Friday, 8 April 2011

How the bubble burst: securities, mortgages and collateral damage

In my previous post I drew attention to the lax attitude to risk of the US investment banks and institutional investors arising from the tacit government backing of investment banking, which led to them taking ever greater financial risks in search of higher and higher returns.  There has been much discussion of the collapse of the "derivatives tower" which brought down Bear Sterns and Lehmann Brothers among others, and there has been major criticism of securities trading in general and mortgage-backed securities in particular.  I think this criticism is unjustified and distracts attention from the real issue, which is the worldwide failure of retail bank lending (more on this in my next post).  So in this post I aim to debunk securities and derivatives trading and show that the investment banks that were brought down through the collapse of the "derivatives tower" were not the cause of the financial crisis, as has been widely reported, but victims of it.

First, some basic facts about securities and derivatives. 

1. Securities.
security is a piece of paper that can be bought or sold through a recognised financial trading exchange.  Securities are issued by organisations such as companies and governments as a way of raising money. Basically these organisations sell pieces of paper which represent shares in the equity of the organisation (share certificates) or loans made to that organisation (debt securities or bonds).  Buyers of securities include pension funds, governments, rich people and even some poorer ones. Once the security has been sold for the first time (new issue), it is traded freely on recognised trading exchanges until the issuing organisation buys back the equity or settles the debt.  When the organisation does this the payment goes to whoever owns the security at the time, not to the original issuer. It's a bit like Pass-the-Parcel - the security is passed around among various investors until the organisation calls it in, when whoever is holding it gets the prize.

Securitisation is the process of pooling contractual debts such as mortgages, credit card loans and car loans and selling bits of paper that represent that debt to various investors.  Suppose for example that Bank X has lent £10,000 each to 100 people for car loans.  The total debt owed to it by these people is £1m.  It could issue a single piece of paper saying £1m, and hopefully someone very rich who doesn't mind having all their eggs in one basket will buy it. Much more likely, though, is that Bank X chops the debt up into £10 portions and sells 10m £10 debt securities.  These are snapped up by institutional and private investors and traded on as and when they wish.  When the debt is redeemed the paper is everywhere and lots of pension funds and individuals get a proportion of the payout.

2. Derivatives
A derivative is a financial agreement whose price is determined by the performance of one or more underlying assets.  Hard assets underlying derivatives are shares, bonds or currencies, but derivatives can also be based upon market indexes (the derivative price is governed by movements in these indexes) and interest rates (the price is governed by changes in interest rates).  The actual derivative itself can be regarded as a type of insurance or a means of offloading risk. 

For example, I might want to buy some of Bank X's debt securities (see above), but I think the issue price is a bit steep.  So I purchase an "option to buy" (known as a CALL option) these bonds when the price falls to a level that I think is reasonable.  I pay a small amount of money for that option - like an insurance premium.  The total amount of option premium plus agreed purchase price for the underlying bonds should be less than the amount I would pay if I bought them at their current price.  If the price of the bonds doesn't ever fall I lose the premium, of course.

Alternatively, Bank X might decide to keep some of its car loan debt - after all, it earns interest. But it doesn't like the fact that the interest rate on these loans is fixed.  After all, the base rate is very low at the moment.  If interest rates rise Bank X will lose out because new loans would be at higher rates.  Really it wants to be able to raise the interest rates on these loans, but it can't because that breaks the contractual agreement. So it finds someone who has the opposite problem (has variable interest rates but wants fixed) and they agree to swap their interest rates.  This is called an interest rate swap. 

Derivatives can be - and often are - issued as securities and traded on exchanges.  The examples I have given here are "over the counter" deals (no securitisaton or exchange trading involved), but the bank could simply issue pieces of paper saying "warrant" (which is an option to buy bonds) or "fixed/floating interest rate swap" and wait for investors to buy them.  It doesn't need to know who the investors are.

Now to return to the "derivatives tower" and its underlying assets.  I have explained how the securitisation process works.  When this is applied only to residential mortgages the ensuing securities are known as "collateralised mortgage obligations" (CMOs).  But of course retail banks have a range of different types of secured and unsecured debt and there is no particular reason to limit securitisation to only one type of debt.  A security created from different types of secured debt, including but not limited to residential mortgages, is called a "collateralised debt obligation" (CDO).  There is nothing wrong with spreading debt in this way - in fact securitisation is a highly effective means of spreading risk among a much greater number of players and thereby reducing the risk to each individual organisation.  But it was CDO trading, above all, that brought down the US investment banks. So what went wrong?

