Thursday 12 October 2017

The (real) truth about our debt

Livewire Markets today posted an article by Stephen Koukoulas from Market Economics, entitled "The Truth About our Debt". The full article can be found at the link below.

However, in my interpretation, the crux of Koukoulas' argument is that we need not worry about the high level of Australian household debt - now approximately 200% of GDP - because looking at the debt is only half of the equation, and the level of household assets has also risen significantly. Australian households, on a net basis (assets less liabilities), are in fact as wealthy as ever, and so - it is argued - fears over Australia's high indebtedness are misplaced.

One of the things I find encouraging about markets is how little most market participants and commentators seem to learn from the past. It gives me great confidence that markets will remain inefficient and, from time-to-time, outright irrational, thereby providing fruitful opportunities for investors that are prepared to think more deeply and learn from the past. For behind every great investment opportunity is a widespread misperception. 

In my opinion, Koukoulas' argument is fundamentally flawed for two reasons, which I explore below. However, what I find most extraordinary is that Koukoulas' perspective is the exact same argument that was used to dismiss concerns about growing debt levels in almost every major debt cycle in the past, including most recently in the years leading up to the 2007-09 GFC (Global Financial Crisis). Everything Koukoulas argues in that piece could have been (and indeed was) equally argued in the US, Ireland, the PIGS, and Iceland pre GFC, and also in 1980s bubble Japan. It is also a frequent argument made in defense of China's debt excesses today.

It is extraordinary how little people are willing to learn from the past, and iteratively course-correct their mental models for how the world works in response to new data. When people behave that way, it is usually due to a combination of ideology and ego. Strongly-held ideologies prevent people from re-examining core assumptions and conclusions in response to new data, or interpreting new data in an unbiased way. The facts are instead contorted or reinterpreted to fit with one's previously-held ideology. There remains a lot of untested or unfalsifiable (i.e. unscientific) ideology in the economics discipline, and it is seemingly immune from empirical data and testing. 

The ego dimension amplifies the above by embedding the interlocutor with an excessive degree of confidence in their own abilities and knowledge, and hence to underestimate the likelihood that they are wrong. If you do not believe you can possibly be wrong, there is little need to consider alternative perspectives or learn from the past or others because, after all, you already know everything. A little bit of humility would go a long way, and contribute to an open mind being maintained.

Fallacies of composition: The micro does not add up to the macro

The root cause of Koukoulas analytical errors is the same as many errors of macroeconomic analysis - falling for fallacies of composition. For many macroeconomic phenomenon, it is simply not the case that what is true at the level of the individual, necessarily scales and is true of the system as a whole. Furthermore, allowing for these fallacies often leads to unintuitive conclusions, and humans often struggle to integrate unintuitive conclusions into their world view, even if supported by evidence.

The most famous example of a macroeconomic fallacy of composition is Keynes' paradox of thrift. For any one individual, saving is a good and prudent thing. However, at a system level, if everyone saves a lot simultaneously, the economy will collapse, because one person's spending is another person's income, and a stable economy relies on the continuing circulation of monetary flows. If somebody decides to save, somebody else must borrow and spend those savings, lest the flow stop and the system fall into a downward spiral leading to a deflationary depression.

Failure to understand the paradox of thrift has lead to severe economic downturns in the past. It was a key contributor to the severity of the 1929-33 Great Depression, as budget hawks mistakenly extrapolated the wisdom of the prudent business or household to the level of government, and insisted on slashing government spending in response to falling tax income to keep the fiscal budget in balance. This ignored the fact that cutting government spending would further reduce aggregate demand and income in the economy at a time when the private sector was attempting to deleverage, resulting in lower income tax receipts, in a vicious cycle. More recently, failure to understand the dynamics of budgetary austerity has also resulted in the IMF and Germany forcing an overly-severe economic downturn on countries such as Greece (which has endured a 30% GDP contraction).

Why Koukoulas is wrong

Turning to the specifics of Koukoulas' argument, as noted, the crux of his case is that while household debt-to-GDP has risen, so have household assets as a percentage of GDP - and particularly housing wealth. We therefore need not worry, as household wealth is at an all time high.

