Tuesday, July 24, 2007

RBC CM Lowers CIBC's Q3 2007 EPS Estimate

The Globe and Mail, Tara Perkins, 24 July 2007

Canadian Imperial Bank of Commerce could be forced to take a hit of about $100-million this quarter because of its exposure to securities related to the U.S. subprime mortgage market, analysts say.

"The continued weakness in securities related to U.S. subprime housing will likely force CIBC to mark down its exposure to the space, as most of its securities are held in its mark-to-market trading book," RBC Dominion Securities Inc. analyst André-Philippe Hardy wrote in a note to clients yesterday.

"We believe that a $50-million to $100-million markdown is possible."

BMO Nesbitt Burns Inc. analyst Ian de Verteuil recently estimated CIBC could take a charge of $100-million to $150-million this quarter.

It is still in the midst of its fiscal third quarter, which runs to the end of this month. Financial results are scheduled to be released Aug. 30.

"Given that CIBC should earn $675-million after tax in the third quarter, such a charge could amount to about 10 to 15 per cent of quarterly earnings," Mr. de Verteuil wrote in a note to clients.

He is assuming the bank's total exposure, or holdings, of this type of securities could be about $1.2-billion. The bank has not disclosed how much exposure it has and CIBC declined to comment on the analysts' reports yesterday.

Any markdown or loss would almost certainly be partly offset by lower taxes and lower employee bonuses, Mr. Hardy said.

He revised his per share earnings estimate down from $2.01 to $1.90. Analysts' consensus is $1.95.

"The bank has not issued a press release on its quarterly results, which is somewhat comforting as the current quarter is only seven days from being over," Mr. Hardy wrote. "In our view, this suggests that the bank's exposure to U.S. subprime housing securities will not cause a material shortfall to earnings estimates."

At this point, CIBC's losses are probably paper losses rather than realized cash losses, Mr. Hardy said.

The paper value of these investments on the bank's books can be volatile, Mr. de Verteuil explained.

The holdings must be valued, or marked, regularly. A weak mark at quarter's end can cause a large mark-to-market - or paper - loss, only to have the position reverse the next quarter, he said.

"Given the poor market currently, this could easily be a problem in CIBC's third quarter, even though holding the positions to maturity may produce no economic [or actual cash] loss," he wrote.

But the idea that any losses would only be on paper is scant comfort, he added.

"The reality is that CIBC purchased a security [or the derivative of a security] that is worth less than when it was originally acquired. In the cold light of hindsight, there must be some negative consequences for the income statement."

Taking a charge this quarter "would be a disappointment for CIBC," which has "done a tremendous amount to de-risk the bank," Mr. de Verteuil added.

CIBC chief executive Gerry McCaughey has made reducing risk a major priority since the bank took a $2.4-billion (U.S.) charge in 2005 to settle a lawsuit related to Enron.

The bank now earns a larger proportion of its profit from bread-and-butter Canadian consumer banking than any of its peers, Mr. de Verteuil said. And, disregarding this situation with the subprime-related securities, it seems to be on track for a strong operating year, he added.

Mr. de Verteuil said he takes comfort from the fact that Bank of Montreal took a "very surprising" $680-million charge earlier this year from its commodity trading operations, and has still performed roughly in line with its peers since.

Blackmont Capital analyst Brad Smith pointed out in a note to clients yesterday that CIBC was the only Canadian bank to make the list of "Top 20 Counterparties by Trade Count" released in Fitch Ratings Ltd.'s recent credit derivative survey. CIBC crept onto the list at No. 20. It did not rank in the top 20 counterparties by notional amount. Morgan Stanley topped both lists.

Fitch's report notes that valuations of credit derivatives are being largely driven by assumptions that may not always reflect market prices, and so "institutions may be understating losses or overstating gains." That "leaves little doubt in our view as to the potential for broad-based mark-to-market losses over coming quarters," Mr. Smith wrote in yesterday's note on the banks and structured finance.

