Pith. sign in

REVIEW 4 major objections 6 minor 67 references

Dynamic Interest Rate Discovery in Decentralized Finance: A Reverse Kelly Automated Market Maker for Risk-Adjusted Lending

T0 review · 4 major / 6 minor · reviewed 2026-07-11 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Inverting the Kelly Criterion yields a closed-form interest rate that keeps DeFi lending pools solvent by pricing each loan’s default risk.

desk verdict Clean expected-return pricing packaged as an EVM AMM with solid convexity math and unusually complete open-source artefacts; the Kelly label is wrong and the solvency MC never tests recovery or PD error. read the letter →

arxiv 2607.04178 v1 pith:Q5CTULPG submitted 2026-07-05 cs.SI

Sai Srikanth Madugula , Peplluis Esteva de la Rosa , Daya Shankar This is my paper · ORCID
classification cs.SI
keywords AutomatedMarketMakersKellyCriterionDecentralizedFinanceDynamicPricingCreditRiskReal-WorldAssetsSmartContracts
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

DeFi lending today sets rates from pool utilization, not borrower risk, so protocols demand heavy over-collateralization and shut out real-world assets such as invoices. This paper inverts the classic Kelly Criterion to obtain a single formula that converts a borrower’s probability of default into the exact interest rate needed to deliver a chosen expected yield to liquidity providers. Because the rate rises convexly with risk and tends to infinity as default probability approaches one, toxic loans are priced out of the market without manual intervention. Monte Carlo stress tests under elevated default regimes show the mechanism holding net yields near the target while conventional utilization curves turn insolvent. The result is a mathematically grounded pricing engine intended to open permissionless credit to under-collateralized real-world assets.

What carries the argument

The Reverse Kelly pricing function r = (y + PD)/(1 − PD): the interest rate that equates expected loan return to a target LP yield y given probability of default PD (zero recovery).

What would settle it

Re-run the Monte Carlo suite with positive recovery rates or with deliberately lagged/biased PD inputs and check whether net LP yield still tracks the 12 % target under the same 15 % average-PD shock.

Watch

Extended reading notes

Core claim

Setting a loan’s expected return equal to a target yield y under zero recovery produces the closed-form Reverse Kelly rate r = (y + PD)/(1 − PD). The function is strictly convex in PD, automatically rations high-risk credit, and, in 10 000-epoch Monte Carlo simulations, keeps simulated liquidity pools solvent at roughly 12 % net yield when average PD jumps to 15 %, while static utilization models generate large negative yields.

Load-bearing premise

Defaults produce total loss of principal and the off-chain AI oracle always supplies accurate, unmanipulated default probabilities in real time.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper proposes the Reverse Kelly Automated Market Maker (rkAMM) for DeFi lending of under-collateralized real-world assets. It derives the closed-form rate r = (y + PD)/(1 - PD) by rearranging an expected-return identity under zero recovery, proves first- and second-derivative convexity in PD, implements the formula in Solidity with WAD fixed-point arithmetic, and supplies an open-science stack (MLflow, DVC/DagsHub, local FinBERT/Llama-3 inference). Monte Carlo simulations over 10 000 epochs claim that the mechanism keeps LP net yields near a 12 % target under elevated PD shocks while a static utilization curve (Aave-style) becomes insolvent (Table 1).

Significance. If the pricing rule and its solvency properties hold under realistic credit-loss and oracle conditions, the work supplies a clean, implementable alternative to utilization curves for risk-adjusted RWA lending and could help open the multi-trillion-dollar supply-chain finance gap to permissionless rails. Strengths that should be credited include the elementary but correct convexity proofs (Eqs. 8–9), the gas-conscious WAD Solidity fragment (Appendix B), the fully open simulation and data-versioning pipeline, and the explicit stress-test comparison in Table 1. These engineering and reproducibility assets are valuable even if the economic framing requires tightening.

