Remarks to the RBC Capital Markets Canadian Bank - OSFI-BSIF

Capital and Procyclicality in a Turbulent Market
Remarks by Superintendent Julie Dickson
Office of the Superintendent of Financial Institutions Canada (OSFI)
to the
RBC Capital Markets Canadian Bank CEO Conference
Toronto, Ontario
Thursday, January 08, 2009
CHECK AGAINST DELIVERY
For additional information contact:
Jason LaMontagne
Communications and Public Affairs
[email protected]
www.osfi-bsif.gc.ca
Remarks by Superintendent Julie Dickson,
Office of the Superintendent of Financial Institutions Canada (OSFI)
to the RBC Capital Markets Canadian Bank CEO Conference
Toronto, Ontario
January 08, 2009
Capital and Procyclicality in a Turbulent Market
Introduction
I hope that everyone was able to take some time to relax over the holidays,
because I think 2009 is going to be as dynamic and challenging and interesting
as 2008.
To kick off the year I thought I would talk about immediate challenges in the
banking industry, and then move to the longer term issue of procyclicality, and
end with a comment on current capital levels in the banking industry.
Immediate Challenges
The main issue, now, is the economy. The current economic situation will
challenge all financial institutions. The economic downturn is the second wave to
hit the system since August 2007. Canadian banks handled the first wave -global financial market turmoil -- relatively well compared to many of their global
counterparts. The second wave -- a serious economic downturn -- is hitting while
financial market turmoil remains ongoing. The second wave will affect both
capital levels and profits at banks. The extent and severity of the hit is hard to
predict, but the Canadian banking system is probably better positioned than most
other systems to deal with this second wave. It is very well capitalized, and has
avoided the need for government injections of capital -- unlike most advanced
systems in the world -- and has been able to tap the capital markets for additional
capital.
But there are challenges, and while we are very focused in the short term on
monitoring and risk assessment of federal financial institutions, we must also
start to think about other, longer term issues. With that in mind, I want to talk
about capital and "procyclicality".
1
Procyclicality
Sir John Templeton noted that bull markets are born on pessimism, grow on
skepticism, mature on optimism, and die on euphoria.
The words pessimism, skepticism, optimism and euphoria certainly make it
easier to understand the part of the cycle in which we may find ourselves, but we
can't use these kinds of words in our rules setting out capital requirements.
Increasingly, people are saying that regulators must step in with special
measures during the optimistic and euphoric phases. The former Governor of the
Bank of Canada, David Dodge, in the CD Howe Benefactor's Lecture (November
18, 2008), said that bank regulators must have a tool to increase financial
institutions' required capital when risky spreads significantly narrow, or when
leverage increases.
He added that we would also need to be able to reduce required capital when
spreads widen significantly beyond some normal level, or leverage contracts
dramatically. This could be done through adjustments to required capital.
Dodge stresses the need for a mechanism that will, in essence, automatically
increase the requirement for capital.
This discussion also goes by the name procyclicality.
There have been a number of discussions about procyclicality and whether
regulatory rules (either accounting rules set by the accounting standard bodies
like the International Accounting Standards Board (IASB), Financial Accounting
Standards Board (FASB) or Accounting Standards Board (AcSB), or capital rules
set by bank regulators like OSFI) feed upturns, and make downturns
worse. These areas could all use some further study.
But this issue is not an easy one.
I do like David Dodge's idea of some sort of automation. It has appeal because
we are talking about regulators stepping in during the optimistic and euphoric
phases which can be seen as wrecking a really good party -- the kind of party
where all the conditions are right -- the weather is warm, the people are new and
interesting, the music is great, the food is amazing, and the drinks are free.
However, in good times there would be many who would argue that any
regulatory decision to increase capital is misguided. How do I know that?
Because any time OSFI changes a capital rule that causes requirements to go
up, we instantly receive a ton of negative feedback from players that have to put
up more money.
2
One can also look at the negative reception Sheila Blair of the Federal Deposit
Insurance Corporation (FDIC) got when she suggested in Fall 2005 that a
leverage ratio -- something we have had in Canada for many years -- be adopted
around the world.
So an automatic switch to rein in euphoria has a lot of merit, especially if that
switch were designed now, while everyone has a new-found appreciation for the
importance of capital. But the devil is in the details. How would an “automatic
rule” be crafted? When exactly should it kick in? What risky spreads would we
choose? How far would risky spreads have to narrow and for how long before the
automatic switch kicked in? How much would capital increase by? Would a 7 per
cent Tier 1 target become an 8 per cent target initially, or more, and over what
time period? Would capital targets continue to ratchet up until a downturn
began? Or would it be a one-time increase until another automatic trip wire was
hit?
While there are a lot of questions that would have to be tackled in designing such
a system, I do encourage a discussion of how and whether to set extra buffers in
good times. A key challenge is that there is a limited ability to predict turning
points in the cycle and the timing of financial instability. Another challenge is that
not all business cycles are characterized by bubbles.
I have taken note of the Bank of Canada’s work to develop a financial stress
indicator that is a good predictor for 1-2 years into the future. While the Bank has
noted that the current indicator was unable to predict the current crisis because
the root causes came from outside Canada, an indicator that has predictive
power would be a valuable tool, if it can be developed.
