The role of
Policy and Financial Regulation
FSA
Congress on Ideas for the Future
16 April
2015
Amsterdam,
Concertgebouw
Dear
Colleagues,
·
Let me start by
thanking you for inviting me to your event. It is at once a very natural event
for somebody like me to attend given my background and interests, but also a very unorthodox one.
And I am not just saying this because some presentations today involved the use
of swords! I am saying this because it is not often that workshops on finance
and economics give morality and ethics
centre stage. I am very impressed that you are taking the time to do that,
and I am certainly in agreement with recalibrating our financial and economic
moral compass.
·
Disclaimer: I
would like to first repeat that while I am formally an employee of both De
Nederlandsche Bank, and currently on secondment at the European Commission in
Brussels, what I will say today are strictly my own views. I would quite like
it for them to become official views but perhaps not quite yet! This is not to
say that neither of the institutions are involved in thinking about at least
some of the issues discussed here. Quite the opposite. But policy making is the
art of compromise and it is impossible to please everybody all the time.
What do banks do?
·
Let me start
with a very simple question. What do
banks do? And more importantly how close is that to what we actually want
them to do?
·
If you open any
textbook on the issue you will find that the role of banks is to finance growth.
In practice what that means, is that they
intermediate between those that have money and those who know what to do with
it. Put another way, they take the money from households (your and my
deposits) and lend it out to investors. The idea is that by exploiting
specialists' knowledge, the money that is intermediated creates values.
Households earn interest; investors are given funds to pursue their ideas and
banks make profits in the process. More importantly the economy works, and
everybody, even those who are not directly involved in this process, win!
Everybody, provided that investments live up to their promise of producing
value. If investments fail, banks are proven wrong in their choices and all
parties involved lose out. This is pretty much what happened in the very recent
past that gave rise to what we all call the financial crisis.
What did we get wrong?
·
What I would
like then to discuss today is what went wrong. What did we miss and could we have done things otherwise? Let me
explain that when I say "we", I certainly mean policy makers, but not
only. I believe that the scale of the crisis implies that no single group of
relevant parties alone, could have been the sole contributor to events. I
personally believe that it is policy makers, the economics profession in its
entirety but also to a very big extent society at large for tolerating
practices that we now know were wrong. While certainly at "liberty"
of making policy mistakes, I simply don't believe that policy makers can force
changes to a society that is not ready, or prepared to have them. I will come
back to that!
·
Irrespective of
what you think has caused the crisis or who is to blame for it, the one thing
that most of us do agree on is that banks
were in the middle of it. I will argue that two things were certainly
wrong. Perhaps more, but two things need in my view to change: our attitude to
risk, uncertainty, and the tools that we use to decide on policy.
Can we measure risk?
·
Most of the narratives on
the crisis tend to agree that at the very start of the crisis is a slowdown in
the US economy. Overextended households found themselves incapable of paying
their mortgages, and banks all over the world with investments linked to those
mortgages started losing money.
·
Then came the collapse of
Lehman brothers, America’s fourth largest investment bank, a fact that sent shock
waves across the whole world. There was a real fear that more banks could fail and
as a result everyone became reluctant to invest. Banks stopped lending to each
other, and the world economy case to a sudden stop. The rest is history.
·
Looking back at this the
fact is that banks failed to understand the riskiness of
investments they were making. This was manifested in the fact that they had
not priced risk correctly, nor had they appreciated the fact that securitising
risk, i.e. cutting it up in small pieces and spreading out around the world,
did not necessarily make it disappear! We know now that on the contrary, it
actually made the whole system more vulnerable as everybody became exposed to
the same risks.
·
So at the heart of the
matter lies the question of risk, how we measure it and how we actually prepare
for it. Given the cost that it generated when risk unravelled, can we rely on
prices to do the job, or is that sense of security simply false?
·
Let me give you an example.
Typically, when we forecast, we present two things: a mean forecast, in other
words our best guess of what will happen, and confidence bands around it to
represent probabilities that the actual variables will land somewhere around
that mean forecast. These confidence banks also tell us that there
is a trade-off between the width of
the range and the probability with which the forecast will land within this
range. The more imprecise the forecast (the wider the range) the more confident
we are about the outcome. Most institutes present the 90% confidence intervals:
this means that the statement that the mean forecast will land in the interval
described, has a 90% chance of being true. My question to you is how happy are
you that this is reliable? And what does it mean that it is 90% reliable?
·
If you look at our current forecasts
for growth that we make, our 90% confidence intervals are of the order of
magnitude of 4-5%. By means of an example, the latest available European
Commission forecast about euro area growth for 2016 is around 2%, and the 90% confidence
interval takes us from around -0.5% to around 4%. Therefore, the statement that
growth will be between -0.5 and 4% in 2016 has a 90% probability of being true.
