Single-coin evidence report · By CoinNudge Research · · Guide updated 2026-10-03 · 20 min read
What is Phala (PHA) coin, and is it worth holding?
Phala (PHA) coin is associated with confidential-computing infrastructure, not a share of Phala Cloud revenue. Market evidence ends at the 2026-09-26 UTC daily open; supply and operating records have separate dates below. PHA’s last dated 90-day spot path gained +183.9% against BTC +40.1%, yet the median of 58 earlier 90-day month-starts was −13.3% and only 14/58 beat BTC. This is a sharp current rally against a much weaker historical distribution, not proof that confidential-AI adoption now pays token holders. Phala reported 398 paid Cloud users and 813 running CVMs at 2025 year-end; it did not publish an aligned PHA-settlement series, so paid product usage is not yet verified token capture.
See the research evidence · Check live market context
Quick answer: PHA coin in three checks
- Phala (PHA) coin is associated with confidential-computing infrastructure, not a share of Phala Cloud revenue. Market evidence ends at the 2026-09-26 UTC daily open; supply and operating records have separate dates below. PHA’s last dated 90-day spot path gained +183.9% against BTC +40.1%, yet the median of 58 earlier 90-day month-starts was −13.3% and only 14/58 beat BTC. This is a sharp current rally against a much weaker historical distribution, not proof that confidential-AI adoption now pays token holders. Phala reported 398 paid Cloud users and 813 running CVMs at 2025 year-end; it did not publish an aligned PHA-settlement series, so paid product usage is not yet verified token capture.
- Historical coverage: 1852 daily opens; 58 complete 90-day monthly starts. Monthly starts overlap and are descriptive, not a future win rate.
- Token-specific question: Does real confidential-compute usage create demand for PHA faster than network reward issuance?
What does Phala’s confidential AI network actually provide?
Phala builds confidential-computing infrastructure using trusted execution environments. PHA is a network utility and governance token, but an AI or cloud-compute announcement is not a measurement of paying PHA demand.
The investor question is: Does real confidential-compute usage create demand for PHA faster than network reward issuance? This report tests what the token does, what it does not entitle a holder to, how supply can change and what actual dated spot paths showed. Price history is not used as proof that product use caused the move.
| PHA research check | Observed or documented answer | Evidence boundary |
|---|---|---|
| Token and project | Phala builds confidential-computing infrastructure using trusted execution environments. PHA is a network utility and governance token, but an AI or cloud-compute announcement is not a measurement of paying PHA demand. | Business success and tokenholder benefit differ |
| Available local/venue daily opens | 1,852 from 2021-09-01 to 2026-09-26 | local archive, potentially retrospective backfill |
| Current chart | 90 matched UTC days vs BTC | Historical opens, not a live quote |
| Provider-estimated circulating supply | 857,048,787 | CoinGecko 2026-09-27T02:55:20.000Z |
Verify: Phala Network official overview · Phala 2024 network and token report · Phala 2025 network report
Which Phala Cloud and worker payments require PHA?
This release does not establish which paid Cloud receipts required PHA or how much PHA customers acquired. Phala’s current network documentation describes TEE-based confidential computation and Phala Cloud deployments. It is important to distinguish provisioned compute, running compute, paid jobs and fees actually settled in PHA. Only the last flow can directly test a token-demand mechanism; a benchmark, partner announcement or deployed app count is not a substitute.
The project’s reports describe a shift toward an Ethereum-based architecture. Migration can alter where staking, governance, mining and assets reside. A historic PHA/USDT price series remains a quoted-holder experience, not a constant-protocol-economics experiment across all dates. The report therefore does not attribute old price episodes to the current cloud design.
The owner of a token, user of a product, operator earning incentives and shareholder of a company can be different parties. The table below keeps their claims separate instead of transferring a product benefit to every PHA wallet.