Eleven U.S. banks, investment banks and brokers reported financial results last week, and on average they beat analysts' consensus estimates by 5 per cent. But their stocks ended the week down 4 per cent, as investors focused far more on risks than near-term earnings, Genuity Capital Markets analyst Mario Mendonca said in a note to clients yesterday.
RBC Capital Markets, 23 July 2007

Investment Opinion

• The continued weakness in securities related to U.S. sub-prime housing will likely force CIBC to mark down its exposures to the space, as most of its securities are held in its mark-to-market trading book.

• We believe that a $50-100 million mark down is possible, based on our belief that the April quarter mark down was in the $25 million range. A mark down would almost certainly be partially offset by lower income taxes and incentive compensation.

• Our prior core cash earnings estimate was $685 million, so the net income hit we expect represents 5% of Q3/07E earnings. We expect the bank to earn a ROE of 23.0%.

• The bank has not issued a press-release on its quarterly results, which is somewhat comforting as the current quarter is only 7 days from being over. In our view, this suggests that the bank's exposure to U.S. sub prime housing securities will not cause a material shortfall to earnings estimates. Our core cash EPS estimate is revised from $2.01 to $1.90, compared to the consensus estimate of $1.95.

• Management has stated that the majority of its exposures to the U.S. subprime real estate mortgage market are in securities that are AAA-rated. AAA-rated securities have seen price weakness, with drops in some ABX indices of 5%, certainly a large decline for AAA-rated securities (and a disastrous drop for holders that employ leverage), but not as drastic as the 50%+ declines in some lower-rated indices.

• CIBC has stated that its unhedged exposure to this sector is well below the US$2.6 billion exposure that media reports have estimated. If CIBC's exposure was indeed that large, we believe that the mark-to-market would have been larger than $25 million in Q2/07.

• We believe that CIBC's losses so far have been limited to mark to market adjustments, rather than real losses. Since CIBC probably holds these securities in trading books, we do not mean to imply that the losses would not be incurred if CIBC sold its positions today but rather point out that loss rates in underlying assets need to rise higher than they are today before holders of AAA tranches begin to lose real money.

Maintain Outperform rating

CIBC trades at 12.0x our estimated 2007E earnings versus a peer average of 12.6 times. We believe that the relative multiple makes CIBC's stock attractive, in spite of a lower revenue growth profile and greater exposure to U.S. sub-prime real estate, hence our Outperform rating. The following points summarize our investment thesis on CIBC shares:

• Potential for upward earnings revisions. We expect the consolidation of FirstCaribbean's results and improving retail revenue growth, combined with flat expenses in Canada, will drive earnings growth that exceeds expectations in 2008. We also believe that Q2/07 retail loan losses could prove to have been abnormally high in the near term.

• 17% increase in the dividend expected in Q3/07. The current dividend rate of $3.08 implies a payout ratio of 38% based on our 2007E EPS, well below the bank's official target range of 40%-50%.

• The bank's multiple should benefit from having the second-lowest exposure to wholesale income. The retail mix should also increase given the acquisition of FirstCaribbean, improving revenue growth in retail businesses and a likely reduction in merchant banking gains.

• Less exposure to potentially rising business loan losses, the area that most concerns us from a credit quality standpoint. Business loan losses for the Canadian banks are currently non-existent, compared to a 17-year average of 80 basis points. CIBC's loan book has the least exposure to business and government loans and the bank holds $10.4 billion in credit default swaps - a large amount relative to its $34.0 billion business and government loan book.

• CIBC's stock is likely to trade on news in the near term. However, the bank's exposure to sub-prime real estate does not prove costlier than we estimate, we believe that those willing to handle short-term volatility will benefit. The disclosure of Q3/07 results on August 30 is likely to shed some light on the issue.


Our 12-month price target of $114 is a combination of our sum of the parts and price to book methodologies. It implies a multiple of 13.0x 2008E cash EPS, compared to the 5-year average forward multiple of 12.0x. Our target multiple is high versus the historical multiple because we believe there is upside to earnings estimates. Our relatively high price to book target multiple of 3.0x reflects the bank's industry-leading ROE and low credit risk. Our sum of the parts target P/E of 12.3x is in line with our target average for the banks, as lower exposure to low-multiple wholesale businesses is offset by slower than average revenue growth.

