Secular stagnation or financial cycle drag?Implications for the banking industry

Claudio BorioHead of the Monetary and Economic Department

Centre for Financial Studies Colloquium seriesGoethe University, Frankfurt, 7 February 2017

2

Questions and thesis

Questions What explains the plight of the global economy? What are the implications for the banking industry?

Compare two different narratives or hypotheses (Demand-driven) secular stagnation (SS) vs financial cycle drag (FCD)

Thesis FCD narrative provides a better explanation Profound impact on banking industry of financial bust and policy response

- Exacerbating longer-term structural problems

Structure of the remarks Summarise in very stylised terms the two hypotheses Argue that the FCD hypothesis is more convincing Assess the condition of the banking industry in this light

3

I - The two hypotheses: a stylised characterisation

Three features of the SS hypothesis World is haunted by a structural aggregate-demand deficiency Pre-crisis financial boom (bubble) was price to pay to keep output at potential The natural (equilibrium) real interest rate is negative

- Low rate needed to avoid a damaging demand-driven deflation

Three features of the FCD World is haunted by inability to constrain financial booms/busts (outsize financial

cycles (FCs)) (G 1)- Joint and long-lasting unsustainable expansions/contractions in credit and

asset prices- Busts cause huge and long-lasting economic damage

Pre-crisis boom was part of the problem, with output above potential The natural (equilibrium) real interest rate is positive and considerably higher

- Overestimate global demand deficiency- Underestimate secular supply-side global factors driving disinflation- Need to define and measure the natural interest rate including financial factors

4

Graph 1: The financial cycle is longer than the business cycle (the US example)

1 The financial cycle as measured by frequency-based (bandpass) filters capturing medium-term cycles in real credit, thecredit-to-GDP ratio and real house prices. 2 The business cycle as measured by a frequency-based (bandpass) filtercapturing fluctuations in real GDP over a period from 1 to 8 years.

Source: Drehmann et al (2012), updated.

5

II - The SS hypothesis: a critique

Evidence for the SS hypothesis Persistently disappointing and low post-crisis growth Stubbornly low inflation despite huge MP efforts Low interest rates way out the yield curve

Three nagging doubts SS initially developed for the US, with large current account deficit Pre-crisis record growth for world as a whole Unemployment now close to historical norms

Specific pieces of evidence that favour the FCD hypothesis Post-crisis recovery not unusual given banking crises and financial bust Evidence that financial booms/busts cause long-term damage to productivity (G 2) Evidence that output was above potential (on unsustainable path) pre-crisis (G 3)

- Estimates based on FC proxies would have shown it also in real time Link between domestic output slack and inflation has been weak for a long time

- Evidence of global (dis)inflationary factors at play (G 4)

6

Graph 2: Financial booms sap productivity by misallocating resources

Annual cost during a typical boom … … and over a five-year window post-crisis

Estimates calculated over the period 1969–2013 for 21 advanced economies. Resource misallocation = annual impacton productivity growth of labour shifts into less productive sectors during a 5-year credit boom and over the periodshown. Other = annual impact in the absence of reallocations during the boom.

Source: Borio et al (2015b).

0.0

0.1

0.2

0.3

0.4

0.5

Other

Other

Resourcemisallocation

Resourcemisallocation

%pts

7

Graph 3: US output gaps: ex post and real-time estimates

For each time t, the “real-time” estimates are based only on the sample up to that point in time. The “ex post” estimates are based on the full sample.Source: Borio et al (2016a).

