According to press reports the IMF may allegedly be increasing its estimate of global bank losses to $4 trillion, a figure consistent with estimates by a variety of independent bank analysts
Apr 10, 2009
But if you think that $2.2 trillion was already a huge figure, the
RGE Monitor new estimate in January 2009 of peak credit losses (available in a paper for our RGE clients) suggested that total losses on loans made by U.S. financial firms and the fall in the market value of the assets they are holding would be at their peak about $3.6 trillion ($1.6 trillion for loans and $2 trillion for securities). The U.S. banks and broker dealers are exposed to half of this figure, or $1.8 trillion; the rest is borne by other financial institutions in the US and abroad. The capital backing the banks’ assets was last fall only $1.4 trillion, leaving the U.S. banking system some $400 billion in the hole, or close to zero even after the government and private sector recapitalization of such banks and after banks’ provisioning for losses. Thus, another $1.4 trillion would be needed to bring back the capital of banks to the level they had before the crisis; and such massive additional recapitalization is needed to resolve the credit crunch and restore lending to the private sector. What are the implications of these figures? Let me explain.
These figures suggest that the US banking system is effectively near insolvent in the aggregate. Of course this does not mean that all or most institutions are insolvent but that many of them will be – at the peak – with a massive capital shortfall. Of course time can heal many wounds and with borrowing costs for banks cost to zero and net interest margin rising some banks may be able to rebuild capital over time in spite of severe losses from marking down the value of loans and securities. But some institutions are so severely damaged that even rising interest margins, forbearance and time will not be enough to let them recover their capital losses; i.e. some – even large –institutions may be insolvent under most likely states of the world. Note that the difference between the RGE estimate of $3.6 trillion of losses and the IMF estimate of $2.2 trillion – in spite of taking the same analytical approach – was that the IMF was looking at current charge-off/delinquency rates while RGE made estimate of peak losses based on a sensible scenario about the US economy (growth, home price declines, unemployment rate).
But it has now reported in the press that the IMF will soon publish new revised estimates of credit losses of $4 trillion;
of this estimate $3.1 trillion is losses on loans and assets originated
by US financial institutions while $0.9 trillion are losses by
European and Asian institutions. Since the RGE estimate is for loans
and securities originated by US institution the appropriate comparison
is between the $3.6 trillion estimate by RGE and the $3.1 trillion of
the IMF. So at this point the IMF estimates and those of RGE are
converging towards a very similar figure. RGE has not made yet an
estimate of losses on loans and securities issued by European and Asian
institutions but we estimated (based on Fed and IMF estimates) that
about 40 to 50% of the losses on securities are borne by non-US
investors. Thus, about $1 trillion of the US originated securities
losses are suffered by non-US institutions because of this
international credit risk transfer. If you then add this $1 trillion to
the IMF alleged estimate of $0.9 trillion losses on loans and
securities originated by Europe and Asia total losses for these
European and Asian institutions would be $1.9 trillion that is not very different from our estimate of $2.6 trillion of losses ($3.6 trillion minus $1 trillion) borne by US investors and financial institutions. One critique of the RGE and IMF estimates of credit losses is that the estimates of losses on securities are based on current mark to market values of such securities. If current prices are depressed because of a large illiquidity premium then long term losses will be smaller. While this argument is partially valid it does not significantly affect the overall estimate of losses. Suppose – to be very very generous to the « liquidity discount » critique – that such illiquidity premium is 30%. Then, our estimate of securities losses of $2 trillion needs to be revised downward to $1.4 trillion. Then, our estimate of total credit losses goes down from $3.6 trillion to $3.0 trillion, a figure very close to the most recent alleged IMF estimate.
