§5 Good Money and Bad Money Revisited

For competing-currency theory to operate in practice, one must first clarify under what institutional premises “bad money drives out good” holds—and why, once legal parity is stripped away, the literature treats the Thiersian direction (good drives out bad) as a distinct mechanism rather than a simple “reversal” of the same law.

“Bad money drives out good” has become something of a shield, invoked against any monetary competition: as if, once multiple currencies coexist, the inferior must prevail, the superior withdraw, and ruin ensue. The phrase is cited far more often than it is carefully read. What Gresham observed was a particular phenomenon under particular institutional conditions; its core mechanism depends on a critical premise—legal parity, an official command fixing the exchange ratio between two monies while market valuations have already diverged from that ratio. Once that premise is removed, the direction of elimination is governed by market valuation and relative-price adjustment—what the literature calls Thiers’ Law, good driving out bad—but with this emphasis: it is another mechanism under different institutional premises, not an “inversion” of Gresham’s law under floating rates. One must first clarify Gresham’s precise meaning, its scope of application, and the trigger conditions for the Thiersian direction, before turning to elimination mechanisms under floating rates, on-chain liquidity, and standard-protocol environments—where the definitions of “good” and “bad” themselves are being rewritten by richer market signals.

Section 1. The Precise Context of Gresham’s Law

Thomas Gresham (1519–1579) was financial adviser to Queen Elizabeth I of England and founder of the Royal Exchange. In 1558 he wrote to the newly crowned queen describing the deterioration of English coinage: under Henry VIII the silver content of coins had been sharply reduced, and the poorer new coins circulated at one-to-one legal face value with older coins of higher silver content. Holders accordingly spent the inferior new coins first and hoarded or exported the superior old ones to the Continent—“bad money” drove “good money” from circulation.

The core mechanism of this observation is that legal parity creates arbitrage: when the official exchange ratio between two monies differs from market valuations, holders exploit the gap by spending the officially “overvalued” money (bad) and retaining the officially “undervalued” money (good). Bad money wins not because “bad” naturally conquers “good,” but because price controls create arbitrage opportunities; rational market behavior follows, and good money disappears from circulation.

This mechanism has recurred throughout history. In Ming–Qing China, copper cash and silver circulated in parallel under official exchange rates, yet market valuations shifted with supply and demand; when official ratios diverged from market valuations, the undervalued metal was hoarded or flowed abroad. During the nineteenth-century U.S. Gold Acts, official parity between gold and greenbacks was maintained while market spreads triggered systematic gold hoarding and export. In every case the trigger was official price control, not the essence of competition itself.

Correct application of Gresham’s law therefore requires checking two conditions: Is there a legal or officially enforced fixed exchange rate? Have market valuations of the two monies already diverged from that fixed ratio? Hayek stressed in The Denationalization of Money: “Gresham’s law will apply only to different kinds of money between which a fixed rate of exchange is enforced by Law”¹—only when both conditions are met does the prediction that “bad money drives out good” hold. Under freely floating rates there is no arbitrage opportunity to exploit; inferior money is reflected in market prices and voluntarily sold by holders.

Section 2. Thiers’ Law: Another Direction of Elimination under Floating Relative Prices

Discussed alongside Gresham under floating relative prices and no forced parity is what is sometimes called “Thiers’ Law,” named for the French economist and statesman Adolphe Thiers (1797–1877); Rolnick and Weber (1986), using Minnesota free-banking-era data, contrasted it with the Gresham mechanism on institutional premises—when discounting is allowed the direction aligns with Thiers; when parity is forced, Gresham embeds². The logic is simple: when monetary relative prices can adjust with the market, money whose purchasing power continually depreciates is actively avoided by holders, relatively stable money is retained, and “good money” prevails in competition.

