International banking has exhibited some divergent patterns, largely stemming from the differential systems or institutional legacies. However, the distinct models still afford enough overlapping areas to lend some extra relevance to select conceptual frameworks yet to be invoked. The present report seeks to strike a careful balance between providing a critical account of an extant empirical literature while reconciling its findings with the theoretical frameworks that might otherwise appear incomplete under reduced dimensionality.
Market Liquidity versus Funding Liquidity Risks in International Banking
By and large, these could be thought of as dual and distinct yet entangled domains. Moreover, either one could be seen in dual ways. For starters, the market’s liquidity as a systemic parameter could be reduced to its depth, which refers to the marginal trading volume or turnover that only affects the price or overall cap on margin, or with negligible effect. The metric of illiquidity, be it for the market at large or for a particular asset, could be seen as the inverse elasticity, a la Amihud index measuring just how material the daily response in the price (with the downside effect such as capital loss) has been relative to the respective daily trading volume. Irrespective of whether or how this is related to value-at-risk (VaR) or downside risk that the risk-averse investor could prove more sensitive to amidst market crash or bank runs, illiquidity suggests a disproportionate percentage response or elasticity.
For that matter, the market at large or in a particular asset or ‘style’ is referred to as ‘thin’ as well as illiquid whenever the absolute trading volume has been small, or the frequency of quotes low as per a specific equity or bond. Somewhat loosely, market liquidity risk could thus pertain to how material an asset loss might prove in the event of fire sales, which lends special relevance to liquidity ratios as part and parcel of Basel III seeking to augment conventional CAMELS. On second thought, it would appear that regular “gap management,” referring to maturity match or immunization based on effective durations, might capture the bulk of the asset composition versus capital structure, albeit without necessarily saying much on asset quality per se.
Funding liquidity risk, by contrast, would pertain to either short-term funding being denied or to costly rollover terms. For one thing, this may directly have to do with working capital management—even though emergency financing or bail-out programs are a stand-alone pillar that cannot even qualify as exceptional items or extraordinary events. Indirectly or implicitly, greater funding liquidity means lesser liquidity pressure on the working capital, which translates into an equivalence of better market liquidity for assets at large irrespective of the formal liquidity ratio. For the same token, higher market liquidity implies smaller spreads as well as haircuts, which renders margin call less likely or frequent, thereby implying less intense regulatory pressure at the funding liquidity ends. In this light, either one could be seen as a self-reinforcing property in a multi-stage setup, depending on the initial conditions or business model chosen at the outset. Likewise, it hinges on the exogenous market terms or structure as part of SCP (structure-conduct-performance) analysis to be addressed later on.
To rehash on the implicit interchange between borrowing and lending liquidity, Brunnermeier’s notion of network type risk could be embarked on when rethinking leverage or gearing as a parameter of market structure.
Business Models in Commercial versus Investment Baking
Whereas the conventional commercial bank garners the bulk of its earnings in the regular business lines accruing as loan interest differential and settlements or transaction fees, the representative investment bank’s operating or contribution margin is accounted for by its underwriting business pertaining to fixed-income as well as equity securities. Their single largest overlapping area pertains to scale efficiency as well as efficacy, as both seek to reap hefty as well as stable proceeds irrespective of how low the interest spread is. These business models will now be treated in greater depth.
Commercial banks may at one point come to see their net margins shrink depending on either the industry competition or the monetary policy. However, much like in the subprime case, small lending interest could be offset by the booming scale amidst bandwagon enthusiasm in loan taking affecting the asset growth. By contrast, settlement and servicing fees would amount to the more stable layer of earnings that may or may not outmatch the lending business, yet in any event reveals direct correlation with the scale of operations. What might be viewed as transaction costs or efficiency detractor by the customers is seen as a major efficacy pillar any commercial bank (as well as a brokerage) would boast. In actuality, though, both parties seek to strike a balance of optimum transaction frequency aimed at preserving the value added as well as its reasonable distribution or sharing.
