An Analysis of Risk at German Banks

Abstract

Cross-sectional and regression analysis of ECB data shows that German banks appear to lag peers on the Continent on most earnings, stability and efficiency measures. However, longitudinal analysis for the onset of the current recession suggests that prudence in not pursuing expansion and high-flying revenues may have protected German banks from disastrous losses.

Introduction

The lingering recession precipitated by flawed mortgage decisions earlier in the decade, spun out of control with trans-Atlantic speculation in mortgage-backed securities and emerged full blown in the second semester of 2007 put banks squarely in the spotlight. The U.S. Federal Reserve, the European Central Bank and national central banks are estimated to have injected at least $4 trillion dollars by buying up “distressed assets” and acquiring preferred stock in banks (Altman, 2009)

The toll on financial institutions was severe and even long-established institutions failed. In the UK, government took over Northern Rock. Arguing that the ripple effects would be far worse if the largest banks closed, the USA government bought out AIG, Freddie Mac and Fannie Mae. The Fed also had a hand in letting Bear Sterns, Lehman Brothers, Merrill Lynch, Washington Mutual, and Wachovia either close their doors or endure sell-outs to peers in the industry. All told, 23 American banks failed in 2008 and almost as many, 21, closed their doors in 2009. (Recession.org, 2009).

Sovereign risk suffered, too. In November 2009, state-owned Dubai World informed its foreign bankers it would have to reschedule over $9 billion in repayments due December 2009 and over the first three months of 2010 (Sharif and Cochrane, 2009).

In large part, the lesson one draws from the sorry tales of derivatives trading –mortgage-backed securities (MBS), collateralized debt obligations (CDO), credit default swaps (CDS) all finally liquidated at as little as five cents on the dollar – was how little attention investors paid to risk appraisal and pricing.

Review of Relevant Literature and Theory

The magnitude of trillion-dollar subsidies, loans and buyouts has spilled over to the political arena. At Bretton Woods in 1944, the victors of World War II and their allies formalized the idea of government “management” of floating exchange rates based on gold. At the Davos World Economic Forum of 2010, Presidents Obama and Sarkozy urged political and banking leaders to mitigate financial-sector risk by calling for more regulation, cutting banks down to size, banning ownership of (or sponsoring) hedge and private equity funds, as well as engaging in so-called proprietary trading, that for their own account. This follows an agreement among the Group of 20 nations just last year to regulate banks more strenuously in point of capital, liquid assets, and executive compensation plans (Harper and Kirchfeld, 2010).

Dissatisfied with the published statements that concealed nearly as much as they revealed — mainly owing to innovation in the market place, speculation in derivatives and the trade in risk-bearing instruments – Altunbas, Gambacorta, and Marques-Ibanez (2009) set out to model (Equation 1 overleaf) the effects of monetary policy on size, liquidity, capitalization, and investor risk perceptions of new loans by building on an initial investigation Ehrman, Chatelain, Generale, Martinez-Pages, Vermeulen, and Worms (2003) conducted on how banks respond to monetary policy. In opting to investigate loan risk, Altunbas et al. presumably meant to measure both the willingness and liquid asset base that let a bank continue to underwrite new loans.

The model permitted Altunbas et al. to realise that when a bank expects a lower frequency of loan defaults, they respond with alacrity to shifting monetary policy environments by making more credit available.

Methodology and Data

The problem at hand – making a judgement on the risk faced by German banks versus others in the EU – is at the core of econometric research methods. For one, the problem is one of forecasting near-term performance of banks with a view to finding out whether any more failures will arise to stymie an incipient recovery from recession. Secondly, a sound solution helps both regulators and investors make better decisions about, respectively, stepping in with closer scrutiny or entrusting long-term fund placements to lower-risk banking sectors.

The general approach taken in this analysis combines findings from the data types available in the database:

• Cross-sectional analyses comparing the usual liquidity ratios relevant to banks;
• Time-series data to test whether assets and returns proved more vulnerable, all other things equal, to the onset of the global recession in the second half of 2007;
• Regression analysis to predict risk status of German banks and those in the EU as a group.

The raw data is secondary in nature, compiled and conforming to panel data since both longitudinal and cross-section analyses are possible.