major comments (4)
  1. §3.1–3.2 and Eq. (7): the derivation maximises arithmetic expected return E[R] = y under LGD = 1; it is not an inversion of the Kelly criterion, which maximises E[log W]. Appendix A’s growth-rate expression even contains ln(0) and is left undefined. The repeated “Reverse Kelly” branding and the claim of asymptotic growth optimality are therefore unsupported by the mathematics actually shown. Either re-derive under a logarithmic objective or re-label the construction as a break-even expected-return pricer.
  2. §6, Table 1 and Appendix C: the Monte Carlo prices every loan with the true PD drawn from the Beta and applies full principal loss on default. It therefore never injects positive recovery, PD estimation error, or oracle latency—the two load-bearing assumptions flagged in §3.2 and §5.2. Consequently the “solvent vs insolvent” contrast does not establish robustness under the paper’s own weakest premises. At minimum, re-run the stress suite with LGD < 1 and with noisy/stale PD feeds and report the resulting yield distributions.
  3. §3.3 and the abstract’s claim of “strictly convex superiority over Aave and Compound’s static utilization curves”: utilization curves price liquidity scarcity, not borrower PD. The comparison in Table 1 therefore pits two different risk models against each other without a common risk measure or capital constraint. Either re-frame the experiment as “risk-based vs utilization-based under identical PD shocks” or supply a matched capital-efficiency metric that isolates the contribution of convexity.
  4. §5.2 and §7: the protocol is asserted to be immune to flash-loan oracle attacks because PD is off-chain, yet the same section and the limitations acknowledge liveness risk and asynchronous updates. No quantitative bound is given on how stale a PD may become before the solvency guarantee of Eq. (7) fails. A simple sensitivity analysis (maximum tolerable PD lag under the Monte Carlo shocks) is needed to make the security claim falsifiable.
minor comments (6)
  1. Abstract and §1: the $2 trillion supply-chain gap is cited from World Bank (2020); a more recent figure or explicit caveat that the paper does not itself measure market size would avoid over-claim.
  2. Figure 2 and Figure 5 captions are descriptive but the figures themselves are not reproduced in the text provided; ensure axis labels and parameter values are legible in the camera-ready version.
  3. Eq. (12) in Appendix A writes PD ln(0); either replace with a recovery floor or delete the appendix if the log-growth argument is abandoned.
  4. Heavy self-citation to Madugula et al. (2025, 2026a, 2026b) is appropriate for the multi-paper programme, but the present manuscript should stand alone; a one-paragraph summary of the PD-oracle assumptions inherited from those works would help readers.
  5. Solidity snippet (Appendix B) hard-codes targetYieldWAD = 0.12e18; document that this is a mutable governance parameter and show the setter with access control.
  6. Typographical: “1e18 (WAD)” is used consistently, but “1018” appears once without the “e” (p. 8); standardise.

Circularity Check

1 steps flagged · score 2.0 of 10

Minor self-citation for Reverse Kelly framing; the rate formula itself is a self-contained algebraic solution for target expected yield and is not circular.

  1. self citation load bearing [Section 2.2 (also Intro/1.1 citing Madugula et al. 2026a)]
    "Esteva et al. (2023) proposed that in decentralized invoice discounting, the allocation is often fixed (the invoice amount). Therefore, the formula must be inverted to determine the required payoff (interest rate) given the known probability of default. Our work mathematically formalizes this inversion for permissionless EVM environments."

    The motivation and naming of the 'Reverse Kelly' inversion as the core pricing mechanism is justified by citation to Esteva et al. (2023), co-authored by present author Peplluis Esteva de la Rosa. The paper further states it formalizes models from Madugula et al. (2026a) by the same team. While the algebra r=(y+PD)/(1-PD) from setting E[R]=y is independently derived and not circular, the claim of a novel optimal Reverse Kelly AMM rests narratively on this self-citation chain without external independent derivation of the inversion concept.