To understand what went wrong, look at the wording of the security. "Collateralised debt obligation" means that the debt represented by the security has collateral associated with it. Collateral is a hard asset which is made available to the lender in the event of default by the debtor. A mortgage is the best-known example of a collateralised loan, where the house bought with the mortgage can be repossessed if the homeowner doesn't keep up with mortgage payments, but collateral could also be shares or valuables.  But what if the value of the asset isn't enough to repay the loan? 

Mortgages have traditionally been regarded as "safe" forms of lending. Mortgage risk is routinely valued at 50% or less, which means that in calculating how much capital to hold to support this lending banks only need to take into account less than half the loan value: the rest is assumed to be covered by collateral.  But the aggressive expansion of mortgage lending in both the US and the UK made these loans far more risky.  When the defaults started to rise massively because mortgage lenders were lending silly income multiples and giving loans of 100% or more(!) of the property value to people who couldn't really afford houses, the property market in the US and UK crashed.  House prices fell and repossessions skyrocketed, especially in the US. Lots of people found their house value fell so much that it didn't cover the amount owed on the mortgage.  Suddenly huge numbers of mortgages were no longer adequately secured - they became high-risk loans.

Investors expect to be paid for taking on higher risks.  Well, this isn't unreasonable, because after all they could lose out if it all goes wrong.  So when the assets underlying a security become more risky, the price of
 the security rises. In my previous post I noted that because investment banks are funded by clearing banks, which are supported by government, both dealers and institutional investors stopped worrying about the risks inherent in the instruments they were trading and looked only for higher returns.  So when the price of the CDOs started to rocket because the underlying mortgages were junk, dollar signs went up in the eyes of both dealers and investors.  Trading in these toxic instruments actually INCREASED as the property bubble burst.  Dealers put together ever more complex structures mixing high-risk mortgage debt with safer types of security to increase the return to those investors - including banks such as HSBC - who didn't have a high risk appetite. This was the so-called "derivatives tower", and it was built from the start on shifting sands.

Eventually, though, the genie came out of the bottle. The US and UK governments were forced to bail out their major mortgage lenders - Fannie Mae and Freddie Mac in the US, and Northern Rock, HBOS and Bradford & Bingley in the UK. Mortgage-backed securities and their derivatives, including the fancy packages dealers put together to attract sounder investors, became worthless and investors lost an AWFUL lot of money. A large number of US institutional investors - especially hedge funds - failed, and some investment banks were bailed out by the US government.  HSBC suffered losses on its worldwide income due to its investment in now-junk mortgage-backed securities, but covered this from its own capital.  It was HSBC that raised the alarm about the failure of the sub-prime securities market - it saw the crash coming but too late to protect itself completely from loss.

I'm not saying that there wasn't a fair degree of stupid and immoral behaviour by dealers and investors in the run-up to the collapse of the residential mortgage market.  But the real villains were the retail banks and mortgage lenders, who had destroyed what had always been a safe long-term lending market by their irresponsible, aggressive and possibly fraudulent expansion of mortgage lending. And the worm in the apple - the mechanism that makes it possible for bankers to mismanage their business so appallingly - is the support that all governments give to their retail banks. I shall return to this in my next post.

And before anyone feels too sorry about the businesses that failed and the dealers who lost their jobs, spare a thought for the thousands of people in the US who lost their homes.  They, and taxpayers in the US and Europe who are now feeling the pinch because their governments bailed out the banks, are the real victims in this unholy mess.

Thursday, 7 April 2011

Bank breakups, red herrings and elephants

The Independent Commission on Banking (ICB) is expected to produce its interim (well, final really except for the horse trading) report on 11th April 2011. The twittersphere, egged on by various newsblogs, is full of people attempting to pre-empt its findings by shouting for the banks to be broken up.  They seem to think that if the banks are broken up into smaller units it will ensure that no bank can ever be too big (or too important) to fail.  And in particular, they believe that if banks are forced to specialise in either retail (clearing) banking or investment banking, there will be no repeat of the financial crisis.  They are wrong.

Let's consider the case of Bank A.  Bank A is a clearing bank.  Its business consists of transactional banking - current accounts, ATMS, cheque processing, automated payments.  It is big, because it needs economies of scale in order to provide streamlined processing, and it has thousands if not millions of customers worldwide.  It accepts deposits from its customers into a variety of interest-bearing savings accounts, and it provides a range of secured and unsecured loans to individuals and businesses.  This is traditional retail banking as we understand it.

Now let's look at Bank B. Bank B is an investment bank.  It trades actively on the international financial markets, both on behalf of its customers who are predominantly institutional investors such as pension funds and "high net worth" individuals (very rich people), and also on its own account - this is called market making.  It may accept deposits into interest-bearing savings accounts particularly from its rich individual customers, and use these funds for trading on the financial markets, but it doesn't lend to its customers.

