 19:53.285
We never did that before.

19:54.115 --> 19:55.565
This allows our stakeholders

19:55.625 --> 19:59.725
to understand in almost real time the dynamic

20:00.525 --> 20:02.725
surrounding in the key policy relevant issues.

20:03.255 --> 20:05.645
We'll be publishing attributed votes as well

20:05.985 --> 20:08.245
and the individual submissions of each

20:08.245 --> 20:09.725
and single board member.

20:10.395 --> 20:12.605
This means that readers will get to see

20:13.305 --> 20:15.365
on decision day the concerns

20:15.365 --> 20:17.565
that motivated each board member

20:18.145 --> 20:20.525
and thus have a much richer, more transparent

20:20.785 --> 20:23.165
and more comprehensive understanding of

20:23.195 --> 20:26.285
what the key issues which are driving our decision.

20:26.865 --> 20:28.485
But at this point I think we need

20:28.485 --> 20:30.005
to have systematic attention

20:30.105 --> 20:31.885
to policy risks and uncertainties.

20:32.225 --> 20:36.205
So the primary challenge facing the central bank is

20:37.225 --> 20:41.165
the inadequate inadequacy in terms of

20:41.885 --> 20:42.925
existing policy framework.

20:43.025 --> 20:46.885
So we think that we should be leading that, that work. The

20:46.885 --> 20:50.245
Second one would be huge investment in human capital.

20:51.065 --> 20:53.765
So we need to have a people of that caliber

20:54.305 --> 20:56.525
who would use analytical tools

20:56.865 --> 20:58.845
but at the same time would come to us

20:59.185 --> 21:00.685
and tell us the narratives as well.

21:01.105 --> 21:04.565
So that is, that is a very com very nice combination of

21:05.105 --> 21:08.245
skills which we're working on with, with our own people.

21:08.955 --> 21:11.965
Then the third will be probably investing in Ali

21:11.965 --> 21:15.085
analytical framework and model models themselves.

21:15.585 --> 21:19.485
So making them better, to include

21:20.215 --> 21:21.405
basic nonlinearities

21:21.465 --> 21:23.845
but also something which is much more complicated.

21:25.705 --> 21:27.525
Fourth will be probably transparency

21:27.585 --> 21:31.725
and accountability in terms of us being accountable

21:31.825 --> 21:33.085
to our general public.

21:33.625 --> 21:36.525
And, throughout you you see the red line

21:36.525 --> 21:39.085
and the way of thinking from our side is

21:39.085 --> 21:43.095
that our public good is to provide people with

21:43.355 --> 21:46.495
as much information so they understand better

21:46.635 --> 21:47.775
how we make decisions.

21:48.595 --> 21:50.855
And probably related to

21:50.855 --> 21:53.535
that also will be societal level capacity building

21:53.725 --> 21:57.535
because by providing that impetus, we're also building,

21:57.535 --> 22:00.095
financial literacy blocks within the society, which

22:00.615 --> 22:03.455
probably would help to build credibility with us.

22:03.875 --> 22:06.615
And having people believing in central bank's policy

22:07.575 --> 22:10.655
probably would be the best reward that we can accept,

22:10.685 --> 22:13.095
like expect the trophy of the day.

00:00 - So let me begin by briefly providing some context first. as a small developing open economy situated in a very precarious corner of the world, we at the Central Bank of Armenia are always aware of the fact that we cannot rely on sheer lack or inertia to achieve the price

00:18 - or macroeconomic stability in the country. So risks are always present more dangerous than others. Some of them are more, more dangerous than the others, and many of these risks do in fact materialize. For us, the essence of good governance and good policymaking is the ability to weather whatever comes on our way and avoid the dark corners as Olivia Blanchard described those. So for us, the question has always been not whether,

00:47 - we can avoid risks altogether, but rather whether we can have the type of framework that helps us to effectively deal with and manage those risks. In other words, when we make forward looking policies, does our framework help us mitigate rather than abbreviate the risks and uncertainties that we face? So this way of thinking about the risks and uncertainty became, probably more crystallized

01:16 - during the covid crisis and the inflationary period that followed the covid crisis. Looking globally under particularly advanced economies, we were clearly in the first real inflationary period for decades, and there existed enormous amounts of uncertainty about what was driving it and how policy owed to respond. This created new communication and credibility challenges for advanced central banks. Problems that many thought have been put in bed since

01:49 - probably 1980s in advanced economies, the post covid period served as a difficult reminder that policy credibility cannot be taken for granted. That policy errors and uncertainty can create real welfare implication that ripple throughout the globe. So, in Armenia, this period made all the more challenging because of the fact that we were facing a complex layering of crises.

