The Ghost of Failed Banks Returns

Sep 27, 2019·Alasdair Macleod

Last week’s failure in the US repo market might have had something to do with Deutsche Bank’s disposal of its prime brokerage to BNP, bringing an unwelcome spotlight to the troubled bank and other foreign banks with prime brokerages in America. There are also worrying similarities between Germany’s Deutsche Bank today and Austria’s Credit-Anstalt in 1931, only the scale is far larger and additionally includes derivatives with a gross value of $50 trillion.

If the repo problem spreads, it could also raise questions over the synthetic ETF industry, whose cash and deposits may face escalating counterparty risks in some of the large banks and their prime brokerages. Managers of synthetic ETFs should be urgently re-evaluating their contractual relationships.

Whoever the repo failure involved, it is likely to prove a watershed moment, causing US bankers to more widely consider their exposure to counterparty risk and risky loans, particularly leveraged loans and their collateralised form in CLOs. The deterioration in global trade prospects, as well as the US economic outlook and the likelihood that reducing dollar interest rates to the zero bound will prove insufficient to reverse a decline, will take on a new relevance to their decisions.

Problems under the surface

Last week, something unusual happened: instead of the more normal reverse repurchase agreements, the Fed escalated its repurchase agreements (repos). For the avoidance of doubt, a reverse repo by the Fed involves the Fed borrowing money from commercial banks, secured by collateral held on its balance sheet, typically US Treasury bills. Reverse repos withdraw liquidity from the banking system. With a repo, the opposite happens: the Fed takes in collateral from the banking system and lends money against the collateral, injecting liquidity into the system. The use of reverse repos can be regarded as the Fed’s principal liquidity management tool when the banks have substantial reserves parked with the Fed, which is the case today.

Having inflated its balance sheet following the Lehman crisis by buying US Treasury bonds thereby increasing bank reserves, from 2011 the Fed started to increase its reverse repo position until 2017. In other words, it was taking liquidity out of the banking system, having previously injected massive amounts of it by means of quantitative easing following the Lehman crisis. From early-2017 to October 2018, outstanding reverse repos then halved, implying liquidity was being added. Since then they have increased by roughly half to $325bn, reducing liquidity.[i]

What spooked market commentators was the unexpected increase in the repo rate, which on Tuesday 17 September suddenly jumped from the previous Friday’s level of 2.19% to as much as 10%. By escalating its repo position, a targeted liquidity injection from the Fed followed as it struggled to maintain control over the repo rate, taking its outstanding repos from less than $20bn to $53bn. The Fed cut its Fed Funds Rate to a target of 1.75-2.0% the following day.

On Wednesday, 18 September the Fed’s repo position increased again from $53bn to $75 bn. Furthermore, on Thursday and Friday respectively the Fed’s repo position remained elevated, reaching $105bn last Monday. Interestingly, overnight dollar Libor declined slightly, in line with the reduction in the Fed Funds Rate, apparently unaffected by the higher repo rates in the US, confirming it is specifically a US problem involving the large banks.

There have been a number of explanations by expert commentators as to why the repo rate rocketed, none of them satisfactory. It reminds one of Verse 29 of Fitzgerald’s Rubaiyat of Omar Khayyam:

“Myself when young did eagerly frequent

Doctor and Saint, and heard great Argument

About it and about; but evermore

Came out by the same Door as in I went.”

Instead, I have a strong suspicion we are seeing the ghosts of past bank failures, most recently in the UK the sorry tale of Northern Rock which I closely observed. For non-British readers, a short reminder: as a licensed bank, Northern Rock was a mortgage lender which got into difficulties in September 2007, before being nationalised the following February. An old-fashioned run with customers queuing outside its branches seeking to withdraw their deposits had alerted the general public to Northern Rock’s problems. It was unable to tap wholesale money markets, because other banks were unwilling to lend to it on an uncollateralised basis.