full rationale

The derivation chain for the core result r = (y + PD)/(1 - PD) begins from the definition of expected return under zero recovery (Eq. 2), sets it equal to target y (Eq. 3), and solves algebraically (Eqs. 4-7). First and second derivatives establish strict convexity independently (Eqs. 8-9). No parameters are fitted to data and then re-used as predictions; the Monte Carlo simply applies the closed-form rate to known simulated PDs and recovers yields near the target by the law of large numbers under the paper's assumptions. Self-citations to Madugula et al. (2025, 2026a, 2026b) and Esteva et al. (2023) supply the broader multi-agent/XAI/oracle narrative and the original suggestion to invert Kelly, but they are not required for the algebraic steps or convexity proofs. The central mathematical claim therefore stands independently of the self-citations. Score 2 reflects only the mild load-bearing role of the co-author citation in motivating and naming the 'Reverse Kelly' construction.

Assumptions & free parameters 3 free parameters · 4 assumptions · 1 invented entities

The central solvency claim rests on a short list of free parameters (target yield, PD distribution shape, hard risk cap) and three modelling axioms (zero recovery, unbiased oracle PD, independent Bernoulli defaults). No new physical entities are invented; the “rkAMM” is simply a named implementation of the pricing identity.

free parameters (3)
  • targetYield y = 0.12
    Hand-chosen hyper-parameter (default 0.12) that directly sets the level of every quoted rate; swept in MLflow but never derived from first principles.
  • Beta(α,β) parameters for PD draws = α=2,β=38 (normal); α=3,β=17 (shock)
    Chosen to produce mean PD ≈ 5 % (normal) or 15 % (shock); control the entire Monte-Carlo outcome distribution.
  • hard PD rejection threshold = 0.30
    Loans with PD > 0.30 are simply refused in the simulation code; this ad-hoc filter, not the closed-form formula, produces the reported 71.9 % approval rate under adverse selection.
assumptions (4)
  • domain assumption Loss given default equals 1 (zero recovery) for every defaulted loan.
    Stated explicitly in §3.2 as a “conservative lower-bound assumption common in unsecured invoice financing”; removes the recovery term from the pricing equation.
  • domain assumption The off-chain XAI oracle supplies an unbiased, real-time PD that cannot be manipulated by on-chain actors.
    Required for the rate to equal true expected loss; defended in §5.2 by divorcing PD from DEX balances, but never proved.
  • domain assumption Defaults are independent Bernoulli draws conditional on the supplied PD.
    Used throughout the Monte-Carlo engine (Appendix C); ignores default correlation that would appear in a real macroeconomic shock.
  • ad hoc to paper Expected arithmetic return, not logarithmic growth, is the correct objective for LP utility.
    The derivation maximises E[R] = y rather than the Kelly growth rate; the “Reverse Kelly” label is therefore a naming choice rather than a derived necessity.
invented entities (1)
  • Reverse Kelly Automated Market Maker (rkAMM)
    purpose: Named packaging of the expected-value pricing formula as a Solidity bonding curve that ingests oracle PD.
    No independent theoretical object beyond the classical break-even rate; the name and the WAD implementation are the paper’s contribution.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Dynamic Interest Rate Discovery in Decentralized Finance: A Reverse Kelly Automated Market Maker for Risk-Adjusted Lending." pith.science (2026). https://pith.science/paper/Q5CTULPG

@misc{pith2026260704178,
  author       = {Pith},
  title        = {Pith review of: Dynamic Interest Rate Discovery in Decentralized Finance: A Reverse Kelly Automated Market Maker for Risk-Adjusted Lending},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q5CTULPG}},
  note         = {Machine review of arXiv:2607.04178}
}
abstract

