Kaal claims by topic: research-methods

236 atomic, individually citable claims from the published work of Wulf A. Kaal tagged research-methods.

A uniform approach to hedge fund valuation is not possible because the variety of hedge fund investments and strategies means some positions, such as non-concentrated positions in liquid securities, are far easier to value than others. 2009
The methodological assumptions of incomplete contract theory improve the analysis of executive compensation arrangements relative to the classical and spot contract models normally used. 2012
This Article reports the first survey study of hedge fund advisers conducted after the SEC's registration effective date, drawing on a population of 1267 private fund advisers who registered before March 30, 2012. 2012
The hedge fund industry's concern with confidentiality and privacy is itself an obstacle to empirical research: it made obtaining a substantial effective sample size for this study difficult, independent of the survey design. 2012
Persistent multi-channel follow-up, by fax, e-mail, and telephone, yielded ninety-four completed surveys, a 7.42% response rate from a population of 1267, which is substantially higher than response rates in prior surveys of this industry. 2012
Mandated disclosure does not automatically produce usable public data: although Form ADV requires advisers to disclose chief compliance officer contact information, the SEC dataset omitted it and contained no e-mail addresses, so researchers could not reach the officers responsible for compliance. 2012
The standard remedies for selection bias are not reliably corrective: simulation studies show that many techniques used to prevent selection bias problems have mixed success rates, can worsen rather than improve estimates, and may skew results under ordinary circumstances. 2012
A treatment and control design is unavailable for studying registration effects, because managers who did not have to register have no exposure to the treatment and therefore cannot generate informative responses about its effects. 2012
Comparison of the responding sample against the full registered population on Form ADV parameters shows the sample is not biased toward any particular subgroup of hedge fund advisers, and gives no indication that respondents differ from nonrespondents. 2012
Prior surveys of hedge fund manager expectations left the central questions unanswered because they were fielded before the registration effective date and used substantially smaller samples, so they measured anticipation rather than experience. 2012
The study's findings are bounded in time: because the data was collected within three months of the registration effective date, the study shows trends and perceptions but does not provide insights on the long-term implications of the registration and disclosure requirements. 2012
Comparative corporate law research is challenging and may include inaccuracies because countries differ in legal history, legal origins, and legal cultures. 2013
Form PF disclosures have not been standardized, and anecdotal evidence indicates that the SEC and the FSOC may be working with contradictory, misleading, inaccurate, and incomplete systemic risk data. 2013
The hedge fund adviser registration requirement under the Dodd-Frank Act creates a discontinuity in hedge fund returns at the registration effective date of March 30, 2012. 2014
Strategic behavior by fund advisers around the assets under management registration threshold produces a strong increase in the measured discontinuity at that threshold. 2014
In simple linear regressions of monthly returns on log AUM across the full sample, the AUM coefficient is statistically significant at the 5 percent level only during March through August 2012. 2014
In the period close to and following the registration effective date, fund size has a positive relationship with fund performance, with positive beta coefficients in March through May and July 2012. 2014
Adding a dummy for AUM above $150 million to the linear regressions leaves no variable significant at the 5 percent level in March 2012, and the dummy is significant only in April 2012. 2014
The linear regression models are jointly valid on the F-statistic but their explanatory power measured by R-squared remains very limited, with R-squared values of roughly 0.003 to 0.014. 2014
Under the sharp regression discontinuity design, the estimated treatment coefficient exceeds one only in March 2012, at 1.104 with a p-value of 0.015, and is close to zero and insignificant in every other month. 2014
The March 2012 discontinuity coefficient is the only estimate with a p-value below 5 percent; all subsequent monthly estimates are statistically insignificant. 2014
The discontinuity in hedge fund earnings at the registration effective date is positive, which is the opposite of what the hedge fund industry expected the Dodd-Frank Act to produce. 2014
The March 2012 discontinuity effect is not persistent and is completely absorbed in the months following the registration effective date for private fund advisers. 2014
Because the compliance probability denominator is very close to one, fuzzy regression discontinuity estimates differ only trivially from the sharp design estimates on the entire sample. 2014