In contrast, the investment banker (which is routinely viewed as anything but a banking entity as far as the Continental model is concerned) will ensure that the entire bond or stock issuance be bought into, whether as an IPO or a private placement. A hedge of the sort will come at a cost to the issuer, which amounts to the underwriter’s subscription and advisory fee. In fact, the range of services, or indeed the layers of the value chain, might embark on extra levels as diverse as, arranging for an underwriting syndicate or selling group, stabilizing the market (which might hint at an interim ‘specialist’ or market maker’s status in the asset in question), and road show marketing ex ante alongside swapping loss mitigation ex post. In effect, this entire scope could be thought of as diversification in its own right, aimed at partially absorbing the possible losses resulting from post-IPO price sags or ‘post-money’ valuation.
Although no commercial bank is allowed to be a full-fledged investment manager (much less a principal investor in equity), there are fewer restrictions on ownership, whereby a single set of shareholders could treat both as business units, which is one way around the BHC (bank holding company) or Glass-Steagall restriction. Its posterior revisions or relaxations could be deemed as realization of the constraint being more damaging to efficiency or stability than binding.
Bank versus Capital Based Financial Systems
Depending on the institutional legacy, economies by and large tend to adhere to either the Continental or the Anglo-American system. The former draws heavily upon bank debt financing over and above paid-in capital and retained earnings, whereas the latter reveals a far more pronounced part as contributed by capital financing, such as bond debt and equity—both traded with a sizable free float. More specifically, the bank system may or may not reveal a high leverage, yet in any event the equity margin is scarcely ever composed of publicly held stocks. The silver lining might be that the gap between book value versus market value is negligible or less than relevant—for equity and debt alike. By contrast, the marking-to-market implications for GAAP or common law based systems might either be inconclusive at this stage or too far-fetched with an eye on Basel III rethinking yet to emerge.
In a sense, the German version could be representative of the Continental or bank model, which builds on a statutory or civil code legal system. The latter could be thought of as far less flexible, with extra stability being the flip side. Dominant forms such as Landesbank point to an implicitly cooperative paradigm, with size hardly making a huge difference whenever the public good or regional development is at stake as a line of convergence.
In sharp contrast with the above, the UK and US systems show nearly equal odds for the polar extremes of ex-ante competitive incentives mapping into explicit collusion or ex-post concentration along more efficiency-laden lines, e.g. scale and scope falling little short of reaping the alleged natural-monopoly benefits.
Although in either case the ratio of branching to size has been somewhat inconclusive or unstable historically, the two diverge drastically as a matter of dominant structures and patterns. Whilst the Continental model builds on few large banks (low concentration ratio or HHI) as well as fewer branches per bank, the common-law alternative proves more of a hub-and-spikes type.
Modern Financial Crises: Drivers & Causes
In light of the aforementioned trade-off between size and networking scale, however, it remains to be seen whichever system boasts greater immunity vis-à-vis crises with an eye toward concentration and endogenous or non-regulatory institutional structure alike. For one thing, insofar as crisis can be coupled with liquidity crunch as proposed above, it is the availability of short-term rollover as well as a high liquidity ratio or balanced gap management that posit the individual bank as less prone to the aftermath of major runs. At the macro-level, high concentration might embed some extra stability too, in that the smaller banks cannot possibly have accounted for the bulk of the system-wide money multiplier or effective, velocity adjusted monetary mass.
On second thought, this refers to the fundamental, money creating capacity as a driver of aggregate demand. Instantaneously, any small bank could turn out to be a major bottleneck on failure to settle payments a few days in a row, in which case a chain reaction mounts with many more following suit, unless emergency funding can readily be tapped into.
International banking could be of special interest, in that they are exposed to currency risks (bordering on twin crises) yet could enjoy access to a variety of hedging as well as rollover facilities to choose from. Although capital constraints of the Tobin type could work both ways on the downside (thus denying stop-loss reversals or prompt and low-cost entry into currency swaps), still far from many economy cycles would end up so synchronized for major contagion to have plagued the global bank’s entire value chain.
Some extra downside co-movement may well be ushered in by large-sized institutional players, notably hedge funds or otherwise ‘style’ investors as in Barberis & Shleifer, which are more likely to group non-Continental type markets into similar categories likely to see contagious spillover. On the other hand, the existence of “concept stocks” as a style strongly preferred by the speculative profiles would suggest a matter of self-selection match and one pillar of non-regulatory screening.