The establishment missed the point. As Gillian Tett wrote in the Financial Times at the time, there were increasing concerns over how Libor was operating.[ii] There was a growing divergence in the rates that different banks were quoting in the various currencies priced in Libor, discriminating against the smaller borrowers (actually, an indication of growing counterparty risk, not a supposed failure of Libor). Furthermore, larger banks were reducing their exposure to Libor by sourcing funds from the treasury operations of large companies and using the developing repo market (which is collateralised, unlike Libor – a further indication of increasing systemic concerns) to maintain their overnight balances instead.

I recall vividly being in RP Martin’s office (then a leading money broker – now part of BGC Partners) in December that year, when all Libor offers mysteriously disappeared, leaving borrowers stranded. Having expected for some time that the credit bubble would come to a head and burst, I took this to be a significant signal of a developing crisis.

The following February, Northern Rock, which had depended on money markets for its financing, collapsed and was nationalised by the government, and the great financial crisis duly followed.

Could the erringly similar repo failure today be the ghost of Northern Rock returning to haunt us in New York? If so, we now have a far larger credit bubble to pop, and the figures in the repo market are in tens of billions, instead of tens of millions. This time it is perhaps less obvious to the general public, because old-fashioned public bank runs are probably a thing of the past.

The crisis brewing in 2007 was attributed to residential property and liar loans in America, securitised into collateralised debt obligations (CDOs), sliced and diced to give the appearance of tranches riskless to investors, while the risk was pushed into smaller equity and mezzanine tranches, retained by the sponsors. If we have a repeat performance of that, it is likely to involve the successor to CDOs, collateralised loan obligations (CLOs). They do roughly the same thing, but with low quality corporate debt.

This is why we must take notice of trouble in the repo market, and not dismiss it as just a one-off. The reason for its failure has little to do with, as some commentators have suggested, a general liquidity shortage. That argument is challenged by the increase in the Fed’s reverse repos from $230bn in October 2018 to $325bn on 18 September, which would not have been implemented if there was a general shortage of liquidity. Rather, it appears to be a systemic problem; another Northern Rock, but far larger. Today we call such an event a black swan.

What is today’s Northern Rock – or is it a Credit-Anstalt?

We cannot dismiss the possibility that a large non-American bank operating through a US-licensed subsidiary is perceived by its peers as too risky as a counterparty. This being the case, the most likely candidate is Deutsche Bank, which may be needing a significant liquidity replacement for fleeing deposits, having just concluded the sale of its prime brokerage to BNP. It is one thing to remove a business from the asset side of a bank’s balance sheet, but another to secure the far larger deposits that go with it.

Last July, Bloomberg reported that when the BNP deal was first mooted, Deutsche Bank clients were pulling out a billion dollars every day[iii]. Presumably, that was manageable, with enough liquidity on the asset side of Deutsche Bank’s very large balance sheet to iron out any difficulties, and it had its access to the US repo market.

Coinciding with current events, the BNP deal was finally signed and announced only last Monday, though it would have been known in New York banking circles last week when the difficulties in the repo market surfaced. Furthermore, large depositors would have almost certainly been made aware of the timing in advance in an effort to keep them onside, and some of them may have chosen to simply withdraw their deposits.

The sums involved could easily be large enough to marry up with the support provided by the Fed through the increased level of its repo exposure. Furthermore, we cannot dismiss the likelihood of the problem spreading to the US primary dealers of other foreign banks, including BNP itself.

For comparison, the time-lapse between the failure of the Libor market and Northern Rock’s nationalisation was less than two months. We cannot know for certain whether the trouble in the American repo market and the obvious difficulties faced by Deutsche Bank are definitely linked, let alone comparable in terms of time and outcome to the Northern Rock experience. But banks, hedge funds and operators of synthetic ETFs will be watching closely.

Synthetic ETFs are comprised of cash, near cash and bonds (which are meant to be liquid but often not), while matching their price performance to an index through derivatives. Having grown to an estimated $4 trillion overall, through synthetic ETFs the industry has accumulated substantial quantities of cash, bank deposits and near-cash at large banks with primary dealerships.

If the repo troubles escalate, there is a danger the investment management industry will start to move these funds from banks perceived to have increasing counterparty and operational risk, with potentially devastating consequences for all involved. Cynics have thought for a long time that the ETF industry would end in disaster for investors, without having a convincing explanation of how it would happen. Perhaps we are now beginning to see early evidence pointing to the ending of the ETF phenomenon, and to therefore be able to anticipate the knock-on effects on financial and derivative markets generally.

