Normal Recessions:
A normal recession can go by a few definitions. For example, a common measure is 2 quarters of consecutive declines in GDP, though some people base a recession off other measures. Regardless, a recession is typically accompanied by a rising unemployment rate, falling investment, declines in retail spending, production and GDP.
Most commonly (almost invariably), recessions are precipitated by tightening of monetary policy after a period in which inflation has moved uncomfortably high for the central bank. In a typical recession, some weak businesses fail and unemployment imposes hardship upon some individuals.
During this period of economic trouble, the central bank significantly eases policy and automatic stabilisers in the form of government spending kick in, which limits the downside. In this case, for most citizens, the combination of easier policy encourages spending, risk taking and new investment to kick-start again and so the recession is short lived. Acute financial pain is common enough, but the entire population is not under duress.
Balance Sheet Recessions:
A different kind of recession occurs when that recession is also accompanied by a collapse of widespread investment in an asset bubble - be it a real estate or stock market, or any other bubble that has captured the attention of the entire economy and is usually debt driven. This is unlikely to occur from bubbles in Gold, Bitcoin, a particular stock, or any investment that does not involve a large portion of the public, but rather those such as real estate which are accessible to all, attractive to all, and have a strong investment narrative - with access to easy credit. Such an event might only happen once every 50-100 years or so - long enough for the next generation to have forgotten the lessons of the past.
Credit extension is an important feature, as it creates an environment where declines in asset prices can severely impair private sector balance sheets - leverage works both ways - leading to deeply negative equity with price falls. If no leverage is involved, then those involved in the bubble typically will not have deeply negative equity, even those that bought late in the bubble might not have negative equity if the purchases were not debt funded - in such a case, net worth may fall, but will not be negative. While asset prices may fluctuate with market sentiment, liabilities (debt here) tend to be fixed or at least fixed with a floating rate - bursting bubbles disproportionately affect the asset side of the balance sheet, leaving liabilities at bubble valuations, and asset values at post-bubble, depressed valuations - creating an impaired balance sheet.
When leverage is involved, even those who bought earlier in the asset bubble may find themselves facing leveraged losses, and cripplingly negative equity. It is this feature that really makes the pain widespread and can make traditional responses to recessions useless. Let's introduce an example investor in the US housing bubble, Tim. Tim's Balance sheet looks like this at the peak of the bubble:
Richard Koo described the psychology of people and corporations undergoing this kind of stressful environment, stating that their immediate goal changes from profit maximisation towards debt reduction. This is important, because a typical response to recession is to reduce short rates in an attempt to kick start consumption and investment - under a typical recession, this will probably work.
If, however, there is widespread pain inflicted by a bursting asset bubble, and those under duress have been imprudently leveraged, the first response to lower interest rates will be to pay down debt faster (debt reduction), rather than any particular increased consumption or investment. Individually this is perfectly sensible behaviour, though in aggregate it means that monetary policy is fairly handicapped to respond to this kind of crisis, and that demand for credit will remain startlingly low. If consumers did not behave in this way, it would actually be a little worrying.
Revisiting Tim for a moment to show how this is the only logical response, let's look again at his balance sheet after his area is hit by a 50% fall in housing values, also assuming he has paid down some of his mortgages and credit card debt, but the weak economy has also caused him to dip in to his cash savings:
Immediately following the fall, Tim is now worth - $154,000. Understandably, this causes discomfort. People naturally want to ease this discomfort by returning to a positive net worth. For Tim have any positive net worth, a combination of the following processes must take place:
- House prices must rise again.
- Tim must use income to pay down debt.
- Tim can save his money as cash/other to offset the loss of asset value (he can also sell the houses at a loss and reinvest).
Number 1 takes a long time. While US house prices rebounded strongly following the introduction of ZIRP, they are not exceeding previous real values for the affected areas even after 8 years. It is unlikely that a person would sit on negative net worth during this time and continue to take on more debt.
Number 3 happens to an extent, but since in a balance sheet recession interest rates head to near zero, and mortgage interest should always be higher than deposit or savings rates, this does not make a very cost effective method of rebuilding net worth. If the savings were invested in shares or more housing, then this may be effective, however the psychological and financial impact of bursting asset bubbles prevents those affected from investing at an opportune time, when these assets are undervalued.
This leaves number 2, the only sensible option for an individual. Since it is more cost effective to pay down debt than to save at zero interest rates, or to wait and hope that asset values exceed the previous peak, Tim would end up putting a lot of effort in to paying down his outstanding mortgages in order to return to a normal positive net worth - even forgoing consumption in order to do so, worsening the crisis. This behaviour in aggregate explains part of why bank loan books have barely grown since the onset of the crisis - there has been little demand for new credit.
The problem extends to corporates also. Of course, many US corporations went bankrupt during the crisis, but the corporate balance sheet recession was not as extensive as that of the consumer, unlike the case in Japan's balance sheet recession. However, the effect on corporates is still very pervasive even when it is more the consumer that has been directly affected, and this feeds in to a lower demand for credit.
