{"id":"3a9fc95f-0f9f-4562-9d5a-4ab56e8e38c8","arxiv_id":"2607.16622","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"In a calibrated proof-of-stake model, equilibrium price is only a very-long-run anchor (≈46-year half-life), passive staking raises price and centralizes ownership, and symmetric speculative cycles transfer tokens from investors to active users.","lead":"An economic model of proof-of-stake tokens finds the token price does have a long-run anchor, but the network would take roughly 46 years to reach it after a shock; it also shows passive staking shifts ownership toward investors, while boom-bust cycles shift tokens back to active users. Why read it: it gives concrete mechanisms for why crypto prices stay far from fair value for decades and how institutional staking can change who controls a PoS network.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The volatility-harvesting transfer in §3.1 is hard-wired into the log-Cobb-Douglas utility's fixed consumption share; the paper does not show it survives a more general Consumer preference specification.","rationale":"The reader's weakest assumption (fixed expenditure share from U1) captures the most load-bearing point. The central redistribution claim is a corollary of an accounting identity imposed by the utility function; the paper acknowledges but does not test its generality. The eigenvalue formula in Theorem 1(ii) is a genuine derivational error, but it does not affect the first-order half-life and is less foundational. A robustness check with a generalized utility would determine whether the model's key qualitative conclusion—redistribution toward Consumers under speculative cycles—holds beyond the log-Cobb-Douglas case. Since the reader already returned CONDITIONAL, this stress test does not change the verdict; it sharpens the condition: the model's second headline result should be reported as conditional on the constant expenditure share.","tokens_in":12185,"tokens_out":10490,"duration_ms":106103,"concrete_test":"Re-run the §3.1 buy/sell experiment with U1 replaced by a Stone-Geary utility U1 = ν log(C−C0)+(1−ν) log V for a small C0 (e.g., 1% of steady-state fiat inflow), re-derive the clearing price and Consumer staking dynamics, and measure the Investor-to-Consumer token transfer over the full symmetric cycle. If the transfer is no longer robustly positive, the volatility-harvesting result is specific to the log fixed-share form.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Equation (1) fixes C=νW_c, so eq. (14) νW_{t+1,c}=Λ_{t+1}+I_{t+1,c} is an accounting identity. Equation (29) then makes the Consumer's net fiat flow exactly equal to -Λ, so the 'contrarian rebalancing' is simply the fixed expenditure share. The token transfer from Investor to Consumer in the buy/sell cycle is a direct corollary of this share, not an emergent property of the market. The paper's §4 lists abstractions but does not explore the mechanism's robustness to a non-constant consumption propensity (e.g., Stone-Geary or CRRA). If C'(W) deviates from ν, the identity breaks and the sign/size of the transfer is undetermined. This matters because the abstract and §3.1.2 present the transfer as a robust implication ('endogenous constant-value strategy', 'Shannon's-demon-like volatility harvesting'), and it is a central piece of the second finding. The eigenvalue error in Theorem 1(ii) is a separate, smaller issue: the printed λ>1 contradicts stability, though the first-order approximation and the 46-year half-life use the correct value.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper builds a two-class open-economy model of a Proof-of-Stake network. In a Consumer-only economy with a log-Cobb-Douglas utility over fiat consumption and crypto holdings, the authors prove global asymptotic stability of a unique steady state (Theorem 1), derive a closed-form price anchor, and calibrate a relaxation half-life of about 46 years. They then introduce an Investor class. Lemma 1 describes a regime in which passive staking raises the token price and shifts staked-token ownership away from Consumers. In §3.1, a simulation of symmetric buy/sell shocks shows that the Consumer's fixed fiat-consumption propensity generates a contrarian rebalancing that transfers tokens from Investor to Consumer while leaving the Consumer's fiat-denominated wealth unchanged, which the paper labels 'volatility harvesting.'","tokens_in":12369,"tokens_out":11346,"duration_ms":104739,"significance":"If the stability and slow-relaxation results hold, they constitute a valuable contribution to the macroeconomics of PoS networks: the 46-year half-life is a striking quantitative claim with practical implications for understanding price deviations from fundamentals. The model is tractable, with explicit market-clearing conditions and a closed-form steady state; the proof of global stability is a genuine theoretical result. However, the second main finding—volatility harvesting and the associated token transfer—is largely a direct consequence of the fixed expenditure share imposed by utility (1). This reduces the novelty and robustness of that claim relative to the presentation in the abstract and introduction. The paper is transparent about many limitations but does not analyze whether the mechanism survives a more general utility specification.","major_comments":[{"comment":"The 'endogenous constant-value strategy' and the Investor-to-Consumer token transfer are hard-wired by the fixed expenditure share in utility (1). The market-clearing identity νW_{t+1,c}=Λ_{t+1}+I_{t+1,c} (14) and its corollary I_c^t−νW_c^t=−Λ_t (29) follow immediately from C=νW_c and the accounting condition; they do not represent an emergent market mechanism. The paper should state this explicitly in the main text and, ideally, test robustness to a non-constant consumption propensity (e.g., Stone-Geary or CRRA). Without such a check, the abstract's claim of an 'endogenous constant-value strategy' is overstated and potentially circular.","section":"§3.1, eqs. (1), (14), (29)"},{"comment":"The displayed expression for λ appears to read λ = 1 + (1/2)y*/(1 + y* + (1/2)(y*)^2), which exceeds 1 and contradicts the global stability proved in part (i). The intended formula is presumably the fraction (1 + (1/2)y*)/(1 + y* + (1/2)(y*)^2), which is approximately 1 − (1/2)y*. The numerical half-life uses the correct value, so this is a typesetting/notation error, but it should be corrected because the theorem as printed is internally inconsistent.","section":"Theorem 1(ii), eq. (28)"},{"comment":"The paper claims that passive institutional staking 'compresses the endogenous staking yield below the Consumer-only equilibrium level.' However, Lemma 1 does not prove this; it assumes 0<y_1<y* as a sufficient condition. While a large Investor stake likely makes S_1>S* and hence y_1<y*, this is not derived from the model's primitives. The result should be stated conditionally, or a proof of yield compression from the Investor's presence should be supplied.","section":"Introduction, Abstract, and Lemma 1"}],"minor_comments":[{"comment":"The statement that increased fiat inflow lowers the long-run equilibrium price is counterintuitive and deserves a more explicit explanation of the supply-side channel (higher inflow → more staking → higher issuance → lower price).","section":"Comment 3, §2.2"},{"comment":"The 'asset-market clearing locus' is stated without derivation. It appears to follow from (21)-(22), but the intermediate steps would help readability.","section":"§2.2, eq. (25)"},{"comment":"Several references are to the author's own unpublished arXiv preprints ([10], [11]). If published versions exist, citing them would improve the paper's verifiability.","section":"General"},{"comment":"Equation (19) defines S_{t+1} implicitly through y_{t+1} on both sides. The text should note this explicitly to avoid confusion about the nature of the transition map.","section":"§2.0.4, eq. (19)"},{"comment":"The post-shock horizon in the simulation is finite; the paper should clarify whether the two-class economy converges to a new steady state or remains displaced indefinitely. The persistent displacement claim would be stronger with a longer-run analysis.","section":"§3.1, Figure 2"}],"recommendation":"major_revision","confidential_remarks":"The paper's first contribution—the Consumer-only stability result and the 46-year relaxation half-life—is solid and well within the journal's scope. The second contribution, however, is presented as a general mechanism while being an almost direct consequence of the log-Cobb-Douglas utility's fixed expenditure share. I would be comfortable with acceptance after the authors either substantially temper the claims about 'endogenous volatility harvesting' or provide a robustness analysis with a more general preference specification and correct the Theorem 1(ii) formula. The current framing risks overstating the novelty."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nHere's my take on Perepelitsa's PoS model.\n\nThe paper is better than the abstract makes it look. The core construction—two classes, Consumers with dual utility and Investors with exogenous fiat shocks—is coherent, and I checked the main algebra: the clearing price equations (12)–(14), the Consumer-only reduction (26), and the steady state (23) all check out. Theorem 1's claim of global asymptotic stability is plausible and the proof via monotone maps is standard. The eigenvalue/relaxation calculation, once you fix the sign error, gives λ ≈ 1 − y*/2, and the resulting half-life of about 46 years is a good illustration of how slowly the physical supply adjusts. The overshooting analogy to Dornbusch is apt. Lemma 1 on passive accumulation is also straightforward and correctly shows a monotone rise in price and fall in Consumer share.