A backtested, 2x levered momentum strategy that rotates into the levered equity ETF with the lowest RSI (momentum) when volatility isn’t extreme, deploying about 70% to that fund and 30% to a short-term Treasuries ETF. If volatility spikes or momentum indicators exceed risk thresholds, it shifts to a bond-heavy, safer mix and uses VXX-based risk controls to catch black-swan scares. It’s a mechanical, backtestable approach to capturing upside in trends while trying to limit losses in shocks.

- Universe: The strategy scans a set of leveraged equity ETFs (examples shown in the code include ProShares and Direxion products like TQQQ, SPXL, SOXL) and a set of bond/short-term Treasury ETFs (eg. BIL, SHY, IEF, TLT, TMF, TMV, TMF). It also uses VXX (a volatility-related instrument) as a risk gauge. - Primary rule (when markets aren’t too volatile): Among the levered equity ETFs, compute a short-term momentum measure called RSI for each (a number that helps judge how strong and extended a move has been). Sort those ETFs by RSI and choose the one with the lowest RSI (the “least hot” among the levered funds). Put the bulk of capital into that fund (70%), and put the remaining 30% into a safe, short-term Treasuries ETF (BIL). The idea is to ride a levered index fund that isn’t yet overbought, while preserving some ballast in cash-like bonds. - Rebalancing rule within this mode: The weighting shown (roughly 70/30) is explicit in the rules. If conditions remain favorable, you stay with that choice; if conditions deteriorate, you shift. The code shows nested checks to see if another levered fund should be chosen next time, always aiming to keep exposure tilted toward the one with (relatively) stronger, but not overbought, momentum. - Risk control / Black swan catcher: There are explicit blocks that reference VXX (volatility futures) and high RSI readings (e.g., RSI thresholds around 80) to catch spikes in fear/volatility. When those are triggered, the strategy steers toward the bond sleeve rather than staying fully in levered equities. This acts as a hedge against sharp market shocks. - Bond-market checks: There are “Bond Market Check” blocks that compare bond proxies (like IEF, TLT, SHY, BIL) against each other and to equity benchmarks (like SPY). These checks use moving averages, standard deviation and price comparisons to decide whether to tilt toward short-term Treasuries or longer-duration bonds in the mix, i.e., adding a risk-off flavor when bonds are signaling safety or risk-off environment. - Secondary modes and refinements: The nested structure includes a “Fund Surfing (2x)” path and variants that attempt to combine multiple signals (momentum, volatility, bond strength) and even a “double pop bot” approach to capture repeated opportunities when markets swing, always with the goal of preserving capital in volatility regimes while still allowing upside in favorable ones. The code also shows some explicit weighting rules (e.g., 66% TQQQ with 34% BIL or 70/30 splits) and a preference for the bottom RSI among the candidate levered funds. - Backtest scope: The notes say “Backtest to 4/28/2009” and that this is a leveraged (2x) extension of Fund Surfing, with additional risk controls. It’s designed to be executed mechanically, not to be hand-tuned on the fly. - What happens in practice: In calm, trending markets, you’re mostly in a levered fund with bond ballast; in choppier or overbought regimes or when volatility spikes, you pull back exposure and overweight bonds. The exact levered ETF chosen can rotate over time based on RSI rankings and other checks, but the general blueprint is: leverage up on favorable momentum, shift to bonds when risk rises, with occasional risk-hedging triggers to avoid big drawdowns.

Levered momentum with bond ballast and volatility hedges. Out-of-sample annualized return ~57% vs SPY ~22%, Calmar ~1.31, Sharpe ~1.15. Captures trend upside with risk controls; note higher drawdowns in shocks due to 2x exposure.