Banking: they do things differently there

Illustration: John Spooner.From today’s Age and SMH column:

A pillar of economic reform is competitive neutrality. We strip government utilities of tax and regulatory advantages over private competitors because we want the best to win, not the most favoured. But banking is a different country. They do things differently there. Our regulatory arrangements pathologically favour the large over the small, and legacy business models over newer ones. This weakens competition, fattens margins and top salaries, and paves the way for future crises.

Where banks attract deposits to fund loans, a more recent business model involves ‘securitisation’: pooling loans into securities, and selling them to wholesale investors. Securitisation enabled non-banks like Aussie Home Loans to break into the banks’ market and force a near halving of margins.

Competing on a level playing field, securitisation might well dominate home lending. It avoids the banks’ extreme asset-liability mismatches – with banks’ short-term borrowing supporting long-term lending. And because they’ve been crafted for the external scrutiny of investors, mortgage-backed bonds (MBS) are more comprehensible to outsiders, whereas complex banking institutions with revolving managers getting bonuses for short-term performance are difficult – arguably impossible – to regulate effectively.

But it’s not a fair fight. When banking markets get skittish they can seize up without a risk taker of last resort. Three centuries of banking evolution, three centuries of crises have given us central banks which go lender of last resort delivering liquidity to distressed banks’ balance sheets when others are heading for the exits. As we’ve seen, securitisation markets seize up too and making them liquid would require similar action, though in this case the risk taker of last resort would provide liquidity to markets themselves, buying MBS during crises and offloading them in better times.

In addition to assistance with liquidity, large banks also enjoy an implicit guarantee of their solvency because, being too big to fail, they’ll be bailed out in a crisis. And that help translates into lower cost funding from wholesale lenders who expect to benefit from bailouts if it comes to that.

Our two banking regulators, the Australian Prudential Regulation Authority and the RBA, focus on “financial system stability” within the ‘official family’ of banks. Competitive neutrality with ‘shadow banking’ or securitisation gets short shrift. As Professor John Kay puts it “the principal objective of regulation appears to be to stabilise the existing structure of financial institutions … this goal is in fact a guarantee of further, and potentially more damaging, crises.”

Sure enough, when the crisis came, the big banks’ got huge public support. Taxpayers guaranteed their retail deposits for free and guaranteed their wholesale loans at fees reflective of credit ratings which incorporated the ‘too big to fail’ distortion, further disadvantaging the small fry.

While no investor in AAA rated Australian MBS ever lost principal during or after the crisis, liquidity in this market nevertheless collapsed. Securitisers needed a buyer of last resort. The Australian Office of Financial Management (AOFM) gingerly provided life-support, but still allowed the securitisers market share of home lending to fall by around three quarters while the major banks snapped up RAMS, Aussie, Wizard and Challenger’s lending arm.

The Bank of England calculates the annual too-big-to-fail taxpayer subsidy for the world’s 29 biggest banks at about $70 billion before the crisis. Today it’s around a third to half a trillion (yes, trillion) coincidentally around the same amount as our four monster banks collective market cap. Are we fixing this by busting them up (scale economies are exhausted well before you get to the majors’ size), or reducing their complexity? Well no.

In trying to minimise the chances of future crises the global regulator of banking regulators, the Basel Committee recommends banks hold buffers of “level one” liquid assets (mainly government bonds) and riskier “level two” assets on which they’ll be expected to draw before turning to their lenders of last resort – you and me.

But not here. Australia’s government debt is too low to provide our banks with enough level one assets and our regulators are concerned that they face similarly slim pickings of level two liquid assets – which include MBS. So the RBA will give them a loan – via its new “Committed Liquidity Facility” (CLF).

And why are there so few MBS in our market? Because, enjoying such exorbitant privileges, mortgages that might otherwise be securitised repose blissfully, implicitly guaranteed, on bank balance sheets. And what collateral will banks pledge to the RBA to draw on its CLF? Why the same mortgages that could be securing MBS!

So as the RBA ramps up its support for banks with policies that increase moral hazard as they depress demand for MBS, the AOFM’s last annual report announced its “increasing focus on the need to encourage a transition towards a . . . securitisation market that is not reliant on Government financial support.”

So much for a level playing field.

This entry was posted in Economics and public policy. Bookmark the permalink.

10 Responses to Banking: they do things differently there

  1. desipis says:

    I thought the reason the CLFs were brought in was:

    “… to reduce the systemic risk of excessive cross-holdings of bank-issued instruments. In our judgment would be undesirable for a bank to meet its liquidity needs by significantly increasing its exposure to the rest of the banking system.”