Historical cases give Thiers’ Law broad support. Across decades of Argentine peso inflation, whenever official exchange controls loosened or were circumvented, dollars circulated in the market at a premium; during Venezuelan hyperinflation the bolívar was rapidly displaced by dollars and cryptocurrency. Zimbabwe’s inflation peak, tax anchor, and “sanctity departed” mechanism are already detailed in Chapter 2, Section 5³—Hanke & Krus (2013), using official and black-market series, estimated the November 2008 monthly inflation peak at roughly 79.6 billion percent (7.96×10^10%); after dollarization in 2009 the domestic currency had functionally exited, while legal discontinuance came only in 2015—the C-money circle (tax nominal) and M-money circle (everyday store of value) can coexist in rift for years³. Here the emphasis is only on the Thiersian reading: bad money is actively abandoned, not Gresham arbitrage under legal parity; only when official and black-market rates diverge for long periods does superficially similar “spend local, keep dollars” behavior appear—one must not conflate similar directions with a “Gresham inversion.” Charles Goodhart distinguishes “market money of account” (M-money, in the Mengerian saleableness tradition) from “state money of account” (C-money, in the Knapp–Innes tax-circle tradition): Zimbabwe and Argentina display a rift in which the C-money circle remains while M-money ranking has been rewritten—taxes still require domestic-currency nominal, yet everyday store of value and pricing have shifted to the dollar⁴. Thiers’ Law describes M-money ranking; it does not automatically negate C-money’s institutional floor. The tension between the two is key to understanding how “voting with one’s feet” and “institutional chains still binding” can coexist.

In these cases legal status did not save inferior money; market behavior ultimately drove out the bad. Whether authentication costs can reproduce Gresham-like outcomes without parity still turns on information costs—legal parity alone does not uniquely determine the direction of elimination.

Free-banking historiography supplies finer-grained corroboration for the Thiersian direction. Analyzing competitive issue, Selgin noted that when banknotes circulate at market discount rather than legal parity, notes of overissuing banks are spent first and notes of prudent banks retained—consistent with Thiersian logic, except that “bad” and “good” are defined by clearing discounts rather than official exchange rates⁵. After the 1865 federal 10 percent tax on state banknotes, Selgin argued from issue statistics that state notes were forced to accept at parity and in practice exited circulation—legislation can rebuild de facto parity within a private system, blocking the Thiersian direction and re-embedding Gresham-style retention⁵. Selgin’s 1986 “legal restrictions theory” generalizes the mechanism: it is not that “multiple issuers necessarily produce chaos,” but that layered rules—parity laws, branching bans, holding taxes—determine whether relative prices can float—point-by-point comparable with Hayek’s proposition at pp. 42–43 that “Gresham applies only under legal parity”⁶. Conversely, once legislation forces acceptance at face parity (or bans discounted circulation), the Gresham mechanism can re-embed in private banking: antebellum state “parity laws” for banknotes are, in Rockoff’s and Selgin’s historiographical reconstructions, precisely the institutional switch for bad notes remaining in circulation⁵. Rolnick and Weber, drawing on nineteenth-century Minnesota experience, further showed that so-called “Gresham’s law” often fails statistically—the key remains whether forced parity exists; under free float, bad money discounts and good money premiums, in the Thiersian direction². Thus the good/bad direction depends not on “whether there are multiple issuers,” but on who fixes relative prices—Hayek’s distinction between Thiers and Gresham at pp. 42–43 of The Denationalization of Money and the empirical detail of free-banking literature can mutually corroborate.

A subtler manifestation occurs at the level of “soft substitution”: residents of high-inflation countries often do not conspicuously “dump the domestic currency,” but protect purchasing power through dispersed holdings—foreign-currency deposits where law permits, gold jewelry, real estate, or commodities. Such behavior may be invisible in macro statistics, yet it is a genuine signal of monetary holders voting with their feet. When central-bank foreign reserves come under pressure from such capital outflow, the response is often tighter capital controls—itself evidence that market pressure for “bad money driven out under free flow” is real.

Section 3. Cryptocurrency Competition under Floating Exchange Rates

Applying the Gresham–Thiers framework to contemporary cryptocurrency, the first finding is that token-to-token pairs in the crypto ecosystem almost never face legal parity—exchange rates among tokens are set in real time by automated market maker (AMM) protocols or centralized-exchange order books, without any institution fixing ratios. Hayek illustrated with the end of Weimar inflation: when exchange rates float freely, “people refuse to accept inferior money and insist on superior money”⁷—the price-control-driven circulation of bad money scarcely exists between pure floating pairs; crypto-asset competition is closer to the Thiersian framework in relative-price mechanism.

Stablecoins do see adoption in cross-border corridors and high-inflation edge cases, yet on-chain real payments remain marginal in global payment flows (three metrics and data in Chapter 1, Section 8)⁸—the Thiersian direction is observable within crypto submarkets but cannot yet be extrapolated as displacement of fiat monopoly. Moreover, dollar-pegged stablecoins (USDT/USDC) functionally play a “1:1 parity” role: issuers promise redemption, and regulation and audit constrain deviation—this is not the same institution as Hayekian “pure floating competition after abolishing forced legal tender”; algorithmic Terra/UST’s 1:1 peg depended on sentiment and mint–burn mechanisms and failed rapidly in confidence crises—showing that parity need not appear as legal compulsion, yet can still produce Gresham-like retention and death spirals, with the trigger shifting from legal command to protocol promise and market belief.