Decentralized Finance (DeFi) lending protocols currently rely on heuristic, utilization-based bonding curves that mandate severe over-collateralization, systematically excluding under-collateralized assets like corporate invoices. This paper introduces a mathematically optimal pricing mechanism for decentralized credit: the Reverse Kelly Automated Market Maker (rkAMM), the core engine of our proposed lending framework. By inverting the Kelly Criterion, traditionally used for optimal bet sizing, we construct a dynamic interest rate discovery protocol that explicitly prices individual loan risk. The rkAMM ingests real-time Probability of Default (PD) streams from an off-chain Explainable AI oracle and dynamically calculates the exact interest rate required to sustain target liquidity provider (LP) yields. We mathematically derive the Reverse Kelly pricing function ($r = \frac{y + PD}{1 - PD}$), proving its strictly convex superiority over Aave and Compound's static utilization curves in managing capital efficiency. Furthermore, we deploy the rkAMM architecture via Solidity smart contracts, optimizing for gas-efficient 1e18 (WAD) floating-point arithmetic. To ensure decentralized transparency, our simulation infrastructure leverages MLflow for tracking yield hyperparameters, Data Version Control (DVC) linked to DagsHub for versioning Real-World Asset (RWA) data arrays, and localized edge-inference via Ollama (Llama-3) and Hugging Face (FinBERT) for zero-cost predictive modeling. Monte Carlo simulations across 10,000 macroeconomic stress scenarios confirm that the rkAMM maintains protocol solvency and stabilizes LP yields at 12-15\% net of expected credit losses. This work provides the foundational financial engineering required to bridge the \$2 trillion global supply chain finance gap using permissionless blockchain infrastructure.

Figures

Figures reproduced from arXiv: 2607.04178 by the authors.

Figure 1
Figure 1. Open-Science Decentralized Tech Stack Orchestration. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Logic flow for the Reverse Kelly Convex Bonding Curve. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Edge Inference Pipeline for Zero-Cost PD Oracles. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: EVM Sequence Diagram utilizing WAD Mathematical Operations. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Monte Carlo state transition logic mapped to Beta distribution macroeconomic shocks. [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

67 extracted references · 3 canonical work pages

  1. [1]

    Aave. (2023). Aave V3 Risk Framework. Aave Documentation

    2023

  2. [2]

    Adams, H., Zinsmeister, N., & Robinson, D. (2020). Uniswap v2 Core. Uniswap Whitepaper

    2020

  3. [3]

    Akerlof, G. A. (1970). The Market for "Lemons": Quality Uncertainty and the Market Mechanism. The Quarterly Journal of Economics, 84(3), 488-500

    1970

  4. [4]

    Altman, E. I. (1989). Measuring Corporate Bond Mortality and Performance. The Journal of Finance, 44(4), 909-922

    1989

  5. [5]

    Angeris, G., & Chitra, T. (2020). Improved Price Oracles: Constant Function Market Makers. Proceedings of the 2nd ACM Conference on Advances in Financial Technologies, 80-91

    2020

  6. [6]

    T., Chiang, R., Noyes, C., & Chitra, T

    Angeris, G., Kao, H. T., Chiang, R., Noyes, C., & Chitra, T. (2019). An analysis of Uniswap markets. arXiv preprint arXiv:1911.03380

    arXiv 2019

  7. [7]

    M., & Wood, G

    Antonopoulos, A. M., & Wood, G. (2018). Mastering Ethereum: Building Smart Contracts and DApps. O'Reilly Media

    2018

  8. [8]

    Araci, D. (2019). FinBERT: Financial Sentiment Analysis with Pre-trained Language Models. arXiv preprint arXiv:1908.10063

    arXiv 2019

Show all 67 references
  1. [9]

    Atzei, N., Bartoletti, M., & Cimoli, T. (2017). A Survey of Attacks on Ethereum Smart Contracts (SoK). International Conference on Principles of Security and Trust, 164-186

    2017

  2. [10]

    Bartoletti, M., Carta, S., Cimoli, T., & Saia, R. (2020). Dissecting Ponzi schemes on Ethereum: Identification, analysis, and impact. Future Generation Computer Systems, 102, 259-277

    2020

  3. [11]

    Ben-Sasson, E., Chiesa, A., Garman, C., Green, M., Miers, I., Tromer, E., & Virza, M. (2014). Zerocash: Decentralized anonymous payments from bitcoin. 2014 IEEE Symposium on Security and Privacy, 459-474