Align paid workload receipts and PHA settlement with worker rewards and net supply over the same month. Separate paid receipts from free-credit or provisioned workloads. A larger user count is not a token-demand numerator; the frozen report does not provide the flow data required to compute one.
| Role | What the source supports | What it does not establish |
|---|---|---|
| PHA holder | Holds a network utility/governance token | Not equity in Phala Cloud or an automatic cloud-revenue dividend |
| Worker or delegator | May earn rule-based network rewards | Rewards, slashing and migration terms depend on current network rules |
| Cloud customer | Can use confidential compute services | A customer deployment is not proof of a passive holder cash flow |
Verify: Phala Network official overview · Phala 2024 network and token report · Phala 2025 network report
How do PHA worker rewards, halvings and migration affect supply?
PHA’s reported cap does not replace current accounting of worker rewards, locks and migration balances. Phala’s own 2024 report states a one-billion PHA maximum and describes worker rewards and halving events. The frozen 2026 provider snapshot estimates circulating units separately. A maximum supply is not proof that all remaining units become freely tradable on one date, and older issuance rates should not be copied into today’s economics.
Net token demand would require current PHA-denominated compute payments, worker rewards, staking/locks, treasury transfers and migration bridges aligned by date. We do not have a complete public, independently frozen series of those quantities in this release. The large price gain is observed; the proposed causal explanation remains unverified.
The 2026-09-27T02:55:20.000Z CoinGecko snapshot lists 857,048,787 circulating PHA, 1,000,000,000 total and 1,000,000,000 maximum. A reported stock difference is not an unlock event, a date or a sell order. Actual vesting rules and transfer records need a separate dated check.
| Supply or valuation field | Frozen value | Use correctly |
|---|---|---|
| Circulating units | 857,048,787 | Provider estimate; not an on-chain audit |
| Total units | 1,000,000,000 | Do not substitute for current float |
| Maximum units | 1,000,000,000 | Missing maximum is not zero |
| Market capitalization | $63,534,335 | Price × provider circulating amount |
| Fully diluted valuation | $74,131,527 | Scenario stock value; not near-term selling amount |
Verify: Original provider snapshot · Phala Network official overview · Phala 2024 network and token report · Phala 2025 network report
How many Phala Cloud users paid, and did their spending reach PHA?
The issuer’s 398 paid users are a dated usage claim, not a measured PHA settlement flow. Phala's issuer-reported 2025 year-end snapshot lists 10,004 total users, 2,113 subscribed users and 398 paid users; paid users were 4.0% of reported total users. It also lists 813 running out of 2,529 total CVMs (32.1%). These are meaningful product-use distinctions, but they are one issuer statement, not an independently audited monthly growth series or current September 2026 count.
The report also describes free Cloud credits and multiple commercial surfaces. Paid Cloud users establish that some customers paid Phala for a product; they do not disclose whether payment required PHA, whether a treasury converted receipts into PHA, how much was paid to workers, or how much new PHA reached circulation. This report therefore does not infer PHA holder revenue from the user and CVM counts. The next verifiable bridge is monthly paid receipts → PHA conversion/settlement → worker rewards and net supply.
| PHA operating / token-capture test | Dated evidence | What it can prove |
|---|---|---|
| Total / subscribed / paid Cloud users, 2025 year-end | 10,004 / 2,113 / 398 | Issuer-reported one-date snapshot |
| Running / total confidential VMs | 813 / 2,529 | Provisioned is not paying |
| Paid Cloud receipts and PHA-settled share by month | Not publicly reconciled | No token-demand estimate |
| PHA worker rewards minus net demand | Not aligned to paid Cloud cohort | No holder-capture conclusion |
Verify: Frozen operating-evidence record · Phala 2025 operating snapshot · Phala Cloud free-credit onboarding
Does PHA’s current surge erase its prior deep drawdowns?
No: a +183.9% PHA 90-day endpoint and −35.7% drawdown within that path can coexist; neither removes earlier losses. The September 26 UTC-open 90-day return was +183.9% versus BTC +40.1%; the asset’s same-window daily-open peak-to-trough drawdown was −35.7%. This is one end-date path, not a 90-day expected return.