For the Thiersian direction to hold, two testable premises must also be met: relative prices can adjust with the market, and quality signals can be verified at low cost. On-chain rule transparency favors the second, yet oracle price feeds, thin-pool manipulation, cross-chain bridge custody, and exchange concentration can still raise verification costs, allowing short-run “bad” assets to linger in circulation via liquidity subsidies or narrative premiums—surface outcomes sometimes resemble Gresham, yet the mechanism is information lag and incentive distortion, not legal-parity arbitrage.

Under this framework, holders’ choice criteria expand beyond the single dimension of “purchasing-power stability” in traditional monetary analysis to multidimensional evaluation. A token’s degree of “goodness” is priced by the market across several dimensions.

Transparency and credibility of issuance rules come first: Is the supply cap clear and verifiable? Do historical issuance records match white-paper commitments? Can governance resist “governance attacks” that alter critical parameters? Bitcoin supporters prize this dimension highly, treating the 21-million-coin cap as core value; holders of other tokens accept more flexible monetary policy in exchange for functional extensibility.

Liquidity depth and market acceptance follow closely: Can one buy and sell on major exchanges at reasonable slippage? Is the token accepted as collateral by mainstream DeFi lending protocols? If a token cannot flow smoothly into major financial Lego components, its usefulness as money or store of value is limited, and market valuation will reflect that constraint.

Depth of ecological application measures what developers build upon it—both a technical metric and a proxy for economic usefulness: the deeper the ecosystem, the more scenes in which the token has real demand; “demand-driven liquidity” is more robust than “incentive-driven liquidity.”

Section 4. Liquidity: The Modern Version of the Gresham Test

Liquidity depth is the most direct “good/bad” judgment test in on-chain environments. Automated market makers (AMMs) externalize liquidity in mathematical formulas: a token pool’s reserves directly determine the price impact of a given trade size. This information is real-time inspectable by anyone—transparency traditional money markets cannot match.

Yet “liquidity” can be driven by real demand or artificially stacked by incentives. The most common form of on-chain liquidity manipulation is “liquidity mining”: projects attract liquidity providers with high token rewards, manufacturing superficially deep pools. Capital floods in when reward rates are high and exits rapidly when they fall—leaving ruins of liquidity collapse and sharp price drops. Early injectors who cash out at high prices profit; later holders bear losses. This is not bad money driving out good in the Gresham sense (there is no legal pricing arbitrage on-chain), but short-run signal distortion manufactured by artificial incentive gradients, systematically transferring gains from later entrants after capital is attracted.

Distinguishing real from artificial liquidity requires comparing several metrics: liquidity relative to total market cap (may be anomalously high under heavy incentives, collapsing after incentives exit); average holding time of liquidity providers (long-term holders withstand premature exit better than reward chasers); historical stability of pools (liquidity that survives different market conditions is more real). Analyzing free-banking “face-parity laws,” Selgin noted that once legislation bans discounting or forces face acceptance, inferior notes linger in circulation—on-chain liquidity mining’s high-yield subsidies are structurally isomorphic: short-run incentives manufacture de facto parity acceptance; when subsidies ebb, inferior projects’ liquidity collapses⁹. The full mechanism of incentive distortion and tokenomics design problems. Over recent cycles the crypto ecosystem has accumulated market intuition for distinguishing the two, but the learning process is costly—DeFi total value locked (TVL) jumped from roughly $1 billion early in 2020 to a peak above $180 billion, then fell below about $500 billion after incentives ebbed in 2022¹⁰; “costly” here has concrete magnitude: large numbers of late, information-asymmetric participants bore the main losses.

Section 5. Bubble Cycles and the Market’s Self-Cleaning Logic

On-chain good/bad competition also exhibits a self-reinforcing cyclical distortion. In high-sentiment periods, new projects’ token reward rates are high enough to attract large liquidity inflows and push prices up; rising prices generate “wealth effects,” attracting more new participants and further lifting prices—a classic positive-feedback structure. In that process, “good” projects with real demand support and “bad” ones that pile up false demand purely via incentives are hard to distinguish on surface numbers, because both prices rise and both display the illusion of “ample liquidity.” This differs from Gresham’s law, which requires legal-parity premises¹; here market sentiment and incentive gradients distort quality signals, postponing good/bad discrimination until after the bubble bursts.

