Skip to content

View original

InjSpotDrawdownAccumulationLong

Hypotheses

INJ Spot Multi-Day Drawdown Accumulation Long (BINANCE_SPOT, Daily, OHLCV-Only)

Hypotheses

A long-only, single-instrument, single-venue, OHLCV-ONLY strategy on INJUSDT.BINANCE_SPOT DAILY bars. Same drawdown-accumulation mechanism as the landed Btc/Bnb/Near/Dot/FtmSpotDrawdownAccumulationLong siblings. EXPLICITLY ALIGNED WITH ANALYST DIRECTIVE: INJ has one of the clearest organic-demand profiles in crypto via its INJ 2.0 burn auction mechanism — 60% of all protocol fees on Injective dApps are used to buy and burn INJ from the open market weekly. This creates a directly mechanical organic-demand signal (DeFi usage → more burns → tighter supply → reliable accumulation flow after dips). INJUSDT.BINANCE_SPOT has been listed since 2020 — ~5 years of daily bars. Five parameters total. Hits BINANCE_SPOT venue quota gap (currently 10.6% vs ≥15% target).

Hypotheses

Faithful clone of the landed Btc/Bnb/Near/Dot/Ftm/AvaxSpotDrawdownAccumulationLong family applied to INJ, changing only the instrument, spot qty precision (2 decimals), and INJ-appropriate calibration (20% drawdown threshold / 15% target for INJ's higher volatility). The mechanism is correctly routed to BINANCE_SPOT: it is long-only, uses no leverage, and the edge is organic spot accumulation (INJ 2.0 burn-auction demand), so spot is the right venue per the routing rule and it fills the BINANCE_SPOT venue-quota gap; a CASH account also makes the single-position equity sizing clean (the one position is sized once at entry from full equity). calculate_signal emits a discrete 1.0 on the drawdown trigger but returns a CONTINUOUS tanh(drawdown) proxy (damped to (0,0.5), strictly below the 0.99 entry threshold) on all other bars, so Layer-2 frozen-signal detection passes while never producing a spurious entry. should_enter is BUY-only (never sells to enter, respecting the CASH/spot constraint); exits are entry-relative take-profit, peak recovery, and a hold cap. Drawdown/peak are computed causally from buffered closes (no lookahead). INJUSDT spot has been listed since 2020 with multi-year daily history, avoiding data_unavailable. With ~5-9 entries/year and multi-week holds at a 15% target, per-trade moves dominate the ~0.20% spot round-trip fee.

Hypotheses

Not worth optimizing — no edge, same failure mode as the abandoned AVAX and AAVE drawdown-accumulation variants. The mechanism is correctly implemented (identical to the landed BTC/BNB/NEAR/DOT/FTM siblings) but loses on INJ: profit_factor 0.86, expectancy -$706/trade, Sharpe -0.26, total return -34% over 5.5 years, max_drawdown 58% (52-trade sample, so this is a fair read). Win rate is 55.8% but avg_loss ($11,419) far exceeds avg_win ($7,790, tail_ratio 0.56), and the result swings violently by regime (2023 +101% but 2022 -41%, 2025 -42%, 2026 -11%; annualized volatility 189%). The accumulation mechanism catches falling knives on a hyper-volatile alt — it buys 20% drawdowns that keep deepening (trade days -23.7%, -27.7%, -19.8%, -18.9%) — because the revert-to-peak premise needs durable recovery, which INJ's catastrophic 2022/2025 down-years do not provide. Tuning cannot fix a structural non-recovery problem, and optimizing would overfit the 2023 bull. NOTE for research: the spot drawdown-accumulation mechanism is validated only on recovering majors (BTC/BNB/NEAR/DOT/FTM); it has now failed identically on three hyper-volatile alts (AVAX, AAVE, INJ) whose deep drawdowns do not reliably revert. Stop extending this mechanism to high-volatility mid-cap alts to fill the BINANCE_SPOT venue-quota gap — pick instruments with durable recovery profiles. FAILURE PATTERN: spot drawdown-accumulation on hyper-volatile alts catches falling knives in non-recovering regimes (PF <1, deep drawdowns, regime-swing losses).