    It seems to me there’s a keenness to forget the lessons of the GFC and ignore the moral hazards that arise from a potential systemic failure of the financial system. I’m not sure its a good idea to bring back the fantasy that competition and market tradeable securities are sufficient to maintain stability in the system.

    And what collateral will banks pledge to the RBA to draw on its CLF? Why the same mortgages that could be securing MBS!

    So what you’re really after is a way for non-bank mortgages to flow into the CLF process? What’s stopping that from happening?

  2. Nicholas Gruen says:

    Thanks despis,

    A fair point, but if there’s systematic risk they’re already borrowing against that part of the system that sits on their balance sheet. If they had exposure to other similar assets, it might not give them much diversification, but it’s hard to believe it actually exposes them to more systemic risk than the current arrangement.

    And my own thinking was relating more to the moral hazard side of the ledger. Easy access to the CLF at what to me looks like bargain prices (0.25% over the cash rate when they use it plus a 0.15% facility fee) would have me salivating to use the facility if I were a bank.

  3. desipis says:

    Allowing “self-securitised” RMBS to be part of the CLF enables the banks to use the reasonably safe asset of residential mortgages as a liquid asset without involving other banks in the process. This minimises the number of failure points in the liquidity supply chain, lowing the risk of one distressed bank putting stress on others and thus lowing the risk of the whole system freezing up.

    I guess what I’m unclear on is why you think that banks using their own mortgages as liquidity via the CLF rather than wrapping their mortgages up as an RMBS for market based liquidity, somehow affects the net demand for RMBS from non-bank lenders. I was under the impression it was the lack of international demand for securities coupled with the vertically integrated (and cartel-like) nature of the local banks that dried up the funds available to non-bank lenders. I’m not sure there’s much the RBA can do to fix that.

    As for the fee and margin, this is from the RBA Speech:

    While at times like the present, liquidity can have considerable value, the Reserve Bank will not be varying the size of the fee through the cycle. Consequently, the facility is to be priced at a level that takes into account the value of liquidity in more normal conditions, as well as in stressed circumstances.
    From the history of the Reserve Bank’s own market operations, we can look at repo rates on some of the eligible securities to try and gauge how much a one month liquidity premium might be worth. The answer is not very much in normal circumstances, generally less than 10 basis points. Moreover, when you take into account the fact that to access the facility, an institution has to pay a penalty rate and, in most cases, with greater haircuts than previously applied, a ‘market-based’ valuation from historical rates would imply a low fee.

    The first bit of that does seem to as though the banks will be getting a bit of a freebie during periods where liquidity had high value. Other than that, I assume that the RBA knows how to do the numbers correctly to get the 15 basis points.

  4. Nicholas Gruen says:

    Hmm, I don’t think I’m following you.

    Allowing “self-securitised” RMBS to be part of the CLF enables the banks to use the reasonably safe asset of residential mortgages as a liquid asset without involving other banks in the process. This minimises the number of failure points in the liquidity supply chain, lowing the risk of one distressed bank putting stress on others and thus lowing the risk of the whole system freezing up.

    Do you agree that requiring banks to hold assets that they might take a loss on before running to mother bank would reduce moral hazard – leading banks to be more careful about drawing on the CLF – which exposes the RBA to higher levels of risk than in normal times?

    So that’s an important aspect of my concern. As for ‘involving other banks in the process’ well if one bank holds some RMBS as a liquidity reserve, they are not “involving other banks in the process. They’re holding assets that they have done due diligence on. They may be originated by other banks or others altogether, but they’re now owned by the bank that is using them to shore up its liquidity when it’s short of liquidity before it heads for mother bank. I can see how it could be objected that pushes up the cost of liquidity – as far as I can see it does. But, in an effort to minimise moral hazard, that’s what other countries’ regulatory structures are pursuing under Basel III.

    I don’t really follow this paragraph either.

    I guess what I’m unclear on is why you think that banks using their own mortgages as liquidity via the CLF rather than wrapping their mortgages up as an RMBS for market based liquidity, somehow affects the net demand for RMBS from non-bank lenders. I was under the impression it was the lack of international demand for securities coupled with the vertically integrated (and cartel-like) nature of the local banks that dried up the funds available to non-bank lenders. I’m not sure there’s much the RBA can do to fix that.

    Certainly the big bank’s lower cost of capital has the effects you’re talking about. But requiring them to hold RMBS would presumably increase demand for them – and liquidity in the market for RMBS. But it’s not a panacea I agree. And the government sharing some of the “risk bearer of last resort” role with the RMBS market – in this case by going investor (buyer) of last resort rather than lender of last resort would be much more effective and direct in meeting the competitive neutrality problems – and it would intensify competition with the banks.

