I write with reference to a discussion in a recent Bloomberg View article, We’re Still Not Sure What Causes Big Recessions.
Debt/broad money supply is a key foundation of asset and human capital values and their supporting GDP flows. Because of this, wealth and debt effects (new loans create deposits) on GDP/income flows should not be considered as separate forces.
Debt in its money supply origination (bank deposits) is a foundation of both GDP flows and asset values and it is when debt, and specifically in the form defined, increases relative to GDP/national income flows that we should pay attention. And we need to pay attention to all flows, not just income flows on risky assets, for example corporate profits which can squeeze out returns on both fixed interest and human capital during periods of enforced low interest rate policy.
Money leverages many activities, and asset values are always to a certain extent in a form of a bubble, but excess leverage, especially during periods where we have structural imbalances and frame transitions creates instability and risks to the financial system.
Frame transitions that we need to watch out for with respect to excess asset focused money supply growth are where drivers of GDP growth are in decline (labour and population demographics, productivity and global transitions impacting the same) requiring lower levels of capital or growth rates of capital accumulation resulting in increasing levels of capital depreciation. In this context monetary frame dynamics should also be contracting or slowing. Frame transitions can be accentuated by increasing income and wealth inequality, something that may also be an emergent property of economic systems during frame transitions. This can also leverage asset prices to prospective GDP flows.
If you borrow to invest, for example, and if borrowing raises both demand for assets and their prices, then the impact of a decline in wealth in the event of an asset price shock will play out to greater or lesser magnitude according to leverage. Asset prices are dependent on system leverage, on the amount of money in the system and, in particular, the amount of money allocated to assets classes within the market portfolio: if the market wishes to reduce % money balances the only way this can be done, apart from transferring money to consumption, is to raise asset prices relative to money. While an individual can rid him or herself of money via asset purchases, the market in aggregate cannot.
Moreover the flows that underpin asset values, such as GDP/national income, are also likewise leveraged to greater or lesser degree and the degree of leveraging of these flows can compound an asset price shock. Highly indebted consumers or corporations, in frame transitions exhibiting declines in national income growth rates, will be more exposed to shocks of any kind. Asset price/wealth shocks are also a function of liquidity demand within the system (itself a function of numerous factors and relationships), asset price to GDP flow divergence and hence system leverage at both the asset and GDP flow level.
So we have leverage in the asset market (from asset focused loans, quantitative easing, low interest rates) and we have leverage of economic flows (consumer credit/deposit based lending and commercial bank lending to corporations as well as debt issued by corporations). Leverage upon leverage at extremes (asset price to GDP/income flows divergence) is what we appear to be left with. And of course let us not ignore the increased velocity of asset focused money and the loan/liquidity spirals of the shadow banking sector. All these flows are linked.
Separating out debt and wealth is like trying to separate out the balloon and the air within it, like ignoring the hand in the glove that punches you.
I have some qualms over the discussion in the Bloomberg View article, We’re Still Not Sure What Causes Big Recessions, if not for the simple fact that it also ascribes the cause of the recession to the financial system without definition:
“…..Almost everyone agrees, at this point, that the Great Recession of 2007-09 was caused by the financial system.”
The financial system, the balloon per se, did not cause the great recession. At a basic level, the many imbalances that had built up within the system (the air, the hand in the glove) were the causes of the financial crisis; they came back through the banking system as debtors defaulted on home loans, loan collateral values declined and off balance sheet vehicles came back onto bank balance sheets: and of course as loan sourced GDP expenditure shut off. What led to the imbalances was more complex.
These imbalances arose amidst the confluence of a number of long running global transitions:
D – Accommodations attempting to offset A,B and C: lower interest rates and higher money supply growth that has boosted debt amidst a slower growth frame and that have led to asset price shocks impacting both the financial system and the frame itself.
E – Inter temporal and structural imbalances: the maturing of developed economy growth and the accelerated growth of developing economies and the over dependence on the convergence of the two within an acceptable time frame.
The “monetary system” created important fuel for the asset imbalances (asset focused money). Central banks were heavily involved: interest rate policy was a big determinant of increasing asset focussed money flows as well as leveraging consumption expenditures. And we should not forget the impact of weak regulation and the ethics of those involved in “fraudulent” loan originations in the US that accentuated issues with the securisation of assets (suggested readings William K Black).
Just to ascribe blame to the system ignores the many agents and routes through which serious imbalances were allowed to accumulate. The system is always going to exist and therefore will always be an entity through which shocks of all sizes are delivered! The system did not cause the shock, the shock was already within it.
Systems do not necessarily create bubbles, energy does and systems do not create collapse but rather they have structural limits to which they can be pushed or shifted out of balance. The relationship between a system’s energy, structure and emergent properties of that structure define the shocks and their impact on its dynamics. In a sense we are always in a bubble in a monetary system where asset prices discount the future, where asset prices are determined by a) a money supply’s asset focus and b) within this MS, preferred portfolio money allocations. This is part of the reason why it is often difficult to determine when a shock is more or less likely, i.e. given the fact that shocks are natural within monetary based financial systems. But we have also had quite active central bank support of monetary expansion since the late 1990s (LTCM/Asian Crisis) that has added yet further fuel to this particular fire, so what is a naturally leveraged habitat has become especially so; perhaps this familiarity and acceptance of the natural volatility of such systems has given us a false sense of security.
