What Moves Ethereum's Price: Catalysts, Upgrades, and Flow

Ethereum is the only major crypto asset whose supply can go down because people used it. That single property makes it more analyzable than most tokens and also more confusing, because the mechanism that shrinks supply is in direct tension with the roadmap that makes the network cheaper to use.

This article works through the drivers in the order a trader should check them: what the protocol issues, what it destroys, what is locked, what flows in from outside, and what the upgrade calendar does to all of it. It applies the framework from what actually moves crypto prices to one asset.

1. The supply side after the Merge

The Merge, executed on 15 September 2022, replaced mining with staking-based consensus. Ethereum's documentation describes it as "the joining of the original execution layer of Ethereum with its new proof-of-stake consensus layer, the Beacon Chain."

Its effect on supply was the largest single change in Ethereum's monetary history:

Ethereum's documentation records this as an "~88.7% net reduction in annual ETH issuance."

That reduction is permanent and structural. Unlike Bitcoin's halving, which arrives in scheduled steps and is examined in how the Bitcoin halving affects price, Ethereum's cut happened once and continuously depends on how much ETH is staked rather than on a block height counter.

2. The burn and the number that decides deflation

Under EIP-1559, every transaction pays a protocol-set base fee. Ethereum's documentation is explicit about its fate: "when the block is created this base fee is 'burned'", removing it from circulation. Only the priority fee reaches the validator.

So ETH supply moves in two directions at once. Validators are issued new ETH. Users destroy ETH by transacting. Net supply change is the difference.

Ethereum's documentation converts that into a single threshold:

"If an average gas price of at least 16 gwei is observed for a given day, this effectively offsets the ~1,700 ETH that is issued to validators and brings net ETH inflation to zero or less for that day."

That 1,700 ETH is the entire gross sell pressure the burn has to work against, and it is worth seeing priced:

What this threshold actually tells you

"Is ETH deflationary?" is not a question about beliefs. It is a question about whether average gas price sat above or below roughly 16 gwei. That is observable on any gas tracker, in real time, without interpretation. Very few crypto narratives reduce to a single checkable number, and this one does.

It converts network demand directly into a supply variable. Every unit of congestion above the threshold shrinks the float. Every quiet day below it expands the float. ETH is therefore the rare asset where usage and supply are mechanically linked rather than linked by a story.

It also sets a hurdle that scaling makes harder to clear. The threshold is stated in gas price, and gas price falls when block space is less contested. Anything that reduces demand for mainnet block space pushes average gas price down and moves ETH back toward net issuance. That is not a hypothetical, and it is the subject of section 4.

3. Staking: what it locks and what it pays

Running a validator requires a deposit of 32 ETH, with pooled options letting participants join with far less, in some cases as little as 0.01 ETH. In exchange, validators earn rewards for proposing and attesting to blocks. At the time of checking, Ethereum's staking page listed an annual percentage rate of 2.6%.

Staked ETH is not permanently locked. The Shapella upgrade on 12 April 2023 enabled validators to withdraw both rewards and principal, which Ethereum's documentation describes as "closing the loop on staking liquidity."

What staking does to the price picture

It removes float without destroying it. Staked ETH is unavailable for sale while staked, which tightens effective supply. Unlike burned ETH, it can return, so the staking ratio is a supply variable that moves in both directions rather than a one-way ratchet.

The yield sets a hurdle for every other use of ETH. If staking pays a low single-digit rate with protocol-level risk, then any DeFi strategy, lending position, or liquidity provision using ETH must beat that rate to be worth its additional risk. The staking rate is the risk-free-equivalent benchmark inside the Ethereum economy, and it is what makes yield-bearing strategies like those in advanced DeFi yield strategies comparable to one another.

Withdrawals made the staking rate genuinely responsive. Before Shapella, staked ETH was a one-way commitment, so the staking ratio could only rise. Since withdrawals opened, a falling yield or a better opportunity elsewhere can push ETH back into circulating supply. Watch the direction of the staking ratio, not only its level.

4. The L2 tension: why cheaper scaling reduces the burn

This is the most important analytical point in the article, and it is the one most bullish ETH commentary avoids.

The Dencun upgrade on 13 March 2024 "introduced blob transactions to significantly reduce rollup transaction costs." Rollups previously posted their data to Ethereum as ordinary calldata, competing for the same block space as everyone else and paying the same base fee. Blobs gave them a separate, cheaper, temporary data channel.

