Banking Antitrust and the Philosophy of Economics: A Popperian Approach Presented By: Navid Nowakhtar.

Banking Antitrust and the Banking Antitrust and the Philosophy of Economics: A Philosophy of Economics: A

Popperian ApproachPopperian Approach

Presented By:Presented By:

Navid NowakhtarNavid Nowakhtar

OverviewOverview

Pragmatic VisionPragmatic Vision What is the Popperian Approach?What is the Popperian Approach? Literature HighlightsLiterature Highlights

Mainstream Literature (Bank Consolidation)Mainstream Literature (Bank Consolidation) Experimental Economics (Monopoly Theory Tested)Experimental Economics (Monopoly Theory Tested)

Qualitative AnalysisQualitative Analysis Statistical Inference on State Banks (ANOVA)Statistical Inference on State Banks (ANOVA) Macroeconomic Policy Implications and CaveatsMacroeconomic Policy Implications and Caveats

Pragmatic VisionPragmatic Vision

Illustrate the efficacy of philosophical Illustrate the efficacy of philosophical perspectives in shaping how economics theory perspectives in shaping how economics theory can be tested.can be tested.

Attempt to draw conclusions regarding Attempt to draw conclusions regarding Regulation of Financial Institutions (Thematic Regulation of Financial Institutions (Thematic Goal)Goal)

Original ContributionsOriginal Contributions Linking Popper to ExperimentalLinking Popper to Experimental Application of Experimental/Statistical Results for Application of Experimental/Statistical Results for

Tentative Qualitative/Quantitative Analysis of Banking Tentative Qualitative/Quantitative Analysis of Banking Antitrust (should laws exist?)Antitrust (should laws exist?)

Determination of Macroeconomic Policy FeasibilityDetermination of Macroeconomic Policy Feasibility

What is the Popperian Approach?What is the Popperian Approach?

Karl Popper: Logic of Scientific Discovery (1934)Karl Popper: Logic of Scientific Discovery (1934) Criterion of scientificability resting on falsity Criterion of scientificability resting on falsity

content.content. Contrast with Friedman (Conventional) Contrast with Friedman (Conventional)

econometric significance is synonymous with econometric significance is synonymous with empirical evidence.empirical evidence.

Samuelson: Foundations of Economic Analysis Samuelson: Foundations of Economic Analysis (1947)(1947) word game between falsifiability and word game between falsifiability and testability, or “operationally meaningful”.testability, or “operationally meaningful”.

The test criterion is more palatable but relays The test criterion is more palatable but relays the same concept.the same concept.

Popperian Approach Flowchart

All Conceivable Statements

Scientific Tautology (Meaningless) Non-scientific (Meaningful)

Contains Falsity Content A U A’ (Forbids nothing) Ethical/Value Related

All swans are white. I am lying or I am not lying. God exists.

Popperian Mandate for Science: Testability.

Mainstream LiteratureMainstream Literature

Antitrust laws: “restrict the size and economic Antitrust laws: “restrict the size and economic power that a single firm may hold in a power that a single firm may hold in a geographic or economic market" (Handa, geographic or economic market" (Handa, pg.129)pg.129)

Focus not on minutiae of law but on testing what Focus not on minutiae of law but on testing what theory purports about monopolies.theory purports about monopolies.

Concerns for Banking: Higher fees (monopoly or Concerns for Banking: Higher fees (monopoly or predatory pricing), lower levels of service predatory pricing), lower levels of service (uncontested markets), credit availability (uncontested markets), credit availability (uncontested market), and “too big to fail” (uncontested market), and “too big to fail” (taxation). (taxation).

Mainstream Literature HighlightsMainstream Literature Highlights

Gilbert and Zaretsky (2003)Gilbert and Zaretsky (2003) Local community market concentration; found that large banks Local community market concentration; found that large banks

lend to small businesses in areas where they have no lend to small businesses in areas where they have no branches; CUs compete with banks; results were divergent as branches; CUs compete with banks; results were divergent as multitude of models explored all have different specifications.multitude of models explored all have different specifications.

