The Zcash ArboretumIV Consensus Guide
Contents
PDF
Theme

7 The economics of the attack

The probabilities of Section 5 assume that the customer is an attacker. Following the paper’s §6, this section derives the payment value below which the attack has non-positive expected profit (Proposition 7.1) and tabulates it in units of block income. Throughout, q<p (“[o]therwise, all bets are off”, §6); the model is deliberately stylised, and Remark 7.3 records its caveats.

The attacker targets k merchants simultaneously with payments of v each, all invalidated by the same private branch; the goods are worth αv to him, 0<α≤1; he abandons the attack after mining o blocks without success; and each of his blocks would earn the block income B if it were included in the accepted chain. If the attack succeeds, his blocks are all accepted, and he keeps the goods, the coins and the rewards. If it fails, he has paid kv, keeps goods worth kαv, and his o blocks, excluded from the accepted chain (orphaned), forfeit oB. An exact profit calculation for this stopping rule would take r to be the success probability of that rule; following the paper, the table below substitutes the unlimited-horizon r(q,n) as a heuristic proxy, and so does not evaluate exactly a strategy that stops after o=20 private blocks.

Proposition 7.1 (Profitability; Rosenfeld §6).

Given success probability r and the gains and losses stipulated above, the attacker’s expected profit over not attacking is

k⁢α⁢v−(1−r)⁢(k⁢v+o⁢B)=k⁢v⁢(α+r−1)−(1−r)⁢o⁢B,

positive if and only if

v>(1−r)⁢o⁢Bk⁢(α+r−1)

(when α+r>1; when α+r≤1 the modelled attack is never strictly profitable). The modelled attack is not strictly profitable whenever

v≤o⁢(1−r)⁢Bk⁢(α+r−1)⁢=α=1⁢ok⋅B⁢(1−r)r, (3)

which with the paper’s choices o=20, k=5, α=1, B=25 BTC is its equation (2), v≤100(1/r−1) BTC.

Proof.

The gain kαv is certain (the goods are obtained either way); the loss kv+oB is paid exactly when the attack fails, with probability 1−r. Rearranging the positivity condition gives the threshold; substituting the paper’s constants gives 20⋅25⋅(1−r)/(5r)=100(1−r)/r. □

7.1 Reward-normalised exposure

The block subsidy is value issued by the protocol in a block, distinct from fees transferred by its transactions. Its scheduled component is base issuance determined by height, before any reserve payout. A funding stream is a required allocation of a share of that subsidy to a designated recipient. The coinbase is the block’s first transaction, which distributes the available subsidy and fees according to the applicable rules. The reserve is the accounting balance of funds removed from circulation and awaiting reissuance.

The income B forfeited by an orphaned block is the amount its miner would have retained: miner subsidy plus retained transaction fees, not the subsidy allocated to other recipients. Under NU7, from activation until the scheduled subsidy next halves, the miner subsidy is 41,666,668 zatoshi per block (Section 8.7), and the miner keeps the fees not removed from circulation, 40% rounded in the miner’s favour; once reissuance begins, the subsidy also depends on the prior reserve balance (Section 8.8). A universal ZEC figure would therefore hide assumptions about fee income and chain state.

Instead, Table 4 measures transaction value in units of the chosen block income B. With the paper’s o=20, k=5, and α=1, equation (3) becomes

vB≤4⁢1−rr.

Once a particular threat model supplies B, multiplying by it gives the corresponding monetary threshold. Treating B as constant over the attack remains a modelling assumption.

q n=1 2 4 6 8 10
5% 36 271 10,327 344,743 10,931,506 336,737,801
10% 16 67 729 6,759 59,475 508,917
20% 6 15 55 167 467 1,262
30% 2 5 11 21 35 57
40% 1 1 2 4 5 6
Table 4: The largest whole-number multiple of block income B with non-positive heuristic expected profit: ⌊4(1−r)/r⌋, computed by script. The table is dimensionless: it neither fixes a ZEC subsidy nor predicts future fee income. Values inherit every assumption of the model; the caveats of Remark 7.3 apply in full.

Within the model, a few confirmations make the attack unprofitable for an attacker with a small share at any plausible payment size; against a 30% attacker, six confirmations make modelled expected profit non-positive only up to about 21 times the assumed block income B. By the geometric decay of Theorem 5.2, the required count grows only logarithmically in the value at stake. These figures are neither exact finite-stop profit calculations nor bounds on deployed rollback risk; changing difficulty and rollback rails require separate modelling.

Remark 7.2 (Majority is a different problem).

At q=p, eventual strict overtake has probability 1, but the walk has zero drift; that fact alone does not give the attacker a lasting monopoly. When q>p, the profit analysis does not apply: the paper notes that a strict majority attacker can double-spend at no cost beyond normal mining, reject every other miner’s blocks to take the whole issuance, and exclude transactions at will, driving honest miners out and entrenching his majority. A majority attack therefore “is best seen as an attempt to destroy” the currency, motivated from outside it (a short position, a competing system), not an attempt to profit inside it. No confirmation policy defends against it; the defence is the cost of assembling the majority, which lies outside this model.

Remark 7.3 (The model’s own caveats).

The paper states each limitation of its model. The give-up point o=20 “is a significant simplification” (an optimal attacker balances completion against compounding losses); the safe values “should be taken with a grain of salt, because of the many modeling assumptions”; and a model resting on mining costs rather than forfeited rewards would make the required confirmations linear, not logarithmic, in the transaction value — “very poor security, hence in a situation where this is relevant, we have already lost anyway” (§6). The bound is significant for its form, a required confirmation count logarithmic in the value at stake, not for its numerical precision.