The order of bubble collapse usually runs inverse to the strength of “real demand”: projects lacking real use cases lose liquidity first and prices go to zero; projects with partial real demand but poorly designed incentives suffer sharp drawdowns; only a few protocols with real use cases, robust incentive design, and strong ecosystems maintain meaningful market caps and liquidity after the cycle turns down. In 2018 total crypto market cap fell from a peak near $800 billion to about $130 billion; in 2022 from a peak near $3 trillion to about $800 billion¹¹—both major bear-market cleansings roughly followed that order, though each cleansing also hit some “good” projects indiscriminately via liquidity crises, imperfectly achieving “bad cleared, good continued.” Market self-cleaning is a directional proposition, not an instantaneous justice mechanism.

The market’s self-cleaning function does exist, but its operating cost is high: large numbers of late, information-asymmetric participants (mainly retail) bear the main losses, while early “prescient” large holders have often already exited at highs. This distributional asymmetry is not unique to blockchain; any speculative new asset class undergoes similar price-discovery pain in maturation—the internet bubble and railway mania both left marks of excess and liquidation. Recognizing the cycle requires a long-term view of on-chain good/bad competition: competition is an effective screening mechanism, yet in the short to medium run sentiment and incentives can sharply distort screening clarity and fairness.

Section 6. Standard Interfaces: Infrastructure for Good/Bad Competition

In Gresham’s era, switching from one money to another required finding exchangers, paying discounts, and bearing authentication risk; switching costs systematically affected which money was used, even allowing inferior money to remain in circulation because of those costs. Today blockchain standard interfaces compress switching costs to the technical-adaptation layer.

ERC-20—the technical standard for tokens on Ethereum—defines the minimal interface set a token must implement: transfer, approve, allowance, balanceOf, totalSupply. Any token following this standard automatically works with all ERC-20-supporting wallets, exchange interfaces, block explorers, and DeFi protocols. That means the technical time from deploying a new token to wallet display and DEX trading can be hours—not the months of compliance required to open a new account type in traditional finance.

Standards lower switching costs, but standards themselves are objects of competition. Token standards on different chains (Ethereum’s ERC-20, TRON’s TRC-20, Openverse’s VRC-20, and so on) share similar design ideas but differ in detail; without interoperability across chains, liquidity islands form, users must maintain different accounts on different chains, and cross-holding frictions remain. Per the Openverse white paper, VRC-20 provides within a given network a minimal interface similar to ERC-20 (transfer, balanceOf, approve, etc.) so wallets and DApps can compose calls—whether that path can achieve ERC-20-scale adoption and audit depth in mainstream ecosystems awaits mainnet and cross-chain interoperability practice, and should not be extrapolated from roadmaps as accomplished fact.

From the perspective of good/bad competition, the significance of standards is this: they make “good” easier to identify and adopt, and “bad” easier to detect and discard. When switching costs are extremely low, markets respond faster to monetary quality, and the cost to issuers of maintaining inferior rules rises sharply. White (1994) stressed that what competitive issue most struggles to close is often unit-of-account convergence—when multiple private monies coexist, which sets the price list?—the same class of problem as ERC-20/VRC-20 lowering switching costs without necessarily solving the unit-of-account issue⁹. This is one technical realization path for the competitive mechanism Hayek envisioned—premised on abolishing forced legal tender and allowing monies to compete freely at market exchange rates¹².

Section 7. Good and Bad under Functional Stratification

Traditional monetary analysis defines “good” as purchasing-power stability, reasonable in a fiat background centered on store of value and everyday unit of account. Yet the crypto ecosystem has functional stratification: different asset types serve different purposes and correspond to different standards of “good.”

“Good” for stablecoins: degree of deviation from the peg target (usually the dollar), adequacy of collateral, depth of liquidity, regulatory compliance status. The May 2022 collapse of algorithmic TerraUSD (UST) remains one of the largest stablecoin failures to date. Its design depended on a mint–burn mechanism with LUNA to maintain the peg, without exogenous hard-asset reserves; when market confidence cracked, selling pressure triggered a UST–LUNA death spiral, and both approached zero in under two weeks, evaporating peak combined market value on the order of $40 billion¹³. This is functional “bad”—a stablecoin’s core function is stability, and UST thoroughly destablized under stress; the peg mechanism became a feedback loop amplifying collapse rather than a buffer. Post-mortems attribute failure to a combination of reflexive pegs, thin-pool liquidity, and misaligned governance incentives, not a single parameter error¹³.

