“Enough is enough”

Growth in the wrong type of lending triggers ECB call for policy shift

The key chart

**Growth rates (% YoY) in average RRE prices and loans for house purchase** (Source: ECB; CMMP)

The key message

In its latest “Financial Stability Review” (November 2021), the ECB calls for a policy shift away from short term support measures towards “mitigating risks from higher medium term financial stability vulnerabilities, in particular emerging cyclical and real estate risks.”

This is a welcome development given the extent to which (unorthodox) policy measures have fuelled growth in the “wrong type of lending” to date, and the negative implications this has for future growth, leverage, financial stability and income inequality.

Macroprudential instruments include capital measures (e.g. higher risk weights) and borrower-based measures (e.g. LTV limits). Their adoption varies across the euro area currently, with six economies adopting a combination of both instruments, nine economies adopting borrower-based measures alone, and four economies having no measures in place (Germany, Spain, Italy and Greece).

Further tightening of existing instruments may be required in several economies where RRE vulnerabilities are continuing to build up, but Germany stands out given current house price and lending dynamics, the extent of RRE overvaluation and the absence of targeted macroprudential measures.

“Enough is enough”

Over the past two years, I have been highlighting the hidden risk that unorthodox monetary policies in the euro area (and elsewhere) were fuelling growth in the “wrong type of lending”. From this, I have argued that the resulting shift from productive COCO-based lending towards less-productive FIRE-based lending (see chart below) has negative implications for leverage, growth, financial stability and income inequality in the future.

**Outstanding stock of private sector lending (EUR bn, LHS) broken down by type and share of FIRE-based lending in total lending (%, RHS)** (Source: ECB; CMMP)

Earlier this month, I also argued that it was appropriate, therefore, to expect new macroprudential measures for residential real estate soon. The ECB agrees (finally). In their latest “Financial Stability Review” (November 2021), the ECB is calling for a policy shift away from short-term support towards mitigating risks from higher medium-term financial stability vulnerabilities including residential real estate (RRE) risks.

**2Q21 RRE price growth (% YoY) plotted against level of pre-pandemic valuation** (Source: ECB; CMMP)

The ECB’s analysis includes three key risk factors:

First, nominal house prices grew at 7.3% in 2Q21, the fastest rate of growth since 2005 (see key chart above)

Second, house price and lending dynamics have been much stronger in many countries with pre-existing vulnerabilities. For example, despite above average degrees of over-valuation pre-pandemic (ie, >4% estimated overvaluation), RRE prices grew at above-average rates (ie, >7% YoY) in Luxembourg, the Netherlands, Austria, Germany, and Belgium in the year to end 2Q21 (see chart above)

Third, there is evidence of a progressive deterioration in lending standards, as reflected in the increasing share of loans with high LTV ratios. The share of new loans with LTVs above 90% reached 52% in 2020 compared with only 32% in 2016 (see chart below)

The ECB also notes “high and rising levels of HH indebtedness”, but this is less of a risk, in my opinion, given that the HH debt ratio of 61% GDP is well below the BIS’ threshold of 85% GDP.

**Share of loans with LTV >90% in total new loan production** (Source: ECB; CMMP)

Macroprudential instruments include capital measures (eg higher risk weights) and borrower-based measures (eg, LTV limits). Their adoption varies across the euro area with six economies adopting a combination of both instruments (green bubbles in chart below), nine economies adopting only borrower-based measures (orange bubbles in chart below), and four economies having no measures in place – Germany, Spain, Italy and Greece (red bubbles in chart below, although Greece not shown).

**RRE price growth plotted against mortgage loan growth 1H21 v 1H20. Size of bubbles represents HH debt ratio and colour represents current macroprudential framework** (Source: ECB; CMMP)

Further tightening of existing instruments may be required in several economies where RRE vulnerabilities are continuing to build up, but Germany stands out given the combination of house price and lending dynamics, the extent of overvaluation and the lack of macroprudential measures.

Please note that the summary comments and charts above are extracts from more detailed analysis that is available separately.

“How much? How productive?”

