Given at the International Financial Services Forum, Dublin
Thank you for inviting me to speak this morning. We are approaching the ninth anniversary of the beginning of the global financial crisis. And we are still talking about it; but not just talking, also publishing books, reports, holding conferences and of course releasing films. There is no doubt more than one reason for the continued interest in the crisis, but as public officials charged with putting into place the measures to prevent a repeat and thus produce a more stable financial system, the work is still in progress and thus for us very much a matter of debate. Today, I want to put that work into perspective, looking at banks and the wider financial system.
I am going to organise my comments around something very basic to the system, the balance sheets of banks. It is a striking fact, that the combined balance sheet of the major UK banks increased by around fourfold between 2000 and 2007. That’s a stark fact: yes, there was an accounting change in that period, but still this was a massive change. And other countries often matched that change – the story of Ireland is sadly well known. Let me stop the story of the past for a moment and draw out one message: when we hear people say things like, “credit isn’t back to pre-crisis growth rates”, “there is a gap in the stock of credit to the economy relative to the pre-crisis trend”, “the level of national output is this much lower than it would have been had the crisis not intervened”, “banks’ so-called market making in financial assets is much less than it was immediately pre-crisis”, just pause and consider that all of these statements imply that the pre-crisis years were sustainable. They were not.
Let’s turn to the balance sheet of banks in a little more detail, starting with the liabilities side. In this respect, banks are different from other firms because they provide deposit contracts. A deposit is a very particular form of debt contract. For all of us the essential feature of a deposit contract is simple – we put our money (our asset) on deposit at a bank, and we expect all of it back, with whatever rate of return is agreed, and we expect to have access to it, in part or whole, in line with the terms of the contract. Some deposit contracts provide for more ready availability than others, and this affects the return on the deposit. Of course, there is also insurance on a deposit contract up to a well publicised level.
Let’s contrast that with non-deposit debt (bonds) and equity contracts. If you invest in either of these, you are not promised all of your investment back, though of course you expect to make a return. A debt investor ordinarily expects the full return of their investment, but viewed from a corporate restructuring perspective this is not assured. I will come back to this. Investment in debt and equity can be managed on a collective basis, as asset management. You may remember the Woody Allen quote that a stockbroker is someone “who invests your money until it’s all gone”. Cruel I know, but it illustrates the difference between deposits, debt and equity.
Turning back to banks, they have deposits, debt and equity on the liabilities side of their balance sheet. The problems of the pre-crisis period in this area of bank liabilities can be summarised quite easily: they had too little equity to absorb losses, and particularly so given the major expansion of balance sheets (more leverage of the equity); they had issued forms of hybrid debt-equity that were supposed to absorb losses after equity but could only do so if the bank became formally insolvent, something that could not be allowed to happen in view of the threat to the financial system; they had taken on a very large increase in short-term wholesale market funding which was assumed by the providers to have the characteristics of deposits. Moreover, short-term wholesale funding was footloose, more so than retail deposit funding.
What have we done to deal with these problems? We have increased the required equity capital held by banks – the first loss absorbing element of their liabilities. We have required that hybrid debt - equity instruments have automatic and transparent triggers which are not exercised only by an event that cannot happen. In the UK, banks have set the trigger for conversion to equity high enough at a 7% common equity tier one capital ratio to seek to ensure that conversion happens well before the bank has run out of capital. This is the right thing to do. We have introduced liquidity regulation designed to enable banks to withstand the loss of more footloose, short-term wholesale funding, which has gone alongside a sharp shrinkage by banks in their use of that funding. In the UK, we are introducing structural reform or ring-fencing which will provide for an internal separation inside the major banks so that the so-called qualifying EU customer deposits must be inside the ring-fence. Being inside the ring fence does not I should emphasise protect against all risk; indeed, the crisis saw the failure of a number of banks that were predominantly of the same nature as a ring-fenced bank will be. But it does keep the balance sheet on which most customer deposits will appear in the large UK banks in a more simple state, and that should be conducive to effective governance and supervision without being burdened with great complexity, and also facilitate recovery and resolution actions should they be necessary.
Also on the policy response to the crisis, we have invested in stress-testing to examine how the balance sheet performs under stress, and thus establish whether the bank is robust in terms of its loss absorbency to extreme but plausible outcomes. This is a central part of what we describe as our forward looking judgemental approach to supervision.
I have not so far mentioned one other very important part of the post-crisis reform package, namely ensuring that banks can be resolved if they fail. This should be done without recourse to public funds or the disruption of the critical economic functions provided by banks, such as continuity of access to deposits. At the heart of the resolution reforms is the concept of bail-in, which is that private creditors of a bank absorb the cost and provide the new equity to sustain the provision of these critical functions in the event that it suffers large losses that cause failure. There is nothing radical about the concept of bail-in, which is a debt-equity conversion should a trigger event occur. That is consistent with well-established principles of company restructuring. But with banks it has to happen very quickly if the trigger event that the bank is no longer viable occurs. We cannot wait for a lengthy bankruptcy process. TLAC (Total Loss Absorbing Capacity) or MREL (Minimum Requirements on Eligible Liabilities) are used to describe the specific liabilities which banks will be required to maintain to ensure there are enough liabilities which can feasibly and credibly be converted. TLAC and MREL are comprised of debt-instruments and equity.
