Kabbalah and Economics — reading Michael Snoyman against MoMaT

Michael Snoyman published a striking post this week: Kabbalah and Economics (15 July 2026), a sequel to his earlier comparison of Kabbalah with Haskell. This time he maps the Kabbalistic categories of giving and receiving — each split by action and intent — onto four economic settings: violence, free markets, socialism, and low time preference.

We read it against our axiomatic monetary theory (MoMaT). Where Snoyman argues normatively and theologically, we argue formally and institutionally — and the structural parallels run deeper than the difference in idiom suggests.

Michael's framework in brief

At the root sits ratzon lekabel — the desire to receive. Creation, on this reading, is built from want, and the spiritual task is to move toward giving. That is hard precisely because receiving is where everyone starts.

Michael grades this into four levels (action × intent):

Action	Intent	Economic reading
Receiving	Receiving	Pure selfishness; violence as the extreme case
Giving	Receiving	Voluntary trade: give the apple, want the money
Giving	Giving	True altruism — rare in practice
Receiving	Giving	Highest level in Kabbalah; Michael sets it aside for the economics comparison

From there he diagnoses three macro systems:

1. Violence — receiving-to-receive. Theft, conquest, slavery. Destructive; few defend it openly today.
2. Free markets (idealised) — giving-to-receive. Voluntary exchange, full information, fair courts, hard money. Productive, but cannot by itself produce selflessness.
3. Socialism — aims at giving-to-give, but in practice reverts to receiving-to-receive under coercion: forced transfer is still taking.

His conclusion: selflessness cannot be imposed at gunpoint. Let selfishness run the macro engine, and cultivate giving in the micro — family, close community. He closes by likening the Kabbalistic masach (the screen that defers immediate gratification) to low time preference, and warns that inflation pushes the other way.

Capitalism in MoMaT: the A–R–G core

MoMaT does not define capitalism by ideology or morality. Its core functionality is A–R–G — three coupled forms of sharing (German: Arbeit, Risiko, Gewinn):

Letter	German	English	What is shared
A	Arbeit	Labour	Division of labour — who produces what, specialised roles, coordinated output
R	Risiko	Risk	Division of risk — who bears production, credit, and settlement uncertainty
G	Gewinn	Profit	Division of profit — how surplus, once validated by demand, is allocated

A (labour sharing): No household produces everything it consumes. Work is specialised, and coordination runs through contracts, wages, and the credit that pre-finances production before any sale returns.

R (risk sharing): Risk never disappears — it is absorbed up a hierarchy. Workers and suppliers prefer stable income and pass production risk to entrepreneurs; entrepreneurs pass credit and liquidity risk to banks; banks pass systemic risk to the central bank. Interest is the insurance premium for that absorption, not a rent on idle metal.

G (profit sharing): Profit appears only when invested production meets real demand and closes the cycle. In the market reading it is the return to those who pre-financed and bore the risk — a risk premium, not evidence of exploitation. But profit is more than a risk reward: it is a share of the common product created by the division of labour, and how that share is cut is a separate, explicit institutional choice.

And here MoMaT makes a claim we do not make lightly. For the first time, profit sharing has an explicit algebraic handle — not only as the reward for risk absorption inside a firm or up the banking hierarchy, but as a calculable share of the common product of shared labour. The accounting layer is built as the coproduct of three Lawvere theories, T_DEB + T_Val + T_QEB: double-entry debit/credit (T_DEB), mark-to-market valuation (T_Val, quantity × price), and the bilateral quadruple-entry bookings (T_QEB). Crucially, equity splits as EK = FE + RE: Financial Equity, a model of T_DEB that stays invariant under any change of valuation (the gauge invariance of the books), and Real Equity, a model of T_Val that depends on the chosen valuation functor.

Distribution lives entirely in the valuation theory and is orthogonal to the settlement invariants. That is what makes the old "dividing the cake" problem — the socialist and Marxian question of sharing the social product, GDP — a well-posed, computable object rather than a slogan. Different valuation regimes (historic cost, fair value, market prices, even plan prices) are different models of the same theory T_Val; each is a different cut of the same cake without enlarging or shrinking it. The GDP split itself is an aggregation–disaggregation adjunction (Σ ⊣ Π) with an explicit distribution vector whose shares sum to one. So the classical dream of sharing the common product is expressible in MoMaT: as an explicit, auditable valuation choice on books that balance under any gauge. MoMaT does not decree the split — it makes the split calculable, bookable, and open to institutional design. (Worked out in the distribution chapter of the monograph.)

