How to Value an Airdrop Points Program

Somebody is going to tell you their points are worth $12,000. Ask how they got the number and you will usually find one assumption doing all the work, and that assumption is a guess about a token price that does not exist yet.

There is a defensible way to do this. It does not produce a number — it produces a range with the assumptions written down, which is the only honest output when four of the five inputs are unknown. Here is the framework and a worked example.

The four terms

## expected value of a points position
EV = share * allocation * valuation * discount

share      = your points / total program points at the snapshot
allocation = tokens distributed to the program, as a fraction of supply
valuation  = fully diluted value of the token at distribution
discount   = time value, vesting, and probability the program pays at all

Every term after the first is a guess. Write each one down as a range before you multiply, and never collapse a range into its midpoint until the very end.

Term 1: your share

You know your points balance. You do not know the denominator, and the denominator is growing.

This is the term people get most wrong, because they compute it from today's numbers. If the program runs for another twenty weeks, total points at the close will be far larger than total points today — and if the earn rate is pool-divided, your ongoing accrual is also getting worse as more participants arrive.

Do it properly:

• Start from your current share of the weekly issuance, not your share of cumulative points. The weekly number tells you what you are earning now; the cumulative number is contaminated by an early period that may no longer be representative.
• Project participation growth. Crudely is fine — flat, doubling over the remaining season, tripling. The point is to make the assumption visible.
• Model your own consistency. A share computed from a heavy week you will not repeat is a fantasy.

The output is a range for your final share. A trader earning 0.10% of current weekly points, in a program that will double in size before it closes, is realistically looking at 0.04% to 0.07% of the final total — not 0.10%.

Term 2: the allocation

What fraction of supply goes to this program? This is unknowable until announced, and comparable programs vary widely. A first-season community distribution somewhere in the range of a few percent to the mid-teens of supply is a reasonable band to work with, and the correct thing to do is state whatever band you use as an assumption rather than pretending you know.

Two adjustments that matter more than refining the band:

• Your program may be one of several. If the venue runs seasons, this season's allocation is a slice of the community allocation, not all of it.
• Not all of it reaches traders. Sybil filtering, wash-trading exclusions and minimum thresholds all shrink the distributed set. If your activity is ordinary this is a modest haircut; if you were manufacturing volume to farm, it can be the entire position.

Term 3: the valuation

The genuinely wild term. A pre-launch token has no price, and the plausible range at listing spans at least an order of magnitude for anything that is not already a household name.

Do not anchor on the best comparable. The set of points programs you have heard of is selected on success — the ones that failed, delayed indefinitely, or launched into a bad market are not the ones that generate threads. Anchoring on the visible winners overstates the median outcome, sometimes by a lot.

Use a band, and put the low end somewhere genuinely uncomfortable. A range of 10× from low to high is not pessimism, it is an accurate description of what you know.

Term 4: the discount

Three separate haircuts, multiplied:

• Time. A distribution twelve to eighteen months out is worth less than the same distribution today. Apply a discount rate that reflects what you could do with the capital and attention in the meantime — in crypto, that is not a treasury rate.
• Vesting. Tokens that unlock over a year are not liquid at listing. Value the vesting portion at a discount to the spot price you assumed, because the price you can actually realise is the one at unlock.
• Probability the program pays at all. Programs get restructured, delayed, folded into later seasons, or quietly retired. This is not a small probability and it should not be modelled as one.

Worked, with the assumptions visible

A trader with 0.06% of projected final program points. Assumptions stated:

Low case: 0.0004 × 0.04 × 150,000,000 × 0.35 = $840 High case: 0.0007 × 0.12 × 1,500,000,000 × 0.75 = $94,500

That is the honest answer: somewhere between roughly $800 and roughly $95,000, a spread of more than a hundred to one. The midpoint of that range is not a forecast — it is an artefact of arithmetic on numbers nobody knows.

The error bars exceed the expected value. That is not a failure of the model, it is the correct result, and it is the single most useful thing the exercise produces. Anyone quoting you a point estimate has simply chosen not to show you this table.

What to do with a range like that

The range is not useless. It answers a narrower question extremely well: should I pay for this?

• If you are earning points incidentally, from trading you would do anyway on a venue you chose on execution quality, the position is a free option. Free options with 100:1 outcome ranges are excellent things to hold. Hold it and forget it.
• If you would have to pay for it — extra turnover, worse fills, capital parked at a venue you would not otherwise use, fees you would not otherwise incur — then compare your certain cost against the low end of the range, not the midpoint. Trading $500,000 of extra notional at 6.5 bps costs $325 with certainty. Against a low case of $840 that may be defensible; against a low case of $80 it is not.
• If your low case is near zero, treat the position as zero in every comparison you make. That is the defensible default and the reason points belong in the upside column, not the income column.

Three failure modes to avoid

Anchoring on a headline outcome. "Program X paid $50,000 to average users" is a fact about program X, selected for being memorable. It is not a prior for your program.

Compounding optimistic assumptions. Four terms, each set to the friendly end, multiply into an answer that is wrong by orders of magnitude. If you cannot state each assumption out loud without wincing, do not multiply them together.

Ignoring what you spent. The certain cost of farming — fees, spread, funding, opportunity cost of parked capital — is measurable today and is usually the only number in the entire exercise that you actually know. Treating it as free because it was "activity" and not "a purchase" is precisely the error the program is designed to induce.

The short version

EV = share × allocation × valuation × discount, with every term after the first a stated range. Project your share against a growing denominator, band the allocation and valuation honestly, discount for time, vesting and the chance the distribution never comes, and multiply the low ends and the high ends separately.

You will get a range spanning one to two orders of magnitude. That is the true answer. Hold points as a free option when they cost you nothing, and price them against the low end whenever they cost you something real.

More structure in exchange points programs, compared; more on putting it beside your fee and campaign lines in how to compare crypto exchange rewards. And for a reward whose value you can actually compute today, a live competition with a stated pool and a visible field is a considerably easier calculation.