A McCrary density test of the assignment variable independently supports the presence of a discontinuity at the $150 million threshold in March 2012. 2014
The March 2012 discontinuity estimates remain statistically significant and stable at larger bandwidths, while very small bandwidths yield confidence intervals containing zero and would not detect any discontinuity. 2014
Conventional, bias-corrected, and robust regression discontinuity estimators all produce coefficients of similar magnitude, between 1.13 and 1.33, each with a p-value below 5 percent, affirming the March 2012 discontinuity. 2014
Unlike the entire sample, whose discontinuity coefficient is near zero except in March 2012, the strategic subsample shows a discontinuity coefficient that is always above zero across the sample months. 2014
Some large advisers reduce their AUM in the months from May to August 2012, which strongly increases the discontinuity around the registration threshold and separates the two groups more sharply, though the effect vanishes late in the sample period. 2014
Difference-in-differences analysis confirms the regression discontinuity results, showing a positive and highly significant treatment coefficient for funds above the $150 million AUM threshold. 2014
The parallel trends assumption required for the difference-in-differences design is satisfied, because pre-treatment performance data show the same time trends for treatment and control groups. 2014
Despite the great volatility of hedge fund adviser returns over the observation period, the empirical evidence for a discontinuity at the $150 million AUM threshold is robust, but the discontinuity does not persist beyond the registration effective date. 2014
High quality private fund data is scarce because the industry's entrenched interest in confidentiality combined with decades of regulatory exemption from registration and transparency requirements left no reservoir of comparable disclosure to study. 2014
Enlarging the sample does not cure selection bias in non-statistical sampling: a bigger sample neither compensates for the bias of non-statistical techniques nor guarantees that the sample is representative. 2014
Further randomization of the sample was not available as a remedy, because respondents drawn from outside the private fund adviser population would never have been exposed to the new disclosure requirements and so could say nothing about them. 2014
The near identity between respondents who reported completing Sections 2 through 5 of Form PF and respondents who reported quarterly filing shows the answers are internally consistent, which the author treats as evidence that the survey responses carry above average reliability. 2014
Despite contacting the entire population of 3669 SEC-registered private fund advisers by fax and e-mail over more than five months, the study obtained only 52 respondents, a response rate of 0.014 percent. 2014
SEC flexibility helps filers through a specific mechanism: it authorizes advisers to apply their own internal methodologies when interpreting and answering Form PF questions and to state their own assumptions, rather than forcing them onto an unfamiliar measurement basis. 2014
Prior scholarship on the corporate governance effects of non and deferred prosecution agreements rests largely on anecdotal evidence and individual case studies rather than on systematic evidence, which is why its conclusions about those effects are unreliable. 2014
Because the population of executed non and deferred prosecution agreements is now large, their real trends and real governance impact are quantifiable and measurable, so policy makers can be given evidence based guidance rather than conjecture. 2014
Where coded categories such as cooperating, disclosure, and internal review fall well short of 100 percent of the sample, the shortfall may reflect a gap in what the agreements record rather than a real absence of those corporate actions, so the coded frequencies understate actual conduct. 2014
The governance trends reported in this study apply only to corporations that actually executed an agreement, so because a disproportionately large number of corporations self correct or self report instead, the findings may capture only the tip of the iceberg. 2014
The cost of Title IV compliance, and the other independent variables used as proxies for compliance cost, are associated with the size of hedge fund advisers as measured by assets under management. 2014
Linear, robust, and non-linear regression models all show positive and statistically significant coefficients, and compliance costs per unit of AUM do not diminish in the sample, so the hypothesis that smaller advisers pay relatively more is not supported. 2014
Least squares linear regression is non-robust to outliers: in the presence of outliers its predictions can be dragged toward the outliers and the variance of the estimates can be artificially inflated. 2014
Because the distribution in this sample is heteroscedastic, treating each data point equally would allocate inappropriate weight to data points in the distribution, which is why the author used robust and weighted regression specifications. 2014
Weighted least squares regression depends on estimated weights, and contrary to the theory behind the method the exact weights are almost never determinable in real applications such as this study. 2014