\n\nThe real soft spot is the \"endogenous volatility harvesting\" result. It is not emergent; it is the fixed expenditure share from the log-Cobb-Douglas utility (1). Because C = νW_c, the market-clearing condition (14) becomes an identity, and equation (29) forces the Consumer's net fiat flow to exactly offset the Investor's shock. So the contrarian rebalancing—selling high and buying low—is baked into the preference specification. The paper does not explore whether a non-homothetic or intertemporally optimizing Consumer, or non-myopic expectations, would preserve the transfer. The limitations section is candid about many abstractions, but it does not flag this dependence. As a result, the abstract and §3.1.2 overstate the robustness of the decentralization implication; at best it's conditional on the utility structure.\n\nA few smaller issues. Theorem 1(ii) as printed gives λ > 1, which contradicts stability; the approximation on the same line uses the correct λ < 1. The numerical half-life uses the correct value, so it's a typo, but it will confuse readers. The introduction says \"approximately 50 years\" while the abstract says 46; they should match. The simulation is not reproducible from the text alone—no code, no data, no error bars. And the half-life itself is not an independent estimate; it's a re-parameterization of the calibrated annual yield target, as the paper's own §4 admits.\n\nBottom line: this is a serious modeling paper with one genuine mathematical contribution (the stability/relaxation theorem) and one overstated interpretive claim. It deserves a careful referee, and the author should be asked to fix the sign error, reconcile the half-life numbers, add a robustness check on the consumption share, and tone down the harvesting language. I'd read it again after those revisions.\n\nIt's a yes for peer review, and I'd probably cite the stability theorem, but not the harvesting mechanism without a big caveat.","headline":"A mathematically solid PoS macro model whose advertised 'volatility harvesting' is a consequence of the assumed fixed consumption share; the stability result and the long-relaxation insight are worth taking seriously.","tokens_in":13055,"tokens_out":4202,"would_cite":true,"duration_ms":42721,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A Proof-of-Stake economy has a unique long-run nominal price anchor, but it takes about 46 years to return to it; passive staking centralizes consensus ownership, while speculative buy/sell cycles can transfer tokens from investors back to","keywords":["proof-of-stake","token price dynamics","steady-state equilibrium","open-economy macro model","volatility harvesting","staking centralization","Ethereum calibration","price overshooting"],"falsifier":"An empirical test would be to observe a symmetric, fiat-neutral staking inflow/outflow cycle and check whether the non-speculative staking address ends up with a larger token balance and whether the token price remains below its pre-cycle level. A second check: if the model's steady state is correct, the price should slowly mean-revert with a half-life near 46 years; observing much faster mean reversion in major Proof-of-Stake networks would contradict the calibration. A structural test would replace the Cobb–Douglas upper-tier utility with a CES utility of elasticity different from one, which","tokens_in":11850,"feed_emoji":"⛓️","tokens_out":7332,"duration_ms":64558,"temperature":0.7,"pith_summary":"The paper builds a macroeconomic model of a Proof-of-Stake network with two agent classes: Consumers, who use the token for transactions and staking and have a fixed propensity to consume fiat, and Investors, who stake passively and inject or withdraw fiat. It proves that the Consumer-only economy has exactly one stable steady state, giving a well-defined long-run nominal price anchor; calibrated to current Ethereum parameters, the system's relaxation half-life is about 46 years, so the price can remain persistently displaced from its equilibrium benchmark. Adding a passive Investor class makes the token price rise monotonically while the Consumers' share of staked tokens falls, implying a centralizing force on consensus ownership. Symmetric buy/sell shocks, however, trigger an endogenous contrarian rebalancing by Consumers—they sell tokens as the price rises and buy as it falls—so a zero-net-fiat speculative cycle ends with physical tokens transferred from the Investor to the Consumer. The paper frames this as volatility harvesting in token units rather than in fiat.","feed_headline":"Proof-of-stake prices can stray from their anchor for 46 years","feed_subtitle":"Passive staking concentrates consensus ownership; active speculative cycles can hand tokens back to utility users.","key_machinery":"The argument rests on the Consumers' top-level Cobb–Douglas utility U1 = ν log C + (1−ν) log V, which fixes the fiat consumption share at ν and the crypto retention share at 1−ν. Combined with myopic price expectations, market clearing becomes the identity νW_{t+1,c} = Λ_{t+1} + I_{t+1,c}, so any Investor fiat flow is exactly offset by the Consumer's net fiat contribution. This identity is what turns speculative buy/sell cycles into a contrarian constant-value strategy. The stability proof uses the monotone transition map S_{t+1}=F(S_t) from the recurrence in equation (27), with the square-root issuance rule y=c/√S, to show global