Clearly these are very different forms of banking activity, and prior to the repeal of the Glass-Steagall Act in the US, they would have been conducted by completely separate institutions.  The UK was not subject to this Act of course and therefore has had an integrated banking system for much longer than the US.  But under current legislation in both countries Bank A and Bank B can be exactly the same bank - let's call it Bank AB.  Banks like to have both forms of activity under one roof, as it were, because then the investment banking benefits from access to clearing facilities for settlement (particularly international payments facilities such as SWIFT), and retail banking benefits from being able to access the international financial markets.

But why would a retail bank benefit from access to the international financial markets? Aren't its lending commitments supported by money from its depositors, and aren't its clearing activities self-supporting (after all, it's only moving money around, isn't it?).  Wrong, on both counts.

In fact all retail banking activities are underwritten by government.  Transactional banking (clearing) is supported by central bank reserves. Deposits benefit from an unlimited taxpayers' guarantee. Lending benefits from the "fractional reserve lending" system in which banks can effectively invent the money they wish to lend and the depositors' guarantee ensures that the central bank will bail them out if it all goes wrong.    The whole retail banking system depends on government support from end to end.  It is not a business at all but a public service.  Breaking it up will simply make it clunky and inefficient - as retail banking was in the US in the 1980s and 90s (I speak from personal experience).

Investment banking, on the other hand, is supposedly self-supporting.  Investors knowingly put their money at risk in the hopes of gaining a return, and similarly the bank itself knowingly accepts risk.  Understanding and managing risk is the whole business of investment banks. 

Banks that integrate retail and investment banking obviously benefit from the government guarantee of retail banking spilling over into investment banking through funding of investment banking activity from retail deposits.  Because banks can invent the money they lend to customers, they don't need to lend out depositors' money directly to customers.  Instead, they use this money to fund their investment banking activities.  Doing this alllows them to make money directly from financial trading using savers' money while at the same time earning interest payments on invented money (loans).  That is lots more money than they would earn if they could only lend out money they have received in deposits.  No wonder the banks like the fractional reserve lending system - and no wonder they don't want investment and retail banking separated.

However, let's suppose that the ICB recommends that Bank AB should be split into its component parts - a government-supported retail bank, A, and an independent investment bank, B.  Would this prevent a future financial crisis?

The trouble is that as I mentioned above, clearing and investment banks can trade with each other.  Supposing that Bank B gets into difficulty because it has invested in some very dodgy financial products which it was hoping to sell later at a profit but the market for them has collapsed.  "Tough luck" would be the normal response, wouldn't it?  Well, no, apparently not.  You see, Bank B settles its trades through Bank A, and Bank A lends to Bank B (interbank lending) to fund its trades. Nothing illegal about this - in fact Bank B, as it is not a clearing bank, has no alternative but to settle through Bank A as it needs access to the international payments system.  Remember I said that clearing banking is supported by central bank reserves?  Bank A's interbank lending activity is effectively underwritten by government, either because it is lending depositors' money (since it now can't trade directly on the financial markets) - and retail deposits are guaranteed by government - or because it is inventing money underwritten by central bank reserves. So in effect Bank B has access to government funds to settle its trades, even though it is supposed to be self-financing.

Suppose that the (unscrupulous) management of Bank B knows that the funding it receives from Bank A is effectively guaranteed by government? Well, it's obvious actually - the bank clearing system can't be allowed to fail, because it's used to pay people's wages, and their tax bills, and their council tax, and their mortgages, get the idea.  And even if it doesn't have enough depositors'money to fund Bank B, Bank A can simply invent some more money, because although there are supposed to be rules requiring it to hold a certain amount of capital reserves to support its lending, it's been years since anyone bothered to enforce them.  So Bank B knows that Bank A will ALWAYS fund it, and it knows that if Bank A gets into trouble because of the support it is giving to Bank B, the government will ALWAYS bail it out.  Knowing that, Bank B can take on ridiculous risks. In fact it doesn't even need to know what the risks are.  All the financial instruments it trades can have a triple-A credit rating, even if they are complete junk.

If you think this scenario is imaginary, or could never happen because no government would be so stupid as to allow unlimited funding of investment banks by clearing banks, you would be totally and completely wrong.  This is EXACTLY what happened in the US - as is evident in this after-the-event analysis by Susan Krieger.  Investment banks took on more and more risk in the search for higher yields for their investors.  Neither they nor the investors had any idea what the real risks underpinning the fancy financial instruments they were trading were, and they didn't care.  When the whole mortgage-backed securities pyramid collapsed, they were bailed out by the Fed, as they expected.  Well, except for Lehmann Brothers, that is.  The US govt allowed this one to fail.  And there were HOWLS of outrage from the other banks.  The US government hadn't honoured its implicit guarantee!  TREASON!