Selection bias is a valid concern in this study because of the lower sample sizes available for Models 1 through 5. 2014
The compliance and administrative costs created by Title IV of the Dodd-Frank Act are associated with the size of hedge fund advisers' assets under management. 2014
All regression models show positive and predominantly statistically significant coefficients, with 18 out of 30 coefficients in the entire sample statistically significant. 2014
The results suggest that the private fund industry may be more robust and less affected by financial regulation than other financial services providers. 2014
While all coefficients are positive in the entire sample and the multi strategy subsample, the negative coefficients in the single strategy subsample suggest that the strategy employed by a hedge fund adviser could change the assessment of the effect of compliance cost. 2014
Even in the single strategy subsample only 9 of 30 coefficients are negative, so the strategy based qualification to the main finding is limited. 2014
Matching the identified Form PF defects against the FSOC's specific uses of that data suggests possible inaccuracies in the FSOC's systemic risk assessment process, although the author disclaims scientific or empirical precision for the analysis. 2014
Stock prices respond significantly and predictably in a positive direction to the DOJ press release announcing execution of a non- or deferred prosecution agreement and to the start of the N/DPA term. 2015
The market does not price the three defining N/DPA events in isolation; it treats the announcement, the start of the term, and the end of the term as sequential and conditional events. 2015
There is no systematic price momentum beyond the three core N/DPA event dates, which the authors read as evidence that the market is reasonably efficient with respect to N/DPA information. 2015
Investors react negatively to impending N/DPA driven changes in the period before the agreement is executed, but once the N/DPA is executed they generally treat it as a positive event for the firm. 2015
The combination of a positive market reaction at the start of the N/DPA term and a negative reaction at its end is evidence that the governance changes N/DPAs mandate actually matter to firm value. 2015
This is the first study to examine stock price reactions to non- and deferred prosecution agreements, using all publicly available N/DPAs across several industries from 1993 to 2015 (N=330). 2015
The event study design is appropriate for N/DPAs because the wrongdoing event is identifiable through the execution of a reasonably standardized agreement and because information about the firm's wrongdoing can change the distribution of stock returns. 2015
The date of the DOJ press release announcing execution is the only reliable announcement date for an N/DPA, so it is the defensible event date for measuring market reaction. 2015
Pre-announcement leaks about a pending N/DPA, including leaks by prosecutors, should not move markets significantly because leaked details carry no certainty or finality as to final terms or fine amount. 2015
Unlike legislative mandates, N/DPA governance changes are preceded by no public debate or publicity, so their effect on firm value is not gradually incorporated into prices and is therefore testable by event study. 2015
Around the DOJ announcement, N/DPA firms show a significant positive cumulative abnormal return trend before day minus five, a significant drop at day minus five, and negative CARs relative to matched competitors from day minus four to day minus one. 2015
After the DOJ announcement, competitor firms' cumulative abnormal returns trend negative from day zero to day twenty five while N/DPA firms' CARs continue on a neutral to positive trend. 2015
Immediately before the N/DPA term becomes effective, from day minus three to day minus one, the market prices the onset of the term as a negative event. 2015
From the first day of the N/DPA term through day twenty five, the period in which mandated governance improvements are in force, the market prices the N/DPA as a positive event for the firm. 2015
The sign of the market reaction reverses at the end of the N/DPA term: where the announcement and the start of the term produce positive CARs, the end of the term produces negative CARs at day zero and from day one through day fifteen. 2015
The data support Hypothesis 2: the market reacts positively at the DOJ announcement and at the start of the N/DPA term and negatively at the end of the term. 2015
The data support Hypothesis 3: the market reacts positively when N/DPA governance changes become mandatory at the start of the term and negatively when they cease to be mandatory at its end. 2015
The study's results may be biased toward low impact N/DPAs, because firms that were acquired, merged, or went bankrupt as a result of an N/DPA had to be dropped for lack of public trading data, and those are precisely the highest impact cases. 2015
The confluence trends identified are correlational, not causal: the author expressly disclaims any claim of cause and effect and presents the peripheral effects as long term possibilities warranting monitoring. 2016
Using self-reported Morningstar earnings data for 3,424 US private fund advisers covering 2010 to 2015 in multiple regression discontinuity designs with robustness checks, private fund adviser registration and disclosure under the Dodd-Frank Act had no significant effect on private fund adviser returns. 2016