Price Target Impediment

Risks to our price target include the health of the overall economy, sustained deterioration in the capital markets environment, loss of domestic market share, a decline in underwriting activity and U.S. subprime CDO markets, and weakening retail credit quality.
BMO Capital Markets, 13 July 2007

Establishing Limits

Concerns on CIBC’s exposure to sub-prime have clearly weighed on the stock. With the bank having made little disclosure as to the potential hit to earnings (other than to say that the “un-hedged” exposure is less than US$2.6 billion), it is scarcely surprising that the market is suspecting the worst. What is interesting is that the market seems to be discounting the stock far worse than a worst case scenario. Clearly, the Enron fiasco still weighs heavily on the market’s perception of how the bank is run.

Before we get into our understanding of the details, it is worth noting that the vast majority of CIBC’s business is domestic, retail-oriented. Indeed, in the first half of 2007, CIBC got a larger proportion of its earnings from retail banking in Canada than any of its peers (see table below). Outside of the CDO situation, CIBC seems to be on track for a strong operating year. We believe that the market has overcorrected and we are reiterating our Outperform recommendation on CM shares.

Our continued preference for CIBC shares is not to say that we are confident that CIBC won’t at some time in the next few quarters take a “mark-to-market” loss on its CDO positions. Indeed, a charge of $150 million (pretax) seems likely in the third quarter if, at quarter end, the market remains as hostile as it currently is. What gives us comfort however is this: Bank of Montreal took a very surprising $680 million charge on commodity options, despite its reputation as a conservative institution. Since that charge, the bank has performed roughly in line with its peers. We are relatively confident that CIBC charge will be less than $680 million, yet its stock has already underperformed its peers by roughly 7% since this “kerfuffle” began.

A Look at Tricadia

CIBC was relatively late to re-enter the game of underwriting Collateralized Debt Obligations (CDOs) - and its timing was atrocious. The underwriting of Tricadia 2007-8 coincided with a sharp sell-off in residential mortgage backed securities (RMBS) - the collateral that often backs these CDO structures. This isn’t to say that CIBC was totally ignorant of the CDO market. It is clear from notes to the financial statements that the bank had ongoing involvement in “warehousing” these securities on behalf of various underwriters or managers. It does, however, seem as though the bank picked the wrong time to become more aggressive in this business.

In April 2007, with the Tricadia deal in the late stages of underwriting and with the well-publicized weakness in the residential mortgage market, demand for high-rated CDO paper declined (from what we can determine, the demand has typically been relatively good for the lower quality, higher yielding parts of these CDOs). As such, CIBC as underwriter was left holding most, if not all, of the US$330 million senior-secured tranche of the deal.

We have not been able to see what collateral/securities are in Tricadia, but from analysis of various deals done by this manager, we can assume that the underlying collateral was RMBS and that some proportion was investment-grade. Given the rating of the tranche held by CIBC (US$330 million out of an entire deal size of US$507 million), we assume that the majority of the collateral was investment-grade at the time of origination. This probably means default rates on the mortgages that back the structure need to be well over 10% before the AAA tranche takes a haircut. Above this level however, there would be fundamental impairment on these securities. As we show in the chart on page 4, which shows the ABX-AAA Index, and despite material concern within the market, indicative bids for AAA rates RMBS securities remain in the high 90s.

Fundamentals are Only Part of the Story

Whatever the fundamentals, CIBC holds its Tricadia and other CDO exposures in its mark-to-market, or trading book. A quick note on the accounting for these positions: Given the lack of transparent markets for some positions, banks occasionally have to mark-to-“model” rather than mark-to-“market”. The CDO market requires this, given the unique nature of each structure. Having said that, indicative prices and/or prices for various indices provide important ingredients in establishing the carrying value of these positions.

As we have said above, these CDO positions are maintained in CIBC’s trading book. This is both good news and bad news. On the good news front, accounting regulations force management to “fess up” quickly to problems. The bad news is that the quarter-end “marks” are important in establishing the P&L, which can result in increased volatility—a weak mark at quarter-end can cause a large mark-to-market loss, only to have the position reverse the next quarter.