In per cent

IMF OECD

Hodrick-Prescott Finance-neutral

8

Graph 4: GVCs and the explanatory power of global output gaps……across countries1 … and over time2

Rela

tive

glob

al fa

ctor

(RG

F)

Rela

tive

glob

al fa

ctor

(RG

F)

AU = Australia; AT = Austria; CH = Switzerland; DE = Germany; DK = Denmark; ES = Spain; FR = France; GB = United Kingdom; IE =Ireland; IT = Italy; JP = Japan; KR = Korea; MX = Mexico; NL = Netherland; NZ = New Zealand; US = United States; ZA = South Africa

1 The cross-country relationship between ITO = (exports & imports of intermediate goods and services)/GDP and the relative globalfactor (RGF) for 18 economies. Each observation shows the over-time average of ITO and RGF 1982-2006. The red fitted line has aslope of 2.09 (significant at the 1% level). Canada (RGF=-3.17, ITO=0.40) is not included. 2 The relationship between ITO=(exports &imports of intermediate goods and services)/GDP and the relative global factor (RGF) for the years from 1983 to 2006. Eachobservation shows the cross-country average of GVCI and RGF in a given year for 18 countries. The red fitted line has a slope of 15.6(significant at the 1% level).

Source: Auer et al (2017).

9

II – The FCD narrative

Current plight: (series of) financial booms gone wrong and an inadequate policy response

Elements of the story MP focused on near-term price stability plus inadequate regulation and supervision

allowed unsustainable financial booms to develop Booms turned to bust and caused major recessions Policy response to recessions and aftermath was inadequate

- Too little balance sheet repair- Too much traditional aggregate demand management and overreliance on MP

Over time, policy effectiveness diminishes and side effects increase- Especially limitations of unusually low interest rates for unusually long- Difficulties in returning to robust sustainable growth- Financial stability risks in non-crisis-hit economies…

• Build-up of financial imbalances in EMEs (T 1) Along the way, both short-and long-term real interest rates decline (G 5)…

- ... and global debt-to-GDP ratios rise

10

Table 1: Early warning indicators for banking distress – risks ahead

Credit-to-GDP gap Property price gap Debt service ratio Debt service ratio if interest rates rise by 250 bp

Asia6 16.5 5.7 2.0 4.4Australia 2.8 3.7 1.1 5.0Brazil 0.9 –33.7 7.0 8.7Canada 17.4 9.0 3.2 7.4Central and eastern Europe

–12.2 9.9 –0.2 1.2

China 28.8 –1.9 5.5 8.9France 2.4 –9.5 1.2 4.3Germany –5.1 15.6 –1.9 –0.1Greece –14.4 11.9India –4.4 1.5 2.6Italy –13.8 –14.1 –0.5 1.6Japan 3.5 16.1 –2.2 0.6Korea 3.2 5.4 –0.4 3.2Mexico 8.9 7.7 0.7 1.4Netherlands –20.7 –11.4 0.6 5.4Nordic countries –2.3 3.5 0.0 3.9Portugal –41.1 13.8 –1.4 1.9South Africa –1.7 –9.2 –0.4 1.0Spain –46.6 –15.2 –3.0 –0.1Switzerland 8.2 7.8 0.0 3.1Turkey 5.5 5.8 7.4United Kingdom –19.5 1.0 –1.4 1.5United States –7.8 5.1 –1.5 1.0Legend Credit/GDP gap>10 Property gap>10 DSR>6 DSR>6

2≤Credit/GDP gap≤10

4≤DSR≤6 4≤DSR≤6

11

Graph 5: Interest rates sink as debt soars

1 From 1998, simple average of France, the United Kingdom and the United States; otherwise only the UnitedKingdom. 2 Nominal policy rate less consumer price inflation. 3 Aggregate based on weighted averages for G7economies plus China based on rolling GDP and PPP exchange rates.

Sources: Borio and Disyatat (2014), updated.

–4

–2

0

2

4

6

170

190

210

230

250

270

85 88 91 94 97 00 03 06 09 12 15

% of GDP

Long-term index-linked bond yield1

Real policy rate2, 3

Lhs:Global debt (public and private non-financial sector)3

Rhs:

12

II – The natural (equilibrium) interest rate

Four points on the natural rate’s level and long-term decline The rate is not observable

- Inferred based on an assumed model of the economy- Inflation is assumed to provide a key signal

If allow also financial imbalances to provide a signal- This is more consistent with the data (G 6)- And results in a higher estimate (G 7)

• Same logic why FC-based measures of potential output work pre-crisis Defining the equilibrium rate without reference to financial stability is incomplete

- How can one argue that an equilibrium rate causes instability? Long-term interest rates can be misaligned for very long periods

- All asset prices can (common source of financial instability)- Should we now think that SS is not a risk as markets have changed their mind?