A third channel is measurement rather than economics: if regulation curbs misreporting, private fund advisers' performance would appear to decrease simply because managers are less able to inflate their monthly returns. 2016
The study identifies the causal effect of Title IV by treating the March 30, 2012 registration effective date combined with the $150 million AUM threshold as an exogenous regulatory shock, so that any discontinuity in returns at the cutoff is evidence of a causal treatment effect. 2016
A fuzzy regression discontinuity design was run to test whether the discontinuity occurred at a date other than March 30, 2012, on the theory that advisers may have anticipated compliance costs in the preceding months. 2016
The sharp and fuzzy regression discontinuity approaches yield results of only minor difference in this setting because the denominator of the fuzzy estimator is very close to one, meaning treatment take up at the threshold is nearly deterministic. 2016
Across an array of robustness tests, the requirements introduced by the Dodd-Frank Act create no significant effect on private fund performance, with all reported RD p-values above the 5% level. 2016
In simple linear regressions of monthly returns on log AUM across December 2011 to December 2012, fund size does not appear to matter for fund returns because only a few coefficients are statistically significant and those remain close to zero. 2016
Fund size shows a negative relationship with performance in the months before the March 2012 registration effective date and a positive relationship afterward, with beta coefficients negative in January to March 2012 and July 2012 and positive in April and May 2012. 2016
In regressions including a dummy for AUM above $150 million, the dummy is statistically significant and positive only in April and September 2012, and the explanatory power of the models measured by R-squared remains very limited. 2016
A single point in time RD design anchored to March 30, 2012 is inadequate on its own because advisers could and did register before the deadline, funds near the $150 million threshold could choose between registered adviser and exempt reporting adviser status, and self-reported Morningstar AUM is not calculated the same way as the SEC's RAUM. 2016
Tighter RD designs using varying window lengths and hand selected control groups confirm the finding of no effect obtained in the broader design. 2016
The discontinuity in the data is evident regardless of the number of bins chosen, even though increasing the number of bins smooths the estimated regression function. 2016
Relative to existing work, this study uses a much larger dataset and a more sophisticated empirical approach, regression discontinuity, and finds no statistical evidence for an effect of Dodd-Frank Act requirements on private fund advisers' performance. 2016
Any conclusion that hedge funds contributed to the financial crisis of 2007-2008 is circumstantial or anecdotal, because the data needed to test it, on leverage, counterparty relations, AUM, and portfolio holdings, were not collected for any substantial period before the crisis. 2016
The opacity of the hedge fund shadow banking system blocks direct measurement of hedge funds' role in the crisis, leaving researchers with indirect measures extracted from existing data rather than primary pre-crisis sources. 2016
Traditional risk-adjusted alphas underestimate hedge fund risk: once correlation risk is controlled for, previously observed significant hedge fund alphas disappear, which makes correlation risk a systematic risk factor for hedge fund returns. 2016
Risk measures that are not adjusted for serial correlation in hedge fund returns can considerably underestimate the true extent of both individual and systemic hedge fund risk, so empirical work in this area must account for autocorrelation. 2016
Systemic risk rankings that place a loosely defined other financial services sector above banking and insurance are of limited use, because the analysis does not clearly identify the firms included in that category even though a substantial portion of them may be hedge funds. 2016
The most useful product of the post-crisis empirical literature for regulators is a set of methodologies for evaluating hedge fund systemic risk and prescribing remedies, especially methodologies addressing the counterparty credit measures of hedge funds and their prime brokers. 2016
Venture capital deal flow, meaning the totality of potential deals and business plans screened by venture capitalists, would provide the optimal assessment of innovation trends, but that data is not available, so realized investment allocations must be used as a second-best proxy. 2016
The study's evidence base is a PitchBook dataset of 77,508 United States venture capital deals involving 37,298 companies from 2005 through 2015, covering all venture capital deals and all venture capital stages. 2016
Regulation is mostly reactive and follows business cycles rather than being proactive; data on venture capital investments lets regulators see where innovation trends are heading and what risks they entail before the disruptive innovation actually materializes. 2016