There are two fundamental errors with this perspective. The first is that it overlooks the reflexive impact that rising debt levels have on asset values themselves. When an increased quantity of money is borrowed to purchase housing assets, it has the effect of pushing out the demand curve and increasing the aggregate purchasing power of property buyers, set against a relatively inelastic housing stock. Not surprisingly, this dynamic pushes up the market value (although not the intrinsic value) of the housing stock. Rising leverage therefore drives rising asset prices, and it also does so disproportionately, as the marginal borrower drives up the marginal price of all housing.

This effect is obvious in the stock market - when system margin leverage rises, it pushes up stock prices, and whenever a deleveraging is triggered/forced, it invariably leads to a sharp slump, and the impact on pricing is typically disproportionate to the absolute amount of leveraging/deleveraging involved. It is therefore to be expected that at the peak of a leveraged housing bubble, household wealth - on paper - will be at record highs! But try realising that wealth in an environment where debt levels are falling - the apparent equity will rapidly vanish into thin air. 

This is a key component of the reflexivity that drives booms and busts that George Soros has spoken about for decades. Rising leverage not only drives up asset prices, but then also increases the degree of apparent collateral, which supports yet further credit creation, while in the process making system LVRs (loan to value ratios) appear artificially low.  In addition, often overlooked is the fact that system leveraging - and the asset price inflation it triggers - is also invariably stimulative to the economy and consumer spending. Consequently, the denominator of many affordability ratios - i.e. income and GDP - is itself often inflated during a leveraged boom. The result is that the level of systemic risk is invariably radically underestimated during times of leveraging, and only becomes evident during the bust when the above movie needs to be played in reverse. In the fullness of hindsight, the boom and the inevitable bust that subsequently awaited it then appears obvious to all.

It is amazing that such elementary points about the functioning of the economic and asset pricing cycle are missed by commentators such as Koukoulas - particularly when (1) we have the experience of the GFC to reflect on within recent living memory; and (2) when investment greats such as George Soros have preached the above for decades, making such wisdom and knowledge available to all that care to listen. Why so many market participants refuse to avail themselves of the wisdom and knowledge of the investment greats, and attempt to reinvent the wheel, is beyond me.

The second major error Koukoulas makes is overlooking distributional effects. It is often argued (fallaciously) by economists that we can ignore debt, because one person's debt is another person's assets. We could all just pay each other back and the debt would be eliminated, but the productive machinery of the economy would not change. Koukoulas makes a derivative of this argument by saying households could pay back all their debt and have a comfortable asset position left over.*

The problem is this: while it is true that one person's debt is another person's savings, the debt and the savings are not distributed equally. Typically, the savings/financial wealth is concentrated amongst the wealthy (who also enjoy a disproportionate share of the income), whereas the debt - particularly during leveraged residential property bubbles - is spread more broadly amongst the middle class (who also have a comparatively smaller share of income). The consequence is that the debt-to-income ratios of the middle class can be much much higher than system aggregates suggest, and the median household wealth calculus much more precarious.

Furthermore, the flip side to there being an asset behind every debt, is that in order to wipe out the debt, you need to simultaneously wipe out an asset/deposit (or, in its place, bank capital). If when disaster struck, the wealthy were prepared to simply give their money to the middle class to allow them to pay back their debts, there would not be a problem, but in practice - barring an unlikely socialist revolution - that is simply not going to happen.

Instead, banks act as the buffer between debtors who cannot pay back their loans, and the wealthy who which to retain their assets/savings. Bank capital absorbs the hit, and if the hit is large enough, the banks become insolvent. This is what happened during the GFC, and it is exactly what will happen in Australia if the property market crashes and the economy experiences a severe recession.


The above is not to say that economic disaster is either imminent of assured - only to say that risks have risen considerably along with rising levels of household debt, and that people are correct to worry about those risks, rather than casually dismiss them in the manner Koukoulas has. This complacent and flawed perspective has lead to financial catastrophe too many times in the past already.


*Incidentally, Ben Bernanke made this same analytical error pre GFC, dismissing the risks posed by a build-up in household debt by arguing that distribution effects would need to be 'implausibly large' for that debt to have any economic significance. It was a profoundly mistaken ideological perspective.