13

Graph 6: The financial cycle helps explain the variation in the output gap and the natural rate

Output gap Natural rate

The leverage gap and debt service gap are proxies for the financial cycle. The graph indicates that the informationcontent of inflation for the output gap (potential output) and for the natural rate is quite limited once the data areallowed to choose between inflation and financial cycle proxies.Source: Juselius et al (2016); based on US data.

14

Comparing interest rates: standard and financial cycle-adjusted

The standard natural rate estimate follows a common procedure, which assumes that inflation provides the key signal. The financial-cycle adjusted estimates allows, in addition, financial cycle proxies to play a role. The dotted line traces what the natural rate could have been in a counterfactual exercise in which monetary policy had leaned systematically against the financial cycle in addition to output and inflation as opposed to follow its actual historical path.Sources: Juselius et al (2016); based on US data.

–3.0

–1.5

0.0

1.5

3.0

4.5

95 00 05 10 15

%

Real policy rate Standard natural rate Financial cycle-adjusted natural rateFinancial cycle-adjusted natural rate (counterfactual)

(Graph 7)

15

III – The banking industry: the good news

In crisis-hit economies, profits have recovered albeit at an uneven pace… Recent losses have in part reflected the cleaning up of loan portfolios.

… And banks have been rebuilding their capital at a brisk pace (G 8) CET 1 capital up from 7 to 12% between 2011-2015 Capital shortfall effectively eliminated Mainly through retained earnings and little impact on credit

In non-crisis-hit economies, profits and balance-sheet strength look generally good (G 9) High profits and rising equity-to-assets ratios Exceptions where FCs have turned

- But no full-blown crisis

16

Graph 8: Banks have strengthened their balance sheets

Bank capital ratios on the rise¹

Capital shortfalls have been met², ³

Profits support capital increases

1 Group 1 banks. 2 Group 1 and Group 2 banks. 3 The height of each bar shows the aggregated capital shortfall considering requirements for each tier (ie CET1, Tier 1 and total) of capital.

Sources: Basel Committee on Banking Supervision; BIS calculations.

17

Graph 9: Profits and capitalisation: EME banks outperform their advanced economy peers

Net income, AEs¹ Net income, EMEs¹ Capital ratios, EMEs³

1 Net income as a percentage of total assets. The calculation of total assets may differ across banks due to different accounting rules (eg on netting of derivative positions). 2 France, Germany, Italy and Spain. 3 Median ratios.

Sources: SNL; BIS calculations.

In per cent

18

III – The banking industry: the bad news

Widespread scepticism Price-to-book ratios below 1 in many advanced economies, esp. crisis-hit ones (G 10)

- Shares have underperformed market Credit ratings tell a similar story (G 11)

- For euro area banks stand-alone ratings are lower than at end-2010- Not much better for UK and US banks

Loss of competitiveness vis-à-vis market-based finance (G 11)- All-in ratings have deteriorated relative to corporates- Increase in market borrowing costs even for similar ratings

• Boost banks’ disintermediation

Where optimism present, need not be always well founded Questions about rosy financial statements in countries with strong financial booms

- Statements are lagging indicators of trouble- Pre-crisis refrain «banks have never been as well capitalised»

Only time will tell

19

Graph 10: Bank price-to-book ratios struggle to recover¹

Advanced economies Latin America Emerging Asia

1 End-of-month data.

Sources: Datastream; S&P Capital IQ; BIS calculations.