This is another way of saying how confident we are in such an outcome.
·
If you look at the
forecasts we were making back in 2007, the 90% confidence intervals were a lot
tighter. We were a lot more certain about what we thought back then would
happen at the same forecast horizon. Not only were we wrong in our forecasts
(given the extent of the revisions), but we were a lot more confident in our
ignorance! And I can assure you, the whole world was caught in this lack of
knowledge. But my point is, how
do we know that we are not in the same position right now? Can we be sure (or
even just 90% sure) that growth in 2016 will indeed be between -0.5% and 4%?
Can we rely on these numbers to design policy?
·
Of course, you could accuse me of picking the
most difficult point in time to demonstrate my concern. If you were to look at
other times, you wouldn't see equal revisions and indeed the 90% confidence
interval would have served us pretty well.
How could we have done it differently? But to this I will say, that it
is precisely at these difficult and exceptional moments that we need accurate and precise forecasts so that we
can prepare for them. Unfortunately, it is exactly then, when we need them
the most, that we cannot rely on conventional methods to provide us with
confidence on how best to prepare for events to come.
·
In my view, the
lesson that the 2007 exercise has taught us is that even though models do serve
us satisfactorily most of the times, there will be times that they will fail
us, and they may even fail us spectacularly. It is in these occasions that
probabilities do not provide any confidence on how certain outcomes will occur.
Relying on them provides a false sense
of security that can lead to wrong policy choices.
Models and the way we use them
·
And that brings me to
models themselves, tools. Economics and
finance have in the past 20 to 30 years become increasingly mathematical and complicated. This is good
because quantitative tools provide a consistent approach to argue a case. However,
this has also pushed us to the comfort of believing that just because models produce
numbers, these numbers are correct! And the tendency has been for models to
become bigger and bigger at the expense of tractability. So we have to decide
between big dimension models that are basically black boxes, where nobody
understands what goes on, or small models that are simple thought experiments but
which may lack in realism. So my first objection is that although I am big
proponent of models, I also believe that models are there to be used but not
necessarily believed, and certainly not all the time. Competent policy makers
will know how and when to make this distinction.
·
But beyond that
I also believe that we have made a number of wrong choices in terms of which
models to use:
·
First we need to
understand that probabilities are
nothing more than frequencies of past events. In that respect they may not be a
good measurement of things to come.
·
Second, if you
look at the way that macroeconomics and finance have evolved in the past 20
years, you would have thought that they are parts of two parallel universes, with no influence on each other. Banks do
not affect the macro environment and macro policy (taxes, interest rates) do
not affect bank decisions. I may be exaggerating a little but only just a
little. Think of your macroeconomics class and ask yourself whether the word banks
featured in any shape or form. With a considerable degree of embarrassment, and
perhaps a little too late for Lehman brothers, the profession is now attempting
to change the way we see that, starting with economics curricula.
·
Third, most of economic and
finance education since the 70s, relied on the premise of rational expectations. Our beliefs about the future are on average
correct, markets are efficient and while individually we may be mistaken in the
decisions we make, collectively we cannot be and certainly not for long periods
of time. Let me use a metaphor to illustrate how restrictive that is. Imagine
that you are an archer and your objective is to hit your arrow at a given
target. Normally what you would do is you would stand behind a starting line
and make a quick calculation of the distance and forecast wind, keeping your
eye at the mark. Rational expectations is the equivalent of tying your bow to a
string, then running to the other side where the target is, holding that string
in your hand. Once at the other end all you do is simply pull the string. Your
only real calculation in then how long your string should be!
·
To be fair, lots of people do
appreciate the limitations of this approach and that the rational expectations
"straightjacket" may not
be an accurate representation of the way we should approach our problems.
However, fear of the alternative, what we call the "wilderness" of bounded rationality, has led us to simply ignore
it. We simply continue to only want to examine how long the string should be
and do not worry about the possible winds that might hit or indeed consider the
possibility of missing the target!
·
Last, one of the
characteristics of modern modelling is in the way that we treat change. Change
is never big and is never sudden. In that respect we cannot allow for sudden changes of sentiment, or
behaviour of the type we saw in 2007.We typically allow for linear progressions
when the interesting and costly problems that we need to know about, are highly
non-linear. Slowly, we are beginning
to appreciate that.
Building robust systems
·
So that leaves
with the question of how to move forward.
In case my previous comments were not suggestive enough, here is what I think we
need to change to provide for robust systems.