    2014

  4. [12]

    Bracke, P., Datta, A., Jung, C., & Sen, S. (2019). Machine learning explainability in finance: an application to default risk analysis. Bank of England Working Paper, (816)

    2019

  5. [13]

    Breidenbach, L., et al. (2021). Chainlink 2.0: Next Steps in the Evolution of Decentralized Oracle Networks. Chainlink Labs

    2021

  6. [14]

    Buterin, V. (2014). A next-generation smart contract and decentralized application platform. Ethereum Whitepaper

    2014

  7. [15]

    Caldarelli, G., & Rossignoli, C. (2021). The Problem of Trust in Decentralized Finance: The Case of Oracles. Journal of Open Innovation: Technology, Market, and Complexity

    2021

  8. [16]

    Chen, Y., Bellavitis, C., & Kamath, R. (2022). Decentralized Finance (DeFi) and the New Paradigm of Trust. Journal of Business Research

    2022

  9. [17]

    M., Jiang, J

    Chiu, J., Davoodalhosseini, S. M., Jiang, J. H., & Zhu, Y. (2022). Bank Market Power and Central Bank Digital Currency: Theory and Quantitative Assessment. Journal of Political Economy

    2022

  10. [18]

    Compound. (2023). Compound Protocol Whitepaper. Compound Finance

    2023

  11. [19]

    & Juels, A

    Daian, P., Goldfeder, S., Kell, T., Li, Y., Zhao, X., Bentov, I., ... & Juels, A. (2020). Flash boys 2.0: Frontrunning in decentralized exchanges, miner extractable value, and consensus instability. IEEE Symposium on Security and Privacy (SP), 910-927

    2020

  12. [20]

    DappHub. (2020). DS-Math: Safe Math Library for Solidity. GitHub

    2020

  13. [21]

    Dika, A., & Nowostawski, M. (2018). Security vulnerabilities in ethereum smart contracts. 2018 IEEE International Conference on Internet of Things (iThings)

    2018

  14. [22]

    Duffie, D., & Singleton, K. J. (1999). Modeling Term Structures of Defaultable Bonds. The Review of Financial Studies, 12(4), 687-720

    1999

  15. [23]

    R., El-Fakdi, A., & Ballesteros-Rodríguez, A

    Esteva, P. R., El-Fakdi, A., & Ballesteros-Rodríguez, A. (2023). Invoice Discounting Using Kelly Criterion by Automated Market Makers-like Implementations. Mathematics, 11(7), 1673

    2023

  16. [24]

    Glasserman, P. (2013). Monte Carlo Methods in Financial Engineering. Springer Science & Business Media

    2013

  17. [25]

    Gudgeon, L., Werner, S., Perez, D., & Knottenbelt, W. J. (2020). DeFi Protocols for Loanable Funds: Interest Rates, Liquidity and Market Efficiency. Proceedings of the 2nd ACM Conference on Advances in Financial Technologies, 92-112

    2020

  18. [26]

    Gudgeon, L., et al. (2020). DeFi Protocols for Loanable Funds: Interest Rates, Liquidity and Market Efficiency. AFT '20

    2020

  19. [27]

    R., Ramachandran, A., & Santoro, J

    Harvey, C. R., Ramachandran, A., & Santoro, J. (2021). DeFi and the Future of Finance. John Wiley & Sons

    2021

  20. [28]

    Jorion, P. (2006). Value at Risk: The New Benchmark for Managing Financial Risk. McGraw-Hill

    2006

  21. [29]

    Kelly, J. L. (1956). A New Interpretation of Information Rate. Bell System Technical Journal, 35(4), 917-926

    1956

  22. [30]

    J., & Werner, S

    Klages-Mundt, A., Harz, D., Knottenbelt, W. J., & Werner, S. M. (2020). Stablecoins 2.0: Economic Foundations and Risk-based Models. Proceedings of the 2nd ACM Conference on Advances in Financial Technologies, 59-79

    2020

  23. [31]