Scott and Dunkleberg (2003): Consolidation has no significant Scott and Dunkleberg (2003): Consolidation has no significant impact on ability of small firms to obtain a loan or the contract impact on ability of small firms to obtain a loan or the contract loan rate on the most recent loan from a commercial bank; loan rate on the most recent loan from a commercial bank; drawback is an increased rate of searching for banking drawback is an increased rate of searching for banking alternatives as a drain on smaller businesses.alternatives as a drain on smaller businesses.

Other issues: branchless banking (Deyoung 2003), economies of Other issues: branchless banking (Deyoung 2003), economies of scale, (Valverde and Humphrey (2004), qualitative task of merger scale, (Valverde and Humphrey (2004), qualitative task of merger analysis (Cyrnak (1998), Walter (2003).analysis (Cyrnak (1998), Walter (2003).

No real case in favor of trusts.No real case in favor of trusts. No mention of experimental economics evidence or institutional No mention of experimental economics evidence or institutional

design as loci of attention.design as loci of attention.

Experimental EconomicsExperimental Economics

Falsify operationally meaningful content that Falsify operationally meaningful content that exists in the theory.exists in the theory.

Corollary, not replacement for econometrics.Corollary, not replacement for econometrics. Agnostic Agnostic Strives to create lab contexts under which Strives to create lab contexts under which

theoretical predictions may occur.theoretical predictions may occur. Vernon Smith (1962): salience, dominance, non-Vernon Smith (1962): salience, dominance, non-

satiation, and privacy.satiation, and privacy. Limitations to date: low incentives, Limitations to date: low incentives,

representative sampling, little replication, lack of representative sampling, little replication, lack of accounting for role evolution (Harrison).accounting for role evolution (Harrison).

Experimental LiteratureExperimental Literature

Coase Conjecture: Demand Withholding:Coase Conjecture: Demand Withholding: Means Means that no monopolist will ever extract full monopoly that no monopolist will ever extract full monopoly power in DA vs. PO.power in DA vs. PO.

Smith (1981), Harrison (1990)Smith (1981), Harrison (1990) demand withholding demand withholding exists in DA with repetition over several periods, rare exists in DA with repetition over several periods, rare in PO.in PO.

Predatory PricingPredatory Pricing: dual markets (Issac and Smith : dual markets (Issac and Smith (1985), Harrison (1988): is a concern empirically (1985), Harrison (1988): is a concern empirically when there exists “escape routes” for firms; when there exists “escape routes” for firms; however, market remains contested in that firms are however, market remains contested in that firms are willing to incur costs of re-entry into the market as willing to incur costs of re-entry into the market as soon as monopolist attempts to charge monopoly soon as monopolist attempts to charge monopoly price. Intertemporal sets of losses may be need to price. Intertemporal sets of losses may be need to full de-contest the market.full de-contest the market.

Experimental LiteratureExperimental Literature Contested marketsContested markets: Competition for the market polices behavior : Competition for the market polices behavior

just as competition within the market does. just as competition within the market does. Coursey, Issac and Smith (1984): contestability with 1 or 2 Coursey, Issac and Smith (1984): contestability with 1 or 2

sellerssellers found that there was significant buyer withholding in all found that there was significant buyer withholding in all cases in a PO market, but could not conclude as to which factor cases in a PO market, but could not conclude as to which factor (buyer withholding or threat of entry) was the disciplining factor for (buyer withholding or threat of entry) was the disciplining factor for sellers. When the market is contested, they concluded that prices sellers. When the market is contested, they concluded that prices would be closer to the competitive equilibrium (CE) result.would be closer to the competitive equilibrium (CE) result.