Ratio

20

Graph 11: Weak ratings erode bank’s funding advantage

Bank ratings¹, ² Non-financial corporate ratings¹Relative funding costs: A-rated banks vs A-rated NFCs³

1 Asset-weighted averages. 2 Numbers of banks in parentheses. 3 Option-adjusted spread on a bank sub-index minus that on a non-financial corporate sub-index, divided by the spread on the non-financial corporate sub-index. Sub-indices comprise local currency assets.

Sources: Bank of America Merrill Lynch; Moody’s; BIS calculations.

21

III – The banking industry: Why such scepticism?

Reason 1: Stock problem – legacy of the financial bust and its aftermath Low asset quality and mistrust (miscounduct and risk-weight optimisation)

Why persistent asset quality issues? A morphing crisis (from GFC to sovereign strains in the euro area) Protracted post-financial boom economic weakness Uneven policy response

- Lessons of early-1990s financial busts not fully learnt• Nordic countries vs Japan

- Why?• Initial mis-diagnosis (liquidity vs solvency)• Harder to resolve: securitisation and international dimension• Main reason: thought could afford to wait

22

III – The banking industry: Why such scepticism? (ctd)

Reason 2: profit prospects In addition to the Fintech challenge and regulatory concerns… … a key factor has been persistent ultra-low (nominal) interest rates

- Result of bust and of MP response to push inflation back towards objectives

Why concerns with persistent ultra-low nominal interest rates? Unprecedently low

- Even negative out the yield curve in some jurisdictions Low rates may temporarily raise profits….

- Capital gains and higher activity ... But if persistent enough hurt net interest margins

- Impact of low rates and flat yield curves- Empirical evidence confirms this: Non-linear relationship at work (G 12)

Such rates raise a business model conundrum Retail model that had performed best through crisis and has gained ground (T 2).... ... is the most vulnerable to ultra-low rates: large retail deposit base

23

Graph 12: Lower interest rates can be a drag on profitability

RoA = profit before taxes divided by total assets; short-term rate = three-month interbank rate, in per cent; slope of theyield curve = spread between the 10-year government bond and three-month interbank rate, in percentage points. Thevertical axis reports the derivative (response) of RoA with respect to the short-term rate (left-hand panel) and the slopeof the yield curve (right-hand panel), in percentage points. The shaded area indicates 95% confidence bands.

The downward slopes indicate that as the level of the interest rate or yield curve slope decrease, the depressive impacton profitability (return on assets) becomes larger.

Sources: Borio et al (2015a).

dRoA

/ d

Shor

t rat

e

dRoA

/ d

Slop

e

24

¹ Based on a sample of 108 banks from advanced and emerging market economies.

Sources: Roengpitya et al (2014).

Table 2: Business models: traditional banking regains popularity¹ Number of banks

Business model in 2007Retail-funded Wholesale-funded Trading Total

Business Retail-funded 53 10 0 63model Wholesale-funded 3 25 2 30in 2005 Trading 2 0 13 15

Total 58 35 15 108Business model in 2013

Retail-funded Wholesale-funded Trading TotalBusiness Retail-funded 57 1 0 58model Wholesale-funded 16 16 3 35in 2007 Trading 3 1 11 15

Total 76 18 14 108

25

III – The banking industry: Why such scepticism? (ctd)

Reason 3: Stock and flow problems have exacerbated structural difficulties Tendency towards excess capacity in the industry

Structural because of the well-known exit difficulties Exit is harder than in other sectors because of costs for rest of the economy

- Safety nets induce over-expansion by reducing market discipline - Cost-cutting is key

Progress uneven and disappointing relative to post-Nordic crisis experience (G 13)- But critical to establish basis for sustainable profitability

26

Graph 13: Uneven progress in bank’s costs adjustments

1 Based on 2015 list of G-SIBs for which cost-to-income ratios were available. 2 For the change in the number of branches, 2007–14.

Sources: OECD; SNL; national data; BIS calculations.