    Liu, X., Wang, Y., & Zhang, L. (2021). Risk Assessment in Supply Chain Finance utilizing Explainable AI. IEEE Transactions on Engineering Management

    2021

  24. [32]

    M., & Lee, S

    Lundberg, S. M., & Lee, S. I. (2017). A Unified Approach to Interpreting Model Predictions. Advances in Neural Information Processing Systems, 30

    2017

  25. [33]

    H., Olickel, H., Saxena, P., & Hobor, A

    Luu, L., Chu, D. H., Olickel, H., Saxena, P., & Hobor, A. (2016). Making Smart Contracts Smarter. Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security, 254-269

    2016

  26. [34]

    C., Thorp, E

    MacLean, L. C., Thorp, E. O., & Ziemba, W. T. (2011). The Kelly Capital Growth Investment Criterion. World Scientific

    2011

  27. [35]

    S., De La Rosa, P

    Madugula, S. S., De La Rosa, P. E., & Shankar, D. (2025). A Multi-Agent AI-Blockchain Framework with Reverse Kelly AMM for Under-Collateralized Real-World Asset Lending. Research Square. https://doi.org/10.21203/rs.3.rs-8135097/v1

    2025 doi

  28. [36]

    S., Esteva de la Rosa, P., & Shankar, D

    Madugula, S. S., Esteva de la Rosa, P., & Shankar, D. (2026a). Explainable AI-Driven Dynamic Loan Pricing on Ethereum: Integration of SHAP-Interpretable Risk Models, Reverse Kelly AMM, and Blockchain Trust Mechanisms. Preprints. https://doi.org/10.20944/preprints202605.1749.v1

    doi

  29. [37]

    S., Esteva de la Rosa, P., & Shankar, D

    Madugula, S. S., Esteva de la Rosa, P., & Shankar, D. (2026b). Data Leakage-Free Explainable AI for Decentralized Credit Scoring: A SHAP-Interpretable Approach to Default Prediction. Preprints. https://doi.org/10.20944/preprints202606.0556.v1

    doi

  30. [38]

    Makarov, I., & Schoar, A. (2022). Cryptocurrencies and Decentralized Finance (DeFi). Brookings Papers on Economic Activity

    2022

  31. [39]

    Merton, R. C. (1974). On the Pricing of Corporate Debt: The Risk Structure of Interest Rates. The Journal of Finance, 29(2), 449-470

    1974

  32. [40]

    M., & Sandner, P

    Meyer, E., Welpe, I. M., & Sandner, P. (2022). Decentralized Finance—A Systematic Literature Review and Research Directions. SSRN Electronic Journal

    2022

  33. [41]

    Molnar, C. (2020). Interpretable Machine Learning. Lulu.com

    2020

  34. [42]

    Nakamoto, S. (2008). Bitcoin: A peer-to-peer electronic cash system. Decentralized Business Review

    2008

  35. [43]

    OpenZeppelin. (2023). OpenZeppelin Contracts Documentation. OpenZeppelin

    2023

  36. [44]

    Paszke, A., et al. (2019). PyTorch: An imperative style, high-performance deep learning library. Advances in Neural Information Processing Systems, 32

    2019

  37. [45]

    M., Xu, J., & Livshits, B

    Perez, D., Werner, S. M., Xu, J., & Livshits, B. (2021). Liquidations: DeFi on a Knife-edge. Financial Cryptography and Data Security, 457-476

    2021

  38. [46]

    Poundstone, W. (2010). Fortune's Formula: The Untold Story of the Scientific Betting System That Beat the Casinos and Wall Street. Hill and Wang

    2010

  39. [47]

    Qin, K., Zhou, L., Livshits, B., & Gervais, A. (2021). Attacking the DeFi Ecosystem with Flash Loans for Fun and Profit. Financial Cryptography and Data Security, 3-32

    2021

  40. [48]

    Qin, K., et al. (2021). CeFi vs. DeFi--Comparing Centralized to Decentralized Finance. arXiv preprint arXiv:2106.08195

