Last updated May 15, 2026
Eight honest counterfactuals using documented historical close prices, plus a live snapshot of the last ninety days. Some of these rotations won. Some didn't. One cost almost everything that mattered.
The point is not regret. The architecture of moving between assets — custodial detour vs. non-custodial swap — determined which of these trades was even available. Flip is the version where the rotation isn't the hard part.
No major has run away from the others in the last 90 days. Rotation is a directional bet; sometimes the honest answer is that nothing obvious is on offer.
Updated weekly. Source: CoinGecko daily closes. Indicative, not advice — as of 2026-05-17.
$10,000 of BTC held vs the same dollars rotated into ETH at the start of the cycle. Both took the run; one took it nearly three times harder. Most BTC holders watched this play out and didn't rotate, because the path went through a CEX or a wrap step — neither worth the friction on a coin they intended to hold long-term anyway.
Sometimes the alt outruns BTC. The cost of not rotating wasn't the math — it was the architecture.
$1,000 of DOGE bought at $0.01 ran to roughly $0.70 in May 2021 — a 70× peak almost no one believed was real on the way up. The rotation that mattered wasn't the entry. It was the exit: $70k of DOGE turned into $70k of BTC, then carried through three years of cycles to roughly $117k by late 2024. Holding the meme through the round-trip ended closer to $30k. The path required either a CEX (KYC, capital-gain event, hot-wallet trust) or a non-custodial swap. Most chose neither, and rode it back down.
The meme is the entry. The hard asset is the exit. The hard part isn't either — it's the door between them.
$50,000 of BTC (~2.5 BTC at $20,500) sent to FTX for trading. Two weeks later, withdrawals froze. The bankruptcy paid back close to 100% of USD claims — but in Nov-2022 dollars, while the same 2.5 BTC held in self-custody would have been worth around $240,000 by late 2024. The principal returned. The rally did not.
The cost of the custodial detour isn't always 'you lose everything.' Sometimes it's 'you watched the run from the wrong side of a withdrawal queue.'
$10,000 of BTC sold to USDC in the week the FTX collapse turned counterparty risk into a market-structure event. The decision wasn't irrational at the time — every CEX with a quote was being treated as a potential next domino, and 'flat in stables' looked like the only correct posture. BTC bottomed within days, then ran nearly six-fold over the following two years. The principal was preserved. The cycle was not.
The architecture makes the rotation possible. It doesn't make it correct. This one cost the cycle to a decision that felt obvious at the time.
$10,000 of BTC rotated into SOL at $13, when SOL had just collapsed from $35 in the FTX cascade and was being written off as terminal. BTC itself returned roughly 6× over the window; the SOL rotation returned over 15×. The trade was only realistically available to whoever could route non-custodially — every CEX deep enough to do the rotation had just imploded.
The bottom is the entry. The architecture is what determines whether you can take it.
$10,000 of ETH at ~$1,700 vs the same dollars rotated into SOL at ~$25, when SOL had clawed back from the FTX cascade but was still trading like a stranded asset. ETH roughly doubled into the next leg. SOL ran eight-fold on a combination of returning fundamentals and the alt-of-alts narrative. Alt-to-alt is the most architecturally penalised swap in crypto — most paths require routing through BTC or stables, two taxable hops and two surface-area windows for things to go wrong.
Alt-to-alt rotation is where the friction tax is highest. That's also where the gap that Flip closes is widest.
$100,000 of BTC rotated into stables at $70k, locking in a 4× from the 2022 bottom. BTC chopped sideways through mid-2024 and ran to ~$95k by year-end. The stablecoin position protected against a drawdown that never came and missed roughly a third of further upside. The rotation wasn't wrong — it was conservative — but rotating is still a directional bet, and directional bets resolve both ways.
Rotation is not always the win. Sometimes it's the bet that didn't pay.
$10,000 of BTC wrapped into WBTC in early 2020 vs the same amount held in self-custody. Both positions ended around $118,000 — wrapping doesn't change the underlying. The difference was five years of extending unsecured credit to BitGo, uncompensated. BitGo remained solvent. The trust cost wasn't realized.
The trust cost wasn't realized. That doesn't mean it wasn't paid.
Numbers above use rounded historical close prices on the stated dates. Indicative, not advice. Past returns do not predict future returns; the only reason to look at counterfactuals is to make the cost of the architecture visible. The architecture is the only part Flip controls.