The line chart normalizes PHA and BTC to 100 at the same UTC start, making relative direction visible without pretending that the two assets have comparable unit prices. Both asset and BTC come from Binance USDT spot daily opens. Intraday low, spread, fees and actual execution can produce a worse realized experience than the line shows.
| Holding window | PHA spot return | BTC spot return | PHA − BTC | Asset path drawdown |
|---|---|---|---|---|
| 7 days | +164.0% | +4.0% | +160.0 pp | −4.2% |
| 14 days | +238.3% | +8.9% | +229.4 pp | −4.2% |
| 30 days | +233.1% | +6.4% | +226.7 pp | −8.6% |
| 60 days | +316.3% | +31.9% | +284.4 pp | −10.0% |
| 90 days | +183.9% | +40.1% | +143.8 pp | −35.7% |
| 180 days | +146.7% | +27.4% | +119.3 pp | −63.1% |
| 365 days | −4.3% | −22.8% | +18.5 pp | −81.2% |
Verify: Full frozen return inputs
How did PHA compare with BTC across 58 monthly 90-day starts?
Across 58 complete 90-day month-start paths, PHA’s median return was −13.3%; 14 beat BTC and 57 had deeper drawdowns. We predefine calendar-first 90-day and 30-day starts plus Monday 14-day starts from the first available daily bar through the last date that permits a complete exit on 2026-09-26. Every candidate must have each asset and BTC daily open in the path. We neither choose a favorable entry date nor fill a missing day with zero. The three cohorts answer different holding-horizon questions; they are not pooled into a single success rate.
The 58 90-day monthly starts are descriptive and overlap. An every-third-month subset has only 20 observations and 5 BTC-beating paths; it is a dependence sensitivity, not an independent win-probability estimate. The 90-day monthly starts median PHA return was −13.3%, BTC’s was +2.7%, and 14/58 asset paths finished ahead of BTC. 57/58 had a deeper daily-open drawdown than BTC.
| Prespecified cohort | Complete / candidates | PHA median | Beat BTC | Deeper drawdown | Reading status |
|---|---|---|---|---|---|
| monthly 30d | 60/60 | −9.0% | 19/60 | 58/60 | Descriptive; dependent windows |
| monthly 90d | 58/58 | −13.3% | 14/58 | 57/58 | Descriptive; dependent windows |
| weekly 14d | 262/262 | −2.7% | 94/262 | 244/262 | Descriptive; dependent windows |
Verify: Full candidate and outcome dataset · Excluded/missing ledger
Did PHA work better after a rising or falling BTC month?
PHA’s median 90-day return was negative in both prior-BTC groups: −9.7% (29 nonnegative starts) and −14.7% (28 negative starts). For the 90-day monthly starts sample, we classify the previous 30-day BTC return using only BTC daily opens before each hypothetical entry. This is a genuinely pre-entry state label, not a split by whether BTC rose after entry. It tests whether the observed relation looks different across prior market environments, but small and overlapping groups cannot establish a trading rule.
The worst BTC-relative 90-day start was 2025-01-01: PHA −74.2% versus BTC −11.8%, with −74.2% PHA path drawdown. The best counterexample was 2024-10-01: PHA +286.3% versus BTC +48.0%. Both belong to the same prespecified candidate universe; neither was selected to define the rule.
10 of 58 90-day observations are printed below (the most recent ten by start date; the remaining 48 are in the complete CSV). A daily-open drawdown misses intraday moves, and a historical endpoint omits spreads, liquidity and tax. The cohort table above includes every eligible horizon group.
| Prior BTC 30d state | n | PHA median | PHA beat BTC |
|---|---|---|---|
| Prior BTC 30d ≥ 0 | 29 | −9.7% | 8/29 |
| Prior BTC 30d < 0 | 28 | −14.7% | 6/28 |
| Prior BTC unavailable | 1 | −22.4% | 0/1 |
Which PHA entry dates support or contradict the headline?
The ten recent PHA starts are examples within the 58-start study, not its whole distribution or evidence that all buyers captured the latest rally. The rows below are the ten most recent 90-day starts for the main sample. Date, terminal return and path drawdown should be read together. A positive final return may still have required surviving a large temporary loss. A negative BTC-relative result does not imply an absolute loss, and vice versa.