In per cent

Cost-to-income remains stubbornly high for G-SIBs1

Cost savings vary 8 years after the onset of the crisis

27

Conclusion Two key takeaways

The world has not been suffering so much from SS but from FCD This has left profound scars on the banking industry

- In financial bust countries: low asset quality, low profitability, excess capacity- In several elsewhere: build-up of possibly unsustainable financial booms

Implications for the global outlook Demand-driven constraints on growth are persistent but temporary

- supply-driven constraints are more important in the medium term World faces a “risky trinity”

- Long-term decline in productivity growth- Increasing debt levels (private and public) globally- Limited room for policy manoeuvre

• Risks of further episodes of financial distress in non-crisis hit countries Implications for policy

Macroeconomic level: take the FC more systematically into account- Develop a macro-financial stability framework

Banking industry level: strengthen regulation and facilitate structural adjustment- Completion of regulatory reforms, notably Basel III, is critical

28

Auer, R, C Borio and A Filardo (2017): The globalisation of inflation: the growing importance of global value chains, BIS Working Papers, no 602, January. Bank for International Settlements (2014): 84th BIS Annual Report, June ______ (2015): 85th BIS Annual Report, June ______ (2016) 86th Annual Report,, June Bech, M, L Gambacorta and E Kharroubi (2012): “Monetary policy in a downturn: are financial crises special?”, BIS Working Papers, no 388, September.

(published in International Finance) Borio (2014a): The financial cycle and macroeconomics: what have we learnt?, Journal of Banking & Finance, vol 45, pp 182–198, August. Also available

as BIS Working Papers, no 395, December 2012. —— (2014b): The international monetary and financial system: its Achilles heel and what to do about it, BIS Working Papers, no 456, September. —— (2014d): Monetary policy and financial stability: what role in prevention and recovery?, Capitalism and Society, Vol 9(2), Article 1. Also available as

BIS Working Papers, no 440, January. ______ (2016a): Revisiting three intellectual pillars of monetary policy received wisdom, Cato Journal, vol 36, no 2. Also available as BIS Speeches. ______ (2016b): The banking industry: struggling to move on, keynote speech at the 5th EBA Research Workshop, 28-29 November. Borio, C and P Disyatat (2014): Low interest rates and secular stagnation: Is debt a missing link? , Vox EU, 25 June Borio, C, P Disyatat and M Juselius (2016b): Rethinking potential output: embedding information about the financial cycle, Oxford Economic Papers, Also

available as BIS Working Papers, no 404, February 2013. Borio, C, B Hofmann and L Gambacorta (2015a): The influence of monetary policy on bank profitability, BIS Working Papers, no 514, October.

Forthcoming in International Finance. Borio, C and M Drehmann (2009): Assessing the risk of banking crises – revisited”, BIS Quarterly Review, March, pp 29–46 Borio, C, M Erdem, B Hoffman and A Filardo (2015a): The costs of deflations: a historical perspective, BIS Quarterly Review, March. Borio, C, E Kharroubi, C Upper and F Zampolli (2015b): Labour reallocation and productivity dynamics: financial causes, real consequences, BIS Working

Papers, no 534, December. Borio, C, B Vale and G von Peter (2010): Resolving the financial crisis: are we heeding the lessons from the Nordics?, Moneda y Crédito, 230, pp 7-47.

Also available as BIS Working Papers, no 311, July. Borio, C and A Zabai (2016): Unconventional monetary policies: a re-appraisal, BIS Working Papers, no 570, July. Forthcoming in R Lastra and P Conti-

Brown (eds), Research Handbook on Central Banking, Edward Elgar Publishing Bruno, V and H Shin (2014): Cross-border banking and global liquidity, BIS Working Papers, no 458, September. Cecchetti, S and E Kharroubi (2015): Why does financial sector growth crowd out real economic growth?, BIS Working Papers, no 490, February. Drehmann, M, C Borio and K Tsatsaronis (2012): Characterising the financial cycle: don’t lose sight of the medium term!, BIS Working Papers, no 355,

November. Filardo, A and P Rungcharoenkitkul (2016): A quantitative case for leaning against the wind, BIS Working Papers, no 594, December. Hofmann, B and B Bogdanova (2012): Taylor rules and monetary policy: a global "Great Deviation"? BIS Quarterly Review, September, pp 37–49. Hofmann, B and E Tákats: (2015): International monetary spillovers, BIS Quarterly Review, September. Juselius, M, C Borio, P Disyatat, M Drehmann, (2016): Monetary policy, the financial cycle and ultra-low interest rates, BIS Working Papers, no 569, July.