Herding could be the behavioral dual of style-induced co-movement, when it comes to incentives structure. Some of the smaller investors might believe, as a matter of satisficing or deliberation cost saving amidst high expected losses, that market inefficiency suggests some of the larger players are more privy to insider information, so their moves could safely be mimicked. The “better fool” stance could be a replica of just another Ponzi scheme, with an otherwise rational belief in the smaller investors’ restricted access to complete information coupled with their flawed stop-loss or position-reversal timing propensity.
Apart from the general incentives working under weak-form efficiency, their ad-hoc impact on the real-estate bubble bursts in 2007-2008 could be of special importance. As it happens, too lenient rollover that retail lending would enjoy on long-term mortgages had at some stage mapped into the moral hazard on the part of the otherwise scrupulous home owners that saw their investment principal, even in residual or net terms, outmatching the decaying market price. Although interim terminal value might have appared less of an issue to those still in need of their sole home, the speculative types having more discretion must have manipulated the bubble at the boom and bust stages alike. On the expectations of soaring haircuts alone, banks saw their collateralized bonds lose value irrespective of the interest rates or yields, which was one other shock that may have been factored in.
One final afterthought would draw on Laeven & Valencia, with reference to the observed trade-off between the smaller direct fiscal burden having developmentally unsustainable debt overhang or otherwise public burden that is hard to amortize or spread. It appears that neither securitization nor Tobin taxes have succeeded at sterilizing excessive losses or downside opportunity costs that are hardly commensurate to the respective period’s GDP equivalent.
Economic versus Regulatory Capital
This level of analysis features duality of another sort, which pertains to the multiple alternate interpretations of categories or metrics that are about as distinct as they are inter-related and possibly overlapping. To begin with, economic capital could pertain to whichever funding sources other than hot money suffice to cover or offset comparably timed losses, expenses, or outlays. For the most part, it is equity (along with retained earnings) and long-term liabilities that are treated as capital irrespective of company profile. However, specific to banking or FI, it is the less risky (or less likely to be claimed first) liability layer such as subordinated debt that readily counts toward Tier 2 capital.
As early as at this stage, the line proves to be fuzzy between economic versus regulatory capital, as risk-weighted schemes and categories are defined and imposed in line with Basel II. Although this is neither GAAP nor IFRS specific, it is binding enough for the purpose of comparison and regulatory screening or licensing. If one were to narrow it down, regulatory capital would refer to a cut-off or lower-bound applying to the CAR (capital adequacy ratio) overall and each particular layer (e.g. 4% to 8% at least). The actual ratio is what pertains to the positive rather than normative side.
Alternatively, the economic capital as a matter of loss absorbing capacity could be inferred ex post or imposed ex ante in terms of VaR or its KPI dual (in line with CAMELS or rating target). Along these lines, economic capital could be referred to as the core of insensitivity or stress test performance—in line with Basel II CAMELS (or its S-pillar). On the other hand, it remains to be seen whether imposing fixed weights makes much sense beyond ease of comparison, or whether their rigidity can be maintained irrespective of the RBC (real business cycle) stages or the bank-specific asset quality.
To extend the critique even further, suffice it to refer to Cihak suggesting a step function approach for mapping CAR into default likelihoods. Although the actual elasticity is to be re-estimated much the way posterior likelihoods are, it is no wonder that the poor prior performance at 12% (failure to detect 88% cases) only improves to about 66%, with the bulk of evidence still faring as inconclusive. Insofar as the study deploys Type II error test, the implied power analysis would suggest a low beta (power in addition to the alpha significance) as per any sticky states or near-zero effect sizes (i.e. the cognate of Z), which call for degrees of freedom or effective sample sizes far in excess of those claimed.
Bank Profitability & Market Concentration: An SCP Account
The SCP (structure-conduct-performance) paradigm, as presumably the underpinning of Porter’s Five Forces, refers to the non-linear inter-linkage between market structure (notably concentration), company strategy or organizational behavior, and the resultant profitability as well as efficiency outcomes for the system at large. When it comes to banking, the relationship of concentration versus profitability (e.g. HHI serving as one explanatory variable for ROA or ROE to be regressed on) could be key to understanding how business models map into actual earnings as well as the other way around, with the market structure acting as the sharing parameter.