02:19 - Many of those that included not just covid, but also we had war and regional geopolitical shifts. So while we, appropriately dealt with these challenges and avoided worst case outcomes in dark corners, I think we start thinking at the same time what the best policy framework would look like in the future. So in that sense, we asked several questions to ourselves, and first would be like in the face of such uncertainty, how could one develop a single interpretation?

02:52 - And I want to stress the word single because most of the central banks, including ourselves, were relying on this Single forecast line, which we call baseline. So, and the forecast that future. So can we forecast the future with any degree of legitimacy? That was the first question. The second one was, what type of communication would effectively convey the richness and complexity of our policy discussions within the board

03:21 - and without minimizing uncertainty that we were facing? So we didn't want to give our general public or professional audience this false assurance that we are the ones who know the best. So there is, there were rich discussions with multiple scenarios in the board, and what we thought is a good idea is to replicate that and to provide this richness to the public so they can make their own decisions. And the third thing that we tried to have, have our heads

03:50 - around was whether we should serve. And that would be probably the, the kind of extension to problem number two, whether we should act as a expert forecaster in this case, or we should effectively manage the risks and uncertainty and be very honest. And humility would be part of that. Also telling that we are not the ones who know all the best, but these are the options that we have discussed. And this what we think

04:22 - how the risk management should look like. This is an analytical work which is based on, comments or approaches made by Alan Greenspan in 2004, I think, if I'm not mistaken. And he called it monetary policy as a risk management. So we take that approach and try to create an operational framework, which is based on models and everything underneath. But the basic idea was to apply to this monetary policy as a risk management framework

04:53 - and to enrich that the term monetary policy as risk management originated from Alan Greenspan, who implemented versions of this approach at the Federal Reserve. However, the fed's use of MPRM largely happens on the internal level, and a lot of the richness and benefits of this approach is lost when it comes to external communication. So for example,

05:21 - they do not immediately publish certain documents or analysis that shed light on the different risks and uncertainties being considered at the board of FYMC. Our approach under FPAs mark two, is to take a lot of the foundational and conceptual elements of NPRM approach, but augmented with significantly more transparency about how we arrive to our decision. So that is critical. So this includes not only following the very highest

05:53 - standards for policy communications, but also publishing Our key communication materials in virtually real time. So for example, our monetary policy report on the attributed minutes of the board. So how board deliberates on on the, the, those issues will be published on the same day as the decision is made. So we are trying hard to do that. We had just first example couple of days ago. It's extremely hard because it puts a lot of pressure on the team

06:22 - who does the stuff plus on the board and including myself because I need to prepare myself also for the press conference. So that takes hell of a time, but we are committed to be completely transparent. So in that sense I think that's, that's a good approach. In terms of application, I think the Federal Reserve Bank of New Zealand is probably the most relevant for us, considering their success with implementing flexible inflation targeting

06:48 - and least regrets policymaking. Additionally, like us, like Armenia, they are also a small open economy and despite being a small country, they managed to set the global standard for central banking practice, which is also something that we aspire to do with this new framework. One good example of Reserve Bank of New Zealand has been their use of policy of least regrets during the covid, which is also an important component

07:17 - of the decision making process in our framework. So in the immediate, onset of the Covid pandemic, the RBNZ communicated policy of least regrets, recognizing early on that they were knowing knowingly taking a very easy monetary policy stance in order to support demand in the economy. During the covid lockdowns, they recognized and communicated that they simply did not know the extent of the decline of the demand that would occur when the economy shut down.