Of course regulators are not the only ones involved in the procyclicality debate.
I would also invite accounting standard setting bodies to assess whether
accounting rules are procyclical. One key accounting issue concerns loan loss
provisions -- regulators would like to see more ability for banks to build provisions
in good times. This is currently very difficult as accounting standard setters prefer
provisions to be based on incurred losses versus losses that you fully expect but
which have not yet happened. This means that loan loss provisions are very low
heading into a downturn, as losses have not yet been incurred. This makes
capital cushions even more important to the extent they have to do double duty:
meeting both expected and unexpected losses.
Existing Features to Moderate Procyclicality
While I hear the word procyclicality a lot more now than I used to, it is important
to note that the system already contains many features that moderate
procyclicality. And the more developed these features were, the better off
3
financial institutions and financial systems have been. This is where I think we
should focus most of our efforts. Features moderating procyclicality include:
•
Strong risk management (as has been noted elsewhere, there is nothing
more procyclical than a badly managed bank or life company).
•
High capital buffers going into a downturn, either as a result of rigorous
stress testing by the institution or as a result of regulatory requirements,
which have definitely moderated cyclicality (OSFI's targets of 7/10 were
well above Basel minimums of 4/8, and provided more cushion to be
drawn down in bad times).
•
High core common share levels in Tier 1 ratios going into a downturn
(because OSFI allowed banks in good times to have 30 per cent of Tier 1
in innovative and preferred shares, we were recently able to increase 30
per cent to 40 per cent in a downturn, giving banks added flexibility to
raise capital at the low point in the cycle).
•
Through-the-cycle estimates of Probability of Default (PD) and Loss Given
Default (LGD) in Basel 2. These key drivers of capital act to moderate
procyclicality, as required capital will not spike up in a recession if the
estimates were based on more than just euphoric or optimistic times but
were painstakingly developed based on expert judgment as well as
historic data incorporating a downturn.
•
A leverage ratio which constrains on and off balance sheet growth and
thus gives more comfort regarding capital adequacy, recognizing that
there are a lot of risks that are difficult to measure and that are not
captured in current capital calculations.
Because Canada scored well in a number of the areas identified above, the
Canadian banking system is considerably less procyclical than many other
systems. However, many of the above features can be improved upon. Stress
testing can be improved, as can risk management, through-the-cycle estimates,
and expectations regarding quality of capital during boom times. OSFI will
continue to focus on these areas.
Limits to the Procyclicality Debate
The procyclicality debate can go too far because at its extreme, it means that
capital should not go up, at all, in a downturn. This would mean that capital for a
given loan should remain constant throughout an economic cycle regardless of
what happens to the borrower's business and credit rating. Similarly, this might
4
mean that segregated (seg) fund capital requirements should not increase
for long dated guarantees even when equity markets fall substantially.1
The Basel 2 Capital Accord, as well as capital rules in general, have become
more risk based, as charging the same capital regardless of risk was seen as
being unwise. A system based on risk does have some degree of procyclicality
because capital will increase when risk increases.
That being said, regulators will remain focused on the impact of procyclicality
because, while there were some measures built into the system to moderate
procyclicality, we need to know whether more can be done.
One good example of where an additional change is needed is with respect to
Value-at-Risk (VAR), used to measure capital required in the trading book. The
current VAR is highly cyclical because banks may base the capital charge on as
little as one year of historical data. The Basel Committee has noted that, because
benign financial market conditions are typically accompanied by low volatility,
they result in low VAR based capital requirements. Periods of stress are
associated with jumps in volatilities and correlations, producing spikes in the
minimum capital requirement. The revised VAR framework will be made public by
the Basel Committee in January 2009.
Misunderstanding about the Current Procyclicality Debate
Dealing with procyclicality has costs when times are good. There is no free
lunch.
Some have interpreted the procyclicality debate as regulators stepping up now to
reduce required capital. This would make sense if capital requirements a few
years ago -- in the optimistic/euphoric stage -- had been increased (even though
Canada set high targets, those targets were not specifically adjusted some time
in 2005 which might have been the time in the cycle where optimism turned into
euphoria).
1
OSFI did act in October 2008 to change required capital for seg funds because we became aware that the
required capital calculation was producing results that were not representative of the risk. Requirements
shot up in the short-term for obligations that did not come due for many years.
When we made the changes to seg fund guarantee requirements we kept the procyclicality debate in mind.
While many economists suggest that seg fund capital requirements by their nature cannot be procyclical,
because stock market movements do not mirror the economic cycle, it is clear that increases in capital
should not mirror stock market declines because markets can move up and down quickly. Thus recognizing
the full amount of point-in-time equity market movements was considered inappropriate for seg fund capital.
That is why the changes to the seg fund capital effectively only recognize about 5 per cent of quarter-toquarter equity market changes (for guarantees with terms five years and greater). This substantially
removes short term equity market performance from the required capital calculation.
5
If required capital had been adjusted upward during strong market conditions,
then one could see a case for reducing required capital now, to further mitigate
procyclicality beyond existing mitigating factors.