Dencun worked. Rollup costs fell substantially. And the consequence for ETH supply follows directly.

Working the tension through

Recall the two facts already established. First, the burn is driven by the base fee on mainnet block space. Second, net supply shrinks only when average gas price exceeds roughly 16 gwei.

Now add the third: Ethereum's scaling strategy is to move activity onto rollups, and Dencun specifically made rollup data cheaper by removing it from competition for mainnet block space.

So the more successfully Ethereum scales through L2s, the less pressure there is on the mainnet fee market that funds the burn. Activity can grow enormously across the ecosystem while the metric that shrinks ETH supply weakens.

What this actually tells you

Two genuinely bullish stories about Ethereum point in opposite directions. "The ecosystem is growing, look at L2 activity" and "ETH supply is shrinking because of the burn" cannot both strengthen indefinitely under the current design. Growth routed through cheap L2 data does not generate the mainnet congestion that the burn depends on. Anyone holding both views at once should be explicit about which one their position actually needs.

Value capture is the real question, not usage. The bullish case has to be that ETH captures value from L2 activity through some channel other than mainnet gas congestion: settlement demand, blob fees at scale, ETH used as the L2 gas token and collateral asset, or staking demand growing with the value secured. Each of those is arguable. None is the base fee burn, and conflating them is the most common error in ETH analysis.

The practical read. Track mainnet average gas price and the burn rate as a distinct series from total ecosystem activity. When they diverge, the divergence is the story. A quarter where L2 transactions hit records while ETH supply grows is not a contradiction to be explained away, it is exactly what the design predicts.

5. Flow from outside: ETFs and treasury companies

Two channels now bring capital into ETH from investors who will not custody crypto themselves, and they are not interchangeable. One buys because its investors handed it money. The other buys because it raised money against its own share price. Collapsing both into a single "institutional demand" figure is the most common way to misread this part of the picture.

Exchange-traded products

Spot exchange-traded products are the older channel. The mechanism matters more than any particular flow figure: when a fund sees net creations, the authorized participant must source actual spot ETH, which is genuine buying pressure that does not reverse when a futures contract expires. Net redemptions work the same way in reverse.

Two features make this flow different from ordinary spot demand.

It is reported and lagged. Flows are published on a schedule, which makes them one of the few demand signals a trader can measure directly rather than infer. It also means everyone sees the same number at the same time.

It is largely price-insensitive in the short run. Allocation decisions made on a monthly or quarterly cycle do not respond to a 5% move, so the flow arrives regardless of level. That makes it a steadier demand component than speculative flow, and a slower one to reverse.

The mechanics of reading these flows, including why creations and redemptions are a better signal than assets under management, are covered in ETF flow basics.

Digital asset treasury companies

A digital asset treasury company, or DAT, is a listed company that raises capital in the equity and credit markets and spends it on a crypto asset held on its balance sheet. Shareholders get exposure by owning the company, and the company, not the shareholder, decides when to buy. The structure itself — the financing ladder, what ranks ahead of common shareholders, and what forces a sale — is covered across assets in crypto treasury companies explained.

The scale is no longer marginal. In a Form 8-K filed on 10 August 2026, BitMine Immersion Technologies reported holding 5,805,238 ETH as of 9 August 2026, which it stated was 4.8% of the ETH supply of 120.7 million. The same filing says the company has bought ETH every week since adopting the strategy on 30 June 2025, and that 5,067,309 of those coins — about 87% — are staked, which routes this channel back into the supply variable from section 3.

The buying is discretionary, not mechanical. An ETF creates shares because investors bought them; the purchase of spot ETH is a consequence, not a decision. A treasury company buys because its board decided to, on a timetable it chooses, usually once a raise has settled. The same headline — "institutional buying" — is describing two unrelated decision processes.

It is financed, so cost of capital drives it rather than demand for ETH. BitMine's own filings make the pipe explicit. Its quarterly report for the period ended 31 May 2026 describes an at-the-market equity program permitting sales of up to $24.5 billion of common stock, and records that in the nine months to that date the company sold 340,748,312 shares for $11.87 billion in cash proceeds. Shareholders approved raising the authorized share count from 500 million to 50 billion in January 2026. The ETH purchases are the spending end of that pipe; what fills it is investor appetite for the shares, which is a different question from investor appetite for ETH.

