Thesis answer
Direct answer
If you want real downside protection without strangling upside, the honest ranking is:
• Protective put — cleanest protection, full upside, but you pay for it. Best when vol is cheap or the risk is specific (earnings, macro event).
• Put spread (debit) — cheaper protection, full upside, but the floor only holds for a defined range. Best when you want to hedge a *normal* drawdown, not a crash.
• Collar — often free or near-free, full downside protection, but caps your upside. Best for concentrated positions you want to hold but de-risk.
• Put spread collar — sells a call to finance a put spread. Cheap, capped both ways, leaves a tail open below the long put strike. A compromise structure.
If I had to pick one default for "I want to stay long but sleep at night": a collar set 5-10% OTM on both sides, 60-90 DTE. You give up some upside, but you eliminate the gap-down risk that actually hurts portfolios.
Evidence that matters
Cost and structure comparison
Structure
Typical Cost
Downside Floor
Upside Cap
Best Use
Protective put
1-4% of notional (varies w/ IV, DTE)
Hard floor at strike
None
Event hedge, low-IV regimes
Put spread
0.3-1.5% of notional
Floor between strikes only
None
Modest drawdown hedge
Collar
$0 to small debit/credit
Hard floor at put strike
Hard cap at call strike
Concentrated holdings
Put spread collar
Often credit
Floor in put-spread range
Cap at short call
Cost-conscious hedging
Long-dated LEAP put
Higher absolute, lower annualized
Hard floor
None
Strategic portfolio hedge
Why each one works (or doesn't)
Protective put:You own the stock, buy a put. Below the strike, dollar-for-dollar protection. Above it, you keep all the upside minus the premium. The problem is the premium — pay 3% for a quarterly hedge four times a year and you've spent 12% on insurance you may never use. Worth it when IV is low or a specific catalyst is on the calendar.
Put spread:Buy one put, sell a lower-strike put. Net cost drops 50-70% vs. a single put. The catch: protection only works between the two strikes. If the stock falls *past* your short put, you're naked again on the way down. Useful when you're hedging a -10% scenario, not a -30% one.
Collar:Buy a put, sell a call against shares you own. The short call finances the put. Done right, it's free. The cost is your upside — if the stock rips, you get called away or have to roll the call up (paying debit). Best for positions where you've already had a run and want to lock in gains without selling (and triggering taxes).
Put spread collar:Sell the call, use proceeds to buy a put spread instead of a single put. Cheaper than a regular collar, but you have a tail gap below the long put — that's where the real damage happens in a crash. A compromise, not a clean hedge.
Long-dated LEAPS put:12-24 month put. High absolute cost, but the *annualized* cost is often lower than rolling quarterly puts. Theta decay is slower far from expiry. Good for strategic, "I want a tail hedge on" exposure rather than event-driven hedging.
Risks and counterpoints
⚠ Hedges have a hit rate problem. Most of the time, the market goes up. Most of the time, your hedge expires worthless. Over 5-10 years, the drag from constantly hedging often exceeds the drawdown it would have prevented. The math only works if (a) you size the hedge correctly, (b) you actually execute monetization when it pays off, and (c) you stop hedging when it's not needed.
⚠ Collars cap you in the wrong moments. The years you most regret a collar are exactly the years markets rip 25%+. Selling a call on NVDA in January '23 felt smart on Jan 2 and terrible by July.
⚠ Put spreads give false comfort. A -10% hedge against a -30% event is no hedge. Know what scenario you're actually protecting against.
⚠ IV is the silent variable. Buying puts when VIX is 30 is expensive; you're paying crash premium. Buying when VIX is 12 is when hedges are cheap. Most people do the opposite — they hedge after the move, when premiums are already inflated.
⚠ Tax friction. Collars on long-term holdings can trigger constructive sale rules if too tight. Protective puts on stock held <1 year suspend the holding period. Worth checking before structuring around concentrated positions.
✓ What works in practice: Hedge specific risk (earnings, single-stock concentration, macro event windows) rather than running a permanent insurance program. The cheapest hedge is *position sizing* — if you can't afford the drawdown, the answer is usually to own less, not to buy puts.
What would change the view
• IV environment: If implied vol is near multi-year lows on your name, protective puts get attractive on a cost-per-unit-protection basis. If IV is elevated (post-event, crisis), shift toward put spreads or collars where you're selling some of that rich vol back.
• Holding period and tax basis: Large unrealized gains shift the math toward collars (avoids realizing the gain). Short-term holdings push you toward simple puts to avoid holding-period complications.
• Concentration: Single-name >20% of portfolio? Collar or zero-cost structure makes sense — you're protecting an idiosyncratic risk. Diversified book? Often cheaper to hedge at the index level (SPX/SPY puts) than name-by-name.
• Defined catalyst vs. ongoing: A known event (FOMC, earnings, election) favors a tight DTE put or put spread covering just the window. Ongoing tail risk favors longer-dated structures.
• Your actual scenario: If you tell me the position (ticker, size, cost basis, holding period) and what you're worried about, I can run real strikes and price out the trade-offs concretely. The abstract comparison only gets you so far.