·
Use
probabilities to describe history. This helps understands both what occurs more
often and what are, for all intends and purposes, rare events. However, when we
plan for the future, designing policies to
address what has occurred the most often is not sufficient. We need to also prepare for rare but
potentially catastrophic events. Imagine if dykes in the Netherlands were
designed for the average yearly rainfall! We build our cities with that in
mind, but we also put a system in place that will save us in the once-in-a-thousand-year
storm. Bank capital requirements and our macroeconomic structure needs to have built-in buffers that will allow the
system to function well in normal times, but will also prevent the system from
collapsing when rare events happen. In this respect what we mean with "rare"
here does not matter. What matters is what type of catastrophe we want to be
prepared for, irrespective of how rarely it happens. It is in this respect that
I believe that we should de-emphasise
the likelihoods with which certain events happen and emphasise the impact they
have when they do occur. Then prepare for them!
·
When it comes
down to our tools, in my view the models that we use need to be both less and more ambitious. Less
ambitious in terms of providing us with clear and elegant answers but also more
ambitious in terms of coming closer to the real and urgent problems that we
face. We cannot limit the questions that we ask as policy makers by the ability
of our models to answer them. I would urge you to use models in your future
work as economic professionals because they give you the discipline to follow a
logical structure and build up a consistent argument. But make sure you know what
the models tell you and more importantly what they do not tell you, as no
single model can answer all questions. You simply cannot use one single hammer
to hit any size nail.
How much can a policy maker deliver?
·
But since in
this event we want to think about greater issues than just economic
practicalities, let me say something about the limits of policy making. At the
start of my talk I said that policy
makers cannot really deliver on much more of what society is prepared to accept.
What do I mean by that? Consider the issue of bonuses for bank executives.
There are good reasons, at least economically, for which paying out these
bonuses is a good idea. More importantly since staff at that level compete at
the international level, coordinating with others matters a lot more than what
decision you actually think is best. If society for itself however, has decided
that this is not acceptable, as it seems to have done here in the Netherlands,
it is not easy for the regulator to argue the merits of the case. Let me give you another example.
·
Consider the housing bubble in the Netherlands but
also other countries, like the US, the UK, Spain, Ireland. Is it a good idea to
burst the bubble and when is it a good time do that? Prior to the crisis it was
very difficult to convince banks to stop lending or households to stop
borrowing to buy overpriced houses. It
would have been as though policy makers were aiming to spoil a jolly good
party! Now, and with the benefit of the very instructive hindsight, society
is a lot easier to persuade about the risks of housing bubbles and how
regulating on the upside may not be a bad idea. If the party runs the risk of
getting completely out of hand, maybe it is not such a bad idea to break it
early enough.
·
I am hoping that
if a new housing bubble starts to emerge, policy makers will be able to act
pre-emptively. But it will not be
because the way they understand the economy has changed; it will be because
society understands the risks differently.
Ethics and their place in policy making
·
Let me finish
with one final thought on ethics and
their place in policy making. There is no doubt that at the very heart of
what we do, lies a system of values, morals that guides us along the way. And
there is in my view a legitimate concern that some of the choices we had made
prior to the crisis, and our response to it, from the amount of risks that we
took all the way to the manner in which the risk was shared, need re-thinking.
Just think of how many banks were rescued by tax payers across the world. We did that because it was necessary to do,
but then we questioned its legitimacy from a moral perspective. Was it fair
that tax payers paid the bill? Did they have an equal share in the profits on
the way up, to justify their contributions in the costs on the way down? Since
then we have put in place a new way of dealing with bank failures, which will
not burden tax payers as much as it did in the past. This is a reflection of
our notion of fairness and how the cost of failure should be distributed in the
society.
·
While it is
important that periodically we take the time to reflect on the values on which
our system operates, we cannot challenge them at the same frequency as the
frequency at which we take decisions. There are important decisions to be made
in real time, and it is just as important that they are made on grounds of economic efficiency and
effectiveness.
·
Let me give a
very actual example. The biggest
impediment to growth in the euro area right now is private debt. Debt in the
hands of households and companies is very big. Some of it is unserviceable.
This means that it cannot be paid back. What do you do next? You can take the
moral high ground and say that a debt is a debt: you pay it back. Or you can
forgive the debt on grounds of distribution and equality. I would much rather that you asked the question: how can I resolve the
problem in a way that will bring back the economy on a path to growth? You
may or may not come down to the same conclusion and irrespective of what it is,
some will not like it. But you would not have appealed to a sense of fairness,
which by the way would be just yours, not necessarily that of the society as a whole.
·
My point is that it should not be the job of the
economist to decide what is fair.
This does not mean that he should not operate based on an established sense of
fairness. Nothing is further from the truth. But, let society, in the shape and
form of governments, decide what is morally acceptable; let economists decide
what constitutes good economic policy; and then let the rule of law make sure
that the two meet.
Thank you.
