What Cardano's Biggest Wallets Actually Paid for Their ADA

Ada · CardanoScope On-Chain Research Data cutoff: 2026-07-14 (UTC)

Research Question

ADA closed at $0.1654 on July 14, 2026 — roughly a quarter of where it traded a year earlier. Every time the price falls another leg, the same question resurfaces in Cardano's discourse: are the wallets holding the largest positions accumulating through the decline, or are they sitting on the same losses as everyone else? Top-holder lists answer "how much do they hold." They don't answer "what did it cost them." This piece uses CardanoScope's tier-level cost-basis model — built directly on db-sync UTXO reconstruction, not a third-party estimate — to answer the second question.

Executive Summary

• As of July 14, 2026, CardanoScope's moving-average cost model puts the blended entry price at $0.594 for mega wallets (≥10M ADA), $0.483 for whale wallets (≥2M ADA), and $0.455 for shark wallets (≥1M ADA) — all three tiers are underwater at the current $0.1654 spot price, by 72%, 66%, and 64% respectively.
• Over the past 12 months, mega and whale wallets grew their aggregate ADA holdings (+8.4% and +5.2%) while their blended cost basis *declined* (-3.4% and -5.5%) — consistent with continued buying through the price collapse, not just holding static balances.
• The shark tier moved in the opposite direction: both wallet count (-7.8%) and total ADA held (-7.1%) shrank, yet its blended cost basis barely moved, ticking up 1.3% instead of falling — a pattern the accumulation story alone doesn't explain.
• The model is a blended tier-level average, not a per-wallet ledger. It cannot distinguish real accumulation from a wallet migrating between tiers, and it should not be read as anyone's actual tax basis.

How the Model Works

CardanoScope classifies stake addresses into three fixed floors by live ADA balance: shark (≥1,000,000 ADA), whale (≥2,000,000 ADA), and mega (≥10,000,000 ADA). This is the same floor structure used by the transfer-tier tracker, applied here to holdings instead of single transfers.

The cost-basis series (whale_cost_basis_daily, model version tier-moving-average-v1) is a moving average, not a snapshot of any individual wallet's purchase history:

• When a tier's aggregate balance *rises* day over day, the added ADA is valued at that day's ADA/USD close (sourced from Coinbase's exchange feed, price_source: coinbase_exchange) and blended into the running average, weighted by the ADA added.
• When a tier's aggregate balance *falls*, the model does not revise the average downward — the remaining balance keeps the prior blended cost, the same convention as a weighted-average-cost ledger where a partial disposal doesn't change the recorded cost of what's left.

Both the holdings series and the cost series are historical reconstructions, not just live scans. CardanoScope's backfill script reconstructs the UTXO set as it existed at a past block height directly from db-sync's tx_out/tx_in history — outputs created before the cutoff block, minus outputs already spent before it — rather than approximating past balances from the current chain state. That is what makes a 2026-vs-2025 comparison meaningful instead of guesswork.

What the Data Shows Right Now

| Tier | Floor | Wallets | Total ADA held | Est. cost basis | Spot vs. cost | |---|---|---|---|---|---| | Mega | ≥10,000,000 ADA | 452 | 20.41B | $0.594 | -72% | | Whale | ≥2,000,000 ADA | 1,041 | 4.17B | $0.483 | -66% | | Shark | ≥1,000,000 ADA | 1,386 | 1.81B | $0.455 | -64% |

All three tiers are sitting on unrealized losses on a blended-average basis at the current $0.1654 spot price. That alone isn't surprising in a year where ADA fell more than 75% — the more informative question is how each tier's numbers moved to get here.

Twelve Months Ago vs. Today

A year earlier (July 10, 2025), ADA traded at $0.6776 — roughly 4.1x today's price. Comparing that snapshot to July 14, 2026:

Mega wallets (≥10M ADA): wallet count fell from 497 to 452 (-9.1%), while aggregate holdings rose from 18.83B to 20.41B ADA (+8.4%). The blended cost basis declined from $0.615 to $0.594 (-3.4%). Fewer wallets, holding more ADA between them, at a slightly lower average entry price — consistent with continued buying concentrated among a shrinking, larger-positioned group, though the data can't say whether that's the same wallets growing or smaller mega wallets being absorbed or reclassified.

Whale wallets (≥2M ADA): wallet count was roughly flat (1,025 → 1,041, +1.6%), aggregate holdings rose from 3.96B to 4.17B ADA (+5.2%), and the cost basis fell further than mega's, from $0.511 to $0.483 (-5.5%). This tier shows the clearest accumulation-at-lower-prices signature of the three.

Shark wallets (≥1M ADA): this tier moved differently. Wallet count fell from 1,504 to 1,386 (-7.8%) *and* aggregate holdings fell from 1.95B to 1.81B ADA (-7.1%) — the only tier that shrank in both dimensions. Yet its cost basis rose slightly, from $0.4496 to $0.4553 (+1.3%), instead of falling like the two tiers above it. Over the same period the metric moved in a narrow $0.44-$0.47 band rather than tracking the price collapse in either direction. A shrinking tier with a stable-to-rising blended cost is not what a straightforward "smaller holders capitulated at a loss" story would produce — that would pull the average down, not up. What's more likely at work is a mix of tier migration (shark wallets crossing into whale as they grow, or dropping out entirely) and continued smaller-scale buying offsetting exits. The model can't separate those effects from the aggregate series alone.

What This Can't Tell You

This is a tier-level analytical estimate, not a ledger of any individual holder's real cost basis, and it isn't designed to be one. Three limitations matter for reading it correctly:

• Tier migration is invisible to the model. A wallet growing from shark into whale territory looks like a shark-tier "exit" and a whale-tier "entry" priced at that day's close — even though no sale happened. This can shift a tier's blended average without any wallet actually trading.
• It can't separate accumulation from custody reshuffling. Exchange or custodial wallets crossing the same balance floors are counted the same way as a long-term holder's wallet. CardanoScope's address-label data only identifies a subset of known exchange addresses.
• Historical cadence isn't uniform. Snapshots are close to weekly for older history and daily for the recent window, so short-term moves in older periods are smoothed more than recent ones.

None of this changes the headline: on CardanoScope's own reconstructed on-chain data, every tier of large ADA holder is currently underwater on a blended basis, and the two largest tiers got there while still adding to their positions through the decline. The smallest of the three tiers is the outlier, and the data raises the question without fully answering it.

The full interactive chart — including 1-month, 1-year, and 5-year ranges across all three tiers — is live on CardanoScope's Whale Tracker page.

--- Sources:

• CardanoScope Whale Tracker — live holdings and cost-basis chart
• CardanoScope Data Limitations — methodology and coverage notes

*[email protected]*