07:53 - Therefore, early on they communicated that a clear risk of this policy is that it would potentially mean higher inflation in the future during the recovery phase, but that they were willing to take this type of risk because the risk of not doing enough to support the economy during the pandemic was unacceptable. So potential might be lost. So they were kind of contemplating between two risks and they came up and they were extremely honest in their

08:23 - communication, telling their general public that they clearly understand that the outcome might not be as what they wish to be, but there are more dangerous risks on the downside where the entire economy could collapse. So those are the few examples that I just want to mention to you, if I may call it innovation. The key innovation of our framework from an operational and procedural point of view is shifting the focus away from a single

08:52 - narrow baseline forecast To a scenario based approach as I mentioned earlier. So during each forecasting ground, our economists, who work with a board very closely, they built the best arguments, the best cases they can make for why the future path of the policy rate should be higher, which we call case a higher than what market expects or lower as we call the case B. Then what is currently priced in. So the financial markets have their own forecast,

09:25 - the one which is higher, we call it case A, which it means basically that we do expect that inflation will be higher for longer or something like that, or downside risks on the demand side, which would be incorporated in case B. So in practice this happens through intense discussions between the team and the board in terms of coming up with a very broad set of relevance scenarios. and then from this broad set of scenarios, we choose two illustrative ones for case A

09:58 - and case B, which are basically used for our external communication. But discussions and decision making itself continue to be motivated by the broader set of scenarios. So though we illustrate and communicate the exemplary ones, but at the same time within the board, within the boardroom, we discuss much broader scenario base that we have at our, under our control. So this achieves a few goals for us.

10:27 - First, through this process, we hope to embrace what is known as adversarial collaboration. Daniel Kahneman has got a paper on that about him and his wife being working on that field. So second, with this new approach, we are much more focused on, on understanding and debating what the policy relevant risks are that could threaten price stability and how we ought to respond to them rather than structuring our policy discussion

10:56 - around the baseline scenario. So this is, probably two major important ones and probably the third is that the framework requires engagement within development of financial markets. Since our scenarios are prepared relative to the current market pricing in the financial markets. So we should be very clear with them where we see the future would like look like the summer of 2021 is probably almost a textbook example of,

11:25 - there being two possible or plausible scenarios that were very different implications from the policy perspective. Our firm perspective is that relying on such a scenario based approach during that time would have been very helpful to prepare the central bank and financial markets for higher interest rates when inflationary risks began

11:47 - to materialize and to have avoided some of the policy errors at that time. So in our framework, is, so we had this scenario with that the case say with a more, sticky kind of inflation where interest rates would need to be higher than market expectations. So this echoed views expressed by many in the field, including our, one of our most famous ones, Larry Summers, who argued that the combined fiscal and monetary stimulus was too big

12:21 - and represented a grave economic error. In particular, the primary concern that this camp voice was that aggregate demand was already pushing up against aggregate supply and thus translating into higher inflation, which would be exacerbated by such high stimulus once the economy fully reopened. And we somehow agreed, but there is also the second part, so the other scenario which is inflation is

12:49 - transitory in us. That scenario, this is similar to probably other arguments voiced at that time that supply side factors connected to the covid lockdowns were the main contributor to the rise of inflation. Therefore, given that was assumed to be a temporary nature of the supply disruptions, inflation expected to peter out over coming quarters, so it will come back. Thus the PR policy rate could be normalized in line

13:19 - with still recovering labor market. And then what we were telling through our framework is that if you go out and communicate with the markets both of these scenarios and trying to tell them the richness of the story, that would've preserved your credibility much more than trying to compress everything this diversity into one baseline, which was believing in transitory and somehow being wrong and then somehow risking your own credibility

13:52 - and tying your hands in terms of policy action. Because previously you over and over have repeated that this is transitory phenomenon and then when you somehow betrayed your first inception, then it's for you hard to go and explain how wrong you were. So let me begin by mentioning that FPA mark one I think was an important and necessary first step for developing the analytical capacity in our own bank, I think in other banks as well.

