Strategy archetypes

Most trading strategies are variations on a handful of well-known archetypes. Recognizing them is useful because it tells you what a strategy is really betting on — and where its risk lives. Parachute's own strategies are all instances of standard archetypes expressed in the shared component vocabulary; nothing here is exotic.

Read this page skeptically. Descriptions of what an archetype aims to do are not claims that it will make money on your account. Public strategy write-ups almost always quote their own unverified performance; treat those numbers as marketing, not evidence. The only evidence that counts here is a backtest of the exact strategy you intend to run — and even that is an approximation of the past.

The two families

Strategies sort into two families by what they bet on:

Directional — a bet on price direction:

  • Trend following — ride a sustained move and exit when it fades.
  • Momentum — lean toward whatever has been moving.
  • Breakout — position for a quiet market becoming a volatile one.
  • Scalping — many small, short-horizon captures.

Relative-value / non-directional — a bet on a pricing relationship or a risk premium rather than direction:

  • Mean reversion — a price snaps back toward its average after an extreme.
  • Pairs / statistical arbitrage — trade a stable spread between related assets.
  • Arbitrage — lock in a price or timing dislocation with little directional risk.
  • Carry / theta harvesting — collect a risk premium whose defining feature is its return shape — small premiums collected often, with exposure to occasional large losses — rather than a directional view.
  • Market making — quote both sides and earn the spread.

Every archetype is the same three-stage pipeline with a different signal filling the Model slot and a different distinguishing filter or hedge. That is the whole point of a shared vocabulary: once you can read the stages, you can read any strategy.

The strategies these map to

Parachute's own strategies are instances of these archetypes. The ones below are how they decompose into the component vocabulary — read as what each aims to do, not a claim about results:

  • Short-Vol Carry (carry / theta) — the reference strategy. A Jump-Diffusion Fair-Value Model prices each contract and trades only where the edge clears the filters, hedged with a cheap Tail Hedge. It aims to trade selective mispricings the model finds, not an "always-on" premium. Its risk shape is short-volatility: many small outcomes, exposed to occasional large adverse moves — which is why the filters and the hedge are load-bearing.
  • Deep-ITM Moneyness Capture (carry variant) — the backbone restricted by a Moneyness Filter to deep-in-the-money contracts. Its identity is instrument selection. It carries the loudest risk caveats of any of these: the premium it targets is small and its losses can cluster, so size caps and hedges matter most here.
  • Price-Time Arbitrage (arbitrage) — targets the brief window where a Kalshi contract price lags fast-moving spot, capping the paired cost with a Hedge Ladder instead of a Tail Hedge. Its distinguishing component is the ladder. Public write-ups of this latency idea exist but their profitability claims are unverified.
  • Calm-Regime Carry (carry variant — planned) — the same backbone as Short-Vol Carry with the Volatility-Regime Filter promoted to the defining gate, so it pauses in turbulent conditions rather than trading through them. A clean example of "same backbone, one filter is the identity."
  • Penny-Longshot (longshot / lottery — not currently running) — small fixed-budget bets on cheap, far-out-of-the-money contracts. It is the complement of buying favorites — the side of the favorite-longshot bias the research identifies as underpriced rather than overpriced.

Where several of these share a backbone and differ by a single distinguishing filter or hedge, that is the shared-vocabulary payoff at work — and it is the direction Parachute is building toward, where you clone a known composition rather than start from a blank canvas.

A worked example: why selectivity, not carry

It is tempting to think you can simply "buy the favorites and collect the premium" — a pure, model-free carry bet. On this venue, that intuition is a trap, and it is worth understanding why because it is the single most common mistake.

The classic favorite-longshot bias — favorites being underpriced and longshots overpriced — is one of the most-replicated results in betting markets. But that bias is strongest where casual participants dominate (horse racing, novelty bets) and weakest in sophisticated, fast markets. An internal measurement on Parachute's own venue found no harvestable favorite premium net of fees — buying favorites at the market here tends to donate the spread, the fee, and the hedge cost. So an unconditional "buy the favorites" strategy is rejected on principle.

The lesson generalizes: on a competitive venue, a durable edge comes from selection and tail control — a model that finds the specific mispricings worth trading, filters that skip everything else, and exits that bound the downside — not from a broad, always-on bet. If a strategy's entire thesis is "collect a premium with no model edge," be very skeptical.

Where to go next