ECB claims from a CMMP perspective

The key chart

**Annual growth (% YoY) in EA private sector lending split between FIRE-based and COCO-based lending** (Source: ECB; CMMP)

The key message

Is the ECB correct to argue that, “monetary policy measures continue to support lending conditions and volumes” in the euro area? Yes, but only up to a point.

On the supply-side, the APP, PEPP and TLTRO III programmes are having a positive impact on banks’ liquidity positions and overall market financing conditions. On the demand side, borrowing costs are at (mortgages) or close to (NFC loans) historic lows in nominal terms and at historically low and negative levels in real terms. So far, so good.

That said, lending volumes are unexciting in relation to recent trends and previous cycles and currently negative in real terms. No compelling volume story here.

More importantly, current policy measures are supporting the “wrong type of credit”.

Less-productive lending that supports capital gains through higher asset prices (FIRE-based lending) contributed 2.5ppt to the 3.2% total loan growth in September 2021. Worryingly, this is part of longer-term trend. While the outstanding stock of private sector loans hit a new high in September, the stock of productive lending that supports production and income formation (COCO-based lending) remains below its January 2009 peak.

This matters for two key reasons. First, the shift from COCO-based lending to FIRE-based lending has negative implications for leverage, growth, financial stability and income inequality (expect new macroprudential measures for residential real estate soon). Second, it re-enforces the importance of an on-going policy response that remains “fiscal, first and foremost”.

“How much? How productive?”

In its latest “Euro area bank lending survey”, the ECB argues that, “monetary policy measures continue to support lending conditions and volumes” in the euro area (EA). Is this correct?

The supply-side

**“Liquidity-providing” monetary policy operations in EUR bn** (Source: ECB; CMMP)

On the supply-side, EA banks report that “the ECB’s asset purchase programme (APP), the pandemic emergency purchase programme (PEPP), and the third series of targeted longer-term refinancing operations (TLTRO III) continues to have a positive impact on their liquidity positions and market financing conditions” (see chart above).

The demand-side

**Composite cost-of borrowing for house purchases and NFC loans in nominal and real terms** (Source: ECB; CMMP)

One the demand side, borrowing costs are at (mortgages) or close to (NFC loans) historic lows in nominal terms and at historically low and negative levels in real terms (see chart above). In September 2021, the composite cost-of-borrowing for house purchases hit a new low of 1.30% (-2.03% in real terms). The composite cost of borrowing for NFC’s was 1.48%, 0.8ppt above its March 2021 low, but a new low in real terms (-1.86%).

How exciting is the volume story?

**Annual growth (% YoY) in private sector lending in nominal and real terms** (Source: ECB; CMMP)

Lending volumes are unexciting in relation to recent trends and previous cycles and currently negative in real terms. Lending to the private sector grew 3.2% YoY in September 2021 both on a reported basis and after adjusting for loan sales and securitisation. In nominal terms, lending growth has been relatively stable since March 2021 but is 2ppt lower than the recent peak growth recorded in May 2020 (5.2% YoY). Lending growth in the current cycle is relatively subdued, however, in relation to past cycles (see chart above). Furthermore, in real terms, lending in September fell slightly when adjusted for HICP inflation.

No compelling volume story here.

Loan growth from the ECB perspective

**Annual loan growth with ECB breakdown by borrower** (Source: ECB; CMMP)

As an aside, the ECB typically classifies lending by type of borrower – households (HHs), non-financial corporations (NFCs), non-monetary financial corporations (NMFCs) and insurance companies and pension funds (ICPFs) – with further subdivisions based in the type of HH borrowing and the maturity of NFC borrowing.

In September 2021, HH lending contributed 2.2pt to the total 3.2% YoY growth, essentially mortgages. NFCs and NMFCs contributed 0.6ppt and 0.5ppt respectively but lending to ICPFs made a slight negative contribution of -0.1ppt.

Loan growth from the CMMP Perspective

CMMP analysis presents an alternative classification based on the productivity of credit use. Broadly speaking, lending can be spilt into two distinct types: lending to support productive enterprise; and lending to finance the sale and purchase of existing assets. The former includes lending to NFCs and HH consumer credit (and other HH lending) and is referred collectively here as “COCO-based” lending (COrporate and COnsumer). The latter includes loans to non-bank financial institutions (NBFIs) and HH mortgage or real estate debt and is referred collectively here as “FIRE-based” lending (FInancials and Real Estate).