Divergence was a key theme in the build up to the crisis in key developed economies: one of the key divergences, in key developed economies, was the rise in consumer debt relative to GDP and incomes. This equates to leverage of the critical flows supporting asset prices and excess leverage increases the sensitivity of these flows to future shocks of any magnitude.
In the US consumer credit (FRB G19) relative to GDP, income and personal consumption expenditures has been on an inexorable rise from the 1990s onwards (A,B,C,D,H below). Mortgage debt growth, in particular home equity lines of credit which were key drivers of expenditure prior to 2008, peaked prior to the crisis but was a key driver of asset prices and consumer growth prior to the crisis (B & Q).
One of the key issues in addressing the impact of higher consumer debt levels is the fact that real income growth (in aggregate) prior to the breaking of the crisis, although trending down, did not appear outwardly weak (F).
This is deceptive: during periods where we have high rates of debt financed consumption, national income flows will naturally increase (credit is spent, becomes revenue, comes out as national income), but the flows belie their dependence on credit growth. These flows however were becoming more highly leveraged and more exposed to the increasing leverage. Higher consumer credit or consumption focused loans to GDP/income flows means that GDP/income flows are more highly leveraged and thus more unstable in the event of a shock to the system. This is one of the key concerns with China’s economy in the sense that current GDP flows are highly leveraged on debt financed gross fixed capital investment.
Higher levels of consumer credit, relative to income and GDP, create instability with respect to expenditure in pretty much the same way that higher investment leverage increases risky asset investment risk. While new credit is important for growth in flows, the existing stock of credit is relevant for the overall stability of those flows. Note the charts showing the increased importance of consumer credit as a % of personal disposable income (A, D) and cumulative personal consumption expenditure relative to wage growth (I).
Post crisis, the reliance of consumer expenditure on consumer credit has risen higher still in the US (A) and while overall consumer debt relative to GDP and income levels has fallen (C), the dynamics of debt relative to GDP growth and income growth rates have not (C & H). A substantial part of the drop in debt was due to defaults on higher risk securitised mortgages (1, 2) and not really consumers significantly and voluntarily cleaning up their balance sheets: chart M shows show the surge in mortgage debt in excess of deposit growth which is an indicator of the securitised lending excesses of the US housing boom. The decline in mortgage debt in the US was more likely attributed to defaults and reduced capacity/demand for loans as opposed to balance sheet repair.
Continuing system stress posed by low wage growth (post 2007), income inequality (for much longer) and high debt levels remains in situ in the US and other countries.
It is also worth pointing out that while asset price/GDP divergence continued to widen post crisis consumer durable goods expenditure has been a different matter: here real prices stand at levels little different from the late 1990s (L).
The divergence between asset prices and GDP or income metric, is another key and strongly related divergence. While asset prices to GDP/income is often pointed to as a positive attribute (K), it must be noted that asset prices are leveraged by debt, by preferred holdings of money within the market portfolio at low interest rates, by quantitative easing, and, as noted, by leveraged economic flows (C+I/Y). That is, given the leverage provided by a monetary based system, it is not the relationship between asset values and income or debt but the relationship between debt and GDP/national income flows that matter for stability and the risks posed by debt and asset price shocks.
Interestingly the growth rate of key GDP flows (in the US) has weakened (V) significantly since the late 1990s and this is replicated in many other advanced economies. Global monetary conditions can also be added to the mix and represent another element that can increase the asset focus of the domestic money supply – I buy USD with Euros/Yen, do not increase the USD MS, but if I buy assets with them I make those USDs asset focussed.
While mortgage debt growth in the US has ceased to be a major contributor towards asset focused money supply growth post 2007, the slack in monetary accommodation has been overtaken by quantitative easing (R) and commercial and industrial lending (T) and of course consumer credit.
Over the period August 2008 to August 2014 the monetary base rose at a rate close to 51% of the rate at which consumer debt in total grew over the 6 years prior to Q3 2007. We know that QE is primarily an asset focused expansion of broad money supply (i.e. bank deposits) so its relative magnitude is likely of a higher order.
Excess asset focused MS will create divergence between asset prices and GDP. This divergence can hold latent demand shocks via a number of routes.
If the economic frame is increasingly dependent on loan financing for GDP/income growth, and if this is not discounted in asset pricing, and especially where such loan financing is likewise increasingly dependent on asset values as collateral (and human capital income flows), then any shock that raises the cost or limits the supply of loans (or impairs the growth rate of income to finance those loans) is a risk to demand and the growth rate of demand and hence to asset prices given that GDP flows support (valuation) and fund debt (interest and capital repayments). One of the major mistakes of the US Fed was to raise rates as high as they did pre crisis, given the build up of debt and weakening debt dependent growth fundamentals.