We would also want some comfort that very high capital buffers, set up in good
times, could be drawn down once a downturn hit. If not, the result would be that
extra costs would have been imposed on the system without much benefit. The
market, as well as global regulators, would have to be comfortable with such a
drop. As has been seen recently, in times of extreme uncertainty, markets might
not react this way.
I would also say that people understandably want assurances that all the players
have learned valuable lessons from global market turmoil and are working to try
to prevent a recurrence ever happening again. However, if we draw upon the
current crisis to develop all the rules that will apply in the normal ups and downs
of an economy, we may end up doing more harm than good.
Capital - What Level is Enough?
Aside from the procyclicality debate, there is also a debate about what level of
capital is enough, whether regulators have much of a role today in setting capital
expectations or whether it is the markets that are deciding capital levels.
It was a unique experience for OSFI to put out a statement on December 22,
2008, saying that we had not pushed for capital increases across the board in the
banking industry, and that in our view, it was the market that was pushing. How
did we get to this point?
Regulators like OSFI tend to set minimum levels of capital as thresholds that
must be met in order for us to allow an institution to continue to take money from
the public. OSFI also sets a higher target level, a level at which other types of
regulatory intervention would likely take place, given that capital is the first line of
defense when something goes wrong. Solid buffers are key to safety and
soundness, as a lot can go wrong in the financial sector.
Institutions hold capital above the OSFI target level in order to avoid hitting the
target. They also operate with higher capital levels based on their unique needs - banks know their business, their balance sheet, and their risks -- and the capital
should reflect that assessment. We have been pushing banks, as part of Basel 2,
to really focus on this.
Institutions also hold capital above regulatory targets in order to retain high
ratings from ratings agencies (with respect to rating agencies, capital is but one
element in the rating, but generally the higher the capital the higher the
6
rating). Lastly, institutions also pay close attention to their stock price and that is
affected by the market’s view of capital adequacy and profitability.
OSFI's view of whether an institution is well capitalized is based on the answers
to a number of questions. The size of the cushion should depend on: what are
the cushions made of (amount of common shares which has more loss absorbing
capacity); what kind of losses might hit the cushion; whether the cushion is based
on through-the-cycle estimates (so that required capital stays relatively constant);
expertise of the bank and quality of risk management in general at the institution
and for significant business lines; leverage; and the outlook for the economy
(both global and domestic), especially in key markets in which an institution
operates.
While the market is also expressing a view on the size of the cushion, I think it is
important to keep in mind that capital cushions are meant to fall when needed -otherwise they are not cushions, they are more like cement.
As well, as suggested above, determining capital adequacy requires an in-depth
assessment of many factors unique to a financial institution and involves more
than just comparing Tier 1 ratios across global banks.
OSFI’s view on capital is not related to what levels are needed to make stock
prices perform better, or to what might be needed for an institution to get a high
rating from a rating agency. Also, capital is but one of the considerations in
determining OSFI’s intervention rating for a financial institution. A lot of other
factors also come into play.
Finally in this stage of the cycle, the downturn, it is also important to carefully
assess the situation and not over-react or under-react History tells us that
changes in risk associated with the economic cycle tend to be incorrectly
assessed. In particular risk is often underestimated in booms (bad loans are
made in good times), and overestimated in recessions (when loans might be cut
back). History also tells us that markets can overshoot. At the end of the day,
these are tough issues but I think we need to look at them in a balanced fashion
to try to avoid undershooting or overshooting.
Conclusion
In conclusion, I will reiterate the main points I have discussed:
1. Immediate challenges: The current economic situation will challenge all
financial institutions. Canadian banks have handled the first wave -financial market turmoil -- relatively well. The second wave -- a
serious economic downturn -- will hit both capital and profits, but the
7
Canadian banking system is probably better positioned than most other
systems to deal with the second wave.
2. Measures to dampen procyclicality: There are a lot of existing measures
to dampen procyclicality and help ensure that capital rules do not amplify
economic cycles, but more thought is required. Some proposals, like
changes to VAR, are straightforward. Others, such as proposals to
automatically increase capital requirements in periods of optimism and
euphoria, are good ideas but extremely difficult to implement. All such
proposals however deserve full consideration.
3. What level of capital is enough? The Canadian banking system remains
very well capitalized and there are no plans to increase the current 7/10
targets set by OSFI -- doing so now would also amplify the economic
downturn2. While markets have been driving capital levels up, it is
important to keep in mind that capital cushions are meant to fall when
needed. As well, determining capital adequacy requires an in depth
assessment of many factors unique to a financial institution and involves
more than just comparing Tier 1 ratios across global banks.
I appreciate the opportunity to speak to you today, in January, the darkest,
coldest month of the year. Warmer weather will come, as it does every year, and
conditions in the financial sector will improve, which is something we can all look
forward to. In the end, everyone in this room wants the same result, a safe and
secure Canadian financial system, one in which Canadians can continue to place
their trust.
Thank you.
2
For further background on this issue see Remarks by Superintendent Julie Dickson, Office of the
Superintendent of Financial Institutions, to the NPI 2008 Life Insurance Forum
8