The study rests on two datasets: SEC Form ADV Part II filings by private investment fund advisers from 2007 to 2014 (N=100392) and the publicly available litigation record on private fund investor due diligence from 1995 to 2015 (N=572). 2016
The Form ADV analysis is limited because the term due diligence carries multiple possible meanings, so counts of the term cannot by themselves distinguish among those meanings. 2016
Litigation research on private fund due diligence requires in depth evaluation of case dockets rather than only published judicial decisions, because opinions are snapshots that do not tell the whole story of a case. 2016
Direct regulation of hedge fund leverage collapses on the details because balance sheet leverage is not an adequate measure of risk and would push funds into off-balance sheet avoidance strategies. 2016
Alternative risk measures such as value at risk have severe measurement problems, so any direct regulation of leverage would be set conservatively and would substantially limit hedge funds' ability to provide market liquidity. 2016
Survey research on private fund advisers is structurally constrained because these advisers traditionally oppose publicity and hold a strong preference for confidentiality and privacy, which makes a substantial effective sample size difficult to obtain. 2016
Neither the 2012 nor the 2015 sample is biased, and the comparison across the two populations is consistent because respondents in both surveys were equally subject to Title IV compliance obligations. 2016
The decline in survey response rate between 2012 and 2015 is itself evidence of the private fund industry's relatively rapid adaptation to the new statutory and regulatory regime. 2016
The survey achieved a response rate of 5.44 percent from a population of 1267 registered private fund advisers. 2016
Because private fund advisers prefer confidentiality and generally oppose publicity, most do not respond to survey questions, which makes obtaining a substantial effective sample size for survey studies of this industry difficult. 2016
Guaranteeing complete anonymity is essential to obtaining a sufficient response rate from private fund advisers, but that guarantee prevents a broader descriptive statistical analysis of the sample. 2016
The authors operationalize unconstrained status by binary coding eleven prospectus characteristics and treating a score of nine or better as unconstrained, producing a final study sample of 84 funds out of 114 funds identified in the Morningstar Nontraditional Bond index. 2016
Industry specific venture capital investment data allows regulators to anticipate regulatory needs in the industries carrying the highest levels of disruptive innovation, and the level of disruptive innovation can be quantified by the venture capital dollars flowing into those industries. 2016
The study rests on a hand coded dataset of 120 private investment funds that use blockchain technology in either their strategy or their operations, compiled by the author and research assistants from web searches and multiple databases. 2017
Raw country level counts overstate American leadership: normalizing by population shows roughly 7.5 million people per blockchain fund in the United States against roughly 2.8 million in the United Kingdom, so the distribution across countries is more even than the map suggests. 2017
The absence of Luxembourg funds from the dataset is most likely an artifact of unavailable public information rather than evidence that Luxembourg funds do not invest in blockchain. 2017
The dataset is incomplete by construction: the 120 fund advisers sampled did not answer all questions and the authors were often unable to obtain information on all questions. 2017
In a DAO, disenfranchised community members cannot be judged by race or cultural biases because performance in an anonymized proposal voting scheme is the only basis for assessment and payment. 2017
Because advisers and third party service providers can flatten out and sanitize the information disclosed in Forms ADV and PF, the resulting disclosures may be less useful to the FSOC and the SEC in determining the systemic risk posed by private funds. 2017
Using a regression discontinuity design around the 150 million dollar registration threshold with five years of performance data on more than 3500 reporting private funds, the study finds no significant effect of Dodd-Frank requirements on private fund performance, with all p-values above the 5 percent level. 2017
Increasing the quantity of data does not dissolve the foundational methodological problems of data: construct validity, measurement, reliability, and data dependencies remain the same regardless of how much data is collected. 2017
Because tokens move in and out of the top 100 daily, the study limits its time series dataset to all available data on the top 100 cryptocurrencies until April 2018, selecting coins by market capitalization before April 2018. 2018
For token issuers in the dataset that did not conduct an ICO, the author used the date of first publicly listed trade as a proxy variable for launch date. 2018
Coding categories frequently allowed a token to fall into more than one category, and where that occurred each category was given equal weight, coded as 0.5 and 0.5 for the corresponding dummy variables. 2018