Capitalism, on this reading, is the machinery that makes A–R–G composable, bookable, and learnable — from a household ledger to central-bank settlement. Markets are one expression of it: voluntary exchange as giving to receive, routed through typed debt and accounting invariants.

The capitalist cycle as hylomorphism

What ties A–R–G together over time is a hylomorphism — the composition of an unfold and a fold. In the semantics of programming languages and category theory (familiar ground for Michael's Haskell readers):

Side	Categorical object	Operation	Standard name	Economic role
Coalgebra	γ : S → F S	unfold	anamorphism ⟨γ⟩ : S → νF	Investment — open the future
Algebra	α : F A → A	fold	catamorphism (α) : μF → A	Demand — close against reality

Unfold (coalgebra / anamorphism). A coalgebra γ sends a current state S to a one-step extension F S — a possible next shape. Iterating γ unfolds a tree of futures. In MoMaT this is credit-financed investment: from today's balance sheet, contracts and credit entail wages, inputs, production steps, and payment obligations branching forward in time. Production stays speculative until it is tested. Marx's G → W → P (money → commodities → production) lives on this side.

Fold (algebra / catamorphism). An algebra α sends a composite F A back to a result A — it collapses structure into a value. In MoMaT this is demand and settlement: sales revenue, profit or loss, debt repayment, and cash settlement fold the production tree back onto the books. Marx's W' → G' (realised commodities → more money) closes the arc. Demand validates investment — ex post, never by decree.

Hylomorphism h = (α) ∘ ⟨γ⟩ is the full loop: unfold a promise into the future, then fold reality back onto it. Failed validation writes off (the fold returns a loss); successful validation closes the books and funds the next unfold. That is how the system learns, cycle by cycle. The wage is paid before the sale (unfold); the sale decides whether the unfold was justified (fold). Risk sharing (R) lives between the two phases; profit sharing (G) is the reward of a successful fold.

Michael's masach — deferring immediate receiving — has a structural echo here: the unfold defers gratification, the fold defers judgment until demand speaks. But MoMaT reads this as mechanics, not moral psychology. It is the typed chronomorphism of a credit economy: promises into the future, validated when settlement arrives.

MoMaT top-level structure: AccCat, DecCat, GovCat

Above A–R–G and the hylomorphism sits MoMaT's top-level architecture — three coupled categories that share one topos (a single institutional world, not three siloed tools). The triangle below is the canonical MoMaT picture; Magic Sauce shows how it is glued together from local open games. Related research maps it onto Ostrom's three institutional layers, and the engine preview renders the same stack as an interactive 3D scene.

The global Topos triangle — Governance, Decisions, Accounting (GovCat / DecCat / AccCat) — coupled by three adjunctions. Local open games compose along their interfaces to build this institutional geometry; see Magic Sauce for the full picture.

Category	Plain name	What it holds	Economic question
AccCat	Accounting	DEB/QEB, T-accounts, settlement, gauge invariance	Do the books balance? — unconditional substrate
DecCat	Decisions	Open Games, strategies, forecasts, play and coplay	What do agents do? — investment, pricing, demand
GovCat	Governance	Contracts, deontic rules, mandates, compliance	Which moves are admissible? — the rulebook

AccCat is the substrate. Quadruple entry and cash as terminal settlement are not optional add-ons to price theory — they are the coordinate system every behavioural theory has to pass through. Nothing in DecCat or GovCat can override arithmetic.

DecCat is the field of play. Firms, banks, and households act as compositional open games — locally egoistic lenses (Smith's bakers and butchers) that learn from forecast errors via Sim ⊣ Est against the realised accounts in AccCat.

GovCat is the rule layer. Loan, wage, and settlement contracts type which games and postings are allowed to run at all. Institutions learn over their rules through Constrain ⊣ MechDesign with DecCat — mechanism design meeting constitutional constraint.

