How to manage a financial crisis

In the case of financial crises, responses should come in three stages – monetary, fiscal and structural.

MONETARY RESPONSE

First, immediate monetary response is needed, potentially including emergency liquidity, targeted bailouts, and short term interest rates.

Providing emergency liquidity (money) allows the financial sector to continue operating day-to-day. During normal times, financial institutions meet their day-to-day liquidity needs (e.g. depositor demands for cash, transactions with other financial institutions) by keeping just a small percentage of their assets on hand as cash or other highly liquid assets.

During the initial panic of a financial crisis though (such as August 2007 in the US sub-prime market), people may very suddenly start worrying about their financial institution’s creditworthiness and general ability to meet these obligations and consequently, will try to withdraw all their money while they can[1]. And because a financial institution will only keep a small fraction of its assets in a highly liquid form to meet its usual day-to-day liquidity obligations (and generally can’t quickly sell sufficient of its longer term assets for their true worth), this can cause a ‘bank run’, where even fundamentally sound and profitable financial institutions – not necessarily through any fault of their own – can suddenly find themselves without enough immediate liquidity to meet their day-to-day obligations[2].

And it’s not just depositors that suddenly increase their demand for liquidity. The institutions themselves become less willing to lend to each other because they are less confident about their ability to obtain finance themselves, making them more inclined to borrow and hoard liquidity.

If these sudden liquidity demands aren’t met swiftly, this sudden illiquidity turns into insolvency, where heavily-indebted individuals and institutions are forced to undertake destructive ‘fire sales’ – the sale of illiquid assets rapidly, at a fraction of their true worth – just to meet their short term liquidity needs. And if these individuals/ institutions are active in a number of markets, this panic can be transmitted far and wide, as liquidity needs in one market lead to the liquidation of assets in another, spreading to other financial institutions (called contagion) and potentially causing the entire financial system to freeze up, unable to operate, causing the broader economy to crash[3].

But a central bank can provide this short term liquidity, specifically in exchange for other assets from the financial sector. And not just to banks. And not just in exchange for typical high-quality bank assets. The GFC – and the preceding years of unprecedented financial innovation – showed the increasing systemic importance of – and therefore increasing necessity of supporting – other non-bank financial institutions (such as investment banks), new non-traditional markets (such as derivatives) and other assets that were of poorer quality and longer maturity than central banks normally accept as collateral. This is what the US Federal Reserve, the European Central Bank, and the Bank of England all did in reaction to the August 2007 liquidity shortfalls. And this just shows how dire the situation was, and how much further central banks were prepared to extend their support to deal with it.

Credibility is very important for a central bank. If the public (and the financial system) believes that the central bank has the will and the means to provide emergency liquidity when necessary, such a panic may be prevented before it even begins – no one will be worried about their financial institution becoming illiquid, so they won’t ‘run’ on it. But even if a bank run does occur, the central bank can nip it in the bud with a quick liquidity response before it spreads[4].

And if this liquidity response (or just their credibility) is successful, central banks may not have to resort to the next step of the monetary response – the ‘lender of last resort’.

The central bank may also need to provide targeted bailouts of systemically important financial institutions, lest their failure drag the entire system with them. This is a central bank’s fundamental role as ‘lender of last resort’. The government itself can also provide bailouts and deposit guarantees during a crisis if central bank support isn’t sufficient. Hopefully, the above general emergency liquidity will remove the need to bail out individual institutions. Furthermore, ideally, such institutions would never have been allowed to get ‘too big to fail’. But if an institution WAS allowed to get ‘too big to fail’, bailouts represent the lesser evil between: 1. Saving an institution that brought their problems upon themselves, potentially encouraging future bad behaviour in the financial sector (moral hazard) or potentially supporting insolvent institutions that end up costing more than they’re worth; and 2. Immediate widespread disaster.

Even institutions that did nothing wrong, and just got caught up in this system-wide liquidity crisis, could justifiably be provided with lender of last resort support – it surely wouldn’t be fair to let them fail over something that wasn’t their fault. Furthermore, confidentiality of central bank support during a crisis can also prevent them in the future. If the public becomes aware that an institution is receiving support from the central bank, it can spark a panic and worsen the existing liquidity crisis if other institutions are subsequently deterred from accepting central bank assistance. At the very least, public awareness could damage the institution’s reputation and increase their finance costs. This is what happened when news of Bank of England assistance to Northern Rock was leaked to the public before the Bank of England intended – which they did – to announce it, resulting in a bank run and the necessity of a government guarantee to quell. Consequently, perhaps such support in the future could be more effectively kept confidential, but only until the crisis has subsided. Transparency after the crisis is important to ensure accountability in central bank policy and/ or the use of public funds.