Given the poor market currently, this could easily be a problem in CIBC’s third quarter, even though holding the positions to maturity may produce no economic loss. This is however scant comfort to us. First, the maturity date of the Tricadia senior secured note is 2052 (about my retirement date), and second, the reality is that CIBC purchased a security (or the derivative of a security) that is worth less than when it was originally acquired. In the cold light of hindsight, there must be some negative consequence for the income statement.

How Can We Estimate the Quarterly Pain?

To date, CIBC has indicated that its 'unhedged' exposure to CDO’s is less than US$2.6 billion, but given its comments on Tricadia, one can assume that it is greater than the US$330 million that made up the original senior secured notes. A wild guess would be to simply use the mid point of these numbers, about US$1.5 billion. Note that management continues to state that the “majority of the positions are AAA”.

Another yardstick is to consider the exposure as disclosed in the bank’s annual report as of October 2006. The bank states that at that time, the maximum loss on its exposure to CDOs was $729 million. If we assume that the current level includes both the Tricadia deal and incremental growth in the business, the number could be in the $1.0-1.2 billion range.

A third way to approach the bank’s exposure is to consider the impact of the weak market for CDO pricing at the end of April on CIBC’s second quarter earnings. Management indicated on the conference call that the weak trading revenues were a result of problems in structured product (read CDO’s) and credit markets. From the trend in the “other” segment of trading revenues, it appears as if revenues were $50 million below normal. Some of this would be due to the weakness in overall credit markets, and some to the CDO problem, but in the most punative situation, the entire figure could be due to the CDO’s. Using the ABX levels across various tranches (assuming a 50:50 split AAA to non-AAA) and using some rudimentary arithmetic, we can reverse engineer the current size of the book to be $750 million.


In conclusion, we believe that an estimate of $1.2 billion of CDO exposure for CIBC sounds reasonable. Assuming this, a 50:50 mix of AAA and non-AAA, and using yesterday’s closing prices (see graph below to see indicative pricing on AAA, A and BBB exposures), we estimate that the loss in the third quarter could be $100-150 million. Given that CIBC should earn $675 million after tax in the third quarter, such a charge could amount to about 10-15% of quarterly earnings. Of course, mark-to-market pricing is just that, so that inter-quarter volatility doesn’t require the bank to disclose any issues until the quarter end. Needless to say, early August will be an interesting time for CIBC investors.

From our perspective, such a charge would be a disappointment for CIBC. It has done a tremendous amount to de-risk the bank – despite the numerous detractors that have focused on items such as single elements of trading and derivatives. This situation again raises the spectre that the bank is accident-prone. However, we continue to believe that investors in CIBC shares, which are trading at roughly a multiple point lower than their peer group, are adequately compensated for the risks.

Another situation that provides us with comfort is the events surrounding BMO’s recent $680 million trading loss. From the day prior to the BMO announcement of its initial estimate of the loss until today, BMO shares have roughly traded in line with the bank group. It is therefore interesting that since the start of the sub-prime debacles, CIBC stock has lagged by 7% even though we believe that the loss will be less than half that experienced by BMO. The one difference is that BMO was relatively clear that it believes it has its “hands around” the problem. We believe that if CIBC was able to actually quantify a charge in the third quarter in the order of $100-150 million, the stock would actually perform well.

One material risk to our analysis is that CDO structures are opaque. We have heard second hand reports that even prospective buyers have been denied details on the collateral backing these structures. If the CIBC’s CDOs are in fact “mezzanine” structures or CLO-squareds, it is possible that their entire holdings could be worthless. We are assuming that they aren’t, and are also assuming that when management says the majority of the securities are AAA-rated they, as holders of (and in some cases underwriters of) the CDO are comfortable that there won’t be complete loss unless there is a widespread melt-down in the U.S. housing market. If we are wrong, the entire $1.2 billion position may be at risk. This would certainly hurt, but given CIBC’s ROE and capital position, we note that it would be handled without material risk to solvency.