    2021 arXiv

  41. [49]

    Schär, F. (2021). Decentralized Finance: On Blockchain-and Smart Contract-Based Financial Markets. Federal Reserve Bank of St. Louis Review, 103(2), 153-174

    2021

  42. [50]

    E., & Weiss, A

    Stiglitz, J. E., & Weiss, A. (1981). Credit Rationing in Markets with Imperfect Information. The American Economic Review, 71(3), 393-410

    1981

  43. [51]

    Thorp, E. O. (2006). The Kelly criterion in blackjack sports betting, and the stock market. Handbook of Asset and Liability Management, 1, 385-428

    2006

  44. [52]

    Thorp, E. O. (1969). Optimal gambling systems for favorable games. Revue de l'Institut International de Statistique, 273-293

    1969

  45. [53]

    W., Liu, Y., & Li, Z

    Tolmach, P., Li, Y., Lin, S. W., Liu, Y., & Li, Z. (2021). A Survey of Smart Contract Formal Specification and Verification. ACM Computing Surveys, 54(7), 1-38

    2021

  46. [54]

    Victor, F., & Lüders, B. K. (2019). Measuring Ethereum-based ERC20 Token Networks: An Observational Study. Internet Measurement Conference (IMC), 113-126

    2019

  47. [55]

    Wang, Q., Li, R., Wang, Q., & Chen, S. (2022). Exploring the Attack Surfaces of Blockchain-based DeFi Systems. IEEE Communications Surveys & Tutorials

    2022

  48. [56]

    Wang, Y. (2020). Automated Market Makers for Decentralized Lending. DeFi Research Institute

    2020

  49. [57]

    M., Perez, D., Gudgeon, L., Klages-Mundt, A., Harz, D., & Knottenbelt, W

    Werner, S. M., Perez, D., Gudgeon, L., Klages-Mundt, A., Harz, D., & Knottenbelt, W. J. (2022). SoK: Decentralized Finance (DeFi). Proceedings of the 4th ACM Conference on Advances in Financial Technologies, 20-36

    2022

  50. [58]

    Wolf, T., et al. (2020). Transformers: State-of-the-Art Natural Language Processing. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, 38-45

    2020

  51. [59]

    Wood, G. (2014). Ethereum: A secure decentralised generalised transaction ledger. Ethereum Project Yellow Paper

    2014

  52. [60]

    World Bank. (2020). Micro, Small, and Medium Enterprise (MSME) Finance Gap. World Bank Group

    2020

  53. [61]

    Xu, J., Feng, Y., & Cousaert, S. (2023). Automated Market Makers: A Systematic Review. ACM Computing Surveys

    2023

  54. [62]

    Xu, J., Paruch, K., Cousaert, S., & Feng, Y. (2022). SoK: Yield Aggregators in DeFi. 2022 IEEE International Conference on Blockchain and Cryptocurrency (ICBC), 1-14

    2022

  55. [63]

    Zaharia, M., et al. (2018). Accelerating the Machine Learning Lifecycle with MLflow. IEEE Data Eng. Bull., 41(4), 39-45

    2018

  56. [64]

    Zeng, M., Chen, Z., & Li, Y. (2022). Dynamic Pricing in Blockchain Lending Platforms. Financial Innovation, 8(1), 1-25

    2022

  57. [65]

    Zhao, Y., & Chen, W. (2021). Blockchain-based Supply Chain Finance: A Systematic Review. International Journal of Production Research

    2021

  58. [66]

    N., Chen, X., & Wang, H

    Zheng, Z., Xie, S., Dai, H. N., Chen, X., & Wang, H. (2020). Blockchain challenges and opportunities: A survey. International Journal of Web and Grid Services, 14(4), 352-375

    2020

  59. [67]

    Ziemba, W. T. (2015). Response to Paul A Samuelson letters and papers on the Kelly capital growth investment strategy. Journal of Portfolio Management

    2015

Pith tools

Reviewed July 11, 2026 · model on record in the stance chip above.