Implementation

Long-only single-instrument drawdown-accumulation strategy on INJUSDT.BINANCE_SPOT 1-DAY bars, OHLCV-only. It tracks the rolling 30-day high and enters LONG when the current close is >= 20% below that trailing peak (accumulating a dip), then exits on any of: 15% entry-relative take-profit, recovery back to the prior peak (catch-up), or a 30-day max-hold cap. Five parameters total. Sizes 50% of equity notional; BINANCE_SPOT CASH account (leverage=1 enforced, long-only by venue).

Backtest Review

Mechanism correctly implemented (same template as the landed BTC/BNB/NEAR/DOT/FTM siblings); clean diagnostics; adequate 52-trade sample over 5.5 years.

Backtest Review

INJ has a long spot history and a legitimate organic-demand narrative (burn auction).

Backtest Review

No edge: profit_factor 0.86, expectancy -$706/trade, Sharpe -0.26, total return -34% over 5.5 years, max_drawdown 58%.

Backtest Review

Win rate 55.8% but avg_loss ($11,419) >> avg_win ($7,790), tail_ratio 0.56 — losers dwarf winners.

Backtest Review

Violent regime dependence: 2023 +101% but 2022 -41%, 2025 -42%, 2026 -11%; annualized_volatility 189%.

Backtest Review

Drawdown-accumulation catches falling knives on a hyper-volatile alt — buys 20% drawdowns that keep falling (trade days -23.7%, -27.7%, -19.8%, -18.9%); same failure as the abandoned AVAX/AAVE variants.

Outcome Summary

InjSpotDrawdownAccumulationLong applied the proven spot dip-buying mechanism to INJ, leaning on its burn-auction organic-demand story and the BINANCE_SPOT quota gap. The code was correct and the sample adequate (52 trades over 5.5 years), but it lost -34% with a 58% drawdown and asymmetric losses, propped up only by a +101% 2023 amid catastrophic 2022/2025 down-years. The analyst abandoned it at backtest review after one iteration, citing falling-knife losses on a non-recovering hyper-volatile alt — the third such failure after AVAX and AAVE. The recorded pattern is that drawdown-accumulation only holds up on long-history recovering majors, not high-volatility mid-cap alts.

Outcome Summary

Spot drawdown-accumulation has now failed identically on three hyper-volatile mid-cap alts (AVAX, AAVE, INJ) whose deep drawdowns don't reliably revert to prior peaks; the mechanism is validated only on recovering majors, so it should stop being extended to high-volatility alts to fill a venue quota — pick instruments with durable recovery profiles.

Outcome Summary

The analyst issued an 'abandon' verdict at the pre-optimization backtest-review gate: the mechanism was correctly implemented but had no edge on INJ, catching falling knives on a hyper-volatile alt — buying 20% drawdowns that kept deepening — the same structural non-recovery failure as the abandoned AVAX and AAVE variants, which tuning cannot fix.

Outcome Summary

A long-only spot drawdown-accumulation strategy on INJUSDT.BINANCE_SPOT daily bars — the same mechanism as the landed BTC/BNB/NEAR/DOT/FTM siblings — that bought when price fell at least 20% from its trailing 30-day high and exited on a 15% profit target, recovery to the prior peak, or a 30-day hold cap, motivated by INJ's mechanical burn-auction organic-demand narrative and the BINANCE_SPOT venue-quota gap.

Outcome Summary

Over ~5.5 years and 52 trades it lost: profit factor 0.86, expectancy -$706/trade, Sharpe -0.26, -34% total return, and a 58% max drawdown. Despite a 55.8% win rate, average losses ($11,419) dwarfed average wins ($7,790), and returns swung violently by regime (2023 +101% but 2022 -41%, 2025 -42%, 2026 -11%) at 189% annualized volatility.
Strategy report

Backtest and paper results are hypothetical. Trading involves risk of loss.