But moral hazard is an issue, and it is why central banks should generally be hesitant to spread their support to non-systemically important institutions, and/ or to assets of lower quality and longer length of maturity – they don’t want to encourage the financial sector to become reliant on it/ expect it in the future, thereby encouraging the sector to take unreasonable risks.

One way to minimise moral hazard risks is to exploit all private sector options first – ensure that there is no way for troubled institutions to obtain the finance they need outside of the central bank. Bear Sterns demonstrated this notion of a systemically important institution needing to be rescued because the private sector was unable/ unwilling until the US Federal Reserve provided a guarantee. Northern Rock similarly had to be rescued by the Bank of England and the UK government when private assistance wasn’t forthcoming.

Another way to reduce moral hazard is for the central bank to lend money against good quality collateral and at a relatively high interest rate, naturally to be paid back over time, to discourage implicit reliance on the lender of last resort, except in an emergency. But as mentioned above, in dire circumstances, in a widespread crisis, limiting collateral to high quality assets may not provide sufficient liquidity, and a high interest rate could actually worsen the situation.

Alternatively/ in addition, central banks could be forced to allow non-systemically important institutions to fail (even innocent ones), if it sends a message to the systemically important ones that lender of last resort support is not a given – therefore, don’t be reckless with your operations. Not ideal, but possibly a necessary sacrifice. It’s a balance of risks – save the innocent and risk incentivising bad behaviour, or sacrifice the innocent and send a message to the potentially guilty.

Support for the systemically important institutions that brought this pain upon themselves can also be made contingent on the departure of the institution’s CEO/ other executives/ board members, as well as imposing some losses on shareholders. Taking partial or entire ownership of the institution is also an option in particularly exceptional circumstances, for the purposes of controlling its restructure (making it no longer ‘too big to fail’), or just closing it down slowly at a rate that the system can handle. This makes lender of last resort support a little like health insurance – it may keep you alive, but it’s still an inconvenience to have to use, so you’re not going to deliberately get hit by a bus, just because you have this insurance.

Thirdly, if this financial crisis spills over into the real economy, a central bank must drop short term interest rates. This allows households, industry and government to meet their debt obligations more easily with lower interest payments, ideally helping to maintain current levels of consumption, investment and government spending. Lower interest rates will also encourage them to borrow more money – ideally to finance productive investments like property, technology, innovation, infrastructure, etc., rather than just debt-fuelled consumption and speculation. All this – in the immediate term – helps to prevent economic activity from crashing.

Keep in mind that these first two monetary responses (emergency liquidity and bailouts) do not generate ‘real’ wealth and employment in the short term, per se – rather, they simply create a ‘floor’ underneath the economy to keep it from tail-spinning. Even lowering interest rates only stimulates the economy indirectly via consumption, investment and government spending. But once this floor has been created and the initial panic has ended, if low interest rates have not sufficiently stimulated consumption and investment, real activity can be kick-started through the fiscal response.

FISCAL RESPONSE

Fiscal response is stage two. If the financial crisis has spread to the real economy, and low interest rates haven’t been sufficient to reverse this, targeted tax cuts, government expenditure, and infrastructure programs represent the catalyst to REAL economic momentum – actual wealth and employment will start improving while the private sector is still too nervous to make the first move. This is because during a recession, fiscal multipliers (how much an extra dollar of government spending will affect broader economic activity) are especially strong.

During normal economic times, $1 spent by the government may only represent $1 of economic activity in the short term (a multiplier of one) – maybe more in the long term if spent on productive infrastructure, maybe less if some is lost in bureaucratic inefficiencies, or if it is just transferred to wealthy individuals/ corporations – because that dollar was simply transferred from the private to the public sector. No new wealth was created in the short term – it was just reallocated. As a result, during normal times, fiscal policy tends to crowd out the private sector. In such cases, the only role this reallocation can play is for social ends, rather than economic ends.

During a recession however, fiscal policy doesn't crowd out the private sector because that spending would NOT have ordinarily occurred by the private sector – that’s why it’s a recession. Consequently, that $1 of government spending generates more than $1 of economic activity (a multiplier of more than one). In fact, the IMF estimates that during such recessionary circumstances, a government that incurs debt to pay for infrastructure spending will actually drive economic activity, and consequently government tax revenue, so much that their debt-to-GDP ratio will actually fall, not increase. In these circumstances, we would not be borrowing from future generations – the debt literally pays for itself, if not through financial profitability, then through economic growth.