14:23 - And, there are many core aspects of FPAs mark one that have stood the test of the time and owed to be preserved and augmented. I think with that being said, the past few years have of course drawn attention to several drawbacks related to FPAs mark one, and that will touch upon just a few of them. So the current approach of monetary policy, the current approaches of the monetary policy, and therefore the communication strategies critically

14:50 - Hinge on the ability of the central bank to project the future and design optimal policy responses. Now, that is a problem because we can easily see that it, it is essentially impossible to make a forecast in times of critical uncertainty. So moreover, dropping such projections in the close of assurance and confidence has the potential to mislead both policy makers and economic agents by providing this false sense

15:20 - of security when we are planning at the same time and as much as the central bank tries to add disclaimers to it, projections, the mismatch between baseline projections and what, what actually materializes can have very real impact on credibility as we were discussing in the case of 2021. Finally relying on baseline single forecast that's, somehow a narrow vision, how risks will materialize and what will happen in the future can create huge problems

15:52 - in terms of policy agility and unnecessary type policymaker hands. So, another thing that we have noticed and I would like to highlight is that FPA mark one has a tendency to be a model centric approach to monetary policy analysis. Although models are very important component of, any solid framework, I have nothing against, solid approach to that. But we do have a rule at the central bank

16:23 - and the rule is that our economist never say model says, so it's their own judgment. They can use models as tools, but that that would be the economist own judgment. So, the greatest issue that we see here is that, over dependency on models and their outputs rather than underlying economics or quality of the e economists running the models and conducting analysis is, is a real problem. So we decided that we need to equip our own economists

16:57 - with tools necessary, but at the same time there should be a solid economic judgment behind everything. And probably you remember what Stanley Fisher said about Bob Solo is, do you know that phrase he used during one of his set? He said, I would rather have Bob solo than an econometric model, but I would rather have Bob solo with an econometric model than without one. So that would be my short answer.

17:24 - So our robust communication framework is arguably the most important aspect of this new framework. So we'll put a lot of emphasis on communication, so that is very important from external point of view, our framework is first and foremost built upon the need of transparently clearly and effectively convey to our key stakeholders how the central bank is thinking about key sources of risk and uncertainty. Using this idea as our main motivator,

17:52 - we built a comprehensive framework for communications, transparency, accountability, which seek to achieve, probably six fundamental principles. Principle number one would be to provide clarity on monetary policy objectives. Number two, offer transparency about decision making process itself. Three would be ensure policy agility in the face of risk, uncertainty and dynamically evolving economic conditions. Fourth, provide markets

18:22 - and stakeholders with a combination of qualitative and quantitative forward guidance. Number five, support and strengthen central bank credibility. And number six, provide full transparency about analytics and data and the toolkit that the central bank uses to conduct the analysis and the support of its own decisions to achieve this principles. we recognize that we have numerous different groups

18:48 - of stakeholders that serve as our target audiences in our communication. I think those group would include the general public first and foremost, but also medium journalists, financial markets specialist, academia, government ifi and so on. So of course the target audience for monetary policy communications I di are very diverse and wide ranging. So which means that we need to work,

19:14 - also on economic literacy for the general public, which is from our perspective is providing the social or public good in that sense. But at the same time we have different approaches for different layers and groups how to communicate. So under FPA mark two, we place a very high degree of emphasis on being highly transparent about how we're thinking about the economy and how we arrive to our decision. In support of this, we have began publishing attributed

19:44 - minutes of the deliberation that led the board's decision on the same day as decision itself is published. We never did that before. This allows our stakeholders to understand in almost real time the dynamic surrounding in the key policy relevant issues. We'll be publishing attributed votes as well and the individual submissions of each and single board member.

20:10 - This means that readers will get to see on decision day the concerns that motivated each board member and thus have a much richer, more transparent and more comprehensive understanding of what the key issues which are driving our decision. But at this point I think we need to have systematic attention to policy risks and uncertainties. So the primary challenge facing the central bank is

20:37 - the inadequate inadequacy in terms of existing policy framework. So we think that we should be leading that, that work. The Second one would be huge investment in human capital. So we need to have a people of that caliber who would use analytical tools but at the same time would come to us and tell us the narratives as well. So that is, that is a very com very nice combination of skills which we're working on with, with our own people.

21:08 - Then the third will be probably investing in Ali analytical framework and model models themselves. So making them better, to include basic nonlinearities but also something which is much more complicated. Fourth will be probably transparency and accountability in terms of us being accountable to our general public. And, throughout you you see the red line and the way of thinking from our side is

21:39 - that our public good is to provide people with as much information so they understand better how we make decisions. And probably related to that also will be societal level capacity building because by providing that impetus, we're also building, financial literacy blocks within the society, which probably would help to build credibility with us. And having people believing in central bank's policy probably would be the best reward that we can accept,

22:10 - like expect the trophy of the day.

Faculti AI can make mistakes