Forthcoming in International Journal of Central Banking. Roengpitya, R, N Tarashev and K Tsatsaronis (2014): Bank business models, BIS Quarterly Review, December. McCauley, R, P McGuire and V Sushko (2015): Global dollar credit: links to US monetary policy and leverage , BIS Working Papers, no 483, January.

http://www.bis.org/publ/work602.htmhttp://www.bis.org/publ/arpdf/ar2014e.htmhttp://www.bis.org/publ/arpdf/ar2015e.htmhttp://www.bis.org/publ/arpdf/ar2016e.htm?m=5|24http://www.bis.org/publ/work388.htmhttp://www.bis.org/publ/work395.htmhttp://www.bis.org/publ/work456.htmhttp://www.bis.org/publ/work440.htmhttp://www.bis.org/speeches/sp151112.htm?ql=1http://www.bis.org/speeches/sp161128.htmhttp://www.voxeu.org/article/low-interest-rates-secular-stagnation-and-debthttp://www.bis.org/publ/work404.htmhttp://www.bis.org/publ/work514.htmhttp://www.bis.org/publ/qtrpdf/r_qt0903e.pdfhttp://www.bis.org/publ/qtrpdf/r_qt1503e.htmhttp://www.bis.org/publ/work534.htmhttp://www.bis.org/publ/work311.htmhttp://www.bis.org/publ/work570.htmhttp://www.bis.org/publ/work458.htmhttp://www.bis.org/publ/work490.htmhttp://www.bis.org/publ/work380.htmhttp://www.bis.org/publ/work594.htmhttp://www.bis.org/publ/qtrpdf/r_qt1209f.htmhttp://www.bis.org/publ/qtrpdf/r_qt1509i.htmhttp://www.bis.org/publ/work569.htmhttp://www.bis.org/publ/qtrpdf/r_qt1412g.htmhttp://www.bis.org/publ/work483.htm

CFS_170207 links.pdfSecular stagnation or financial cycle drag?�Implications for the banking industryQuestions and thesisI - The two hypotheses: a stylised characterisationGraph 1: The financial cycle is longer than the business cycle (the US example)�II - The SS hypothesis: a critiqueGraph 2: Financial booms sap productivity by misallocating resourcesGraph 3: US output gaps: ex post and real-time estimatesSlide Number 8II – The FCD narrativeTable 1: Early warning indicators for banking distress – risks aheadGraph 5: Interest rates sink as debt soarsII – The natural (equilibrium) interest rateGraph 6: The financial cycle helps explain the variation in the output gap and the natural rateComparing interest rates: standard and financial cycle-adjusted III – The banking industry: the good news Graph 8: Banks have strengthened their balance sheetsGraph 9: Profits and capitalisation: EME banks outperform their advanced economy peersIII – The banking industry: the bad news Graph 10: Bank price-to-book ratios struggle to recover¹Graph 11: Weak ratings erode bank’s funding advantage III – The banking industry: Why such scepticism?III – The banking industry: Why such scepticism? (ctd)Graph 12: Lower interest rates can be a drag on profitabilityTable 2: Business models: traditional banking regains popularity¹ III – The banking industry: Why such scepticism? (ctd)Graph 13: Uneven progress in bank’s costs adjustmentsConclusionSlide Number 28

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 150 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.00000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 600 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.00000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/CreateJDFFile false /Description >>> setdistillerparams> setpagedevice