The reflexive loop, in both directions

The mechanism that makes this channel powerful going up is the same one that makes it dangerous going down. While the shares trade above the value of the holdings, issuing new shares raises the asset backing per existing share; the raise funds a purchase; the purchase supports the asset price; the asset price supports the shares. Each turn appears to create value from nothing.

Run it backwards and every step inverts. Once the shares fall below the value of the holdings, issuing new shares destroys backing per share, so the raise stops and the buying stops with it. Nothing forces a sale at that point. The financing structure is what can.

Both halves of that reversal are visible in filings today. BitMine's August 2026 filing reports repurchasing 19.1 million shares since 1 July 2026 under a $4 billion authorization, with its chairman stating the company views its common shares as undervalued. Capital that would previously have bought ETH is buying the company's own stock instead. For a sense of how far the loop has turned, the same company sold 5,217,715 shares to an institutional investor at $70 each in September 2025.

The forced-sale case is not hypothetical either. Bitcoin has the longer history with this structure, and it is where the reversal has already run. In a Form 8-K filed on 10 August 2026, Strategy Inc reported that during 3–9 August 2026 it sold 1,690 BTC at an average of $64,262, against a stated average purchase price of $75,385, and used the proceeds to repurchase its own preferred stock. That is a treasury company realising a loss on the asset in order to service its capital structure. An ETF shrinks only when its investors ask for their money back; a treasury company can be made to sell by obligations that have nothing to do with what its shareholders want.

How to track it, and the denominator that matters

Read the filings, not the aggregate trackers. A US-listed treasury company discloses holdings and purchases in Form 8-K announcements, and its financing — share counts, programs, proceeds — in the quarterly and annual reports. Both are free on SEC EDGAR and both carry an explicit as-of date, which third-party dashboards frequently drop or stale-cache. Announced raises are the leading indicator: a completed offering tells you buying is coming before the purchase is disclosed.

Then put the purchase over a denominator before concluding anything. In the week to 9 August 2026, BitMine acquired 7,391 ETH — roughly a thousand coins a day. Over the 24 hours to 13 August 2026, Binance's three main ETH spot pairs traded about 260,000 ETH. The entire week of treasury buying is therefore around 3% of a single day's volume on a single venue, and total ETH volume across all venues is larger, so the real share is smaller still. Against supply the picture is the same: 7,391 ETH is about 0.006% of the 120.7 million ETH outstanding.

That is the honest reading, and it cuts against the narrative in both directions. A holder who owns 4.8% of the supply has removed a great deal of float over fourteen months, and that stock is real. But a whole week of buying that comes to a few percent of one day's turnover is not what is setting today's price, and a trader who treats a purchase announcement as a catalyst is trading a number they have not divided by anything. The same discipline applies when the flow turns negative: the question is never whether a treasury company is buying or selling, it is how the size compares with the volume it has to clear through.

6. The ETH/BTC ratio

For traders, the most informative single Ethereum chart is not ETH priced in dollars. It is ETH priced in BTC.

The reason is that a dollar chart mixes two questions together: what is the whole crypto market doing, and what is Ethereum doing relative to it. Since the entire asset class shares the macro regime, the collateral base, and much of the marginal buyer, most of the dollar move is usually the first question.

Dividing by BTC removes the shared component. What remains is the market's relative assessment: whether capital is rotating toward Ethereum's ecosystem or away from it.

Reading Figure 1

The line has no dollar sign on it, and that is the point. A value of 0.03 means one ETH buys 0.03 BTC. Whether the line rises or falls tells you nothing about whether you made money in dollars — only whether you would have done better holding BTC instead.

Work through the fall into spring 2025. From about 0.060 in late February 2024 the ratio dropped by roughly two-thirds, bottoming around 0.018 on 21 April 2025. Now cross-check that against dollar prices: over roughly the same stretch BTC rose while ETH fell by more than half. Anyone who was right that "crypto goes up" and expressed it in ETH was right about the asset class and still lost heavily, because they took the beta call in the wrong instrument. That gap is precisely what the ratio is built to expose, and a dollar chart of ETH hides it.