On the one hand, insofar as higher concentration is coupled with greater relative market power or bargaining power, the largest banks are in a position to charge higher interest on smaller loans while paying lower interest on deposits. Apart from the upside efficacy, they must have embarked on larger and more efficient scales as well as scope, with the break-even being attained and outmatched on either each standalone product or on the whole, along the lines of lean value creation. There may be some additional efficiency loss items out there confronting the retail customers, yet the concentrated system will likely prove so stable as to create money in a more sustained mode, with the money multiplier and velocity being less volatile (which carries over to GDP and its growth).
Inasmuch as higher concentration could be a matter of X-efficiency (with superior asset management and capital match likely to have secured faster organic growth in the retained earnings to accommodate asset growth), not only could large banks prevent adverse selection, they moreover are more poised to cater for the real sector by channeling funds into productive investments rather than being concerned with incessant hedging.
That said, a few caveats would seem due, in reference to Berger’s model and the role of TBTF (too-big-to-fail) bail-out remedies. The latter appears to be distorting the incentives while crowding out market discipline with cushion that may at best fend off speculative attacks as a matter of target zone support signaling. At the same time, it would posit superior ranks with rating agencies as the sole prerogative of the arbitrarily big banks, bearing in mind that the bulk of Moody’s and Fitch scores are just that—levels of support beyond regular fundamentals or CAMEL performance.
On the other hand, insofar as the latter is what informs the host of right-hand side performance variables other than the left-hand side ROA or ROE, it happens that this fudge aggregate (no matter how complete or arbitrarily cherry-picked on) could reveal multi-collinearity thus rendering their individual significance low. Although the latter might not be very relevant given their control-variable nature, the overall F-score likewise appears less than critical in light of the low R-squared. At the very best, this might refer to an incomplete set of main factors or sheer lack of interactive ones.
Alternatively, the inherently non-linear inter-linkage could be at odds with OLS regressions. Along similar lines, the deployment of logistic functions alongside dummy explanatory variables as one way of capturing the downside odds could violate the homoskedasticity requirement keeping the OLS BLUE, insofar as the residuals reveal a structure (a Bernoulli type distribution utterly at odds with constant variance).
Key Determinants of Bank M&A
Whereas industry concentration cannot amount to ex-ante intended or targeted outcomes on the part of the banking entities irrespective of their colluding propensity, it may well have resulted ex post for better or worse—or indeed been fostered by the policy makers supposedly acting as benevolent social planners rather than corrupt lobbyists or crony rent seekers. Target operating performance may well have been central to shaping strategy along the SCP lines, with operating synergy possibly stemming from scale and scope alike. If anything, it is the resultant capitalization of the target that may at the very least have acted to boost the scale while augmenting the key ratios, CAR not least (as the book value must have added up much the way its market-value counterpart did).
When it comes to the BCG trade-off between size versus growth, neither may have been pursued directly. On the one hand, M&A may have affected the affiliation or ownership yet not necessarily the overall branching. In other words, the short-term expansion has not been organic, even though longer-term growth might have been targeted with an eye on new products and pilot projects that small and resource-strained banks cannot possibly tap into. Again, higher profitability must have mapped into faster growth—unless the larger consolidated entity positions itself as a ‘cash cow’ referring to a low retention or reinvestment ratio as applied to its ROE.
The banking sector as part and parcel of the financial services industry has revealed a rather peculiar SCP framework, with its international dimensions adding on further complexity alongside automatic stabilizers. Not only do large banks indulge in indiscriminate diversification globally (which layer of opportunity is denied to small rivals by definition), they appear to contribute to superior allocation outcomes as accruing to the general equilibrium at large, even amidst sector-specific friction or efficiency slacks left over. In fact, these might be attached with some excessive risks in their own right, whenever a full-fledged perspective is invoked beyond the empirical models that treat the missing domains as but control variables.