Now, in this nightmare scenario there is NO legal relationship between Bank A and Bank B.  There doesn't need to be.  They can trade with each other openly and unrestrictedly.  So breaking up banks that do both retail (clearing) and investment banking is a complete red herring.  It will achieve precisely nothing.  The retail banks will remain "too important to fail" and underwritten by government, and the investment banks will continue to be funded by them.  And there will be another financial crisis within a few years, and more taxpayers' money poured into failing banks.  Because there is effectively an open conduit of government funds to investment banks indulging in high-risk financial trading.  Unless this is closed, investment banks will continue to gamble on the financial markets with impunity, and pay themselves obscene amounts of money for doing it.

One option that has been mooted is to prevent retail banks from lending to investment banks (and vice versa, presumably) - Mervyn King is thought to prefer this idea.  The price retail customers would pay for this would be lower returns to bank depositors (as if they aren't low enough already) and probably a much more restricted range of retail savings accounts - in fact, rather as things used to be before the massive expansion of financial services in the last 20 years.  However, the investment banking effect would be much more interesting. Investment banks have to borrow on the international money markets to fund settlements. But the money markets would be much less liquid than they are at present because retail depositors' funds would no longer be available to them. Evidently the cost of funding settlements would soar. This would significantly reduce the retuns to investors, both individuals and institutional investors such as pension funds.  Pension fund returns are already pathetic. How much lower can they go before people give up putting money into pensions and stuff the money under the mattress instead? (See my previous blogpost on savings).

An alternative would be to continue to alllow clearing banks to fund investment banks, but in a much more transparent and regulated way.  This seems to be the preferred option of the Fed's Ben Bernanke. The issue here is really the extent of regulation and its enforceability.  Imposing much more extensive requirements for reporting of market risk and liquidity would permit the funds conduit to remain open - provided that the regulations were rigorously enforced.  But regulation of banks has historically been far too lax. The regulators don't enforce the rules that already exist, not least because of major conflicts of interest.  How can someone who is CEO of a bank also be on the board of the body regulating that bank?  That is exactly what happened in the UK, and the bank concerned - HBOS - was eventually brought down through years of fraud and mismanagement to which the regulator turned a blind eye.  Personally I am not convinced that regulators can be sufficiently knowledgeable and independent - and resourced - to undertake the task of enforcing detailed risk management procedures and reporting in all investment banks. 

And finally - the elephant in the room.  People like to believe that it was the evil derivatives traders on the financial markets who brought down the banks and caused the financial crisis.  Retail banks of course are much too well-behaved to do anything to cause them to fail, so they can be trusted with an unlimited taxpayers' guarantee across all parts of their business, can't they?  Let's go back to the good old days of traditional retail banking and get rid of these fancy financial markets. Who needs them, anyway?

Wrong again.  ALL FOUR of the banks that failed in the UK collapsed not because of risky trading on the financial markets but primarily because of very aggressive retail lending funded by short-term borrowing and  insufficiently supported by capital reserves.  In other words, in the UK it was the RETAIL banks that failed, not the investment banks. 

I believe that we must address the issue of lax regulation of retail banking.  Aggressive expansion of business activities funded through short-term debt is a high-risk strategy in any business, and in a retail bank - which as I said above is a public service, really - is absolutely unacceptable.  The FSA should have noticed that Northern Rock was funding its mortgage book in the overnight money market - completely inappropriate use of short-term funding to support long-term liabilities and a clear warning sign that the company was seriously short of liquidity.  Nor should HBOS have been allowed to leverage up its lending so much that its capital reserves only accounted for 3% of its total liability. And as for RBS - it was so overexposed across virtually all areas of its business that its eventual failure was inevitable, and I frankly wonder how it survived as long as it did.  Behind all of this was a conflicted and ineffective regulator and a Government that was only too happy for banks to make more and more money, never mind the risk, because it wanted the increased tax income to fund its spending programmes. Totally insane.

The present Chancellor seems to think that handing responsibility for bank regulation back to the Bank of England (BoE) will solve the regulation problem. I guess he's too young to remember BCCI, but you'd think someone would have told him about it. Personally I don't think the BoE is the appropriate body to regulate and supervise retail banking, though it may be ok for investment banking provided this industry is denied access to BoE funding (let's not have any MORE conflicts of interest, please!).  I would like to see a completely independent regulatory body - an OFFIN for retail banking, recognising that it is actually a public service in private ownership and therefore should be regulated in the same way as the privatised utilities. And I would like to see investment banking separately regulated by a body made up of people who REALLY understand financial risk management.

Obviously I don't know what the ICB will recommend.  I would like its recommendations to be comprehensive, achievable and effective, but I suspect that they will be partial and ineffective. 
I really hope, for everyone's sake, I am wrong.