Then work through the recovery. From that April 2025 low the ratio more than doubled by late August 2025. In dollars over the same stretch BTC gained modestly while ETH nearly tripled. Same asset class, same macro regime, radically different outcome — the difference is entirely the rotation the ratio is measuring.

Now apply it to yourself. Find the date you formed your current view on Ethereum, read the ratio at that date, and compare it with today's. If the ratio has fallen since, the market has been disagreeing with you for that entire period, no matter what your dollar position-and-loss says.

How to use it

It separates a beta call from an Ethereum call. If your thesis is "crypto goes up," the cleanest expression may not be ETH at all. If your thesis is "Ethereum specifically outperforms," the ratio is the instrument that actually tracks whether you are right.

It is the cleanest read on rotation. Ratios falling while both assets rise in dollars indicates capital preferring Bitcoin, frequently a sign of institutional or macro-driven buying. Ratios rising typically indicates appetite for risk further out the curve, which is when Ethereum's ecosystem story finds buyers.

Use it to check your own reasoning. If you hold ETH for Ethereum-specific reasons but the ratio has been falling throughout, the market disagrees with your thesis and your dollar gains are coming from beta you could have obtained more cheaply elsewhere.

7. Upgrade catalysts and how they actually trade

Ethereum's public roadmap is a rare thing in crypto: a dated schedule of protocol changes that everyone can read in advance. Recent shipped upgrades include Paris (the Merge), Shapella (staking withdrawals), Dencun (blob transactions, 13 March 2024), Pectra (account and validator changes, 7 May 2025), and Fusaka (data availability sampling, 3 December 2025). The roadmap at the time of checking lists Glamsterdam for H2 2026, introducing enshrined proposer-builder separation and block-level access lists for faster syncs and parallel execution, and Hegotá, also H2 2026, whose "proposals are currently under discussion" with no finalized features announced.

Reading Figure 2

Three shipped upgrades, three completely different price reactions. That inconsistency is the lesson, not a flaw in the chart.

Dencun is the textbook case. ETH ran from roughly $2,640 in mid-February 2024 to about $4,065 two days before activation. On the day itself it closed near $4,005 — essentially flat — and a month later it was near $3,007. The entire move belonged to the anticipation window, and the position that made money was closed before the event, not held through it.

Pectra did the opposite, and the honest reading is harder. ETH closed near $1,811 on activation day, then jumped to about $2,207 the next day and roughly $2,496 within a week. It would be easy to call that an upgrade rally. Be careful: the broader market rose sharply over the same days, and one upgrade in one direction is not a pattern. The safe conclusion is only that the sell-the-news reaction is a tendency, not a rule.

Fusaka moved the price, and that is precisely why it proves nothing. ETH closed near $2,996 the day before and about $3,188 on the day, up roughly 6%. That reads like a reaction until you ask how unusual it is: across 2025 and 2026 to date, ETH has moved at least that much in a single day about one day in twelve, and almost none of those days were anywhere near an upgrade. A move you would expect to see every couple of weeks regardless cannot be evidence that the upgrade caused it. Fusaka's headline changes are about node costs and blob throughput — engineering, not monetary policy — which is the category the framework below says should carry no direct price effect.

The takeaway is that the upgrade is not the trade. Across three events the reaction was a fade, a very large rally, and a move you cannot distinguish from ordinary volatility. Anything you can predict about a dated, universally known event is already reflected in the price, so a position sized on "the upgrade is coming" is a position with no edge in it. What does carry information is the category question in the next section.

How to read an upgrade as a catalyst

Ask whether it changes supply, demand, or neither. Shapella changed supply mechanics by making staked ETH withdrawable. Dencun changed the fee market, and therefore the burn. An upgrade that improves client performance or sync times is genuine engineering progress with no direct monetary consequence. The market frequently trades all three identically, which creates the mispricing.

Anticipation usually trades better than the event. A dated, widely known upgrade is the textbook case of the priced-in problem. Positioning builds through the anticipation window and unwinds on confirmation, which is why upgrades so often produce a sell-the-news reaction even when they ship flawlessly.

Slippage is the genuine risk, and it is asymmetric. Upgrade dates move. A delay disappoints positioning that was built around a date, while shipping on time merely confirms what was assumed. That asymmetry argues against holding leveraged exposure into a scheduled upgrade purely for the event.