Because a balance sheet recession typically follows either a credit funded investment and/or consumption boom, the after-effect involves very low consumption growth and little reason for investing in greater capacity - if consumers are reducing purchases and scared of debt funded consumption, then more capacity does not make sense. Similarly, if there was an investment boom, it is likely that the country is already over invested in capacity/residential housing/supply, so that further investment after the bursting of the bubble does not make sense. In both cases, there is not great demand for new investment, and after a credit fuelled consumption boom there is also little desire to repeat imprudent purchases. The combination is that credit demand, again, remains very low. Actually, the true case post-GFC is most likely made worse by the fact that China continued investing in more capacity after Western demand collapsed, again, making it even less necessary for Western based corporations to invest domestically in more capacity. To recap why a balance sheet recession, especially, causes low rates for an extended period:
- A balance sheet recession makes existing debt holders pay down debt to get back to positive net worth.
- Both those indebted, and those not yet in debt are scarred by the experience of a bursting asset bubble, and are unlikely to take on debt in the immediate aftermath.
- Monetary policy is not very potent under these conditions, so interest rates move to zero for an extended period without much fear of rampant inflation.
- Lowered investment demand in the wake of the crisis keeps demand for credit lower than otherwise.
- Lowered credit fuelled consumption lowers the demand for credit.
- Lowered consumption spending reduces inflationary pressure, and excess capacity relative to demand prevents a supply side inflationary episode. Inflation is nowhere to be found, so yields will not rise substantially based on this fear.
- The private sector has a strong demand to save in risk-free assets such as Treasuries, as they fear the stability of the banking system and are scared out of risky assets, pushing down rates on government debt. Sometimes this pushes shorter maturity bonds in to negative yields - this seems irrational at first glance. If the banking system is facing systemic threats, it is rational to pay a slight premium on safe government bonds as opposed to risking deposit loss in the banking system. Similarly, cash stored under the mattress is typically not as safe as government backed bonds might be, so a negative yield might be considered equivalent to paying for insurance.
The Public Sector:
The solution when the private sector will not respond to monetary policy changes is for the public sector to step in by running a deficit. This is a source of a lot of angst amongst casual observers - it appears at first glance that the country is going bankrupt as public spending spirals out of control. Strangely though, interest rates do not seem to rise (see Japan, the US now or in the great depression).
Declining rates in the face of record deficit spending indicates a line of causality that starts with the private (and foreign) sector's demand, which induces the public sector to supply - that is, the private (and foreign) sector demands the government issue more debt for holding in portfolios, foreign central banks to hold in reserve, and for fiscal stimulus and so on. The public sector then supplies this deficit spending (either automatically through unemployment benefits etc. or through stimulus packages). Because there is a revulsion to becoming indebted, the government is likely to always spend less that would be optimal in a perfect world, and so yields continue to fall for the market to clear.
Higher interest rates are always threatened, but since this does not occur, many theories and insults are directed at authorities for keeping rates too low and punishing savers. Of course, rates would be just as low without intervention, probably lower if stimulus responses were not enacted, but this gets in the way of a good bit of vitriol.
This reactionary public spending is the source of the Widow-maker, because it involves a rapidly rising debt/GDP ratio - which invokes a fearful response. But, it is different to when undeveloped, profligate governments spend freely without good reason (often due to corruption), and it is different to when emerging markets and Euro area countries have heightened default risk - in these cases shorting sovereign debt may work because there are structural failures in the monetary system that allow for default to take place under stress. The US, like Japan is a candidate for a Widow-maker trade to hurt traders because it is well developed, with a flexible monetary system and almost zero default risk.
The main risk with countries like Japan and the US is inflation, which has so far been very difficult to achieve. Were it to actually happen (which both governments and central banks would like very much), this would not automatically make the Widow-maker trade work. Unfortunately for those trying to short the debt of these countries, the central bank always has the ability to set yields across the entire curve. This is actually easier in an inflationary environment than in a deflationary one, as the central bank can place a bid under any price and offer to buy any quantity at this price to hold yields steady. No trader will fight this, because the central bank can never run out of reserves, and so the yield will most likely remain at target until the central bank changes policy. The central bank is limited mainly by how much inflation it will allow before tightening. High inflation is not optimal, but it is the lesser evil for policy makers to choose.
Should inflation move higher (say 5-10%), the central bank could hold treasury rates at 2-3% and allow inflation to reduce the severity of the public and private sector debt burden. Eventually they would be forced to raise rates and loosen restrictions in order to combat inflation, but probably not before the once scary debt burden has been sufficiently diminished - at which point, the rising rates will not immediately cause any funding issues and the higher rates will discourage further indebtedness. Though I will discuss this in more depth in another post, this vaguely describes the post-war period in which government debt was proportionally larger as a share of GDP than today. While it may sound implausible that bond vigilantes will not show up to punish the government, this inflation/yield targeting has already been enacted last century, and so we know roughly what the outcome is likely to be.
Part 2 will be drivers of interest rates, and monetary policy/central bank policy in more detail.