PHA’s current headline path and the distribution of older starts answer different questions. For a decision today, verify live fundamentals, possible supply transfers, current spread and your own maximum loss before inferring anything from a past cohort.
| UTC entry | Horizon | PHA return | BTC return | PHA − BTC | PHA drawdown |
|---|---|---|---|---|---|
| 2025-09-01 | 90d | −56.5% | −16.1% | −40.4 pp | −60.1% |
| 2025-10-01 | 90d | −60.5% | −23.5% | −37.0 pp | −68.3% |
| 2025-11-01 | 90d | −34.6% | −22.8% | −11.8 pp | −50.8% |
| 2025-12-01 | 90d | −50.5% | −25.9% | −24.6 pp | −51.6% |
| 2026-01-01 | 90d | −2.8% | −22.1% | +19.3 pp | −52.1% |
| 2026-02-01 | 90d | +2.6% | −0.6% | +3.3 pp | −38.2% |
| 2026-03-01 | 90d | +75.1% | +9.7% | +65.4 pp | −38.2% |
| 2026-04-01 | 90d | −14.1% | −11.8% | −2.4 pp | −43.9% |
| 2026-05-01 | 90d | −35.4% | −16.2% | −19.2 pp | −62.1% |
| 2026-06-01 | 90d | −35.3% | +6.2% | −41.5 pp | −49.5% |
Verify: Download all cohort rows
Is PHA spot turnover enough to support a large exit?
No: the frozen $1,211,729 median daily turnover does not show bids available for a large sale. PHA’s archived Binance Spot USDT daily candles have 30 valid quote-turnover days in the last 30. Their median daily quoted turnover is $1,211,729. This is an amount traded throughout a day, not the amount available at the current bid, and it does not prove a $10,000 or institutional-size order can be executed near a headline price.
The evidence in this release does not include a synchronized, venue-specific level-two book for PHA. We therefore do not report a fabricated current spread, ±1% depth or expected slippage. Before taking action, inspect the live order book, trade size, market impact, available pairs and withdrawal conditions on the actual venue.
Before an actual PHA trade, compare intended size with fresh bids/asks on the same venue and estimate entry plus exit fees and adverse slippage. Repeat after a material price move. If the fill estimate is missing or exceeds the loss budget, reduce proposed size or wait. This is an execution-check method, not a fill simulation performed in the frozen study.
| Execution question | Observed answer | Missing check |
|---|---|---|
| 30-day median daily quote turnover | $1,211,729 | Not executable depth |
| Venue | Binance Spot USDT | Check the intended trading venue |
| Live bid/ask spread | Not sampled for this release | Capture a fresh order book |
| $10,000 round-trip costs | Not estimated | Include exchange fees, slippage and funding if leveraged |
Verify: Current multi-market liquidity study · Dated venue candle inputs
Are Phala AI claims about users, workloads or token demand?
A confidential-AI claim must identify paid workloads and the PHA payment mechanism, not just technology attention. A useful Phala community debate asks whether privacy-focused AI customers repeatedly pay for workloads, what assurances the TEE design provides, and how fees flow to PHA. Reposts of GPU partnerships or general AI-token momentum are attention indicators, not independently counted paying customers.
This release has no verified count of unique social users or independently filtered sentiment. A trending placement or message volume could include promotions, repeated posts or ticker collisions. We use community discussion to identify claims to test, not as a measured demand or tokenholder-income variable.
Verify: Phala Network official overview · Phala 2024 network and token report · Phala 2025 network report
Is PHA worth holding after a rapid rally?
PHA’s dated rally supports investigation, not assuming paid AI adoption overcame issuance and historical path risk. The evidence does not justify an unconditional “buy” or “sell.” The stronger case would be: Current workload receipts, PHA settlement and net fee demand can be independently reconciled against issuance and bridge balances while liquidity remains healthy. The weaker case would be: The rally is mainly AI-sector rotation; workloads do not translate into measurable PHA flows, or migrations and emissions complicate the advertised scarcity case.
A practical holding checklist is to define a time horizon, the project-to-token mechanism, next supply milestone, position size and acceptable drawdown. Decide what observation would falsify the thesis. Even the larger historical cohort is descriptive, overlapping and vulnerable to survivorship and regime changes.