The top 25 ICO jurisdictions in this study are identified from ICO WatchList data, which ranks countries by the number of ICOs launched and reports how much was raised through those ICO projects. 2018
The comparative summaries of national regulatory responses are necessarily incomplete and require much more analysis before they can be determinative for jurisdictional choice. 2018
The United States was excluded from the comparative jurisdictional analysis because its regulatory setup for cryptocurrencies and DLT business was still too uncertain at the time of publication. 2018
Successful proof of stake experiments running today cannot be used to infer that their protocols are truly secure, because the current participant population is atypically altruistic; confidence must instead come from sound reasoning about incentives. 2018
Legacy insurers will not underwrite smart contracts that sit outside the traditional legal framework, and even if they eventually enter the DApp market their actuarial risk assessment methods may be only partially compatible with rapidly evolving decentralized products. 2019
Big data is often not the output of instruments designed to generate valid and reliable data suitable for scientific analysis, so foundational data quality problems persist regardless of how much data is collected. 2019
Even if legacy insurers do enter the DApp market, their products may not fit the need, because actuarial methods for risk assessment under traditional insurance metrics are only partially compatible with rapidly evolving decentralized products. 2019
Any risk assessment of hedge funds as counterparties is necessarily incomplete, because there is no common measure for calculating leverage and exposure and because fund trading strategies are dynamic. 2019
The degree of decentralization of an open source project is determined by the degree of hierarchy in its governance: a flat structure with consensus decision-making and no power disparities between developers is the most decentralized form. 2021
Concentration produces instability and dispersion produces stability: the more concentrated a structure is, the more unstable it is, because little changes can cause major structural change, while a spread out pattern remains spread out when disturbed. 2021
A hierarchy obstructs solving novel problems because it prevents the identification of talented people at the lower rungs and resists the changes needed to construct a new hierarchy for organizing the effort. 2021
Decentralization makes truth discovery more accurate, more reliable, and more efficient, because averaging results across diverse contributors, like ten news companies reporting a story or ten researchers using ten instruments, is more accurate than repetition by a single source. 2021
The strength of a decentralized organization is measured by summing the power of each member in their individual autonomy, modified by the group's ability to organize and effect its goals in the larger society. 2021
Because digital asset valuation methodologies vary significantly, the tradeoffs among them leave digital asset managers with meaningful valuation discretion. 2022
As of 2022 there is no agreed upon reliable valuation method for cryptocurrencies. 2022
Stocks and cryptocurrencies look similar enough, both traded on markets at fluctuating prices, to invite similar regulation, but they diverge in their potential for abuse, their nature, their acceptance, and their use. 2022
The adjusted net asset approach revalues balance sheet items toward current fair value but still fails to capture intangible assets, because intangibles are not represented on the balance sheet. 2022
The paper's empirical basis is a dataset of DAOs selected by the assets held in their treasuries, drawn from across different industries. 2023
Each DAO in the dataset was given a score between zero and ten by the analyzing teams on each of six factors: Decentralization, Work to Earn, Attack Resistance, Regulatory Compliance, Governance, and Organizational Communication. 2023
Federated learning lacks theoretical guarantees of reliability and robustness, which makes its behavior unpredictable in practical applications. 2024
Reputation based governance allocates decision power by past contribution and community standing, which promotes transparency and trust, but reputation is difficult to measure objectively. 2024
The study evaluates each DAO on six factors: Decentralization, Work to Earn, Attack Resistance, Regulatory Compliance, Governance, and Organizational Communication. 2024
Each DAO in the dataset was scored from zero to ten on each factor by analyzing teams, using only publicly available information and the organization's whitepaper where one existed. 2024
A decentralization score of 10 is stipulated to mean a fully decentralized organization with anonymous participation, minimal barriers to entry, and well distributed power; lower scores indicate concentration of power. 2024
The author concedes that the scoring metric is imperfect and that some scored attributes may have changed by the time of publication, presenting it instead as a structured approach to evaluating what drives DAO success or failure. 2024
Deep learning models adapt to changes in data distribution far less readily than human learning does, which limits their reliability once the operating environment diverges from the training data. 2024