Two adjunctions — and the third they imply

The triangle carries two genuine learning loops, one fast and one slow. The third edge — GovCat ↔ AccCat — is not a third loop but their composition: the round trip that makes audit well defined.

Adjunction	Edge	Clock	What moves
Sim ⊣ Est	DecCat ↔ AccCat	Fast (operational)	Decisions become bookings; realised accounts drive the next decision — plan vs actual under fixed rules
Constrain ⊣ MechDesign	GovCat ↔ DecCat	Slow (systemic)	Contracts define admissible play; persistent gaps revise the rules — covenants, mandates, mechanism design
Pull ⊣ Push	GovCat ↔ AccCat	Composite (audit)	Implied by the two above — one full round trip, both directions

Top-down (Push): Gov → Dec → Acc. A rule becomes a decision, a decision becomes a booking — each step adds detail (covenant → journal entry). Bottom-up (Pull): Acc → Dec → Gov. Books roll up into decisions and then into a governance verdict — each step summarises and drops detail. DecCat sits inside; GovCat and AccCat face the boundary — contracts on one side, cash and balances on the other.

The GovCat ↔ AccCat edge is that same chain read as a single arc:

1. Down: contracts mandate what must be booked — what the rule layer requires to appear in the ledger.
2. Up: the books reveal whether compliance was achieved — and here the audit gap is measured.

That gap — what contracts require booked versus what is booked — surfaces only on this Gov–Acc round trip. It is the core audit quantity, and it decomposes into two causes:

• Wrong execution (right rules, wrong decisions) → fix the operational loop (Dec ↔ Acc).
• Wrong rules (rules that cannot be executed, or no longer fit reality) → fix the systemic loop (Gov ↔ Dec).

One number, two causes. Audit is then not a separate department sampling vouchers after the fact; it is the composed adjunction pointing to where the institution must learn next. It also settles what data to gather between governance and accounting: not everything twice, but the mandated booking set (from Gov, pushed down) against the realised posting set (from Acc, pulled up), with DecCat as the channel where execution variance shows up.

Michael's macro/micro split has a direct homologue: selfish play in DecCat, rule-bound settlement in AccCat, admissibility in GovCat. His worry about coercion is a GovCat pathology — forced identification of agents, rules that cannot glue local action to global demand. His idealised market is a DecCat achievement under GovCat rules (voluntary contract, fair courts), with outcomes booked in AccCat — and the audit loop is where you see whether the three still compose.

Elinor Ostrom's constitutional / collective-choice / operational stack is the same trinity under another name — see the table in Related research. Compositionality across the three layers is exactly what lets polycentric governance work without a single planner.

Where MoMaT agrees

Want as the starting point

Michael's ratzon lekabel matches our core dynamic once you strip the theology: agents want, scarcity persists, and demand validates investment. The capitalist cycle is the hylomorphism above — credit and contracts unfold (coalgebra), sales and settlement fold (algebra). Failed validation writes off; successful validation closes the books and funds the next cycle. Want is not a moral defect but the driver the institutional geometry has to absorb.

Free markets as "giving to receive"

Michael's idealised market — no coercion, voluntary trade, honest information — maps cleanly onto the A–R–G core: labour is specialised (A), risk is borne by those who pre-finance (R), and profit rewards validated production (G). His apple seller and our baker are the same figure: the act is giving, the intent stays self-interested, and the system channels that interest productively through the A–R–G machinery.

We say the same in categorical language: Smith's invisible hand as local egoistic lenses in DecCat, and capitalism as a decentralised macro-learning system — agents update from forecast errors while institutions adapt on two levels, within rules and over rules in GovCat.

Selfishness is channelled, not abolished

Michael is explicit: charity and volunteering in market societies are rarely "pure" in the Kabbalistic sense — status, tax relief, and feeling good all count. Markets do not try to eliminate selfishness; they redirect it.

MoMaT does not moralise this away either. The three-category stack replaces intent-typing altogether: AccCat asks whether the books balance, DecCat models what agents do strategically, GovCat holds the rules. The question is never purity of intent but whether demand validates investment and who bears which risk in the hierarchy (firm → bank → central bank).

