The life of a banking system analyst can be lonely indeed. One tends to both offend and support all various interests at different times. I’ve been accused on MB of being a supporter of both Karl Marx and Ayn Rand whilst being a rent seeker for the non-Mega Bank Australian ADIs.
I can honestly say I’m not a follower of any particular economic theory or have a barrow to push for any particular vested interest. I simply analyse how the financial system works with a view to identifying weaknesses and strengths. I’m not aligned to the dogma that is associated with every movement whether economic or otherwise. Movements to become popular dumb down their messages to simplistic bites that can be drummed into many but with the necessary detail and true meaning lost in order to advance a simple but often dangerous meme.
Dumb downed messages may allow the masses to develop a chant, but they can become very dangerous as they are easily twisted to support almost any human or organisational behaviour.
Modern Monetary Theory seems to chant that borrowings of a central government in its own fiscal currency do not matter as the government can always print to repay the debt. In a single sentence that statement can sound like quite a sensible proposition. However, when I try and work through the rhetoric and dogma that is MMT, that single proposition is used to support all types of illogical chants. One thrown at me on MB was that offshore borrowing by Mega Bank in a foreign currency is about term not funding, as it actually does not raise any net funds. Give me a break!
Speaking as a banker who has raised funds offshore for banks and for my own finance business, I can assure readers that its about funding not just for the organisation but also the country. As for the last few years foreign investors have been buying Commonwealth debt which has resulted in significantly increasing deposits and reducing Mega Bank’s need for increasing its own offshore funding this is not a trivial matter.
There is another meme that seems to abound that if an organisation raises funds offshore in a foreign currency and swaps back to $A, then that is equivalent to borrowing directly in to $A. It’s not.
Money is still owed in a foreign currency and must be repaid in that currency. The borrower however, has created a derivative where a party agrees to pay or receive the difference FX rate between $A and the foreign currency. The derivative creates risk for the swap counter parties but does not change the underlying debt transaction. In cases where the debt is never repaid but just continually rolled over (as the majority of Mega bank debt), I’d argue that creating the swap, which also needs to be continually rolled over, is an unnecessary risk transaction that distorts the FX rate but does not eliminate the underlying FX obligation for the borrower or for the country.
To melt or not to melt
A special piece of dogma is the slow melt of the Australian housing market which seems to be entrenched on this site as the likely future outcome. However, this proposition is never supported with any deductive logic. A slow melt is a massive assumption that Australia’s younger generations are happy to purchase expensive housing with little hope of future capital gains to offset the added debt costs.
What’s even worse of course is the assumption that these generations should pay a large share of the costs of the over leveraging the country has taken on. The biggest losers of a sudden resetting of housing prices will be those that have both created the over leveraging and have benefitted from it. These Australians should pay for their excesses and not rally around an argument that paying the price of the past should be shared as that’s somehow good for the economy. Its as bad as banksters arguments that they should be paid excessively because they’re job creators.
However, the biggest false meme in financial risk management is that history or past performance can be relied upon to as a guide to manage future risk. Sounds reasonable as a general rule but like all dogma, such propositions can and are twisted for highly dangerous consequences.
Recently I had the distressing experience to be the first on the scene at a car accident with a double fatality. The accident occurred right outside my property, somewhere on the NSW’s coast. It was a single car accident on a sunning afternoon with speed playing no part. A very bizarre situation that at first seemed to defy explanation for such an accident by a driver that had driven the same road many many times in similar circumstances.
I had a great deal of time to analyse the situation waiting for the services in this remote part of the coastal Australia. When you see a situation in its raw state without the disruption that occurs when help arrives, you can learn a lot. Rather than the accident being freakish, it was apparent that the many small things conspired to be in place so that the young girl who tragically died may as well have had a loaded gun with a hair trigger pointed at her head simply waiting for the bump that would set it off. There is no blame in my assessment rather its was a stark reminder to look at everything when you are putting someone who is totally reliant on your actions for their health and well-being and not simply rely on the premise that if something has not happened then its unlikely to happen.
