Revenge of the Debt Zombies — or — “What are Banks for Anyway?”

What are banks for?  Typically, banks are described as intermediaries that take deposits and lend them out, earning what is called net interest margin on the gap between what is paid on the savings and what is earned on loans.  From where I stand, this description is wrong on three counts.

First, it suggests that banks need deposits to make loans.  In truth, banks create deposits in the process of granting a loan. Rather, the quest for deposits is motivated by two overriding profitability concerns, namely securing customers that can generate fee and loan revenue — bank customers are notoriously “sticky” so once you have them, you have them — and second, minimizing the outflow of net payments from customers to other institutions. These net outflows create liabilities in the overnight markets through clearing and settlement systems such as the ACSS and LVTS.   Net outflows entail borrowing costs. The name of the game is to minimize these costs by increasing the flow of deposits into your institution.  Large banks like RBC benefit from their extensive network of branches and large customer base — chances are, their customers write cheques to other RBC customers leading to no net outflow and no overnight borrowing requirement.

Second, the traditional story implies that lending volume has something to do with the cost of funds.  There is some truth in this proposition but I would argue that the greater truth is that lending is a demand-driven process shaped by expectations and changing asset valuations (or at least perceived valuations), which is why borrowing in the US is currently in the toilet.  Demand just isn’t there.   The real constraint on lending then is not deposits but credit-worthy customers and capital requirement / leverage rules imposed by regulators. More on this in a moment.

Third and finally, the traditional story misses the real function of private banks, which is to solve an information problem in the purest Hayekian senses.  That is, banks are or should be specialists in risk assessment and risk taking.  They should know their client, understand the local market and have their pulse on the broad economy.  Arguably, if properly structured, they can and should do this better than other entities such as governments.  In other words, the proper role of banks should be underwriting — lend money, hold the debt, and bear the risk.  Which is a long-winded way of getting to the main point of this post.

In the 1990s and 2000s, banks here and abroad tried to increase their lending capacity by offloading mortgage and other loan assets (lowering leverage ratios) through securitization.  Securitized assets were believed to be “off the books” — no longer the bank’s liability.  This freed up capital for new lending in two ways.  First, and it bears repeating, it removed the loans off the books.  Second, it allowed the bank to realize a profit on the loans immediately (albeit discounted relative to the hypothetical value of a buy-and-hold scenario), presumably juicing retained earnings and capital overall.   Of course, we now know that the banks generally did a lousy job of underwriting these loans — of doing their due diligence — because they believed they could offload them almost immediately.  They would not be the ones who were caught without a chair when the music stopped.  Or so it seemed.  The Barron’s piece cited below suggests otherwise.

The upshot of all this is that IF investors successfully put back these investments onto the banks, then this undermines the case for securitization — accountants will be hard-pressed to justify putting securitized assets “off balance sheet.”   This can only be a good thing as it will force the banks to get back to doing what they should do best — understanding and underwriting risk.  We should wish these investors well.


(Barron’s – Jonathan R. Laing) – Credit Risk/Losses

A new threat has emerged for U.S. banks that could seriously affect their earnings power over the next few years: the potential liability facing bankers from the (US)$2 trillion in subprime, alt-A and option-adjustable rate mortgages that they underwrote and sold to investors. The losses on the mortgages will be horrendous before the dust settles – over (US)$700 billion according to some experts.

And now investors – from the federal housing giants Fannie Mae and Freddie Mac to major bond managers like PIMCO and BlackRock – are fighting back. They are seeking to put back the mortgages to the banks from whence they came and force the banks to eat much of the mortgage losses. Their argument hinges on the arcane contract principle of representations and warranties. Namely, did the mortgages go bad because of the unanticipated nationwide collapse in home prices (a so-called exogenous factor) or are the banks responsible for the mess because they “misrepresented” to the mortgage purchasers the shoddy quality of the mortgages they put in securities and pools?

Already some buyers, such as Fannie, Freddie and the monolines have enjoyed a modicum of success in their putback efforts because they have stronger contractual language in their mortgage deals and thereby superior leverage in demanding documentation of all the mortgage loans underlying the securities.

But the biggest putback battles figure to come from the investors themselves in the subprime, alt-A and option ARM nonagency or “private label” securitizations who have suffered grievous losses. How much of a loss the big banks will suffer as a result of investors’ putbacks is anyone’s guess. The law in this area is largely untested. The banks figure to put up an epic battle because the stakes are so large and no federal bailout of big banks seems likely in this political climate. They are on their own. The guesses on the bank private-label securities losses range from as low as (US)$23 billion to worst-case scenario estimates of (US)$180 billion.