Socialism: right goal, wrong mechanism

This is the strongest overlap.

Michael argues that socialism aims at giving-to-give but delivers receiving-to-receive: people "give" at gunpoint. Scarcity remains, "need" stays undefinable, production incentives weaken, and status competition simply moves to non-monetary channels.

We reach the same conclusion structurally rather than mystically. Central planning fails as a knowledge and coherence problem — Hayek's point, which in our framework is a sheaf pathology: local information does not glue into a global plan. When a central organ cannot see real demand, even well-intentioned intervention misallocates — the historical "ton ideology," where the plan counts tons instead of useful nails. Central-bank liquidity is necessary but never sufficient; production must still find real demand. Forced identification of agents in the macro game is the categorical counterpart to Michael's coercion diagnosis.

Capitalism is not intrinsically evil

Michael does not demonise markets. They are productive but, in his framing, do not fulfil "the purpose of Creation."

We push further toward demoralisation: capitalism is the A–R–G system — labour, risk, and profit sharing made explicit in the books. Exploitative systems are built by people, not by the mathematics of accounting. Distribution (G) and stability (R, settlement) are orthogonal in MoMaT: once quadruple entry and gauge invariance are in place, the split of the common product becomes an explicit, computable valuation choice — the dividing-the-cake question made well-posed — rather than something smuggled inside a single word, "money."

Where MoMaT diverges

Hard money

Michael's ideal market assumes hard money — money that privileged parties cannot create cheaply. He ties the masach to low time preference and faults inflation for pushing toward immediate consumption.

We argue the opposite on structural grounds. Gold- and scarcity-backed regimes live in a finite initial algebra: reserves deplete and stabilisation turns brittle (gold-depletes in our stability proofs). Fiat systems, by contrast, are a final coalgebra of the central bank: liquidity can be emitted to meet settlement need — the institutional reading of "whatever it takes." We share Michael's worry about privileged cheap money creation; our answer is institutionalised central-bank elasticity under accounting invariants, not a retreat to commodity scarcity as the stability anchor.

Interest is not time preference

Michael treats saving as low time preference — deferring consumption, "giving to your future self."

In MoMaT, interest is not time preference. Interest payments are insurance premia for risk absorption along the hierarchy: firms pay banks for portfolio and liquidity risk, banks pay the central bank for lender-of-last-resort insurance. Deferral (Verzicht) does appear — but in the hylomorphism (invest now, validate later), not as a psychological screen laid over desire.

Inflation

Michael treats inflation as something that discourages holding cash and encourages near-term consumption.

We treat it primarily as a valuation phenomenon — input prices, wages, resources — read through the valuation algebra, not as a purely monetary phenomenon in the Friedman sense. Central banks do not "control inflation" by rationing reserves alone; taking every price movement for a liquidity problem is itself a category error.

Moral intent vs typing

Michael's four levels are ethical: action and intent, pure and mixed motivation.

MoMaT replaces that grid with typing: giral money, reserves, and cash (G → R → C); claims versus terminal settlement; quadruple entry across paired double-entry systems. Charity may be impure in Michael's sense — for us the only diagnostic is whether the catamorphism of demand validates the anamorphism of investment, and whether the institution learns from the result.

Teleology

Michael's telos is selflessness and affinity with the divine. Ours is Oikos: continuity, learning, institutional coherence — Aristotle's household law made reflexive through accounting. The guiding motif of our program is a place where want and reality speak the same language, reached when dreams have become bookable — validation through demand and balance sheets, not refinement of the soul.

Charity is a public good — the missing Ostrom step

This is where we think Michael's argument takes a wrong turn — and where MoMaT quietly supplies what he is reaching for.

Michael wants giving-to-give: charity, selflessness, care beyond self-interest. In the language of economics that is not a private good traded in a market — it is a public good, non-rival, hard to exclude, and chronically under-provided by self-interested exchange. His own framing half-concedes it: markets are the macro engine for selfishness, while giving has to be cultivated separately in the micro.

Yet he still asks the free market to carry giving — "giving to receive." That is precisely the step an economist would not take. Markets allocate private goods well and provide public ones poorly; the free-rider problem is not a moral failing but the incentive geometry. Pure market dynamics under-provide charity for the same reason they under-provide clean air or flood defence.

