Contents
Banking Siege – Overview 2
Valuations – Unbelievable Levels 4
Recommendations – Strong Buy 5
Market Capitalization Declines vs. Writedowns – Severe 6
Short Interest Ratio Declining – Short Covering 6
Balance Sheet Risk – Exposure to High-Risk Assets
Low Exposure on Absolute & Relative Basis 7
Canada vs. U.S. – Two Very Different Markets – Canada's Structural Advantage 9
Loan Loss Provisions – Trough to Peak – Very Manageable 9
Bank Earnings Momentum – Turning Positive in Q4/08 11
Bank Financing – Aggressive Funding 12
Credit Market/Interest Rate Indicators 12
Bank Relative Performance – Bear Markets, Recessions 13
Periods of Greatest Bank Share Underperformance – Revisited 13
Appendix 1 – List of Exhibits 75

Banking Siege
B U L L M A R K E T H A S B E G U N – B U Y B A N K S
The banking siege is into its tenth month and counting, although showing signs of easing. Canadian bank
stocks continue to be extremely volatile, reacting to every gyration in the credit markets. Bank stocks have
been under siege since the credit crisis began in August 2007, recoupling with U.S. banks at a time when
we believe decoupling would be in order.
Bank stocks underperformed the TSX in 2007 recording its third worst performance against the market in
50 years. The TSX was dominated by a few stocks – Potash, RIM, and Alcan – not dissimilar to the
Nortel-dominated 1999 market. The only two years of weaker relative bank performance was 1979 (TSX
relative earnings growth 41%, commodities) and 1999 (Nortel). Bank stock underperformance two years
in a row is very rare (three times since 1967). However, banks are lagging the overall market again thus far
in 2008. The bank index is down 3% year-to-date versus the market’s increase of 6%, with BMO lagging
down 12%. Canadian banks have not kept pace with the heavily resource-weighted TSX.
Canadian bank market capitalization declines have been extreme compared to writedowns and relatively
low exposure to high-risk assets. The banking siege has presented one of the best buying opportunities in
decades to participate in the major bull market in bank stocks which we believe began March 17. The
long-term risk return track record for bank stocks has been impressive. The bank index has increased more
than 20% in a calendar year 15 times since 1967 while declining more than 10% only four times (1967,
1981, 1990, 2007).
The bank index has rebounded 16% from the lows reached on March 17, The Bear Stearns Rescue. The
Bear Stearns Failure/Rescue, we believe, was the bottom and was a very important step for the market in
grinding through the financial and credit crisis. In fact, a financial failure is usually needed before any
stability is achieved. Prior to the recent bounce, the bank index was down 16% year-to-date.
Canadian bank stocks have been impacted, not surprisingly, by the negative global sentiment towards the
financial services sector. Investors have a heightened awareness of systemic risk and are pricing for it.
What we underestimated was the global players’ exposure and leverage contained within the U.S. subprime
market and how negatively it could impact credit markets and liquidity and the degree of investor
panic towards the sector indiscriminately. We suspect that the volatility and panic has been enhanced by
large pools of capital in hedge funds (short interest), especially influential in periods of fearful and illiquid
markets. Also, we believe Canadian banks were side-swiped by CM’s undue U.S. sub-prime exposure,
creating enhanced general balance sheet fear for all Canadian banks. In addition, the non-bank ABCP
debacle and liquidity crisis tarnished Canadian credit markets, providing further fuel to the fire despite the
major banks having extremely modest involvement.
We believe sentiment, although negative, is starting to shift as global banks are taking another round of
major writedowns with release of their first quarter calendar 2008 earnings and as additional capital is
starting to flow into the sector. Thus global banks are taking their hits and shoring up their capital. We also
expect that fiscal Q2/08 will be the last quarter of noticeable writedowns for Canadian banks – although
writedowns for Canadian banks (except CM) have generally not been material, with capital levels,
funding, and liquidity remaining very strong. Future writedowns are expected to be modest, more orderly,
and manageable and more in the context of the ongoing businesses without significant balance sheet
disruption. However, as the banking sector recovers from the credit and liquidity malaise, investor
attention is shifting to the economy and bank earnings levels through the slowdown/recession.

Canadian banks are expected to weather a recession, with trough ROE or on any other comparable
profitability measure at 15% plus, higher than any previous cycle. In fact, we believe return on equity will
remain materially higher for longer than market expectations. The rumours of the collapse of bank ROEs
are greatly exaggerated. Canadian bank earnings over the past five years generally have held up
significantly better than U.S. banks, we believe, due to a number of factors, including: less reliance on
carry trade and the yield curve; less reliance on a mortgage refinancing boom in Canada; and less reliance
on structured product and securitization for earnings and the stability and lower risk stemming from the
solid, national oligopolistic banking market in Canada.