And once the private sector (households, industry) sees real economic activity returning, their confidence rises and they take over as the driving force of the economy.

STRUCTURAL RESPONSE

The third stage is a structural response. Once real activity has returned to the economy following stage two, this will allow government, industry and/ or households to start gradually paying down the debt levels that triggered the crisis in the first place. Importantly, government debt reduction is NOT the priority during a financial crisis – not when the above monetary response has made it possible to borrow money cheaply, and especially not until the economy is back on track. Government debt reduction during a slump will literally make things worse – not just for the economy, but for the debt situation itself. By further slowing economic activity and therefore, government revenue, austerity during a recession can cause even more money to have to be borrowed than was saved by the initial austerity efforts (recall the fiscal multipliers above), driving debt-to-GDP (and potentially even the total debt dollar amount) up, not down.

Only when the economy is back on track can it support general efforts to reduce debt levels. Debt needs to be paid off using economic growth (or at most, austerity during STRONG economic times when the economy can handle it).

Furthermore, governments and central banks need to adequately supervise and regulate the financial sector – bank and non-bank financial institutions alike – to ensure a similar crisis does not happen again. Full recovery is impossible (or at least short-lived) if the underlying causes of the crisis are not dealt with.

This response can include higher capital reserve requirements, including liquid assets, for financial institutions to ensure they have adequate buffers to meet their day-to-day liquidity obligations during volatile times and not just normal times. Major crises may still require central bank emergency liquidity and lender of last resort support, but that’s okay – that’s what central banks are there for. Forcing institutions to keep too much liquidity in reserve will impede their ability to actually use it for productive long term investments.

The breaking up of ‘too-big-to-fail’ institutions, while it may result in some efficiency losses in the financial sector, will mean that these institutions won’t need to be bailed out during a crisis, because their failure wouldn’t jeopardise the whole system.

There also must be a creation of regulatory/ oversight organisations whose job is to monitor the financial sector as a whole, identify potential problems/ excesses, and make/ enforce recommendations to address them. These bodies don't need to be the central bank/ lender of last resort, if there are concerns over conflicts of interest, but if they are separate, information needs to flow readily between them to ensure timely and appropriate action during a crisis – this was a failure with respect to Northern Rock.

Another monetary response that was more unique during the GFC than any previous financial crisis was the need for international cooperation between central banks and national governments. Central bank support in one country doesn’t help a financial institution with vast global operations that utilise numerous currencies. For example, during the GFC, European banks experienced a lack of liquidity in US dollars. Once this shortcoming was observed, the US Federal Reserve, the European Central bank and the Swiss National Bank allowed cross-country swap arrangements to meet these international liquidity needs. Central banks also needed to make their policies consistent with each other regarding the (looser) collateral requirements they were imposing on financial institutions, so international financial institutions couldn’t ‘game’ between different central banks’ policies. This kind of cooperation is only going to become more important in the future[5].

Finally, an appropriate central bank exit strategy is needed – they need to wind back the support they provided during the crisis. The earlier liquidity responses by central banks are not long term strategies – they need to be used only during crises to buy time for the system to be properly repaired, and must be reversed over time. This means that all the assets which central banks acquired in exchange for their liquidity support need to be allowed to mature over time and not rolled over/ renewed. Their collateral requirements also need to be re-tightened to discourage moral hazard. But again – only once the system has recovered and been repaired.

These three responses – monetary, fiscal and structural – keep the system from tail-spinning, kick-start real economic activity and confidence again, and address the source of the problem to avoid a repetition.

In coming blog entries, I will discuss how Europe, the US and Australia measured up against these responses, and how this compared to the 1929 Wall Street Crash and subsequent Great Depression.

---

[1] Ironically, this is actually perfectly rational behaviour – if doubt arises as to the future liquidity of the financial system, it’s perfectly rational to sell first before liquidity dries up and market failure occurs, even though the market failure may have been avoided if no one panicked in the first place.

[2] The very nature of banking – long term assets but short term liabilities – leaves them vulnerable to bank runs.

[3] Although widespread collapse of the financial sector can also occur without contagion if all institutions are hit by the same external macroeconomic shock, e.g. oil prices, war.

[4] Being careful not to trigger excessive inflation or a currency collapse – capital controls during a crisis are a potential option to reduce this risk.

[5] Unless major anti-globalisation movements around the world are successful.