The economist's answer to public goods is neither "let the market do it" nor "let the state command it." It is Elinor Ostrom's answer: polycentric governance — many overlapping centres of authority, local rules that compose, with neither a single market price nor a single planner in charge. Ostrom showed empirically that commons and shared provision are sustained this way, in the space between market and state.

And Ostrom's institutional stack is already MoMaT's stack:

Ostrom rule level	MoMaT category	Role
Constitutional	GovCat	who may set rules — mandates, charters, deontic norms
Collective-choice	DecCat	strategic choice of policy and mechanism within those rules
Operational	AccCat	day-to-day action and the bookings it leaves behind

This is not an analogy imposed after the fact; it is the same three-layer geometry — governance, strategic decision, operational accounting — that A–R–G and the audit loop already run on. See the mapping in Related research.

So the argument closes on itself. Polycentrism is what actually delivers Michael's charity — public-good provision between market and state. But polycentrism only works if the three layers compose: overlapping jurisdictions are coherent only when GovCat, DecCat, and AccCat glue level by level (household → association → municipality → state), and only when the same stack carries both private market production and public-good production on one set of books. That compositionality is exactly what MoMaT provides — the sheaf-gluing of local open games into a global institution, shown in Magic Sauce.

Read this way, MoMaT does what Michael wants — but by the route he rules out. Giving-to-give is not something free markets secrete as a by-product; it is a public good that has to be institutionally engineered, and the engineering is polycentric composition of exactly the governance / decision / accounting layers MoMaT makes formal. His telos survives; only the mechanism changes — from "the market will provide" to "compose the institutions that provide."

One sentence each

Michael: Markets are the right macro engine for selfishness; selflessness is micro-work; state coercion perverts the goal.

MoMaT: Agreed on the engine — but it runs on typed debt-accounting with a fiat liquidity coalgebra, not hard money, and giving-to-give is a public good, delivered by Ostrom-style polycentric composition of governance, decision, and accounting rather than secreted by markets; the decisive question stays whether demand validates investment and whether institutions learn.

Understand it first, then use it for good

There is a bigger point underneath all of this. Capitalism is one of the great achievements of humanity — a sophisticated machine for sharing labour, risk, and profit at planetary scale, learning cycle by cycle from whether demand validates investment. Treating it as intrinsically evil is as unserious as treating it as automatically just. Both moves skip the only step that matters: understanding what the machine actually is.

Most of the disappointment with markets comes from asking them to deliver what markets structurally cannot — public goods, charity, kindness, the giving-to-give that Michael rightly prizes. Markets are magnificent at allocating private goods and hopeless at provisioning public ones; blaming them for the second is like blaming a ledger for not being a prayer. Stop bashing capitalism for failing at a job it was never the right tool for, and a different possibility opens up: keep the engine for what it does superbly, and compose it — polycentrically, à la Ostrom — with institutions that provide what markets cannot.

That is the whole point of MoMaT. First understand capitalism precisely — not as sin and not as salvation, but as a typed, bookable, learnable architecture of labour, risk, and profit sharing, with distribution a computable choice rather than a hidden fate. Then, on that understanding, design it for the purposes we actually want it to serve. Once the books balance under any valuation gauge, once the three layers compose, once distribution is an explicit dial rather than a buried assumption, kindness and charity stop being sentiments we hope survive the market and become institutions we can engineer on the same foundations. Understand the achievement first; then use it for good. MoMaT is built to do exactly that.

Further reading

• Michael Snoyman — Kabbalah and Economics
• OiC.OS — Magic Sauce — Topos triangle and local open games (the TikZ figure above)
• OiC.OS — Engine preview — interactive 3D view of the same three layers
• OiC.OS — Money Theory — shortest summary of the axiomatic framework
• OiC.OS — Technology — three layers for CFOs and treasurers
• OiC.OS — Related research — Ostrom ↔ GovCat / DecCat / AccCat
• OiC.OS — MoMaT — the research program and open questions
• OiC.OS — Axiomatik der Ökonomik (PDF) — working draft of the German monograph

Comments and collaboration: contact@oicos.systems.