It should also be noted that the RGE figure may be an underestimate
rather than an overestimate - of peak losses. Because of the lack of precise data we have not included losses on two important categories of assets, municipal bonds and sovereign bonds of emerging market economies. Given the likely sharp increase in default rates by state/local government (in 1991 at the time of the last real
estate-driven recession muni bonds were trading like junk bonds) and given that sovereign spread are now very high losses on these two categories of assets are likely to be very significant. Also, at the time of our January 2009 estimates we assumed that the US unemployment rate would peak in 2010 only to 10% (an estimate now shared by the OECD for the US and most advanced economies), But there are now signs that the unemployment rate will peak closer to 11% if not higher (and the current unemployment rate is already above 15% if you include partially-employed workers and discouraged workers who left the labor force).
The RGE and IMF estimates of very large credit losses are consistent
with the analysis of most independent analysts of the banking system,
respected and authoritative figures such as Meredith Whitney (recent victim of cheap and unwarranted shots against her by a WSJ commentator), Chris Whalen and Mike Mayo.
As recently reported by Bloomberg Mike Mayo predicts that US loan losses may exceed Great Depression levels:
CLSA analyst Mike Mayo assigned an « underweight » rating to U.S. banks, saying loan losses may exceed Great Depression levels and the government may be forced to take over large lenders.
Financial shares and major U.S. stock indexes dropped after Mayo advised clients to sell banks including Winston-Salem, North Carolina-based BB&T Corp.
and Cincinnati’s Fifth Third Bancorp. Mayo said in a report today that
he assigned « underperform » ratings to Bank of America Corp. and JPMorgan Chase & Co., the two biggest U.S. banks by assets.
« While certain mortgage problems are farther along, other areas
are likely to accelerate, reflecting a rolling recession by asset
class, » said Mayo, who joined CLSA from Deutsche Bank AG last month. « New government actions might not help as much as expected, especially given that loans have been marked down to only 98 cents on the dollar, on average. »
The 46-year-old Mayo gained recognition in 1999 at Credit Suisse AG for correctly taking a bearish stance on bank stocks when other analysts remained bullish. After being fired from Credit Suisse, he joined Prudential Equity Group in 2001, where he earned a reputation for criticizing investors and companies who tried to curb objective analysis. At Deutsche, Mayo had « sell » or « hold » ratings on all 18 companies he covered, according to data compiled by Bloomberg... Nationalization of banks remains a possibility because government policy remains unclear, Mayo said on a conference call after releasing his report.
Existing government efforts aimed at boosting bank capital don’t « preclude regulators from taking harsher action, » Mayo said. « I don’t want to be a partner with the government in investing in bank stocks. » Mayo said he expects loan losses to increase to 3.5 percent, and as high as 5.5 percent in a stress scenario, by the end of 2010. The highest level of loan losses in the Great Depression was 3.4 percent in 1934, according to the report. Mayo’s estimate matches the prediction he made on March 10 for Frankfurt-based Deutsche Bank.
Mortgage-related losses are about halfway to their peak, while
credit-card and consumer losses are only a third of the way to their expected highest levels, according to Mayo, who declined to comment beyond the report. CLSA is an affiliate of New York-based Calyon Securities.
The nation’s largest banks may be transitioning from a financial crisis marked by writedowns of capital to an economic crisis featuring large loan losses, Mayo wrote. The U.S. government cannot provide much relief because its actions will lead to either banks having to raise new capital or toxic assets remaining on banks’ balance sheets, Mayo wrote.
Mayo said solutions to the banking crisis will take time, as the increase in risk happened over a decade or more. CLSA’s underperform rating reflects the expectation that the stock will underperform the local market by 0 to 10 percent, while a sell rating expects it to fare worse by more than 10 percent, according to the report.
Seven Deadly Sins
Mayo said banks engaged in « seven deadly sins »: greedy loan
growth, gluttony of real estate, lust for high yields, sloth-like risk management, pride of low capital, envy of exotic fees, and anger of regulators. Mayo’s « underweight » rating applies to the entire sector.
Meredith Whitney also remains very bearish on US banks:
Meredith Whitney, who left Oppenheimer & Co. in February to found Meredith Whitney Advisory Group LLC, said in a Forbes interview that banks will continue to write down their mortgage assets as home prices decline further than lenders expected. Home prices are not done falling and will ultimately drop 50 percent from their peak, Whitney said today in a CNBC interview.