GNNs are vulnerable to adversarial attacks that target both node features and graph structure, and their lack of interpretability remains a major obstacle to applying them to real world problems. 2024
Impact measurement in Impact 2.0 has no settled standard: consensus on measurement methodologies and metrics remains elusive, which produces divergent approaches and hinders comparability across programs and sectors. 2024
The impact measurement consulting business follows a distinctly centralized approach, and without the crowd wisdom and community audit that WEB3 Impact 3.0 supplies, such consulting practices are subject to single points of failure. 2024
In a quantum economy, core economic variables such as supply, demand, price, and utility are not fixed quantities but quantum states that can occupy several configurations at once until an observation or measurement resolves them into one value. 2024
The classical law of supply and demand fails on three specific grounds: supply and demand curves cannot be measured independently, economic interactions are intrinsically probabilistic, and goods and financial transactions are discrete rather than continuous. 2024
Existing quantum economic models carry unresolved defects that the field must address: some lack a realistic connection with financial markets, and others strip out the features that make the formalism quantum in the first place. 2024
Headline net job creation figures conceal the real disruption: employers anticipate structural labor market churn of 23 percent of jobs over five years, so a positive net balance of created over destroyed jobs understates how many workers must move. 2024
Many existing studies of automation and job loss rest on flawed assumptions and weak data, and their seemingly precise figures conceal those defects, so their headline numbers should not be taken at face value. 2024
The classical law of supply and demand fails because supply and demand curves cannot be measured independently and because economic interactions are intrinsically probabilistic rather than continuous and deterministic. 2024
A central methodological obstacle for quantum economics is that complex social phenomena such as social power and mental energy resist reduction to exact equations, which leaves the framework without a consistent set of units for subjective forces. 2024
Quantum economics models must be subjected to empirical testing and to rigorous head to head comparison with classical models across different economic contexts and datasets before their validity and generalizability can be established. 2024
Tokens solve the quantification problem in quantum economics by serving as measurable units of value and governance: token denominated voting power in a DAO makes social influence and decision making power countable, supplying the consistent set of units the framework lacked. 2024
When a training dataset disproportionately represents one region or demographic group, the resulting model produces skewed and sometimes inappropriate outputs once deployed in unfamiliar settings. 2025
Prediction on a test set of existing judgments is not the same task as predicting outcomes for a party mid-litigation, because the precise formulation of facts used by such models emerges only once the judgment has been issued. 2025
Institutional alignment through engineered consequence is architecturally superior to exogenous constraint, and the superiority derives from three structural features: scalability, robustness to gaming, and capability complementarity. 2026
Institutional alignment resists gaming because the mechanisms that produce alignment are identical to the mechanisms that produce economic success: an agent cannot game its way to high reputation without actually performing competently and honestly. 2026
Citation weights should be treated as approximate signals of attribution rather than precise measurements, and the governance system must be designed to function robustly under that inherent imprecision. 2026
The citation honesty equilibrium is robust because it does not require agents to quantify precise citation weights; it requires only that citation patterns be directionally honest, which is what the validator assessment evaluates, thereby addressing the quantification impossibility. 2026
Possibility-space quality is captured by four functionals, coverage, precision, robustness, and calibration, and value in Computative Economics derives from their composite rather than from any single dimension. 2026
Existing national accounts are built to measure terminal human consumption and therefore fail to capture intermediate value created in agent-to-agent transaction chains, so the measurement apparatus requires reconstruction. 2026
Because AI agents can generate unlimited Sybil identities at near zero cost, defense must come from multi agent validation with quality based slashing, which imposes economic penalties scaling with the sophistication needed to produce competitive quality output. 2026
Economic institutions that rely on lagging indicators such as price signals, employment data, and GDP reports cannot detect AI driven transformation, because the transformation propagates faster than the monitoring systems built to observe it. 2026