Results and Discussions

Cross-Sectional Analysis

Over the period covered by the ECB database, German banks have been riskier and this is indicated, first of all, by:

• A higher rate of dissolutions (32%) and fewer active institutions overall (65.7%, see Table 1).
• An average asset base at under $4 million, compared to three times as much for banks elsewhere in Europe (Table 2).
• At $1.985 million, average loan portfolio in a German bank is just one-third the prevailing size outside the country.
• Other earning assets stood at just one-fourth the trans-European average (Table 2).

Nearly half the banks in this analysis are German (45,870 versus 40,590 elsewhere in Europe) and yet the former are distinctly smaller, as attested to by the ability to generate deposits and other sources of funds:

• Typically, German banks reported Average Deposits & Short term funding at just $2.6 million, less than one-third the European average (Table 3).
• Partly because so many institutions already reported negative balances, German banks averaged just $878 million in other funding sources, half the $1.6 billion prevailing in typical European banks (Table 3).

Other measures of strength include Tier 1 capital (equity) and reserves against losses:

• The equity measure in the ECB data presumably refers to Tier 1 capital, the fundamental criterion of financial strength to cover for unexpected losses. This comprises “core capital” (both common stock and disclosed reserves or retained earnings) and some types of non-redeemable non-cumulative preferred stock (Bank for International Settlements, 1997; BIS, 1998). On this basis, German banks typically had less than a fourth the equity-based solidity of other European banks (Table 4).
• Contingent liabilities or “off-balance sheet items” came to an average of just $400 million in German banks, around one-twelfth the $4.8 billion other European banks usually boasted. This bespeaks both significantly lower learning opportunity but also a lower risk profile from having to put the bank’s own resources at risk when third parties default on such off-balance sheet items as: “…standby … (and) irrevocable letters of credit that guarantee repayment of commercial paper or tax-exempt securities; risk participations in bankers’ acceptances; sale and repurchase agreements; and asset sales with recourse against the seller; interest rate swaps; interest rate options and currency options” (Dun & Bradstreet/All Business, 2006).

Other indicators of risk:

• Another caveat on the quality of loan portfolio of German banks is that, on average, loan loss provision (LLP) amounts to 0.58% of total loans outstanding versus just 0.55% for other European banks (Table 7). This bespeaks risk since LLP is counted as an expense item, a reserve for loans that turn sour owing to customer default, loan rescheduling or renegotiation of interest, etc. (Investopedia LLC, 2010).
• The capital (adequacy) ratio stands at 11.6 in the rest of Europe versus just half that much in Germany (Table 7). The Basel I and II covenants set varying ways to risk or weight the asset mix of a bank and do not therefore provide a uniform standard for this key bank ratio. Regardless of whether the U.S. Federal standard of 8.0 (Hummel, 2008) applies to the European situation, it is clear that German banks as a group are undercapitalized for the assets they hold.

For the rest, one finds that average net interest margin is precisely the same for Germany and the rest of Europe: 2.91% (Table 7). Since this is tantamount to gross revenue for any bank – the difference between interest received on loans made out and paid on deposits and other sources of funds (Bitner and Goddard, 1992) – one realizes that German bankers are about as perspicacious as elsewhere on the Continent when it comes to setting interest spreads. This ought to be a neutral factor were it not for the fact that German banks are saddled by many drawbacks, as this analysis has already shown.

German banks are also weaker in turning assets under their management into earnings. The ROAA metric in the database shows that pan-European banks usually generate earnings equivalent to 91% of average assets compared to just a third as much for German banks (30%, Table 7). This bespeaks a German banking sector that is relatively inefficient vis-à-vis peers in the EU (Bank Failure Resource, n.d.).

Vulnerability to Recession: 2007-2008

Perhaps there is something to be said for prudent management that verges on the ultra-cautious. When the effects of the speculation in sub-prime mortgages and the contraction in America crossed the Atlantic, the number of German banks still reporting dropped by one-fifth and average published net income went into the red (-$26.4 million, Figures 1 and 2, overleaf). On the other hand, the effects elsewhere in Europe were graver.