Returning to the subject of bank relationships, a more worrying comparison between Deutsche Bank and the Northern Rock episode could be with the Credit-Anstalt crisis of May 1931. It was the largest bank in Austria, just as Deutsche is the largest in Germany, a far larger country with a more important economy. Then in Austria and today in Germany, European economies were tipping into recession, forcing large losses onto their banks. Following the 1931 crisis, within months not only Austria but other European countries endured financial distress, the gold exchange standard began to disintegrate, and the international flow of goods was disrupted by growing protectionism as governments tried to batten down the hatches.

The flight of foreign creditors triggered by these events rapidly turned a major crisis in a minor country into a major crisis for all Europe and beyond. Today, if the same fate were to happen to Deutsche Bank, not only would it be on a far larger scale, but there is the additional question of the gross notional value of its derivatives book of nearly $50 trillion and the future of the euro itself. Is it any wonder, if Deutsche is indeed at the centre of last week’s repo crisis, that other major banks, have decided to step back and refused to accept its collateral in a repo?

The other major banks appear to have left the Fed to pick up the pieces by taking over the repo market. Another potential problem is China, with the Financial Times reporting only eleven days ago that Chinese groups are shedding $40bn in global assets, with a sub-heading that warned US divestments are soaring[iv]. Then there is the unexpected escalation of domestic funding requirements faced by Saudi Arabia in the wake of the attack on her oil refining facilities, almost certainly being covered by the sale of dollar balances in New York.

This confirms that some of the liquidity problems exposed by the repo market may be due to a reduction in dollar balances by both foreign corporations and governments, contrary to a wide-held belief that in a crisis, foreigners should be scrambling to buy dollars. It could throw an unexpected spotlight on US banks, including those with foreign ownership, with direct and indirect Chinese and Saudi connections. Though as mentioned below, with $307.9bn withdrawn in the year to July, foreign withdrawals appear to be a more widespread problem than exposed by current events. Whether it is the major force behind the repo crisis should be considered in the light of the dollar’s performance on the foreign exchanges, which has been remarkably steady in recent weeks.

Collateralised and leveraged loans may be to follow

The course of a credit crisis often starts with an initial shock followed by the uncovering of deeper problems. Almost everyone is taken by surprise by the initial shock, not realising its importance as a signal for a change in bankers’ attitudes to risk. The expression of purely technical reasons for the disruption in the repo market assume that what was known before still applies and there are no other factors involved; just an error of judgement by the authorities in managing markets. This is likely to be a mistake because markets are dynamic, and we can identify three separate reasons why no one should be complacent:

1. A slowdown in the US economy, yet to be reflected in backward-looking statistics, leads to a reduction in corporate cash levels and a drawdown of revolving credit to finance accumulating inventory. US banks may be already seeing evidence of this in some sectors.

2. There has been a reduction in dollar balances by foreign corporations and governments held through correspondent banks (note that in the twelve months to July 2019, there have been net withdrawals of $307.9bn[v]). Bankers will have been assuming that this is a temporary phenomenon, given the dollar’s reserve status. That hope is now being dispelled.

3. American banks are becoming more cautious of counterparty risk in wholesale money markets generally.

Following the current repo hiatus, a combination of all three is likely to lead to a change in the thinking of commercial bankers, with the first two fuelling the third. As to whether the problem is regarded as temporary or rings serious alarm bells, we need to dig more deeply into the marginal loan business which could tip the banks into a collective crisis, even if the immediate repo problem subsides. An obvious candidate is CLOs and uncollateralised leveraged loans.

According to the Bank for International Settlements, outstanding collateralised loan obligations are split with approximately $1.2 trillion in US dollars, and $200bn equivalent in euros[vi]. The dollar exposure accounts for half of all leveraged loans in the US financial system, so the total size of the US leveraged loan market is more like $2.4 trillion, which compares with the book value of total equity capital for commercial banks in the US of $1.95 trillion. While direct bank exposure to CLOs is estimated at only $250bn, they are bound to have the lion’s share of the rest of the leveraged loan market, giving them a total exposure of up to $1.5 trillion without indirect exposure being taken into account. Most of American banks’ equity capital is therefor at risk.