  5. desipis says:

    Do you agree that requiring banks to hold assets that they might take a loss on before running to mother bank would reduce moral hazard – leading banks to be more careful about drawing on the CLF – which exposes the RBA to higher levels of risk than in normal times?

    Yes, but I don’t think that’s the whole picture. There’s two different risks the RBA needs to manage. There’s the asset orientated risk that you describe, where the RBA is left holding dud assets, but there’s also the system orientated risk of draining the market of its liquidity and causing a liquidity crisis. After the GFC it’s this second risk that is (arguably) of greater concern and how the non-bank lenders got screwed.

    My understanding of the need for the CLF is based on two things (and correct me if I’m wrong with these):

    a) There’s not enough (government, or otherwise) bonds around to serve as liquid assets for the banks. Thus if mortgage based assets are excluded from being used to meet the LCR the banks would have trouble getting sufficient assets.

    One way they could meet the LCR requirement would be to sell bonds to each other. This would potentially overexpose the banks to each other, creating the risk that if one go, they all go (solvency crisis). Thus, mortgage based assets need to be included somehow into the liquidity mix.

    b) While the banks have plenty of mortgage based assets, the Australian market isn’t liquid enough to have the assets to considered as liquid assets (without the CLF).

    Having the banks dump their mortgage based assets on the market (‘involving other banks’) during times of stress would absorb existing liquidity (cash) from the market, potentially cause asset values to drop and the market to freeze up. This would make all the assets held by other banks for liquidity no longer useful for that purpose and in turn cause them to become stressed (i.e. a liquidity crisis).

    Having the RBA provide liquidity only once markets are unable to (‘last resort’) would mean requiring the RBA provide liquidity to everyone. This would mean being at greater risk, which means a more significant moral hazard than if they’d dealt with the stressed bank directly.

    By promising to give loans against mortgage based assets when liquidity is needed (the CLF), the RBA is enabling the banks to use them as sufficiently liquid assets to meet the LCR. Dealing directly with the RBA for liquidity removes the dependency of the banks on the liquidity of the market and prevents one stressed bank from affecting the others (via the market).

    While the CLF does transfer a certain level of risk from the banks to the RBA, I think it’ll be much easier to price (and hence recover compensation for) this risk than it is to price the risk of the banks’ actions causing a liquidity crisis. Perhaps the price is low, but I’ll leave that question to the people with access to the data.

  6. Nicholas Gruen says:

    Thanks despis,

    I think you lay out the issues well and I think we agree about most of the mechanics of the mechanisms. I think you end up somewhere where I’m very uncomfortable, still, it’s where the ‘official family’ have ended up routinely since the crisis.

    The reasoning is that the RBA and the prudential regulation system basically applies to banks. The banks get all their liquidity assistance and in return they are subjected to regulation – the shadow banking sector avoids both. There’s also a nice division of labour here because the RBA does liquidity and monetary policy and APRA does regulation and solvency. But as we all know, indeed as Guy Debelle said a while back

    Clearly, the fact that a liquidity issue has risen may well be a signal that the market has concerns about solvency. Hence the central bank certainly should pay heed to that signal from the market.

    Still, there are exceptions (This is the point that Debelle went onto make.) And in those exceptional situations the bank then doles out liquidity – but does so only to the official family. Not only this but the penalty rates it applies, are very minor – 0.35% in the case of the CLF. That seems to ramp up moral hazard too much to me.

    And there’s a feedback loop. The policy of favours only to the official family continually undercuts the shadow banking system when it needs support most. That places the assets the system would need to become more liquid on bank balance sheets, preventing the market in the asset developing the liquidity it needs to become a liquid asset class.

    If you look at the logic of Basel III, the way I see it level one capital will be truly liquid. If anything its asset value will rise during tight spots – so long as the sovereign is credible. By contrast level two capital will come with its own penalty regime. If you get to using it to shore up your liquidity – you lose money – perhaps quite a bit of it as you sell into a falling market. I see this as healthy to a point, though of course it imposes some costs on the public good of liquidity.

    On the one hand higher fees, maybe much higher fees, would reduce moral hazard and enhance competitive neutrality with non-bank finance. Banks could then work out what liquid assets they’d like to hold to manage such liquidity problems without regulators specifying what those assets should be. On the other hand if the assets they are forced to thus invest in are not that liquid, this degrades the public good of liquidity and increases its cost. After the trauma of the last five years I’d take the hit on the cost of liquidity and back off the moral hazard. This is the choice that other countries are making under Basel III.