By contrast, MakerDAO’s DAI is overcollateralized with on-chain assets, with liquidation written into smart contracts and triggered automatically when the collateral ratio falls below a threshold. On 12 March 2020 (“Black Thursday”), ETH plunged roughly 30%, triggering mass liquidations and DAI premiums into roughly the $1.06–1.12 range; Maker community post-mortems recorded zero-price auctions and emergency governance parameter adjustments¹⁴. Under extreme stress DAI roughly held the peg, but deviation and repair costs show that the overcollateralized path is more robust, not frictionlessly immune—“good” is the result of relative comparison and institutional learning, not an absolute guarantee.

“Good” for governance tokens: fee income actually generated by the protocol and the share accruing to token holders; whether governance weight prevents concentrated control; real ecological activity (not incentive-driven).

“Good” for store-of-value assets: credibility of scarcity commitments; censorship resistance; degree of network decentralization. Bitcoin supporters prize these three, especially the second and third—because traditional assets (gold, fiat deposits) can be seized or confiscated, and Bitcoin’s censorship resistance under self-custody is a distinctive advantage.

The implication of functional stratification is that good/bad competition does not occur on a single axis but unfolds simultaneously across multiple dimensions. Comparisons across asset types are apples-to-oranges; the definition of “good” must first specify the object of evaluation and the use case.

Section 8. The Philosophy of Algorithmic Stability: Limits of Pure Rules and Hybrid Paths

After the stablecoin crisis, the crypto ecosystem reached a relatively mature consensus: pure algorithmic stability (no exogenous collateral, fully dependent on arbitrage mechanisms and dual-token structures), without sufficient external demand support, readily falls into death spirals under confidence crises. This case of “pure-rule failure” puts an expensive question mark beside the view that “if rules are clever enough, collateral is unnecessary.”

Yet the core disagreement concerns what rules can and cannot substitute. Rules can precisely define execution logic under specified conditions, but robustness depends on whether those “specified conditions” cover extremes. TerraUSD’s fatal flaw was that its arbitrage mechanism assumed always-present LUNA demand—itself sentimental and able to vanish suddenly; when sentiment reversed, the rules executed correctly, but in the direction of accelerating collapse. The deep problem of rule design is how external assumptions fail under extreme conditions.

The robustness of overcollateralized paths (such as DAI) comes from independence of reserve value: even under extreme pessimism, so long as collateral value (usually ETH or other mainstream assets) holds above some level, liquidation still protects holders. This is a hybrid of “rules + external anchor”: rules supply execution certainty; the external anchor supplies an independent value base. Neither is dispensable—rules without independent collateral leave the system fragile to external shocks; collateral without rules leaves execution dependent on human intervention and loses the value of decentralization.

Fiat-backed stablecoins (such as USDC) push the hybrid to an extreme: their “rule” is that the issuer holds dollar reserves and accepts audit; stability comes from traditional-finance credit backing. This is not the purest “trustlessness” at the on-chain technical layer, but it is the most stable at the user-experience layer. Three paths coexist in the market, serving holders with different needs—precisely the multidimensional definition of “good” in practice.

Section 9. Network Effects and Benign Lock-In

Outcomes of good/bad competition are shaped not only by quality but jointly by first-mover advantage and network effects. Money’s value partly derives from wide acceptability, which rises with the number of users—a classic network-effect structure. In markets with strong network effects, winner-take-all is a common output, not pluralistic steady states.

Bitcoin’s “digital gold” positioning has faced multiple “technically superior challengers” over more than a decade, yet as of 2024 still held roughly 50%–55% of total crypto market cap¹⁵. Ethereum’s smart-contract platform status, amid recurring “ETH killer” narratives, has long accounted for over 60% of DeFi locked value across the ecosystem¹⁰. Both have known technical limits yet still maintain dominance—network effects self-reinforce: more users → more liquidity → more developers → more applications → more users. This concentration aligns with the concern in Chapter 4, Section 9, “scale effects and new monopolies”: Thiersian victory of good money does not automatically equal pluralistic steady states.

Network effects can “lock in benign standards” or “lock in inferior equilibria.” If first-mover advantage happens to belong to a superior design, network effects help: they accelerate adoption of good money and raise high barriers for later inferior substitutes. If first-mover advantage happens to belong to an inferior design, network effects delay upgrade—as with the QWERTY keyboard, technically superior alternatives exist, but switching and coordination costs keep the existing standard stable. In money the problem is especially sharp, because monetary network effects are far stronger than keyboard layouts and coordination costs of switching are higher.