Harrison and McKee (1985) computer-simulated buyers to control Harrison and McKee (1985) computer-simulated buyers to control for demand withholding, using a larger sample, algorithmically for demand withholding, using a larger sample, algorithmically controlling for risk neutrality, and also comparing 2-seller and 3-controlling for risk neutrality, and also comparing 2-seller and 3-seller contested markets with monopolies. They defined an indexseller contested markets with monopolies. They defined an index M M such that as Msuch that as M1, actual profits approached theoretical 1, actual profits approached theoretical monopoly profits. This study produced stronger results with monopoly profits. This study produced stronger results with respect to the subsidy amount for a regulated monopoly. This respect to the subsidy amount for a regulated monopoly. This may have been due to stronger results due to abnormal subject may have been due to stronger results due to abnormal subject intelligence.intelligence.

Experimental LiteratureExperimental Literature

Other forms of regulationOther forms of regulation Loeb-Magat decentralized (Cox and Issac 1987): Loeb-Magat decentralized (Cox and Issac 1987):

which lets the seller keep all consumer surplus, but which lets the seller keep all consumer surplus, but does not allow the seller to price discriminate, giving does not allow the seller to price discriminate, giving the seller an incentive to lower the price to C.E in the the seller an incentive to lower the price to C.E in the decreasing cost case.decreasing cost case.

Franchise Bidding: Harrison and McKee Franchise Bidding: Harrison and McKee (1985):sellers bid for the right to be the monopolist (1985):sellers bid for the right to be the monopolist

Definitive results as to a pure successful Definitive results as to a pure successful regulatory tool were not discovered, boding well regulatory tool were not discovered, boding well for the use of the DA and its efficiency. for the use of the DA and its efficiency.

Qualitative AnalysisQualitative Analysis

After careful review, tentative (i.e. in full After careful review, tentative (i.e. in full recognition of aforementioned experimental recognition of aforementioned experimental limitations) evidence exists for the following:limitations) evidence exists for the following:

1) Demand Withholding will police the increase 1) Demand Withholding will police the increase in higher service fees.in higher service fees. The banking "trust" will not be able to sustain The banking "trust" will not be able to sustain

charging monopoly level services fees for very long in charging monopoly level services fees for very long in every market. Competitors such as credit unions, every market. Competitors such as credit unions, S&Ls, and finance companies notwithstanding, if S&Ls, and finance companies notwithstanding, if small and large firms alike are able to bid out their small and large firms alike are able to bid out their fees in a more free (DA) environment, the downward fees in a more free (DA) environment, the downward sloping demand curve for bank services will police sloping demand curve for bank services will police activities without the need for any regulations. activities without the need for any regulations.

Qualitative AnalysisQualitative Analysis

2) Predatory pricing, and/or concentrated market 2) Predatory pricing, and/or concentrated market power is transient in nature.power is transient in nature. Assuming that there is an escape route for the capital Assuming that there is an escape route for the capital

flowing into the banking industry, predation by the flowing into the banking industry, predation by the trust will not be fruitful for very long, as groups of trust will not be fruitful for very long, as groups of investors will be willing to absorb the cost of re-entry investors will be willing to absorb the cost of re-entry into the market as soon as the trust attempts to into the market as soon as the trust attempts to exercise the full-force of its market position. exercise the full-force of its market position.

Dominant-firm equilibrium may result: the smaller firm Dominant-firm equilibrium may result: the smaller firm simply takes as given whatever price is set by the simply takes as given whatever price is set by the larger firm, and chooses the quantity that maximizes larger firm, and chooses the quantity that maximizes profit. profit.

Qualitative AnalysisQualitative Analysis

3) Markets remain contested without government intervention 3) Markets remain contested without government intervention other than property rights protection.other than property rights protection. If the concept of contested markets, coupled with demand If the concept of contested markets, coupled with demand

withholding is not enough of an impetus for investors to withholding is not enough of an impetus for investors to incur re-entry costs into the market, then there simply is incur re-entry costs into the market, then there simply is not enough withholding in place to warrant re-entry, and not enough withholding in place to warrant re-entry, and the trust can be inferred to be reasonably satisfying the the trust can be inferred to be reasonably satisfying the demand that exists in the marketplace.demand that exists in the marketplace.

4,5)Alternative forms of regulation are underdeveloped, and 4,5)Alternative forms of regulation are underdeveloped, and credit availability is a non-issue (mainstream literature).credit availability is a non-issue (mainstream literature).