  • Are there such investors here? Investors who are trying to put CMHC Mortgage Backed Securities back into the Banks?

  • Good post Arun.

    Not sure of all the magnitudes but in the US and Europe off-balance sheet shadow banks reached enormous proportions. This was averted in Canada because regulators required all assets to to be on the balance sheet.

  • Hi Keith,
    Thanks for that. Not sure about the requirement to keep all assets on the balance sheet. There was and is a pretty vigorous securitization market in Canada (see this site for some discussion:

    This would seem to imply to me that securitized assets were, as a rule, moved off the books. See also here for a discussion that supports that interpretation:

    Happy to be proven wrong though.

  • Two types of loan off-loading have to be distinghished:
    1) off-loading of loans to an SIV (structured investment vehicle) which is an off balance sheet entity that is often sponsored by its creator -a private bank.
    2) off-loading to outside investors like PIMCO, BlackRock.

    In 1), SIV could not possibly accuse the original creator of the loan -the private bank- of defrauding it as the private bank is the parent company of the SIV! So this is almost an internal transaction except that the SIV was of course off balance sheet, which allowed the bank to circumvent capital rules and increase profits. The private bank however had to bring back the SIV on its balance sheet when the MBS market collapse and take huge losses like Citigroup did.

    In 2), the transaction is with an external party. Therefore, issues of fraud or misrepresentation are huge.

    In both cases, this is typical “control fraud” from the banking sector as it allowed exponential growth in profit in the short term through higher loan volumes (for a given level of bank capital). Maximising loan volumes is all about putting the emphasis on the “return” component of the risk/return tradeoff and as low an emphasis as possible on the “risk” component. This is how NINJA type of loans came to exist.

    Absent binding capital requirements and risk control, there is no limit to bank profits in the short term. This speaks to the importance of tight supervision and regulation.

  • Thanks for that Qc.
    Perhaps you could confirm, but I believe that regulators in Canada required that capital requirements apply to SIVs of Canadian banks, limiting (not necessarily eliminating) excessive leverage and associated risk. Could be wrong though.

  • Keith:
    The regulatory maximum leverage ratio in Canada did include some off balance sheet items, but it is my understanding that it did not take into account securitisation or SIV. Note that the regulatory maximum leverage ratio in Canada was raised from 20 to 23 in 2000 (23 was allowed if Banks met certain conditions). Even if we include securitisation, the average leverage ratio of Canadian banks never reached 23. It did go above 20 between 2006 and 2009 though.

    Also, note that SIVs are off balance sheet until they are not. Canadian banks did sponsor SIVs. These SIVs typically got funding through issuance of short term debt like commercial papers with the sponsoring bank often providing back-stop facilities. When the commercial paper market froze during the financial crisis, the sponsoring bank was suddenly on the hook. First, if it owned some of the SIV debt on its balance sheet, it obviously had to writedown its value. Second, even if the bank did not own SIV debt, the use of the back-stop facility by the SIV meant that the leverage ratio of the sponsoring bank was suddenly increasing -even if the bank did not consolidate the SIV on its balance sheet at the time. This de facto brought the SIV exposure under the limelight. Of course, when you are on the hook as the provider of a back-stop facility to a SIV, it is great to have the CMHC offering to purchase MBS as it allows you to shrink back your balance sheet and/or avoid distressed sales of high quality assets.

    It is my understanding however that Canadian banks sponsored SIVs did not have much subprime exposure which ultimately limited losses.

    Two good documents on this:
    Bank of Canada
    “The regulatory measure of leverage in Canada is the ratio of total balance-sheet assets and
    certain off-balance-sheet items to total regulatory capital (adjusted net Tier 1 and Tier 2 capital).
    The off-balance-sheet items in this measure cover all direct contractual exposures to credit risk –
    including letters of credit and guarantees, transaction-related contingencies, trade-related
    contingencies, and sale and repurchase agreements. These off-balance-sheet exposures – which
    currently account for about 75 per cent of all off-balance-sheet assets of banks (excluding
    undrawn liquidity lines and commitments) – are valued at their notional principal amounts.”

    “OSFI notes that during the recent financial turmoil, DTIs increased their balance sheet assets during times of stress in respect of previously off balance sheet assets that no longer qualified to be derecognized and in respect of securitization conduits that were no longer exempted from consolidation under CGAAP. Lessons learned in the recent financial turmoil are that certain securitization structures did not transfer as much risk out of the FREs as expected.”

Leave a Reply

Your email address will not be published. Required fields are marked *