Add this to a weakening growth dynamic (productivity, population growth/demographic changes, increasing income inequality) and higher debt levels imply higher levels of system instability and fragility: ever larger reactions to even smaller shocks.
Monetary stimulus in a weakening growth frame provides only temporary relief to debt burdens and asset valuations while encouraging an increased asset focus of money supply and additions to debt to GDP ratios. If the growth rate of nominal flows continue to weaken, you require successively lower interest rates to maintain the necessary gap (to sustain asset values) between the nominal flows and the discounted flows to support asset pricing given system leverage. But this ignores the compression of flows outside this dynamic . Note the asset liability issues with respect to funding future liabilities at low interest rates (1, 2, 3, 4, 5, 6)
Shocks to asset prices and demand can therefore come from the impact of continuous downward revisions to growth expectations. Monetary accommodation depends on a prior trend re-establishing itself otherwise, as we have seen, debt to GDP and asset valuations to GDP risk continuing to diverge. Divergence implies an accumulating potential future demand shock by virtue of the combination of asset based (in all its forms: QE, loan based, low IR portfolio adjustments) and expenditure based leverage in a low growth frame with high transitional risks. Moreover, high asset prices in a low growth frame with transition risks (demographics changes, income inequality, productivity risks) may also be exposed to overly optimistic return expectations: unrealistic return expectations represent a future demand shock transmitted by lower returns on assets supported by asset focussed money supply and unconventional monetary policy.
Higher levels of debt change the structure and dynamics of the economic frame, especially where the growth rate of the frame is in transition and monetary policy has attempted to counter the transition. The relationship between money growth and economic growth has changed: from simple accommodation of supply side expansion to engineering demand over and above the rate at which the frame seems capable of providing. The very fact you have excess asset focused MS growth and asset price divergence implies issues with the frame, its balance, impairments and conflicts with transitions.
The primary interest rate conflict is not between inflation and growth, but between asset prices and a potential asset price shock to growth and the financial system. Increasing income inequality and weak wage growth keeps the US and other economies within a debt/asset value/IR bounded endogenous money supply chokehold. A successive series of debt/asset bubbles and interest rate lows are not a succession of unrelated incidents but a tightening of an extremely dangerous grip.
As discussed in a number of previous posts, slowing growth rates and contracting frames (population decline/aging populations) require different levels of operating and working capital. Trying to stimulate demand, money supply growth and asset values to counter changes in trend dynamics impairs the monetary and productive capital adjustment the economic system would be expected to go through. It is not a necessary condition in a frame adjustment for asset prices to adjust drastically lower as long as money supply and productive/working capital depreciate in line with demand dynamics.
Usually as new bank loans are granted, money supply expands, and as loans are repaid money supply should contract. If we pay back our loans over time, and all other relationships remain in balance, we should not necessarily have instability. Likewise as population and production expand so should assets and money supply, in the present system, and vice versa for contractions in frame. Organic MS/productive capital expansion and contraction is not an issue in and of itself.
At turning points in population growth and productivity growth, and for up ticks, we should see a short term anomaly in money supply growth, as new money growth exceeds, at a faster rate, the rate at which old loans are paid back. The extent to which the supply of non productive assets/productive capacity expands to accommodate the money supply increase will determine the extent to which we see asset/debt to GDP/GDP growth differentials opening up. These short term dynamics are shocks in a sense to the financial, asset and economic system, but the system should be able to naturally accommodate them. Likewise as economies decelerate, population growth rates slow or decline and productive capacity peaks, we will see a contraction in money supply. In this case the shock is caused by adjustment of asset values/supply and productive capacity to the new frame. Capital depreciates in a contracting frame, debt declines as does the stock of broad money supply growth. As long as we do not attempt to either increase money supply growth or artificially support asset prices, shocks should be within the ability of any given financial/economic system to accommodate. But that is not what “we” appear to have done!
Capital accumulation in a growing frame and capital depreciation in a declining frame should not cause financial crisis on its own. Attempting to engineer higher money supply growth in a contracting frame, as discussed in previous blogs, impairs and obstructs the transition of frames, risking divergence between asset values and flows and hence creating latent asset price/debt based shocks to the system. In a stable frame money supply/assets to GDP should also be relatively stable.
Unhealthy changes in the ratio of asset values to GDP and money supply to GDP are more likely when we have a forced shift in dynamics during transitions (two trains going in opposite directions on the same track): for example aggressive easing of monetary policy whenever the financial and economic system is exposed to a shock (due to accumulated excess) as we have seen especially since the late 1990s onwards (LTCM/Asian Crisis/early 2000s/ and post 2007). The system never resets and the imbalance is allowed to continue – defining the transitions and the flows associated with those transitions is probably a more important analysis than an attempt to define whether wealth or debt effects are dominant per se. Transitions impact and also obscure frame dynamics.