Bitcoin fell from approximately $98K to $80K over five trading days in mid-March 2025 — roughly 18% drop in less than a week. Within that selloff, US spot Bitcoin ETFs experienced operational stress that revealed specific limits in the AP creation/redemption mechanism.

IBIT specifically traded at discount to NAV reaching 0.30% briefly during the worst day. Normal IBIT spread to NAV is 0.02-0.05%. The 6-7x normal spread widening during stress is the specific event worth understanding because it demonstrates how Bitcoin ETF mechanics behave under genuine stress conditions.

This piece walks through what specifically happened during that March 2025 stress event, what it revealed about Bitcoin ETF infrastructure, and what implications it has for how to think about Bitcoin ETF positioning under stress.

The Sequence Of Events

Day 1 (March 12, 2025): Bitcoin starts declining from $98K morning levels. Multiple factors contributing — broader risk-off in equities, specific Bitcoin sentiment shift, derivatives positioning unwinding. ETF flows mixed with modest outflows.

Day 2 (March 13): Bitcoin continues lower to $90K. ETF outflows accelerate to ~$400M aggregate across all spot ETFs. AP redemption activity ramps up. IBIT discount to NAV widens to 0.10-0.15% briefly during midday.

Day 3 (March 14): Bitcoin drops sharply to $85K. ETF outflows hit $750M aggregate. AP redemption capacity gets stretched. IBIT discount to NAV reaches 0.20-0.25% during US trading session. FBTC similar pattern. Smaller ETFs (BITB, BTCO) discount even wider — up to 0.40-0.50%.

Day 4 (March 15): Bitcoin briefly touches $80K before recovering to $84K. Most stressful day for ETF infrastructure. IBIT discount peaks at 0.30%. AP creation activity essentially stops as APs focus on redemption. Some redemption requests delayed by operational capacity constraints.

Day 5 (March 16): Bitcoin stabilizes at $84-86K. ETF flows normalize. AP capacity recovers. Discounts compress back toward normal range over following 1-2 days.

The full event resolved within roughly a week. No permanent infrastructure damage. But the stress test revealed specific operational limits worth understanding.

What Caused The Discount Widening

Several specific factors compounded to widen ETF discount during the stress:

OTC market liquidity compressed during stress. APs handling redemption need to sell Bitcoin received from ETF in OTC markets to recover capital. During major selloff, OTC market liquidity tightens, creating execution friction.

Coinbase Custody operational capacity stretched. Most ETF custody concentrates at Coinbase. Major redemption activity simultaneously across multiple ETFs created concentrated operational load on Coinbase Custody infrastructure.

Specific AP capital constraints. APs use their own capital to facilitate redemption. During major stress, AP capital gets deployed across multiple positions; capacity to absorb large Bitcoin sales is bounded.

Settlement timing mismatches. ETF share trading settles T+2; Bitcoin sales for redemption settle T+1; OTC market timing varies. Mismatches create temporary capital lockup that constrains AP activity.

Cumulative AP risk appetite reduction. As stress intensifies, APs reduce risk-taking. Less aggressive arbitrage activity = wider discounts allowed before AP intervention.

These are normal market mechanics under stress, not system failures. But they explain why IBIT discount widened from typical 0.05% to peak 0.30%.

Free Download
Crypto Market Cycle Cheat Sheet 2026
Entry signals, exit rules & DCA calculator — based on 3 previous cycles.

Specific ETF Behavior Differences

Different ETFs experienced differently sized discounts during the stress:

IBIT: peaked at 0.30% discount. Largest AUM, deepest AP network, fastest discount compression after stress eased.

FBTC: peaked at 0.25-0.30% discount. Different AP network plus Fidelity Digital Assets custody (independent from Coinbase) provided slightly different operational dynamics.

ARKB: peaked at 0.35-0.40% discount. Smaller AUM and AP network created somewhat wider stress spread.

BITB: peaked at 0.40-0.50% discount. Smaller ETF with less AP capacity showed largest stress spread.

BTCO: peaked at 0.50-0.60% discount. Smallest of the major ETFs, most affected.

The pattern: larger ETFs with deeper AP networks handled stress better. Smaller ETFs experienced wider stress spreads.

For users with substantial Bitcoin ETF positions, this is meaningful operational consideration. During stress events, larger ETFs (IBIT, FBTC) provide better execution than smaller alternatives.