Generationally, I believe we have the same responsibility to our children and all descendants to protect them from our actions when they are fully reliant on the situation we leave for them. That’s the simple explanation of why I continue to expose the weaknesses in the Australian financial system. Its not about predicting whether disaster will occur or not, its about whether the risk is acceptable or not, for not only the current set of adults but future generations.
An idea is a dangerous thing
This brings me to a piece of research that I think is excellent but analagous to the use of models within Australia’s Mega bank: ‘The Formula That Killed Wall Street’? The Gaussian Copula and the Material Cultures of Modelling” (link goes to a PDF download) by Donald MacKenzie and Taylor Spears
Its well worth a read and you don’t have to understand the math to understand the issues and risks it exposes. In Australia, perhaps we could substitute “Local Councils” for “Wall Street”
In short, certain models were used by investment banks and rating agencies to determine the risk and to manage CDOs (collaterised debt obligations). Because these models were fundamentally flawed, particularly in relation to asset backed security (ABS) CDOs, massive losses occurred within the holders of the securities generated. These losses were critical in triggering the GFC.
Let’s look at the authors description of their research.
The paper draws upon this history to examine the articulation between two distinct, cross-cutting forms of social patterning in financial markets: organizations such as banks; and what we call ‘evaluation cultures’, which are shared sets of material practices, preferences and beliefs found in multiple organizations. The history of Gaussian copula models throws light on this articulation,
The authors research is focused not on the technical aspects of the models but why were they used, why they were used when flaws were widely understood and why were they adopted across the market?
I have a great deal of experience in ABS structuring and understand the risks and how they are put together. When I first saw a proposal for an ABS CDO, it was obvious to me, that the correlation assumptions of the underlying collateral were faulty and based on a corporate debt model without regard for the underlying mortgage loans. My view was of no consequence as rating agencies used the models and AAA rated securities were created in the hundreds of billions for the unsuspecting investors, and here’s why.
As well as being partially embedded in communication amongst banks and hedge funds, the Gaussian copula also played a pivotal role in a crucial organizational process: determining traders’ bonuses. The critical issue here was how and when profit should be ‘booked’. My interviewees reported a universal desire amongst traders for the profits on a credit derivatives deal (most of which lasted for between five and ten years) to be recognized in their entirety as soon as the deal was done – as ‘day-one P&L’ – and so to boost that year’s bonus as much as possible.
The researchers discovery above is pivotal in organizations adopting what may be legitimate algorithyms in an illegitimate manner no matter what the consequences. The model allows traders to book a total calculated profit today, but the model also provides the legitimacy of good risk assessment, an astounding situation when the faults were well know. The whole game was simply based on the hope that the blow up doesn’t happen on the traders watch and damn the consequences for those caught with the risk.
The result of the embedding of Gaussian copula models in governance via ratings was the large-scale ‘gaming’ of them and of the other models employed by the ratings agencies. The crisis was caused not by ‘model dopes’, but by creative, resourceful, well-informed and reflexive actors quite consciously exploiting the role of models in governance. ‘[T]he whole market is rating-agency-driven at some level’,
Is this a unique situation? No! besides the fact that Gaussian copula models are still be used, just ask the London whale, Australia has its own version of model gaming which may have devastating consequences for the innocent.
As I’ve posted on previously, Mega Bank uses IRB models to determine risk weighted assets and therefore capital. These models use probability of default (PD) of loans based on historical performance as a primary driver in determining the risk weighted assets. In relation to mortgages these models are used to justify very low levels of capital, less than 2%, with devastating future consequences.
Any decent quant would realise that basing PD on historical performance in a rising house price market or over a period which is not long enough to cover different economic circumstances is invalid. This makes no difference, Basel II rules give legitimacy to the methods and low capital requirements allows for higher returns on capital and the ability to write more loans, ensuring the bankers are rewarded today. The similarity of the situation with the misuse of Gaussian copula models should not be missed.
I see it as my responsibility to at least highlight the risks in the system that should be obvious to an independent analysis. Its whether anything will be done before the fall out is heaped on the innocent is the question.