Recursive agent to agent productivity gains are invisible to national accounts because they involve no monetary transactions, no employment, and no market exchange as traditionally defined, even though the final output may register in GDP statistics. 2026
The architecture defines a staged progression (wild, substrate, evolution) through which a cohort passes, with each stage activating a specified set of the architecture's surfaces and each transition marked by an administrative boundary whose state guarantees are verifiable. 2026
Because stages hold the population, task distribution, and harness fixed while varying only the institutional layer, differences between stages isolate the institution's contribution. 2026
There is no validation pool, no REP, no staking, no slashing, no citation graph, and no deliberation: each task is a one-shot call, and each agent's output leaves no institutional residue. 2026
Work assignment remains exogenous, so the stage varies the institution rather than the labor market. 2026
The reporting layer preserves pool-level evidence, run-level controls, and periodic institutional summaries. 2026
A failed control stops advancement rather than merely annotating the run. 2026
Settlement, record, and report derive from a common content-addressed evidentiary base, so the substrate's account of itself can be checked against institutional state by authorized reviewers. 2026
The evaluation of multi-agent systems built from large language models has, to date, been an evaluation of capability under incentive-free conditions. 2026
The result reported here arrives as a two-act empirical cycle, and the cycle is as much the contribution as the numbers. 2026
In the first act, a discovery campaign tested a registered primary hypothesis, that structured deliberation improves the net discrimination of validation pools, and found it null: Youden's J moved +0.0250 with a confidence interval straddling zero. 2026
The program treated those exploratory effects as hypotheses rather than findings. 2026
Deliberation reduced over-approval by 0.1610 (95% CI [-0.2091, -0.1128]) and reduced unanimity by 0.2542 (95% CI [-0.2984, -0.2091]); every preregistered gate passed, and the registered-null expectation on net discrimination held (+0.0606, CI crossing zero). 2026
The separate incentive-layer treatment remains a registered forward program, and no result for it is asserted here. 2026
The design therefore holds the labor force and work assignment fixed while varying the institutional condition. 2026
Nine metrics, labeled A through I, carry the comparison: resolution accuracy, work-product quality, time to consensus, agent retention, independence rate, latency efficiency, reporting accuracy, participation depth, and calibration. 2026
First, the hypotheses, the metric definitions, the correction procedure, and the analysis plan were fixed and preserved before the relevant condition ran. 2026
First, it introduces an anchored discovery-to-confirmation cycle: a null registered primary reported without qualification, exploratory signals promoted only through preregistration, and a confirmation rule applied mechanically. 2026
Second, it reports a preregistered, replicated estimate of what structured deliberation does inside a reputation-bearing validation institution under controlled cohort conditions. 2026
The present program's contribution is therefore not another benchmark but a treatment: a reputation-and-incentive layer imposed on a conventional, heterogeneous, ground-truthed task population, with the layer's presence or absence as the experimental variable. 2026
Each operationalizes a form of agency cost with a definitional foundation that predates LLMs by decades, and the Article insists on the external anchoring because a metric battery whose definitions depend on the paper’s own claims cannot falsify those claims. 2026
The confirmatory analysis fixed the directional hypotheses, bootstrap estimator, exclusion rules, randomization audit, data-quality gate, and confirmation rule before confirmatory computation began. 2026
A public timestamp proves existence no later than the attesting record; it is not, standing alone, proof of every later operational event. 2026
In the treatment arm, validators exchange structured assessments before a binding adjudication; in the matched control arm, the same work is adjudicated without that exchange. 2026
All other treatment-relevant conditions are held fixed. 2026
Neither hypothesis claims deliberation makes the pool smarter in net. 2026
The completed E1 and E2a evidence concerns a historical research apparatus. 2026
The current reference implementation extends the research lineage but creates no new empirical result for this Article. 2026
The registered forward program is a prospective research design, and any proposed production system is a separate object requiring its own evidence. 2026
Holding assignment fixed isolates the institutional effect on work outcomes from any effect on work selection. 2026
The result was null: +0.0250, 95% CI [-0.0502, +0.0985]. 2026
Both primary hypotheses are confirmed under the preregistered rule: the pooled intervals exclude zero in the hypothesized direction, and each contrast is negative across every independent campaign unit. 2026
Net discrimination, the discovery run’s failed primary, remained null in confirmation: +0.0606, 95% CI [-0.0086, +0.1303], crossing zero exactly as the pre-registration predicted it would. 2026