convergence to the unique steady state and to compute the rel","core_discovery":"The central claim is that the nominal token price in a Proof-of-Stake economy is anchored by a unique, globally asymptotically stable steady state determined by Consumer fiat inflows and the network's issuance and burn rates. The price is a fast variable that jumps immediately to clear markets, while the physical staked supply adjusts slowly through issuance and burning, producing a relaxation eigenvalue λ≈1−y*/2 and, under Ethereum-like calibration, a 46-year half-life. Consequently, the observed price can remain far from its evolving fundamental benchmark for decades—an inherent property of the network, not just a symptom of speculation. The second claim is that passive institutional staki","pith_inferences":["The 46-year half-life is tied to the square-root issuance rule and the fixed expenditure share; with a flatter issuance curve or an elastic consumer demand, the price anchor would be much stronger. A natural extension is to compute the half-life under alternative issuance schedules.","The 'constant-value strategy' is not an intentional choice by Consumers but a mechanical consequence of the Cobb–Douglas top-level utility. If consumers had non-homothetic or forward-looking preferences, speculative cycles could transfer tokens in the opposite direction, so the decentralization result is sensitive to preference specification.","A testable signature of the volatility-harvesting mechanism is that after a symmetric fiat-neutral buy/sell cycle, the Consumer's staked fraction rises while the token price remains below its pre-cycle level; this could be checked against on-chain staking and exchange-flow data.","Passive and active speculative capital have opposite effects on consensus ownership, implying that policies shaping capital flows (staking derivatives, ETF structures) could be designed to amplify volatility-harvesting and mitigate passive-accumulation centralization."],"forward_implications":["If the calibration is representative, Proof-of-Stake token prices can sustain multi-decade deviations from their equilibrium benchmark after fundamental shocks, so short-run prices are a poor guide to long-run network value.","Passive institutional staking, by compressing the staking yield, can monotonically raise the nominal token price while concentrating staked-token ownership away from transactional users—a centralization channel independent of validator technology.","Symmetric fiat-denominated speculative cycles are wealth-neutral in fiat but token-accretive for Consumers, meaning active speculation can shift control of consensus toward utility users.","A higher Consumer fiat inflow raises the price only in the short run; because issuance outstrips burning during adjustment, the long-run equilibrium price is inversely related to the inflow rate.","The network behaves like an open economy with overshooting: fast price response and slow stock adjustment produce persistent disequilibrium even without rational speculation."],"fun_headline_variants":["PoS price anchor eludes for 46 years","Staked supply lags, so prices overshoot for decades","Passive staking shifts ownership; active cycles return it","Token price inertia: 46-year half-life in proof-of-stake"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The entire mechanism—the 46-year half-life and the Investor-to-Consumer token transfer—rests on the Consumers' fixed fiat-consumption share from a Cobb–Douglas utility and on myopic price expectations; if real users adjust their consumption propensity or form forward-looking expectations, the price anchor and the volatility-harvesting transfer could weaken, reverse, or disappear.","fun_headline_variants_meta":{"raw":{"variants":["PoS price anchor eludes for 46 years","Staked supply lags, so prices overshoot for decades","Passive staking shifts ownership; active cycles return it","Token price inertia: 46-year half-life in proof-of-stake"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00142,"raw_usage":{"total_tokens":5576,"prompt_tokens":761,"completion_tokens":4815,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":505,"completion_tokens_details":{"reasoning_tokens":4745}},"tokens_in":505,"tokens_out":4815,"duration_ms":36748,"temperature":1.0,"reasoning_tokens":4745,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T20:28:07.307001+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"An empirical test would be to observe a symmetric, fiat-neutral staking inflow/outflow cycle and check whether the non-speculative staking address ends up with a larger token balance and whether the token price remains below its pre-cycle level. A second check: if the model's steady state is correct, the price should slowly mean-revert with a half-life near 46 years; observing much faster mean reversion in major Proof-of-Stake networks would contradict the calibration. A structural test would replace the Cobb–Douglas upper-tier utility with a CES utility of elasticity different from one, which","supporting_citations":[],"review_version":1}