We view Canadian banks’ balance sheets as strong. Dividends, we believe, are safe – we expect dividend
growth to moderate in 2008 before returning to double-digit growth levels as stability is restored in the
credit markets. Canadian banks’ dividend growth prospects remain intact, in contrast to a number of
International banks that have had to cut their dividends on average 64%.
Bank first quarter operating earnings were solid with extremely high return on equity of 22%, although
operating earnings did decline 2% YOY. Reported earnings for the bank group declined 14% year over
year, excluding CM. We expect modest negative operating earnings momentum in Q2 and Q3 of 2008
before the bank group returns to positive earnings momentum in Q4. The bank group has achieved a
record 15% plus earnings growth for the past five years straight. Thus we expect three quarters of modest
negative earnings momentum in 2008 before returning to high single-digit/low double-digit growth in
2009, reflecting the solid business models that are expected to continue to generate excess returns.
Canadian banks now have substantially lower loan/risk concentration (exhibits 21 and 23) than at any
time in the past 25 years on an absolute basis and relative to U.S. banks. Despite relatively strong
earnings, low exposure to high-risk assets (detailed in section titled “Balance Sheet Risk – Exposure to
High-Risk Assets – Low Exposure on Absolute & Relative Basis”), modest writedowns, and generally
solid fundamentals, Canadian bank stocks have seen major declines in their market capitalizations by a
magnitude not seen before. This may be the largest discount for systemic risk in history.
Bank cumulative writedowns including second quarter estimates have been relatively modest, representing
one and a half quarters of bank earnings (exhibit 15), with writedowns at RY, TD, and BNS nominal.
Excluding CM, cumulative writedowns are less than one-quarter of earnings.
Canadian bank market capitalization excluding CM has declined by 15.7x (writedown ratio) the level of
cumulative writedowns, including estimated writedowns for Q2/08. This is excessive compared to
writedown ratios in the 2x range in the Lesser Developed Countries (LDC) loan crisis of the 1980s and the
commercial real estate (CRE) debacle of the early 1990s. The writedown ratio of 15.7x for Canadian
banks is also excessive versus the 7.3x for a number of global banks (exhibit 14). Details are contained in
the section titled “Market Capitalization Declines vs. Writedowns – Severe.”
We believe negative sentiment, fear, and short interests (detailed in the section titled “Short Interest Ratio
Declining – Short Covering”) were major share price drivers. Canadian banks in past cycles have not
incurred any meaningful short interest. Thus we suspect short interest has exacerbated the share price
declines of Canadian banks despite their having low exposure to high-risk assets and modest writedowns.
However, short interest has been declining considerably over the past several months, another favourable
sign. The short interest ratio at CM and BMO has declined recently to 7.7x and 5.2x, respectively, from
peaks of 14.1x and 7.2x reached in late summer of 2007.

Banking Siege May 2008
4
It is interesting that the Canadian and U.S. Bank Index correlation (R2) has been historically very high at
96%, but has significantly decoupled over the past five years with low correlation of 15% based on what
we believe are significant differences in operating performance. However, since the current credit crisis
began in July/August 2007, the correlation has increased to 85% as the market has not differentiated
between Canadian and U.S. banks despite significant differences in the housing/mortgage market and
banking environment, balance sheet risk, earnings, and profitability. Canadian and U.S. bank indices have
recoupled at a time when we believe decoupling would be in order. The macro environments in the U.S.
and Canada are so different. The U.S. has a serious sub-prime mortgage crisis, whereas the sub-prime
market in Canada is small and benign. Equity in household real estate in Canada is at a 25-year high
versus the U.S. at a 25-year low. The differences are highlighted in the section titled “Canada versus
U.S. – Two Very Different Markets – Canada’s Structural Advantage.”
So as the banks survive the siege from the credit/liquidity/sub-prime chaos, we switch gears and look to
earnings levels over the next few years in the face of a gloomy economic outlook and deterioration in
credit quality.
The potential for materially higher loan losses given the recession outlook has become a focus for
investors. Loan loss provisions for the bank group in 2007 were $2.8 billion or a modest 27 bp of loans.
We believe that 2006 was the trough at 22 bp, similar to troughs in 1988 and 1997. We expect loan losses
to rise steadily over the next five years; however we believe that loan losses will peak at much lower levels
than historical norms due to major loan mix shifts and significantly less loan concentration. The impact to
earnings is also expected to be much lower than in the past due to business mix shifts, lower leverage to
credit, higher capital levels and larger earnings base.
Previous loan loss provision peaks were in 1983 at 100 bp, 1992 at 161 bp, and 2002 at 106 bp. These
peaks corresponded with major corporate loan concentration, which the banks appear not to have this
cycle. Thus we believe loan loss provisions will peak in the 53 bp to 76 bp range. This would equate to
loan losses of approximately $7 billion at the peak using the mid-point or an additional after-tax charge
of $3 billion from a 2007 base over the next three to four years. The market capitalization decline of
$40 billion, we believe, has already fully discounted higher loan losses. We arrive at peak loan losses by
attributing loss ratios by portfolio and by individual bank. The section titled “Loan Loss Provisions –
Trough to Peak – Very Manageable” provides greater detail on potential credit losses at the peak of
the cycle.