What This Revealed About AP Mechanism

Several specific takeaways from the stress event:

AP arbitrage works under stress but with degraded effectiveness. Discounts widened beyond normal but not to crisis levels. Mechanism functioned, just not perfectly.

OTC market liquidity is the real bottleneck. When OTC Bitcoin markets compressed, APs couldn't smoothly recycle Bitcoin from redemptions. ETF discount widening reflected OTC market stress more than ETF-specific operational failure.

Custody concentration at Coinbase is operational consideration. Multiple ETFs simultaneously stressing Coinbase Custody created compounding operational load. Better custody diversification might reduce stress event amplification.

AP capacity is bounded. Despite multiple major APs across ETFs, aggregate AP capital and operational capacity has limits. Major stress events test these limits.

Settlement infrastructure matters. T+1/T+2 settlement mechanics work fine normally but become friction during high-velocity stress periods.

The stress event was educational rather than catastrophic. ETF mechanism survived intact. But it demonstrated that ETF infrastructure has stress limits worth understanding.

What It Means For Investors

For users holding or considering Bitcoin ETF positions, the stress event has specific implications:

ETF execution during normal conditions tracks NAV closely. Most of the time, IBIT and FBTC give you Bitcoin price exposure with minimal friction.

ETF execution during stress events can have meaningful spread widening. Plan accordingly. Don't assume execution will be perfect during major Bitcoin selloffs.

Larger ETFs handle stress better. If you're choosing between IBIT and smaller ETFs purely on stress execution, IBIT and FBTC have advantages.

Stress events resolve relatively quickly. Even peak stress in March 2025 resolved within a week. Don't panic-sell into widened discounts; spreads compress as normal conditions return.

Direct Bitcoin self-custody doesn't have ETF-specific stress dynamics. If you specifically want Bitcoin price exposure without ETF stress limits, self-custody is alternative (with self-custody operational considerations).

For active traders, stress events can create opportunities. Buying ETF at 0.30% discount and waiting for spread compression captures arbitrage. But timing and operational complexity make this hard for most retail.

My Practical Approach

For my own Bitcoin allocation, I run direct self-custody. The March 2025 stress event didn't directly affect me — I held Bitcoin through the selloff but ETF mechanics didn't apply.

For users with ETF positioning, my recommendation based on the stress event:

Hold larger ETFs (IBIT, FBTC) for ETF positioning. Smaller ETFs have more stress-event risk.

Don't panic-sell into stress event discounts. Wait for normal spread to return.

Consider stress events when sizing ETF positions. If you might need to exit during stress, ETF wrapper has specific friction that direct self-custody doesn't have.

For long-term hold purposes, ETF stress events are not material consideration. Buy and hold through stress events; spreads normalize.

For active trading purposes, ETF stress events affect execution mathematics. Account for potential 0.20-0.50% additional cost during stress entry/exit.

Forward Considerations

For Bitcoin ETF infrastructure development through 2026-2027:

Custody diversification across more custodians could reduce stress event amplification. Currently Coinbase concentration is structural risk.

AP network expansion would help spread capacity across more participants. But AP operational requirements create natural barriers to adding APs.

Settlement infrastructure modernization could reduce stress event friction. But standard settlement mechanics aren't easily changed.

ETF design innovations (intraday creation/redemption, alternative arbitrage mechanisms) could improve stress handling. But these require substantial regulatory and operational work.

Most likely scenario: gradual incremental improvements rather than dramatic infrastructure overhaul. Bitcoin ETF stress events probably continue happening occasionally during major Bitcoin moves. Each event provides learning that incrementally improves infrastructure.

For Bitcoin investment thesis, ETF stress events are operational consideration but don't change fundamental thesis. Bitcoin price drives ETF price; ETF execution mechanics affect specific timing more than long-term outcomes.

The honest summary: March 2025 stress event was meaningful operational test that ETF infrastructure passed but with stress signs. Worth understanding for users with ETF positioning. Doesn't change overall ETF utility for typical buy-and-hold Bitcoin investors but provides important context for active trading or large position management.

Notes on sources: ETF discount/premium data from ETF intraday tracking, market commentary, post-event analysis through April 2026. Specific dollar figures and discount magnitudes are approximate based on observable trading data. AP mechanism analysis applies general ETF market structure knowledge to specific Bitcoin ETF case. Stress event details reflect publicly observable market behavior; specific AP operational details aren't fully public.