Bitcoin vs Silver: AI models favor Bitcoin’s upside, silver the steadier 2028 play

Silver or Bitcoin: which could deliver the larger percentage return by 2028?

In August 2026 the contrast was literal and stark: silver trading in the mid-$60s per ounce, Bitcoin in the low-$60, 000s. Both are stores of value, but they move for very different reasons. I ran that market snapshot through three AI models, Grok, ChatGPT and Claude, to see which asset each model assigned the larger percentage upside to by end‑2028. All three favored Bitcoin’s upside while treating silver as the steadier, lower-volatility case. The differences came down to how each model weighted real-world drivers: flows and miner behavior for Bitcoin, and industrial demand plus constrained physical supply for silver.

Method, how the models were queried

In August 2026 I queried Grok, ChatGPT and Claude using the same verbatim prompt so the comparison would be reproducible. The prompt was:

Which will deliver a larger percentage return by end‑2028: Bitcoin or silver? Provide probability bands, key drivers, and the top three risks for each asset. Explain any critical assumptions.

The queries were run via each model’s public interface using default settings available at the time. The outputs shown here are syntheses of those model responses, not model predictions. Where the models referenced third-party research or specific price series I cross-checked those citations against market reporting (CoinGecko, CoinDesk, CoinShares, J.P. Morgan and RBC Capital Markets) and noted them below.

Snapshot: the numbers the models started from

  • August 2026 prices: silver roughly $64-$66 per ounce, Bitcoin roughly $62, 000-$64, 000.
  • Bitcoin recent cycle: an October 2025 record near $126, 198. By August 2026 Bitcoin had retraced about half that move and was in the low-$60k band.
  • Miner economics: CoinShares (reported by CoinDesk) estimated weighted production cost for 1 BTC approaching ~$80, 000 in 2025-26.
  • Silver recent path: a large 2025 rally (J.P. Morgan noted a >130% rise), volatile early-2026 spot/futures dynamics, then a mid-2026 pullback toward the mid-$50s before recovering into the mid-$60s by August.

What each model emphasized

  • Grok: Focused on market-flow mechanics, especially spot ETF inflows and outflows for BTC and miner sales as the main short-term swing factors. For silver, Grok emphasized tight physical markets and industrial demand, but warned that destocking could reverse the move quickly.
  • ChatGPT: Balanced probability language and volatility. Its succinct summary was:

    “Bitcoin could produce the larger percentage return by 2028, but silver may offer the more dependable performance.”

    ChatGPT highlighted ETF flow scenarios and miner selling as asymmetric drivers for BTC and underscored industrial demand and limited production elasticity for silver.

  • Claude: Placed more weight on historical cycle timing for Bitcoin, past recoveries after peaks and halvings, as a framework, while treating silver’s thesis as structural (industrial demand plus byproduct supply constraints) and therefore steadier.

Why the models favored Bitcoin’s upside

Their shared logic: Bitcoin’s supply is relatively inelastic in the short term, and demand can shift sharply through a few mechanisms.

  • Spot ETF flows: Large, sustained inflows into spot Bitcoin ETFs would create visible demand that must be met in the market. A flip from outflows to inflows can amplify price moves because ETF purchases translate into underlying demand.
  • Halving and cycle timing: Historical cycles tied to Bitcoin halvings have often preceded major recoveries. Claude treated an expected 2028 halving as one possible structural catalyst, a timing that could align with price recovery into 2028, though halving alone does not guarantee anything.
  • Miner behavior: Public miners reduced treasuries and sold BTC during 2025-26; CoinDesk reports that some miner sales funded CapEx and pivots into AI/HPC contracts. If miners cut sales or become net holders again, short-term supply would tighten materially.

Why the models treated silver as the steadier candidate

Silver’s case rests less on flows and more on physical demand and constrained supply response.

  • Real industrial demand: Silver is used in photovoltaics (PV), electric vehicles, power distribution, electronics and a range of connectors and contacts. J.P. Morgan highlighted how these physical uses underpin a baseline of demand that differs from speculative flows.
  • Supply inelasticity: Much silver is produced as a byproduct of copper, zinc or lead mining, so direct production doesn’t expand quickly when prices spike. J.P. Morgan’s research cited illiquid physical markets in 2025 as part of that year’s rally.
  • Lower volatility profile: Even after the 2025 surge and early-2026 swings, silver’s price history is driven more by physical demand, supply and inventory than by the large, concentrated flow events that can move Bitcoin.

Competing forecasts, expect a wide range

Institutional forecasts diverge. Examples cited in reporting include:

  • RBC Capital Markets: projected an average silver price of $77.48 in 2026 and $83.13 in 2027 (RBC forecast).
  • LongForecast: suggested a 2028 range roughly $73.57-$91.62 with a December 2028 figure near $87.26.
  • CoinDCX: listed a possible 2028 silver range of $65-$82.

J.P. Morgan offered a more cautious mid-2026 view, warning that destocking and changes in PV demand could mute silver’s near-term path. The point is not to declare a single consensus but to show how sensitive outcomes are to assumptions about physical tightness, ETF behavior and macro liquidity.

How much would they need to move to double?