V A L U A T I O N – U N B E L I E V A B L E L E V E L S
In our view, bank valuations are already discounting a recession. The banks’ trailing P/E multiple has
declined to 11.0x. The low was 9.4x on the Bear Stearns rescue of March 17, 2008, slightly above the Asia
crisis bottom in 1998 of 9.0x and below 10.9x multiple during the Telco/Cable meltdown in 2002.
Bank P/E multiples appear extremely compelling at 11.0x and 10.0x our 2008 and 2009 estimates. Our
estimates have some modest vulnerability to continued gridlock in the credit markets and the slowing
economy. We continue to believe that the 2007/2008/2009 stress testing of bank earnings will lead to
another round of P/E multiple expansion just as the stress testing of bank earnings in 2002 telco debacle
led to a major P/E expansion to 15x, from 11x. The credit crisis and negative sentiment towards global
financials have caused Canadian bank P/E multiples (exhibit 49) to diverge from the 20-year trend line.
We believe the trend line for bank P/E multiples continues to be 16x and the current stress testing is in fact
needed to push the multiple through the previous high. The 16x target multiple is based on a 5% 10-year
bond yield. If bond yields were to increase to 6% based on inflation concerns, the target multiple would be
in the 14x range.

Bank P/E multiples are at a 6% discount to U.S. banks despite higher return on equity, higher capital,
lower exposure to high-risk areas, and a more favourable banking and economic environment. The bond
market and the currency markets have revalued Canada favourably versus the U.S. over the past 15 years.
But not the equity markets … yet? If we chart the Canadian/U.S. bank relative P/E multiples (exhibits 55
and 56) versus the Canada/U.S. bond market spreads and Canadian/U.S. exchange rate we see a large
divergence. We conclude that the equity markets have not revalued Canadian bank P/E multiples in line
with bond and currency markets.
We believe, based on underlying fundamentals of Canadian banks and the Canadian banking environment,
that Canadian banks should be trading at a 10% premium to U.S. banks, in line with shifts that have
already occurred in the bond and currency markets.
We believe the banking model in Canada is more efficient with the absence of monolines and dominance
in retail, wealth management, and capital markets. The banking system has higher underwriting standards,
and a superior mortgage/housing system and framework. The underlying fundamentals have led to higher
profitability both on return on equity (exhibit 41) and return on risk-weighted assets (RRWA – exhibit 42)
and superior capital levels (exhibit 43). The banking and economic environments, we believe, will remain
much stronger in Canada, given the major structural advantage Canada has. (See section “Canada versus
U.S. – Two Very Different Markets – Canada's Structural Advantage.”)
Bank dividend yields relative to bonds continue to be at unheard of levels at 115% or 5.2 standard
deviations above the mean. The peak was at an unbelievable level of 7.0 standard deviations on March 17.
I don’t believe we will see this level again. The market is clearly discounting dividend cuts. We believe
the probability of a dividend cut for Canadian banks to be extremely low with payout ratios in the 45%
range, Tier 1 capital ratios at 9.9%, and low exposure to high-risk areas. The market’s fear of dividend
cuts by Canadian banks has been fostered by significant dividend cuts announced by a number of non-
Canadian banks (exhibit 58).
Bank dividend yield relative to other equities is also at historical highs. Bank dividend yield relative to the
overall equity market is two standard deviations above the mean. The only time in history it has been
higher was during the Nortel bubble of 1999. In our view, bank dividend yields are also compelling
against pipes & utilities and income trusts.
The Bank risk premium also appears extremely attractive at 5.73%, one of the highest levels in
history, despite high-quality balance sheets, strong capital and liquidity positions, and solid core
earnings and profitability.
R E C O M M E N D A T I O N S – S T R O N G B U Y
We continue to recommend aggressively buying bank stocks at these levels. As fear subsides, we expect
share prices to move up sharply. Strong fundamentals and compelling valuation remain.
We continue to recommend maximum allowable weightings in bank stocks. We have only Buys and
Strong Buys in the bank group with no Sells or Holds on an absolute return basis. We believe a major bull
market in bank stocks began March 17, 2008. High-quality balance sheets, strong funding and liquidity,
high profitability, compelling valuation, and the major structural advantages from our banking system and
economy support our very bullish stance towards bank stocks.
We maintain our 1-Sector Outperforms on RY and CWB; 2-Sector Performs on TD, BNS, CM, and NA;
and 3-Sector Underperforms on BMO and LB.
Our order of preference is: RY, CWB, TD, BNS, CM, NA, BMO, and LB.