Record the Phala architecture and date to which a claim applies. Seek matched paid receipts, settlement currency, PHA conversions, worker rewards and bridge balances. Migration-era prices describe changing economics, not a backtest of today’s network design. If only user counts are available, the revenue-to-token bridge remains unverified.
| Decision gate | Evidence to request | Failure sign |
|---|---|---|
| Token-specific value capture | Does real confidential-compute usage create demand for PHA faster than network reward issuance? | Only project growth is shown |
| Supply and float | Official allocation plus actual transfers | Unmeasured release reaches market |
| Historical opportunity cost | 58 complete 90-day monthly starts versus BTC | Single rally is treated as a general rule |
| Execution | Fresh venue depth, size and fees | Quoted price cannot be filled |
| Thesis invalidation | The rally is mainly AI-sector rotation; workloads do not translate into measurable PHA flows, or migrations and emissions complicate the advertised scarcity case. | No observable reassessment trigger |
Verify: Frozen PHA operating evidence · Complete PHA paired-price study
How can the PHA result be replayed and where can it fail?
Calculation single-coin-paired-paths-20260927-1.0; daily-open endpoint 2026-09-26 UTC; coverage 2021-09-01–2026-09-26. The asset’s historical source is local archive, potentially retrospective backfill. This is retrospective market research, not a claim that CoinNudge observed every older bar in real time.
PHA uses Binance PHA/USDT and BTC uses Binance BTC/USDT; both are paired at the same UTC daily open. Recent endpoints use 7/14/30/60/90/180/365 daily-open intervals only when all intermediate opens exist. Cohorts use every calendar month-first 30/90-day start and every Monday 14-day start from first available bar, with complete paired paths only. No strategy, fees, spreads, taxes, staking or causal token-use inference. Overlapping windows are dependent; fewer than 20 rows is exploratory. Historical archive/API retrieval is not proof the data was locally known at the original event time. Returns equal end open divided by start open minus one. BTC-relative differences are percentage points. Drawdown is the minimum of path value divided by its prior peak minus one. Candidate starts and all outcomes are preserved; a missing path is null, never a zero return.
The release exposes the calculation inputs, field dictionary, 380 observed cohort rows, 0 excluded candidates, source supply snapshot and SHA-256 input/output identities. Data from CoinGecko and project documents have their own timestamps and are not inserted as features known at a historical start. An independent rerun should match the downloadable rows before interpreting the prose.
Editing updated 2026-10-03: clarified section answers, holding decisions and evidence scope only. September frozen datasets, formulas, sample counts, charts and method version remain unchanged; the editing date is not a new market observation.
| Reproduction file | Purpose | Boundary |
|---|---|---|
| Research JSON | /research-data/hot-coin-profiles/2026-09-27-v1/pha.json | Summary, version and hashes |
| All outcome rows | /research-data/hot-coin-profiles/2026-09-27-v1/pha-observations.csv | Every candidate that passed paired coverage |
| Missing/excluded starts | /research-data/hot-coin-profiles/2026-09-27-v1/pha-exclusions.csv | Explicit absent pairs; no invented zeroes |
| Asset daily inputs | /research-data/hot-coin-profiles/2026-09-27-v1/pha-daily-inputs.csv | Unrounded venue observations |
| BTC daily inputs | /research-data/hot-coin-profiles/2026-09-27-v1/btc-daily-inputs.csv | Same-date comparator |
| Field dictionary | /research-data/hot-coin-profiles/2026-09-27-v1/field-dictionary.json | Deterministic formulas and status |
| Release manifest | /research-data/hot-coin-profiles/2026-09-27-v1/manifest.json | SHA-256 of frozen files |
Verify: CoinNudge frozen paired-price study · All matched outcome rows · All asset daily inputs · Same-date BTC daily inputs · Missing-pair ledger · Frozen CoinGecko supply snapshot · Field dictionary and calculation contract · Phala Network official overview · Phala 2024 network and token report · Phala 2025 network report · Frozen operating-evidence record
Take the next step with this evidence.
Choose ongoing Telegram monitoring or inspect structured data in your own research workflow.