    Finally, I’m not sure I go for the idea that supporting the market itself is so much worse policy – or that it’s (ultimately) riskier. In all other areas of policy we’d plump for competitive neutrality and think that if a job’s worth doing, it’s worth doing properly. Providing the assets are well understood, I’d be quite happy for the state to be market maker of last resort in an important asset class. Indeed we did it with RMBS via AOFM, but we did it reluctantly and without much conviction. The RBA ‘smooths and tests’ the foreign exchange market. Do we know why? I don’t really think it’s very well articulated why we do that. It’s a pretty liquid market after all. (I’m actually in favour of the RBA trying to be a fundamentals trader in the FX market, buying low and selling high, but ‘smoothing and testing’ is pretty different to that.) But the case for buying low and selling high in RMBS seems pretty strong to me.

    So, to try to conclude, I’d impose Basel III type obligations on our banks re level one and two capital. Higher fees for the CLF, maybe much higher fees, would reduce moral hazard and enhance competitive neutrality with non-bank finance. Banks could then work out what liquid assets they’d like to hold to manage such liquidity problems without regulators specifying what those assets should be. That would bring more RMBS off bank balance sheets and onto the RMBS market – making it more liquid. At the same time, where banks were forced to sell RMBS and this was creating liquidity problems in RMBS, that’s the time for the authorities to buy low and later sell high.

  7. Nick De Cusa says:

    “a) There’s not enough (government, or otherwise) bonds around to serve as liquid assets for the banks. Thus if mortgage based assets are excluded from being used to meet the LCR the banks would have trouble getting sufficient assets.”

    Completely ridiculous. The claim here is that government debt serves some sort of rational purpose.

  8. desipis says:

    Nicholas, I think I have a better understanding of where you’re coming from. I’ve also improved my understanding in this area generally, so I appreciate this thread.

    I agree that the banks are getting access to the CLF cheaply. I do think the RBA directly dealing with the RMBS would be a fairer way to provide liquidity, but I’m wary of just how long and deep the RBA would need to get into the RMBS market to do so. The GFC left quite a large hole.

    The RBA ‘smooths and tests’ the foreign exchange market. Do we know why? I don’t really think it’s very well articulated why we do that. It’s a pretty liquid market after all.

    I had always assumed it did so because the wider economy is affected by short term fluctuations in the exchange rate. Smoothing it reduces the impact of noise on anyone involved in international trade. It’s worth pointing out that foreign currencies will hold their value independently of the Australian economy.

    If the economy hits some real trouble, people might start defaulting on mortgages and RMBS would lose value. If the RBA were to take on a significant amount of RMBS assets, it would essentially be using the government’s strong credit rating to leverage itself into the mortgage market, all the while pumping up a potential real estate bubble. Would the advantages be worth that risk?

    Your broader approach seems to be for the RBA to shift into a more substantive banking role. Which brings up the question, why did we sell off the CBA in the first place?

  9. Nicholas Gruen says:

    “I’m wary of just how long and deep the RBA would need to get into the RMBS market to do so.” I’m not proposing going beyond AAA grade paper. If there’s a dearth of that paper around, then that has to be taken into account. And I’m not arguing against the CLF as a backstop, just that we should put more things in its way before banks use it. And doing so will help develop the very liquidity we’re after. There’s cumulative causation here.

    “I had always assumed it did so because the wider economy is affected by short term fluctuations in the exchange rate. Smoothing it reduces the impact of noise on anyone involved in international trade.” I’d say that the Australian economy suffers from serious misalignment of the currency and that that’s why the RBA should be a fundamentals trader in our dollar in FX markets. That happens to coincide with the RBA’s and Australia’s interest in making money on the trade. If it can do this in the process of ‘smoothing’ very well and good. Market movements are, after all, fractal, so it may be possible – and beneficial to do this other than at what are guessed to be the top and bottom of the market. But ‘testing’? What the hell is that about? Perhaps the RBA have set all this out and I’ve missed it, but it seems the policy is not clearly articulated – but indeed I’m in favour of them being a fundamentals trader – which is not very different from what I want them to do in the RMBS market.

    “Your broader approach seems to be for the RBA to shift into a more substantive banking role. Which brings up the question, why did we sell off the CBA in the first place?” I’m reasonably comfortable with selling the CBA, just because, as is the way with these things, it had become pretty indistinguishable from its competitors in structure and motivation, so why would one bother keeping it in public hands? The point here is that central banking evolved before the age of securitisation and also before the internet. The changes I’m talking about here update it to the age of securitisation and there’s plenty of rethinking that could be done around the many and various disintermediation opportunities offered by the internet. I sketched one such idea out – though there are some other important ones – here.

Leave a Reply

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

Notify me of followup comments via e-mail. You can also subscribe without commenting.