Note that COCO-based lending typically supports production and income formation while FIRE-based lending typically supports capital gains through higher asset prices.

Supporting the “wrong type of credit”

**Annual loan growth with CMMP breakdown by productivity of use** (Source: ECB; CMMP)

Viewed from a CMMP perspective, current policy measures are supporting the “wrong type of credit”. Less-productive lending that supports capital gains through higher asset prices (FIRE-based lending) contributed 2.5ppt to the 3.2% total loan growth in September 2021 (see chart above).

**Annual loan growth broken down by productivity of use since September 2006** (Source: ECB; CMMP)

Worryingly, this is part of a longer-term trend. As can be seen from the chart above, higher volumes in the pre-GFC period were more balanced with more-productive COCO-based lending accounting for 56% of total outstanding loans. Today, that share has fallen to 48%.

**Outstanding stock (EUR bn) of COCO-based lending** (Source: ECB; CMMP)

As noted in August 2021, while the outstanding stock of credit hit a new high in September, the stock of productive COCO-based lending (€5,470bn) remains below its January 2009 peak (€5,517bn). In other words, the aggregate growth in lending since early 2009 has come exclusively from FIRE-based lending which now accounts for 52% of the outstanding stock of loans (see chart below).

**Outstanding stock (EUR bn) of COCO-based and FIRE-based lending** (Source: ECB; CMMP)

Conclusion

The ECB is entitled to argue that monetary policy measures have supported lending conditions and volumes. However, current lending volumes are unexciting in relation to previous cycles and negative in real terms. Policy is also supporting the “wrong type” of credit – fuelling FIRE-based lending rather than productive COCO-based lending that supports production and income formation.

This matters for two key reasons. First, the shift from COCO-based lending to FIRE-based lending has negative implications for leverage, growth, financial stability and income inequality. During the COVID-19 pandemic, some national authorities eased macroprudential measures for residential real estate (RRE). This week, however, the ECB argued that further macroprudential measures should be considered where RRE vulnerabilities continue to build up. (Watch this space.) Second, it re-enforces the importance of an on-going policy response that remains “fiscal, first and foremost”.

Please note that the summary comments and charts above are extracts from more detailed analysis that is available separately.

“Tough to get too excited”

Recovery and reflation trades require more substantial foundations

The key chart

**Growth in M3 (% YoY) and contributions (ppt) from M1 and PSC** (Source: ECB; CMMP)

The key message

It is tough to get too excited about the messages coming from the euro area’s (EA) money sector at the start of 3Q21.

Broad money growth is almost 5ppt lower than its January 2021 peak. The positive news here is that households are saving less, indicating that uncertainty levels have fallen. The less positive news is that growth in private sector credit has also fallen to its slowest rate since December 2017. Total lending is growing only 0.8% YoY in real terms and is falling -1.3% YoY in real terms if we exclude lending to HH (mainly mortgages).

In short, while the message from the money sector remains positive for (already overvalued) house prices in the euro area, the wider message is that both a sustained recovery and reflation trades require a more substantial foundation.

The six charts that matter

**Growth in EA broad money % YoY** (Source: ECB; CMMP)

It is tough to get too excited about the messages coming from the euro area (EA) money sector at the start of 3Q21. Growth in broad money (M3) slowed to 7.6% YoY in July 2021, almost 5ppt below the January 2021 recent high of 12.5% (see chart above). The positive news is that this reflects a reduction in the deflationary forces that drove M3 growth during the pandemic.

**YoY growth rates in M3 and M1 since 2001** (Source: ECB; CMMP)

Narrow money (M1) which contributed 7.7ppt to the total 7.6% YoY M3 growth has slowed from 16.5% in January 2021 to 11.0% in July 2021 (see chart above). In short, EA households are saving less.

**Monthly flows (EUR bn) of HH deposits during phases of pandemic** (Source: ECB; CMMP)

Overnight deposits still account for 6.8ppt of total M3 growth, but in aggregate household monthly deposit flows fell to €23bn in July 2021, below the 2019 average monthly flow of €33bn and the smallest monthly flow since June 2019 (see chart above).