Receive qualifying events in Telegram.
Use the ready-made alert checklist, message explanations and daily briefing. Check the supported categories and coins before subscribing; every Research page is not a separate alert category.
Check the alert list · View alert plans
Inspect the data before choosing a pack.
Prepared research packs supply scoped features, separate outcomes where available and quality flags. Buy a fixed edition once and download directly; API use is optional.
Check data pack samples · Explore the full data catalog · Compare data plans
Each purchased pack includes its specified fields and published date range. This page's full dataset is not automatically included. Other data families require coverage, date, field and source-use confirmation before purchase. Telegram Pro is a separate 30-day service; historical packs are one-time purchases.
Frequently asked questions
What is PHA coin used for?
Phala builds confidential-computing infrastructure using trusted execution environments. PHA is a network utility and governance token, but an AI or cloud-compute announcement is not a measurement of paying PHA demand. Documented use is not a guaranteed token price or holder income.
Is PHA coin worth holding based on this evidence?
Phala (PHA) coin is associated with confidential-computing infrastructure, not a share of Phala Cloud revenue. Market evidence ends at the 2026-09-26 UTC daily open; supply and operating records have separate dates below. PHA’s last dated 90-day spot path gained +183.9% against BTC +40.1%, yet the median of 58 earlier 90-day month-starts was −13.3% and only 14/58 beat BTC. This is a sharp current rally against a much weaker historical distribution, not proof that confidential-AI adoption now pays token holders. Phala reported 398 paid Cloud users and 813 running CVMs at 2025 year-end; it did not publish an aligned PHA-settlement series, so paid product usage is not yet verified token capture. This is dated non-personalized research, not a trade instruction.
How did PHA perform versus Bitcoin?
Across the frozen 90-day same-UTC daily-open window, PHA returned +183.9% and BTC returned +40.1%. Historical paths do not forecast the next period.
How much historical PHA data is available?
This release covers 1852 asset daily opens from 2021-09-01 through 2026-09-26 on Binance Spot USDT. Backfills and public API retrieval are retrospective, not original-date CoinNudge observations.
Does confidential AI demand automatically make PHA rise?
No. Verify paid workloads, actual PHA settlement and net supply flows at matched dates; a use-case announcement alone cannot establish price causation.
Is PHA supply capped?
Phala has reported a one-billion maximum, but current circulating supply and reward flows are separate questions.
Why does the report show a rally and a negative historical median?
The last 90 days are one selected end-date path; the historical study includes all eligible month-starts, many of which lost value.
Do Phala paid Cloud customers prove demand for PHA?
No. The issuer reported 398 paid users at 2025 year-end, but no matched monthly PHA settlement, conversion, rewards and net-circulation series was publicly verified for this release.
Can I verify the PHA calculations?
Yes. The full JSON, 380 cohort rows, daily input files, field dictionary, exclusion ledger and manifest are linked on the page.
What evidence should I check before holding PHA?
Record the Phala architecture and date to which a claim applies. Seek matched paid receipts, settlement currency, PHA conversions, worker rewards and bridge balances. Migration-era prices describe changing economics, not a backtest of today’s network design. If only user counts are available, the revenue-to-token bridge remains unverified.
Sources and verification
- CoinNudge frozen paired-price studyVersion, full summary, all observations and hashes
- All matched outcome rowsComplete monthly and weekly cohort ledger
- All asset daily inputsUnrounded source opens, turnover and provenance
- Same-date BTC daily inputsBTC comparator candles
- Missing-pair ledgerExplicit rejected candidates, empty if none
- Frozen CoinGecko supply snapshotProvider-defined supply and timestamp
- Field dictionary and calculation contractPrecise formulas and missing-value semantics
- Phala Network official overviewTEE/cloud architecture and product scope
- Phala 2024 network and token reportPHA max supply, worker rewards and historical halving disclosures
- Phala 2025 network reportMigration and compute-capacity context
- Frozen operating-evidence record
Follow PHA conditions with transparent market context.
Use this historical report to frame your questions, then check current liquidity, supply and Telegram alerts separately.
Explore alert coverage