The unemployment rate also has exceeded banks’ projections and could lead to further loan losses, Whitney told Forbes. Banks « by and large » will show profits in the first quarter before provisions for loan losses, Whitney said on CNBC.
Also, as reported by Forbes:
In a report to clients of the Meredith Whitney Advisory Group,
the closely followed bank stock analyst reiterated her bearish position on the financial sector and the economy in general. Though some are forecasting a recovery in late 2009 or 2010, Whitney believes that banks have still not properly reserved against greater than expected losses in home prices. Her earnings forecasts for 2009 and 2010 are almost across the board lower than consensus.
In Whitney’s narrative financial firms will relive the worst struggles of 2008 because housing prices in the major markets will fall much further than expected. Bank of America, HSBC and even the resilient JP Morgan Chase will have to increase reserves as real estate losses mount unabated. Home price expectations for the banking industry play a critical role in their entire accrual accounting methodology. « When a bank carries both a loan and a loss reserve against such a loan on its balance sheet, where that bank expects home prices to bottom is a key assumption much like unemployment or interest rates, » Whitney warns.
Her report, titled, « The Agony of Incrementalism » forecasts home prices to fall by more than 66.0% of current bank assumptions in the 10-City Case-Shiller Index.
« Increased liquidity drove home prices higher, » Whitney explained, « and contracting liquidity will drive home prices lower. » She pointed out that 70.0% of homeowners need leverage to buy and stay in their homes, therefore an overall declining mortgage market will put pressure on prices.
As home prices fall and as unemployment rises, banks will have
to retain earnings to fund greater reserve funds as part of a cycle
that Whitney says has « no end in sight as both forecasts continue to rise quicker than expectations. » It’s a « never ending game of catch up, » she says because the banks have been underestimating losses ever since the credit crisis began a year and a half ago. The average bank thinks the total decline in housing prices was going to be 30% at the end of the first quarter. Now, says Whitney, they’re thinking more like 37% -- still behind reality.
Similarly, Chris Whalen of Institutional Risk Analytics has recently pointed out that economic losses for US banks could be as high as $4 trillion:
Remember that the maximum probably loss (« MPL ») shown in The IRA Bank Monitor for the top US banks with assets above $10 billion, also known as Economic Capital, is a cash number representing the amount of incremental capital the banks may require to absorb the losses from a 3-4 standard deviation economic slump, such as the one we have today. If you include the subsidy required for the GSEs and AIG, the US Treasury could face a collective funding requirement of $4 trillion through the cycle.
Do Ben Bernanke and Tim Geithner really believe that they can sell such a program to the Congress? To put it in perspective, the $250 billion in the Obama Budget for additional TARP funds will not quite cover Citigroup (NYSE:C).
Finally, given these negative assessments of the prospects for US banks by independent analysts and given forecasts of credit losses by the IMF that are now allegedly close to the ones of RGE
what should we make of the recent sharp rally of banks and financial stocks; of the recent statements by banks such as Bank of America and Citi and other banks that they are making profits in Q1 (before provisioning); of the better than expected earnings results of Wells Fargo that led to another stock market rally on Thursday; and of the recent press reports suggesting that most if not all banks will pass the stress test?
The detailed answers to those questions will be provided in my next blog in the next few days. In brief: banks are benefitting from close to zero borrowing costs; they are benefitting from a massive transfer of wealth from savers to borrowers given a dozen different government bailout and subsidy programs for the financial system; they are not properly provisioning/reserving for massive future loan losses; they are not properly marking down current losses from loans in delinquency; they are using the recent changes by FASB to mark to market to inflate the value of many assets; they are using a number of accounting tricks to minimize reported losses and maximize reported earnings; the Treasury is using a stress scenario for the stress tests that is not a true stress scenario but rather a benchmark of what the economy is likely to look like in 2009 and 2010; a true stress scenario would have considered a much more serious economic downturn..