Collateralised or not, leveraged loans are bank loans to highly indebted corporations with high interest servicing costs barely covered by earnings, and mostly rated at less than investment grade. In an economic downturn these are the businesses that are the first to fail, and underlying asset quality is already reported by the BIS to be deteriorating. Furthermore, as global interest rates and bond yields have fallen towards and into negative territory, the demand for higher yielding CLOs has increased and the underlying quality decreased. The debt to earnings ratio of leveraged borrowers securitised in CLOs has risen and CLOs without maintenance covenants have grown from 20% in 2012 to 80% in 2018.

In its report, the BIS warns that there are additional spill-overs that could arise from disruptions in market liquidity, a statement that is particularly apt considering the current disruption in the repo market. Given the involvement of hedge funds, fixed income mutual funds and bank loan funds, when the credit cycle has more obviously turned there will almost certainly be a rush to sell these CLOs, likely to lead to fire sales with a potential to cascade losses in the manner seen with residential property CDOs eleven years ago.

Consider, for a moment, the position of a typical large US bank and the changing commercial motivations of its directors. Following the Lehman crisis, lending margins to non-financial corporations never really compensated for the risk of extending credit to anything other than large corporations and consumers prepared to pay credit card rates of interest. As the economy gradually recovered, loans to investment grade borrowers increased. Along with higher yields and with a AAA rating attached, lending to sub-investment grade borrowers became increasingly available through CLOs. Once the CLO ice was broken, even better yields were available by lending directly to sub-investment grade borrowers, the key being improving economic prospects underpinning the borrowers’ earnings. Furthermore, the bank’s competitors were also allocating increasing amounts of credit towards borrowers of this sort, so it is nearly impossible for our typical large bank not to follow them.

So far, all lending will appear to conform to the bank’s lending risk criteria, assuming of course that economic prospects are improving. The moment that stops, the directors of the bank will feel exposed and try to contain, then reduce their exposure to loan risk. In this respect, the change in the Fed’s interest rate policy is the clearest signal of an economic slowdown and rings the bell on the soundness of lending assumptions. This also includes considerations of systemic risk, in other words the risk of lending money to other banks and financial institutions deemed to be exposed to both CLOs and other leveraged loans.

As with all humanity, the rapid transposition from greed to fear afflicts bankers as well. If anything, given their tight group-thinking it is especially acute, turning on a dime. The expectation that the Fed was going to cut its Fed Funds Rate could act as a catalyst for fear, instead of laying concerns to rest over lending margin prospects. And bankers have good reason to be extremely concerned when they cast their attention towards geopolitics, the global and domestic economic outlook, and the growing threat of negative interest rates. And here, the news is not encouraging.

Geopolitics and the destruction of global trade

When President Trump embarked on a policy of penalising China with tariffs, the general assumption in financial markets was that a settlement would be achieved before long. Instead, the situation has deteriorated and realistically is nowhere nearer resolution. The effect of the trade dispute has been not only to harm both parties but has resulted in collateral damage as well.

Germany, whose fastest growing market was China, has been driven into recession, with last Monday’s purchasing managers’ index headlined as “simply awful”. With Germany being the locomotive pulling along all the other Eurozone members, this is already leading to deepening concerns for the Eurozone’s outlook and a resumption of asset purchases by the ECB (quantitative easing) is now due in November. It is also very bad news for Germany’s hard-pressed banking community, represented in New York by Deutsche Bank.

US banks will undoubtedly be increasingly aware of the negative impact of Trump’s tariffs on international trade through the credit demands of their customers. Now they see America drifting towards a new conflict in the Middle East against Iran. Saudi’s oil production has been hit by drones and missiles, allegedly from Iran, and the oil price could well increase substantially as a result. It could even drag Russia and China into a conflict. Hong Kong has been paralysed by riots, with China suspecting American provocation.