The key point here is that this important reform requires a very clear distinction among debt contracts between forms of subordinated debt contracts which will bear losses in resolution first and deposit and other senior funding contracts on the liabilities side of the bank balance sheet. This may seem like a very simple point, but past arrangements meant it was unclear, and the moral of the painful story is “don’t leave the providers of debt funding in any doubt about their creditor status.” The lack of clarity about the instruments on the liabilities side of bank balance sheets and the order and mechanisms by which they would absorb losses was an important issue in the crisis. Looking ahead, clear disclosure of creditor hierarchies at a legal entity level will be a critical component of the resolution regime, so that all creditors know where they stand before resolution occurs.
Let’s turn now to the other side of a bank’s balance sheet, the assets side. The massive increase in balance sheet size pre-crisis accompanied what came to be known as the “search for yield”, which could be re-named the “search for risk which turns out to be unsustainable”. On the banking, as opposed to trading book side of the assets of major banks, the search for yield often took the form of loans which included too much equity-like risk because the equity stakes of the owners of the companies were too small. This was not equities themselves, but rather equity components embedded in loans. Commercial property lending in the UK is a good example, as it was in Ireland.
Commercial property prices in the UK have over time tended to vary more than, say, prime residential house prices and that points to a need for larger equity-like component of funding to absorb losses from that variability. The search for yield however meant that banks were more happy to lend against equity risk of this sort. Since the crisis in the UK we have seen a shift in commercial property finance towards a larger share coming from funds which explicitly take this equity-like risk. This strikes me as a good thing in terms of reducing the exposure to such risks of banks, whose liabilities tend more to be in the form of deposits. I sometimes hear comments that banks are losing the race to new innovations such as peer-to-peer lending for the supply of what I would describe as finance with a heavy equity component. I think we need both in a well organised financial system because as banks have a predominance of deposit liabilities with the characteristics I described earlier they are not on their own natural equity providers.
This brings me to the last part of the story, namely the growth of non-bank asset management in its broadest sense as the size of bank balance sheets has tended to reduce post-crisis. It is striking that if you were familiar only with a chart of the evolution of global assets under management over the last twenty years, you should be excused for failing to spot that a global financial crisis had occurred.
It follows from what I have said that this shift from banks to non-banks makes sense as we seek to achieve a clearer demarcation of the types of liability or funding contracts and the assets best suited to go with them. But it only makes sense if a few conditions are met, two in particular. First, that there is no lack of clarity about the nature of the assets held under management. Thus the label on the can is an accurate description. Take as an example the recent failure in the US of the Third Avenue Focused Credit Fund (focused on what a comedian might say). The purpose of the Fund was to invest in high yield debt. The label seems to have made that clear, and it wasn’t therefore in the form of a AAA label which obscured the reality. The failure of this Fund has not made major ripples all on its own. Why? I would argue because there was no obvious lack of clarity around the assets, and this is a reminder against the re-appearance of opaque instruments and complex tranching.
The second condition to meet is that there is no illusion about the liquidity of the assets. It is critical that investor expectations are well adjusted to the prevailing liquidity conditions. A lot of work is under way to assess the market liquidity implications of the expansion of assets under management, and this is important for our remit in financial stability. An important part of this work is to do all we can to ensure that investors understand the characteristics of the assets they hold, and that the liquidity promised by funds to their investors mirrors the market liquidity of the underlying investments.
In conclusion, the objective of the financial reform programme is clearly not to return to the unstable pre-crisis years, however attractive some of the statistics of that period look in isolation. The response has been to strengthen the loss absorbency of the banking system but also to enable greater clarity between the different forms of bank liabilities consistent with the economic terms of the contracts. This may seem like dry balance sheet stuff but the benefits are major. First, greater clarity on liabilities should encourage more appropriate matching with assets in a way that was absent in the run-up to the crisis - some assets suit a deposit contract, some do not. Second, resolving failed banks without the use of public money depends on a very strong delineation of bank liabilities to make clear what can be bailed in. This is a key to unlocking the answer to the too big to fail problem. Third, as the Bank of England announced in December, we are now in a position to clarify the expected steady state regime for bank capital and because we can distinguish going concern capital and thus loss absorbency from resolution or gone concern absorbency, we can put a lower number on the former. Clarity is therefore a good thing when it comes to balance sheets. Put like that, the wonder is that so much was unclear in the previous system. And, finally, the same principle holds for assets under management. Make sure investors know the characteristics of the assets they hold and that they are not promised access to their investment on terms or with a promise (explicit or implicit) that is out of line with what financial markets could deliver. Thank you.