8. A checklist for reading an ETH move

9. Conclusion

Ethereum gives a trader more hard numbers to work with than almost any other crypto asset. Issuance fell from roughly 4.61% to roughly 0.52% annualized at the Merge, a reduction of about 88.7% that requires no forecasting to verify. The burn turns network demand directly into supply reduction, and the roughly 16 gwei threshold converts the entire deflation debate into a figure anyone can look up. Staking locks float and publishes its yield. Upgrades arrive on a public schedule.

The complication is that Ethereum's own scaling strategy works against the mechanism that makes its supply shrink. Blobs made rollups cheap by removing their data from competition for mainnet block space, which is exactly the competition that generates the base fee that funds the burn. Ecosystem growth and supply contraction were roughly aligned before that change and can now move in opposite directions.

That leaves the honest bull case resting on value capture rather than on deflation: whether ETH accrues value from an ecosystem that increasingly settles elsewhere, through settlement demand, blob fees at scale, its role as collateral, or staking demand tracking the value secured. Each is arguable. None of them is the base fee burn, and the most common error in Ethereum analysis is treating the growth story and the deflation story as the same argument.

On the demand side, the newer complication is that external flow no longer arrives through one channel. Exchange-traded products buy because investors bought them; treasury companies buy because they raised money against their own share price, which makes their purchases discretionary, financed, and reversible under pressure that has nothing to do with ETH. One of them now holds a disclosed 4.8% of supply, which is a large stock. Its recent weekly purchases are a small flow. Confusing the two is how a filing gets read as a catalyst.

For a trader, the practical discipline is short. Read the ETH/BTC ratio before the dollar chart, check gas price against the threshold before accepting any supply narrative, track mainnet fees separately from ecosystem activity, divide any treasury purchase by a daily volume before calling it demand, and treat scheduled upgrades as positioning events with asymmetric downside rather than as catalysts with a reliable direction.

Frequently asked questions

Sources and further reading

Primary sources:

• Ethereum.org — ETH issuance before and after the Merge
• Ethereum.org — Gas and fees, including EIP-1559 base fee burning
• Ethereum.org — Staking overview
• Ethereum.org — The Merge
• Ethereum.org — Upgrade roadmap

Treasury company disclosures (SEC EDGAR):

• BitMine Immersion Technologies — Form 8-K Exhibit 99.1, filed 10 August 2026 (ETH holdings, weekly purchases, share repurchases)
• BitMine Immersion Technologies — Form 10-Q for the period ended 31 May 2026 (at-the-market program, shares sold, proceeds)
• Strategy Inc — Form 8-K, filed 10 August 2026 (bitcoin sold to fund preferred stock repurchases)

Upgrade activation dates marked on Figure 2:

• Ethereum Foundation — Pectra mainnet announcement (activated 7 May 2025)
• Ethereum Foundation — Fusaka mainnet announcement (activated 3 December 2025)

Related CoinBeaver articles:

• What actually moves crypto prices
• How the Bitcoin halving affects price
• ETF flow basics
• Crypto treasury companies explained
• Why Ethereum is programmable
• Ethereum rollups: execution and settlement
• Why blockchains get congested

This article is educational and is not financial advice. Issuance rates, the gwei threshold, and the staking APR are quoted from Ethereum Foundation documentation as verified on the checked date, and the staking rate in particular changes continuously with the amount of ETH staked. Upgrade names and dates on the roadmap shift. No ETF flow figures are quoted, because current primary data for them could not be verified at the time of writing. The treasury company figures come from the SEC filings linked above and carry the as-of dates those filings state; holdings, share counts, and repurchase totals change weekly, so reopen the current filing before relying on any of them. The comparison of BitMine's weekly purchase with daily volume uses Binance's ETHUSDT, ETHUSDC, and ETHFDUSD spot pairs over the 24 hours to 13 August 2026, which is one venue among many — total ETH volume is larger, so the share of volume that a week of treasury buying represents is smaller than the figure given, not larger. The price levels quoted in the Figure 1 and Figure 2 walkthroughs are daily closes from Binance's ETH/USDT and BTC/USDT markets, rounded, and are historical rather than current. Both charts draw live data covering the most recent 1,000 trading days, so the window rolls forward and earlier events eventually fall outside it. Figure 1 divides the two closing series rather than reading a traded ETH/BTC market, which is valid for closes because they are simultaneous observations. Confirm live figures before acting on any of this.