6) “Too Big to Fail” is a consequence of the Federal pacifier. 6) “Too Big to Fail” is a consequence of the Federal pacifier. Banks, much like other private corporations, should have to Banks, much like other private corporations, should have to actually implement risk management strategies regardless of actually implement risk management strategies regardless of their size and market share so as to reasonably ensure their their size and market share so as to reasonably ensure their long-run solvency (value judgment in an economic context).long-run solvency (value judgment in an economic context).

Statistical Inference on State BanksStatistical Inference on State Banks

Examination of branching restrictions to determine Examination of branching restrictions to determine whether banks are only nominally differentwhether banks are only nominally different compare vs. hypothetical of “national banks”.compare vs. hypothetical of “national banks”.

Double-edged sword for proponents of the “State Double-edged sword for proponents of the “State View": If state banks are significantly different, how View": If state banks are significantly different, how exactly have they contributed to homogeneity of exactly have they contributed to homogeneity of standards of living across the U.S.? If state banks standards of living across the U.S.? If state banks are only nominally different, what would have been are only nominally different, what would have been the difference between 50 state banks and one the difference between 50 state banks and one national bank over time?national bank over time?

ANOVA was conducted on information contained in ANOVA was conducted on information contained in Tables CB10 and CB11 (bank loans and bank Tables CB10 and CB11 (bank loans and bank investments) as listed under Bank Statistics on the investments) as listed under Bank Statistics on the FDIC website for historical through 2002. FDIC website for historical through 2002.

Statistical Inference on State BanksStatistical Inference on State Banks

Two separate metrics were used to determine statistically Two separate metrics were used to determine statistically significant differences: 1) a measure of how the means for significant differences: 1) a measure of how the means for each state over time differed, if at all, through ANOVA, and 2) each state over time differed, if at all, through ANOVA, and 2) how much co-movement existed across state lines for different how much co-movement existed across state lines for different categories over time, through the generation of correlation categories over time, through the generation of correlation matrices. Values provided by the FDIC were converted to a matrices. Values provided by the FDIC were converted to a percent of total so as to make the data comparable. percent of total so as to make the data comparable.

Before delving into the ANOVA results, it must be noted that Before delving into the ANOVA results, it must be noted that the F-stat "has been shown to be rather robust against the F-stat "has been shown to be rather robust against heteroscedasticity" (Lindman, 1974, p.33 of Hausman) but not heteroscedasticity" (Lindman, 1974, p.33 of Hausman) but not robust against high-mean variance correlation. robust against high-mean variance correlation. High-mean High-mean variance correlationvariance correlation, coupled with the potential existence of , coupled with the potential existence of outliers in the data, can result in seriously misleading results. outliers in the data, can result in seriously misleading results. To hedge against this, all mean variance correlations for each To hedge against this, all mean variance correlations for each data set are reported; some ANOVA results are inconclusive. data set are reported; some ANOVA results are inconclusive.

Statistical Inference on State BanksMean-Variance Correlation

ANOVA Lns. Real Estate 0.044564Source of Variation SS df MS F P-value F crit

Between Groups 12.10885 49 0.247119 19.40408 7.5E-131 1.361105Within Groups 22.92378 1800 0.012735

Total 35.03263 1849

Mean-Variance CorrelationANOVA Inv. US Agency_Corp 0.432291

Source of Variation SS df MS F P-value F critBetween Groups 2.968652 49 0.060585 1.249054 0.117286 1.361105Within Groups 87.30806 1800 0.048504

Total 90.27672 1849

What did our Statistical Models Add What did our Statistical Models Add to the Experimental Evidence?to the Experimental Evidence?