**Growth in PSC (% YoY 3m MVA) since 2001** (Source: ECB; CMMP)

The less positive news is that credit demand is also slowing. Private sector credit contributed only 3.5ppt to M3 growth and the YoY growth rate slowed to 2.9% YoY (3MVA) the slowest growth since December 2017 (see chart above).

**Trends in growth in lending and contribution from lending x HH** (Source: ECB; CMMP)

As noted in “Strip out HH lending”, current lending is predominantly less-productive FIRE-based lending rather than productive COCO-based lending. Total lending grew 0.8% YoY in real terms in July, but fell -1.3% YoY in real terms excluding HH lending (see chart above). HH lending contributed 2.2ppt to the total 3.1% nominal growth in lending in July 2021 (see chart below).

**Drivers of PSC growth** (Source: ECB; CMMP)

Conclusion

In summary, while the message from the money sector remains positive for (already overvalued) house prices in the euro area, the wider message is that both a sustained recovery and reflation trades require a more substantial foundation.

Please note that the summary comments and charts above are extracts from more detailed analysis that is available separately.

“Strip out HH lending”

…and PSC growth is falling in real terms in the EA

The key chart

**Real YoY growth rates in total lending and total minus HH lending** (Source: ECB; CMMP)

The key message

Strip out lending to households (mainly mortgages) and private sector lending in the euro area (EA) is falling in real terms. Why does this matter?

A resynchronisation of money and credit cycles in the euro area is one of three key signals indicating a sustained economic recovery for investors and policy makers alike. The ideal scenario would see a reduction in the deflationary forces that drove M3 growth during the pandemic (i.e. HH money holdings) combined with a recovery in productive lending to the private sector. To date, we have witnessed progress in the former but not in the latter.

Monthly flows of HH deposits in April, May and June 2021 were below pre-pandemic levels, for example. Lending to the private sector is slowing too, however, from 4.7% at end 2020 to 3.0% at the end of 1H21. Base effects play a role here as EA corporates borrowed €243bn in the immediate “dash for cash” in March, April and May 2020. For context, net borrowing over the subsequent 12 months to June 2021 totalled only €103bn.

Significantly, total lending minus HH lending (predominantly productive COCO-based lending) contributed only 0.8ppt to the 3.0% YoY lending growth in June 2021. In contrast, HH lending contributed 2.1ppt, the bulk of which is in the form or mortgages.

In other words, current lending is essentially less-productive FIRE-based lending rather than productive COCO-based lending. While this may provide further support for house prices, a sustained recovery requires a more substantial foundation.

The six charts that matter

**Growth (% YoY) in broad money and private sector credit** (Source: ECB; CMMP)

A resynchronisation of money and credit cycles in the euro area is one of three key signals indicating a sustained economic recovery for investors and policy makers alike.

**What drove M3 growth during the pandemic?** (Source: ECB; CMMP)

The ideal scenario would see a reduction in the deflationary forces that drove M3 growth during the pandemic (i.e. HH money holdings) combined with a recovery in productive lending to the private sector. To date, we have witnessed progress in the former but not the latter.

**Monthly flows (EUR bn) in HH deposits** (Source: ECB; CMMP)

To date, we have the former but not the latter. Monthly flows of household deposits in April, May and June 2021 were below pre-pandemic levels. So far, so good.

**Nominal YoY growth rates in total lending and total minus HH lending** (Source: ECB; CMMP)

Lending to the private sector is slowing too, however, from 4.7% at end 2020 to 3.0% at the end of 1H21.

**Monthly flows in NFC borrowing in EUR bn** (Source ECB; CMMP)

Base effects play a part here as EA corporates borrowed €243bn in the immediate “dash for cash” in March, April and May 2020. For context, net borrowing over the subsequent 12 month to June 2021 totalled only €103bn.

**Drivers of PSC growth by type** (Source: ECB; CMMP)

Please note that the summary comments and charts above are extracts from more detailed analysis that is available separately.

“Herd immunity?”