There is no universal answer to this tension, but there is a concrete design implication: if superior standards are to displace suboptimal ones, switching costs must be acceptable and transitional bridges must exist between new and old standards—one cannot expect all users to switch synchronously. Protocol interoperability, cross-chain bridges, and multi-chain wallets are technical means of lowering network-effect lock-in costs.

Section 10. Regulatory Arbitrage and Exogenous Constraints on Good/Bad Competition

Good/bad competition occurs not only inside markets but is deeply shaped by regulatory frameworks. Differing jurisdictional attitudes toward crypto assets create regulatory-arbitrage space, and that space itself affects competitive outcomes. If a well-designed protocol cannot gain institutional adoption in major markets because of regulatory uncertainty, while a weaker competitor operating freely in a lax environment gains liquidity, the regulatory framework distorts market-screening signals.

SEC enforcement actions against multiple crypto projects in 2022–2023 deemed some tokens unregistered securities¹⁶. That determination depends not on technical design or economic-mechanism quality, but on the Howey Test—whether there is investment of money, a common enterprise, and expectation of profits from others’ efforts. A carefully designed decentralized token under protocol standards may be deemed a security because of early token-allocation methods; a weaker economic mechanism that successfully decentralizes may evade securities-law application. Legal screening and market screening may therefore point different ways—the regulatory framework itself is an exogenous variable in good/bad competition; one cannot assume on-chain Thiersian mechanisms operate in a vacuum.

There is no simple solution to this contradiction, but there is an institutional exit: clear legal frameworks can reduce competitive distortion from uncertainty. The EU Markets in Crypto-Assets Regulation (MiCA), phased in from 2024, sets unified disclosure and reserve requirements for stablecoins and CASPs (crypto-asset service providers)¹⁷—practitioners treat it as a case of “rule clarity,” even where some clauses are stricter. The more predictable the rules, the less market screening is distorted by regulatory arbitrage alone; whether MiCA in practice lowers distortion still awaits enforcement and cross-border coordination tests, and should not be presumed a sufficient condition for competitive discipline.

Section 11. Design Problems of Competition in the Protocol Era

Discussion of good and bad money in the protocol era reduces to several interlocking design choices: how issuance rules trade off transparency and revisability; how liquidity incentives reflect real demand rather than short-run arbitrage; how standard interfaces compress switching costs low enough; how network effects avoid freezing first-mover advantage into permanent monopoly. These problems have no general answers detached from concrete protocols. Competition is no longer only a philosophical proposition but a daily on-chain contest of liquidity, adoption, and code quality—every contract call, every new protocol adopted, is a real-time market vote on the good/bad question.

Gresham’s history tells us that even when competition exists, the wrong institutional framework produces distorted outputs of “bad winning, good losing.” The on-chain ecosystem continues to evolve in incentive design, governance structure, and standardization; formation of the “right institutional framework” is a learning process requiring repeated market validation and rule iteration. That process cannot be shortened, but it can be learned consciously—each failure case is an information output about “what rules are infeasible,” valuable though costly.

From Gresham to Thiers, from Scottish free banking to the TerraUSD collapse, the story of good/bad competition spans five centuries; technical forms change while incentive logic does not. The pursuit of “good” is real; vigilance toward institutional conditions that give “bad” short-run advantage must not relax.

When full information, low switching costs, and enforceable rules are in place, good money tends to accumulate advantage, while bad money faces continuing liquidity pressure and user attrition. But these three conditions must be actively built and maintained; they are not public goods the market automatically produces—the responsibility of protocol designers is to construct information and incentive environments that let markets distinguish good from bad effectively¹⁸. Between “relative prices can float” and “bad money is driven out” lie three further thresholds: liquidity depth, verification cost, and compliant accessibility; on-chain experiments provide observation windows, yet the core position of global retail and sovereign monetary institutions has not switched.

Competition itself is a means; the end is to discover and promote better monetary institutional arrangements: competition under price controls drives out good money; competition under free float drives out bad; the on-chain ecosystem’s “institutional framework” includes technical standards, incentive design, and governance structure—not only legal provisions. Overestimating the automaticity of Thiers’ Law and underestimating the marginal contribution of rule verifiability to discipline equally distort judgment—the prudent conclusion is: direction can be argued; timing cannot be prophesied.