There exists evidence to support the assertion that in There exists evidence to support the assertion that in general, states across the U.S. have fundamental general, states across the U.S. have fundamental differences in how the apportion out both their loans and differences in how the apportion out both their loans and their investments. In addition, despite a few highly their investments. In addition, despite a few highly correlated categories, the majority of investment and correlated categories, the majority of investment and loan statistics contain states that are deviant, or who loan statistics contain states that are deviant, or who have little to no correlation in their actions over time with have little to no correlation in their actions over time with the actions of other states. Deviants vary over category. the actions of other states. Deviants vary over category. These statements must be tempered with the caveat that These statements must be tempered with the caveat that some of the deviations may be due to inherent some of the deviations may be due to inherent topographical or climatic or infrastructural heterogeneity topographical or climatic or infrastructural heterogeneity across the U.S.across the U.S.

Notice the intermingling of econometric and experimental Notice the intermingling of econometric and experimental methods—corollary roles of the two cannot be stressed methods—corollary roles of the two cannot be stressed enough.enough.

Macroeconomic Policy Implications Macroeconomic Policy Implications and Caveatsand Caveats

Policies in light of tentative evidencePolicies in light of tentative evidence 1) An end to branching restrictions in any shape for any 1) An end to branching restrictions in any shape for any

financial institution.financial institution. This is justified on the basis of the concepts of demand This is justified on the basis of the concepts of demand

withholding, contested markets, the transience of withholding, contested markets, the transience of predatory pricing, the existence of escape route markets predatory pricing, the existence of escape route markets for capital, and the establishment of no quantitatively for capital, and the establishment of no quantitatively conclusive evidence that state banks have contributed in conclusive evidence that state banks have contributed in any way to homogeneity. any way to homogeneity.

2) An end to antitrust language and the regulatory costs 2) An end to antitrust language and the regulatory costs associated with the subjective process of regulation.associated with the subjective process of regulation. All resources that the government and its subordinate All resources that the government and its subordinate

entities are currently wasting on regulation should be re-entities are currently wasting on regulation should be re-apportioned to seeing to the fruition of our final policy apportioned to seeing to the fruition of our final policy recommendation.recommendation.

Macroeconomic Policy Implications Macroeconomic Policy Implications and Caveatsand Caveats

3) Move from a PO banking market to a DA 3) Move from a PO banking market to a DA market, similar to the futures market, as a market, similar to the futures market, as a necessary conditionnecessary condition for policies 1 and 2 above. for policies 1 and 2 above. The process by which this could be handled is a The process by which this could be handled is a

major avenue for further researchmajor avenue for further research, but the central , but the central design would center on the DA and/or electronic design would center on the DA and/or electronic dispatching, as is currently found in the deregulated dispatching, as is currently found in the deregulated electricity Independent System Operator (ISO) electricity Independent System Operator (ISO) markets, with the addition of brokerages bidding in markets, with the addition of brokerages bidding in favor of the consumer. The dispatch would act as the favor of the consumer. The dispatch would act as the invisible hand's guide in maximizing efficiency and invisible hand's guide in maximizing efficiency and minimizing dead-weight loss and the take-it-or leave it minimizing dead-weight loss and the take-it-or leave it structure we deal with when attempting to do structure we deal with when attempting to do business with banks.business with banks.

Macroeconomic Policy Implications Macroeconomic Policy Implications and Caveatsand Caveats

CaveatsCaveats The macroeconomy is not governed by economics, but by politics. The macroeconomy is not governed by economics, but by politics.

As such, we must offer the following potential roadblocks to further As such, we must offer the following potential roadblocks to further investigation of these topics, bearing in mind that there may be investigation of these topics, bearing in mind that there may be more limitations and compound variables that may be overlooked.more limitations and compound variables that may be overlooked.

The Congressional Committee structure.The Congressional Committee structure. State and Federal Lobbyists.State and Federal Lobbyists. Municipal Political Action Committees (PACs).Municipal Political Action Committees (PACs). Intrastate nationalism.Intrastate nationalism. Resistance to massive change from policymakers.Resistance to massive change from policymakers. Growing fiscal irresponsibility on both ends of the political spectrum.Growing fiscal irresponsibility on both ends of the political spectrum. The unwavering growth trend of the size of government over time.The unwavering growth trend of the size of government over time.

Questions?Questions?