The thing that finally made hedging click for me was realizing that selling and hedging look similar from a distance but are two different tools. If I sell my spot, I owe tax on the gain, I lose my cost basis, and I have to time a re-entry that I will almost certainly get wrong. If I open a short against the same position instead, my downside is neutralized for as long as I hold it, my long stays where it is, and I pay a small carrying cost. Neither is free. But they fail in different ways, and knowing which failure I am signing up for is most of the work.
A perpetual short is the cleanest way to do this because a perp has no expiry to roll and its price is tethered to spot through funding. I short roughly the same dollar amount I am long, and the two positions move against each other. If BTC drops ten percent, my spot loses value and my short gains almost exactly that much back. I have parked the position. The catch is that "almost exactly" hides a few decisions that decide whether the hedge works.
Sizing the hedge: full versus partial delta
The first decision is how much of the position to neutralize. A full delta hedge matches my short notional to my long notional so the net exposure is close to zero. If I hold roughly fifty thousand dollars of spot, I short roughly fifty thousand of perp. Now I am flat. I do not make money if the market rips, and I do not lose money if it dumps. I have converted a directional bet into a parked position that costs funding to hold.
Most of the time I do not actually want to be flat. I want less exposure, not zero, usually because I still think the position is fine long term and I am hedging a specific event. That is a partial hedge. If I short half my notional, I have cut my downside in half and kept half my upside. Think in target delta. If I want to keep thirty percent of my exposure, I hedge seventy percent. Write down the delta you actually want before you size anything, because "hedge it a bit" turns into guessing at three in the morning.
One trap is leverage math. A perp lets me post a fraction of the notional as margin, so a five thousand dollar balance at ten times leverage controls fifty thousand of short. The hedge size is the notional, not the margin. People under-hedge because they anchor on the cash they put up rather than the exposure they opened. Size to notional, then keep your margin buffer fat enough that a rally does not liquidate the short before your spot catches up.
Proxy hedging alts through BTC or ETH
The messy part is that most portfolios are not one clean BTC bag. They are a spread of alts, and many of those either have no liquid perp or have one so thin that shorting it moves the price against you. So instead of shorting each name, I short a proxy. Alts tend to move with BTC and ETH, just more violently, and I use that relationship to hedge a basket with a single liquid short.
The number that matters is beta, roughly how much the alt moves for a given move in BTC. If an alt historically moves about one and a half times as hard as BTC in the same direction, its beta is around 1.5. To hedge ten thousand dollars of that alt, I short about fifteen thousand of BTC, not ten, because the alt will fall further in a selloff. Do this across the basket and sum it up.
- For each holding, estimate its beta to your proxy from a decent stretch of history, ideally a window that includes a real drawdown and not just a calm uptrend.
- Multiply each holding's dollar value by its beta to get its BTC-equivalent exposure.
- Sum those beta-adjusted values and short that total in the proxy perp.
The honest caveat is that beta is not a constant. In a normal market alts and BTC drift apart and the proxy hedge is loose. In a real crash correlations snap toward one and everything falls together, which is exactly when you want the hedge tight, so the proxy works best when you need it most. But there is always basis risk. Your alt can bleed on its own bad news while BTC sits still, and the short does nothing there. A proxy hedge protects against market-wide moves, not against your specific coin breaking.
Funding is the price of admission
Holding a perp short is not free, and the meter running in the background is funding, a periodic payment between longs and shorts that keeps the perp price pinned to spot. When longs are crowded, funding is positive and longs pay shorts. When everyone is piled into shorts, funding goes negative and shorts pay longs.
For a hedger this cuts both ways. If I short into an overheated, greedy market, funding is usually positive, which means I get paid to hold my hedge. That happens more than people expect, because hedges are most tempting exactly when the market feels frothy. If I short into fear, funding may be negative and I am paying to stay hedged, which quietly erodes the position. Before I open a hedge I check the funding rate and do the rough arithmetic of what a week or a month of it costs against the notional. Usually it is small relative to the downside I am covering. Occasionally, in a very stretched market, it is large enough to change the plan.
When to hedge instead of sell, and how to unwind
Hedging earns its keep in a few situations. When selling would trigger a tax event I do not want yet, a hedge neutralizes risk without realizing the gain. When the position is illiquid or large enough that dumping it would tank my own exit price, a liquid perp short lets me get flat instantly while I unwind the spot slowly or not at all. And when I have a defined window of worry, an unlock, a macro print, a stretch where I am traveling and cannot watch, a hedge buys peace for a known cost.
Unwinding is where people hurt themselves, usually by treating the two legs as separate trades. The legs are supposed to offset, so I close them together or nearly so. If I panic and buy back the short after a sharp drop because it looks like the bottom, and the market keeps falling, I have re-exposed my spot at the worst moment. That is whipsaw, and it turns a working hedge into two bad trades. My rule is to decide the exit condition when I open the hedge, not in the middle of the move. Either the event has passed, or price has hit a level where I am comfortable being long again. When that is met, I lift the short. Keep a margin buffer to the end so a last rally does not liquidate it a day before you meant to close.
Done this way the hedge is boring, which is the point. I know my worst case, roughly what carrying it costs, and in advance what makes me take it off. The mistakes all come from improvising one of those three after the position is already on.