Resilience and risks in global housing

The key chart

**Trends in global house prices since the GFC** (Source: BIS; CMMP)

The key message

Anyone looking for evidence of COVID-19 “herd immunity” need look no further than global housing markets!

House prices rose 4% globally in 2020 in real terms, the fastest rate of growth since the GFC. Prices rose 7% in advanced economies, compared with a more modest 2% in emerging economies. House price resilience during the pandemic reflects many factors: a recovery in HH incomes thanks to continued policy support; lower borrowing costs; reduced supply as construction activity slowed; temporary tax breaks; and perceptions that housing was/is a relatively safe investment.

The combination of rising prices and an uncertain macro backdrop has kept measures of overvaluation elevated. In the euro area, for example, above average increases in house prices occurred in Luxembourg, Slovakia, Estonia, Portugal, Denmark, Austria, the Netherlands and France. With the exception of Estonia, estimates suggested overvaluation in each of these countries before the start of 2020, notably in Luxembourg, Denmark and Austria. Similarly, the Bank of England indicated unease about the UK housing market recently (1 June 2021) after the Nationwide Building Society said that prices were growing at their fastest pace since 2014.

Current EA housing and lending dynamics reflect Minsky’s hypothesis that, over the course of a long financial cycle, there will be a shift towards riskier and more speculative sectors. The flow of funds towards property and financial asset markets (FIRE-based lending) is increasing at the expense of more productive flows to the real economy (COCO-based lending). FIRE-based lending in the EA hit a new high of €5,905bn in April 2021 and accounts for 52% of total lending with negative implications for leverage, growth, stability and income inequality.

Resilience and risks in global housing

Anyone looking for evidence of COVID-19 “herd immunity” need look no further than global housing markets! House prices rose 4% globally in 2020 (in real terms) according to latest BIS data release, the fastest rate of growth since the GFC. Prices are now 21% higher than their average after the GFC (see chart below).

**Real price change in 2020 plotted against real price change since the GFC** (Source: BIS; CMMP)

Prices rose 7% in “advanced economies” (especially New Zealand, Canada, Denmark, Portugal, Austria, Germany, US) compared with a more modest 2% in “emerging economies.” The resilience of housing markets reflects many factors: a recovery in HH incomes thanks to continued policy support; lower borrowing costs; reduce supply as construction activity slowed; temporary tax breaks; and the perceptions that housing was/is a relatively safe investment.

**EA trends – 2020 price change ploted against valuation at end-2019** (Source: ECB; CMMP)

The key risk here is that the combination of rising prices and an uncertain macro backdrop have kept measures of overvaluation elevated.

In their latest Financial Stability Review, for example, the ECB notes that “house price growth during the pandemic has generally been higher for those countries that were already experiencing pronounced overvaluation prior to the pandemic (see chart above).”

The largest/above average increases in house prices during 2020 in the EA occurred in Luxembourg (17%), Slovakia (16%), Estonia (9%), Portugal (9%), Denmark (9%), Austria (7%), the Netherlands (7%) and France (6%). With the exception of Estonia, ECB estimates suggest that house prices were overvalued in each of these countries before the start of 2020, notably in Luxembourg (39% overvalued, not shown in graph above), Denmark (16% overvalued) and Austria (15% overvalued).

On the 7 June 2021, the BIS will release 4Q20 credit and affordability data which will provide further insights into the risks associated with housing trends in the EA and the rest-of-the-world.

**The rise in FIRE-based lending in the euro area** (Source: ECB; CMMP)

In recent posts, I have noted an adaptation of Hyman Minsky’s hypothesis that states that over the course of a long financial cycle, there will be a shift towards riskier and more speculative sectors.

Minsky’s theory can be applied to the house price trends described above and to HH lending trends described in previous posts. Minsky’s “shift” is reflected in the decline in bank credit to the real sector (COCO-based credit) and an increase in funds flowing towards property and financial asset markets (FIRE-based credit).

FIRE-based lending in the EA hit a new high of €5,905bn in April 2021 and accounts for 52% of total lending with negative implications for leverage, growth, stability and income inequality.

Please note that the summary comments and charts above are extracts from more detailed analysis that is available separately.