Clearly, the conflict between America and China has escalated well beyond just tariffs, making it difficult to visualise how the damage to global trade can be corrected. The economic outlook is therefore set to deteriorate further, with no end to it in sight. From a banker’s viewpoint, a global recession is the greatest threat to his business as a financial intermediary between failing borrowers and nervous depositors. He can only survive by taking anticipatory action to avoid potential losses.

Some bankers will have been clinging to the hope that the Fed, by reducing interest rates and if necessary, reintroducing quantitative easing, will rescue the US economy from outright recession and that economic growth will resume. Without doubt, this is the advice being given to management by in-house economists, unfamiliar with today’s destructive dynamics of tariffs combining with a failing late-stage credit cycle. These conditions were last seen in 1929, when Smoot-Hawley tariffs coincided with the end of a long phase of credit expansion. However, there is little statistical evidence so far that the US economy faces anything more than a pause in economic growth, which is why stock prices and other collateralised assets have held their values.

The reality is that a credit crisis cannot be avoided, only deferred. It is also hard to see how zero interest rates reduced from current levels can be enough to rescue markets that, on the evidence from the repo market, are beginning to price growing counterparty risk into interbank loans. Recent experience and central banking models suggest that dollar interest rates should be reduced by at least four or five per cent to stabilise the situation, putting them deep into negative territory. And as for negative rates, there is no development more likely to drive depositors into gold, silver and other media to escape from the taxation of negative rates on deposits.

Outcomes and their timing

Having been warned that not all is hunky-dory in Repo-land, the more forward-thinking bankers will begin to foresee the risks this reality will bring. One hopes that Deutsche Bank does not suffer a fate similar to that of Austria’s Credit-Anstalt in 1931 with the consequences that followed, but one cannot rule the possibility out, given that Germany is already in recession and the outlook for her weakened, undercapitalised banks is exceedingly grim.

That being the case, a new banking crisis is not only in the making, for which the repo problem serves as an early warning, but it could escalate quite rapidly. Given the rethinking that must be taking place in the boardrooms at all the major US banks, bankers will be looking at not only their exposure to Deutsche, but also the implications for their wider lending exposure to other American counterparties, particularly those owned by foreign banks.

Understanding there will be a transition of attitudes from investing in CLOs and leveraged loans to a concern over their soundness is the key to realising how a credit crisis evolves. This time, as well as a mountain of derivative contracts, there is the further problem of synthetic ETFs, many of which are sponsored and managed by the same bank. For example, Deutsche Bank controls 42 ETFs in the US market alone, worth $14.6bn, all of which appear to be synthetic.

Assuming this analysis is correct, there is probably not much the Fed can do, other than react to events. Like all central banks, the Fed relies on models that cannot incorporate the changing attitudes of market participants. Just imagine, if the Fed did spot a developing crisis in advance and called in the major bankers in an effort to get them to help stabilise the situation, they would likely leave the meeting with the clear impression things are worse than they thought, and their clear duty to their shareholders is to liquidate all positions at risk.

If it took two months between Libor freezing in December 2007 and Northern Rock being rescued by the UK government and if that timing is replicated today, a new banking crisis will hit in November. It could easily take longer to materialise, but there’s no guarantee it won’t escalate even more rapidly than that.

[i] See Federal Reserve Statistical Release H.4.1, Factors Affecting Reserve Balances: for the latest information on both repos and reverse repos.

[ii] FT, 25 September, 2007:

[iii] See,familiar%20with%20the%20matter%20said.

[iv] See

[v] See Line 29 in the table at

[vi] See BIS Quarterly review pages 11-14:

The views and opinions expressed in this article are those of the author(s) and do not reflect those of Goldmoney, unless expressly stated. The article is for general information purposes only and does not constitute either Goldmoney or the author(s) providing you with legal, financial, tax, investment, or accounting advice. You should not act or rely on any information contained in the article without first seeking independent professional advice. Care has been taken to ensure that the information in the article is reliable; however, Goldmoney does not represent that it is accurate, complete, up-to-date and/or to be taken as an indication of future results and it should not be relied upon as such. Goldmoney will not be held responsible for any claim, loss, damage, or inconvenience caused as a result of any information or opinion contained in this article and any action taken as a result of the opinions and information contained in this article is at your own risk.