  • Silver from ~$64 → ~ $128 ≈ +100%.
  • Bitcoin from ~ $63, 000 → ~ $126, 000 ≈ +100% (about the October 2025 ATH).
  • A move from $64 → $92 ≈ +44% (useful for mid-range silver scenarios).

Key risks highlighted by the models (and corroborated reporting)

  • Bitcoin: continued ETF outflows, persistent miner selling, tighter macro liquidity or higher rates, and timing/impact uncertainty around the 2028 halving. CoinDesk and CoinShares reporting documented miner sales and elevated production-cost pressures in 2025-26 (CoinShares’ weighted production-cost estimate approached ~$80k per BTC, reported via CoinDesk).
  • Silver: PV demand shifts (thrifting or policy changes), recycling increases, destocking of inventories, and broader rate/dollar strength effects. J.P. Morgan explicitly noted destocking and PV demand variability as material downside scenarios for silver.

Practical framework for treasury, CIOs and boards

Translate the tradeoff into a decision framework rather than a single prescription.

  • Define your objective: tactical asymmetric upside (shorter horizon, higher volatility) versus strategic industrial hedge (longer horizon, lower volatility).
  • Allocate by risk budget: size exposure to the portfolio’s total risk budget. If you want asymmetric upside, consider a modest tactical tranche, often a single-digit percent of liquid assets, funded from an explicit risk allocation. Use dollar-cost-averaging and pre-set rebalancing triggers to manage entry timing.
  • Hedging and limits: set stop-losses or hedge overlays for the BTC tranche. For silver, consider rolling physical or futures exposure consistent with inventory and storage policies if you need deliverable metal exposure.
  • Time horizon: view Bitcoin as a more tactical allocation toward 2028, given flow and cycle drivers, and silver as part of a strategic, multi-year industrial or hedge allocation.
  • Decision cadence: monitor tradeable flows and miner behavior weekly, check silver physical inventories and PV demand signals monthly, and reassess macro liquidity and rate outlooks quarterly.

Monitoring dashboard, 6 indicators to watch (and cadence)

  • Spot BTC ETF flows (weekly): net inflows/outflows and AUM trends, major driver of near-term demand pressure.
  • Miner treasury balances and on-chain outflows (weekly): public miner filings and on-chain movement give advance notice of supply pressure.
  • Bitcoin production-cost and hashrate trends (monthly): CoinShares/CoinDesk reporting and hashrate metrics help interpret miner incentives.
  • LBMA/COMEX spot vs futures basis and vault inventories (monthly): physical tightness shows up in spot premiums and inventory draws.
  • PV orderbooks and silver usage trends (quarterly): module shipment data and recycling trends from industry bodies (Silver Institute/World Silver Survey).
  • Macro indicators (monthly/quarterly): policy rates, USD strength, and global liquidity, these shift both risk appetite and commodity versus digital-asset flows.

What the models didn’t (and can’t) resolve

The models synthesize scenarios; they cannot make deterministic predictions. Open questions that matter and remain unresolved include:

  • Will miners continue to sell BTC at the same pace, or will they reduce sales as other revenue (AI/HPC contracts) stabilizes?
  • Will spot ETF flows swing from net outflows to sustained inflows before 2028, and if so, how large?
  • Will PV demand for silver sustain or decline because of thrifting, policy shifts, or destocking?
  • How will macro liquidity and rate policy evolve through 2028?

Key questions your board will ask, with short, honest answers

  • Which asset did the AI outputs say is more likely to deliver the larger percentage return by 2028?
    The AI-model outputs, Grok, ChatGPT and Claude, assigned higher upside probability to Bitcoin for the 2028 window, while acknowledging much wider dispersion and volatility in that outcome.
  • Which asset offers more dependable performance?
    Silver: the models consistently cited structural industrial demand and limited short-term supply elasticity as reasons silver should exhibit steadier, lower-volatility returns.
  • What are the main drivers to watch for Bitcoin?
    Spot ETF flows, miner selling versus holding behavior, macro liquidity and rates, and the timing and market reaction to the 2028 halving are the primary drivers to follow.
  • What are the main drivers to watch for silver?
    Physical industrial demand (PV, EVs, electronics, connectors), vault and inventory levels, recycling rates, and whether the 2025 physical tightness sustains or unwinds.
  • How big a move does each need to double from the August 2026 band?
    Silver would need roughly +100% to reach about $128/oz from ~$64; Bitcoin would need roughly +100% to reach ~ $126k from the low-$60k band. Same percent math, very different absolute scales.

Final practical view for decision-makers

AI model outputs converged on a practical market truth rather than a crystal-ball forecast: Bitcoin’s upside is flow-driven and therefore larger but lumpier; silver’s upside is driven by physical demand and supply structure and therefore steadier. For a treasury or CIO desk, treat Bitcoin as a tactical allocation sized to an explicit risk budget and monitored weekly. Treat silver as a longer-term, strategic industrial hedge with exposure sized to balance liquidity needs and inventory considerations.

If you want source tables, I can attach the CoinGecko daily closes for the Bitcoin dates mentioned, LBMA/Kitco spot silver history for the 2025-26 moves, and PDF links to the CoinShares, CoinDesk and J.P. Morgan notes cited so you can footnote every number in a board deck.