Regression Analysis: Germany and the Rest of Europe

Taking the viewpoint of a risk-averse investor desiring to minimize risk before taking a position in a German bank, it would be feasible to employ OLS to gain insight into the factors that impinge on earnings more than others. The first step is to select the independent variables that correlate better with published net income statements. Table 8 in the Appendices shows that higher correlation loadings are observed for: size of equity (r = 0.50), the “Other Operating Income” contribution to total revenue (r = 0.49), total deposits (r = 0.40), total loans (r = 0.39), total revenue (r = 0.38), securities held (r = 0.36), total assets (r = 0.35) and total interest income (r = 0.34).

Proceeding with the regression run filtered solely for German data, we obtain a model based on the calculated coefficients is as follows:

Net income = 1.211 + 0.251 (equity) – 0.01 (other operating income) – 1.26 (deposits) + 0.21 (loans) + 5.34 (revenue) + 0.61 (securities) + 0.05 (assets) – 4.89 (interest income)

Explaining about 33% of the total variance in the dependent variable net income, this computed model means that bank revenues may well be the single most powerful predictor of net income since the bank bottom line will rise by 1 for every increase of 5.34 standard deviations in bank revenues. Far from being a case of autocorrelation, it must be recalled that a bank can book plentiful revenues but keeping the lid on expenses and cost of funds is what translates to better net income.

Buying and holding on to securities is the second most important variable.

The model suggests that deposit growth and interest earned (chiefly from loans) are inversely related to growth in net income. The simplest explanation is that simply generating more deposit or loan business is not enough. A canny investor therefore looks at the quality and efficiency of bank management if he is to reasonably expect a sustained stream of earnings from equity participation.

All in all, this model looks robust for ANOVA results (Table 9) that test the null hypotheses H₀:β2=β3=β4=β5=β6=β7=0 against H₁:βj≠0. The F value is so high as to yield a significance statistic p < 0.001. This means the non-zero beta coefficients could have occurred by chance alone fewer than once in a thousand country data compilations.

Conclusions and recommendations

Even at the level of country banking sectors, the ECB database permits a much greater range of analyses than this concededly short paper can accommodate. For now, one observes that Germans banks have been riskier for being afflicted by a higher rate of dissolutions and being distinctly smaller in point of assets, loan portfolio and other earning assets. Modelling selected variables as determinants of income, the most influential IV’s seem to be total revenue and size of securities portfolio. Nonetheless, the model captures only a fraction of the variability in net income because, as Borio and Zhu (2008) point out, market innovation and an evolving capital regulatory framework (Basel II) are material factors in bank risk behaviour, too.

References

Altman, R. C. (2009) The great crash: A geopolitical setback for the west. Web.

Altunbas, Y., Gambacorta, L. and Marques-Ibanez, D. (2009) Bank risk and monetary policy. (Working Paper Series no 1075) Frankfurt am Main: European Central Bank.

Bank Failure Resource (n.d.) Return on equity, return on assets, and bank earnings performance. 2010. Web.

Bank for International Settlements (1997) Basle capital accord: International convergence of capital measurement and capital standards (1998). Web.

BIS (1998) Instruments eligible for inclusion in Tier 1 capital. Web.

Bitner, J. W. & Goddard, R. A. (1992) Successful bank asset/liability management: a guide to the future beyond GAP. Hoboken (NJ), John Wiley & Sons.

Borio C. & Zhu H. (2008) Capital regulation, risk-taking and monetary policy: a missing link in the transmission mechanism? BIS Working Papers, No. 268.

Dun & Bradstreet/All Business (2006) Barron’s Educational Series: Off-balance sheet items. Web.

Ehrman, M., Chatelain, J.-B., Generale, A., Martinez-Pages, J., Vermeulen, P. & Worms, A. (2003) Monetary policy transmission in the Euro area – evidence from micro data on banks and firms. Journal of the European Economic Association, 1, pp. 731-742.

Harper, C. and Kirchfeld, A. (2010) Obama proposal to curb banks dominates, divides Davos debates. Web.

Hummel, W. F. (2008) Banking basics. Web.

Investopedia LLC (2010) Loan loss provision. Web.

Recession.org (2009) FDIC bank failure & watch list. Web.

Sharif, A. & Cochrane, L. (2009) Dubai World seeks to delay debt payments as default risk soars. Web.