Ride the trend.
Respect the exit.
Compound
for decades.
Tideline is a systematic strategy across major US equity indices that adjusts market exposure to protect capital in drawdowns. Rules first, discipline always, no guesswork.
When the market breaks,
Tideline steps aside.
Growth of the strategy versus a passive buy & hold benchmark. The story is in the selloffs: buy & hold bleeds through every drawdown while Tideline pulls exposure back.
Every figure is computed from institutional-grade backtested results and updated weekly with real data.
Performance since inception versus SPX index. Data as of .
Hypothetical performance: backtested simulation, not live trading. See methodology and disclosures.
Sometimes it will trail a raging bull. That's the point.
When markets run hot and euphoric, Tideline will sometimes hold back while buy & hold sprints ahead. By design. A strategy built to protect capital cannot also chase every last percent of a stretched rally, and trying to do both is how investors get hurt.
Discipline is the product. The monthly note tells you exactly what the model did and why, so when Tideline steps back, you understand the reasoning instead of second-guessing it. Staying invested through the plan is where the results come from.
“The goal isn't to win every month. It's to still be standing, and compounding, after the months that break everyone else.”
Three rules, followed without exception.
Rules-based allocation
The model allocates across major US equity indices by rule, a systematic split driven by market conditions, never a hunch or a headline.
De-risks when stretched
When the market runs far above trend, systematic models reduce exposure, pulling back from stretched conditions before they unwind.
Capital protection first
Shallow drawdowns are the priority. The strategy is designed to survive first and compound second, protecting capital over chasing every rally.
Every move, on the record.
When the model changes its allocation, we publish it, timestamped, the day it happens, never edited after the fact. Our signal history is a permanent, append-only record you can audit yourself. No hindsight, no quiet edits, no cherry-picking. This is what systematic really means: the record speaks, not us.
The systematic part is the whole point.
Every Tideline strategy is researched and built by systematic machine-learning models, then overseen by a person. Those models study decades of market data, form and test hypotheses, and put what they learn through backtests and through stress, robustness, and survival testing before anything is allowed to trade. What survives is expressed as fixed, rules-based strategies. The models do the research and validation; the rules do the trading, the same way for every subscriber. That is why the strategy is systematic: there is no discretion to talk it out of a rule once it is set.
Borja Peñalver
Founder of JBPS Capital and portfolio manager for Tideline. He sets the strategy's mandate, oversees the models, and is accountable for what runs in production.
LinkedInStart free. Upgrade when you're ready.
For following along and learning the discipline.
- ✓Monthly allocation summary
- ✓Email updates
For investors following the model in their own account.
- ✓Flagship signal line: every model state change, delivered in real time
- ✓Regime Dashboard: the system's live state at every close: current regime, how stretched the market is, how close we are to the next signal
- ✓Weekly “what the model did & why” note
30-day free trial; a payment method is required at signup but you will not be charged until the trial ends. Unless you cancel before the trial ends, your subscription begins automatically at $49 per month and renews monthly until you cancel. Cancel anytime from your account settings; cancellation takes effect at the end of the trial or current billing period. No pro-rata refunds for partial periods. A confirmation email with these terms will be sent at signup.
The strategy would run directly in your own brokerage account. You never wire us money. Your capital stays with your custodian.
- ◇Direct execution in your account
- ◇You hold custody at all times
Managed accounts will be offered separately through JBPS Capital LLC upon completion of its investment adviser registration, which is currently pending. Until that registration is effective, JBPS Capital LLC does not offer, and is not soliciting clients for, any advisory or managed account service. The Tideline signal publication is not an advisory service and is offered independently of any future managed account offering; subscribing to Tideline creates no advisory relationship and confers no priority or right with respect to any future offering.
A backtest is a claim. We treat it like one.
Most strategy backtests fail quietly. Not because the idea was wrong, but because the measurement flattered it. Frictionless fills. Free leverage. A hundred variants tested, one shown. We built our research process specifically to make that impossible for ourselves. Every strategy we publish has survived the same gauntlet. Most research candidates don't.
Simulated like it's real
All headline results come from an institutional-grade, event-driven simulation engine, not a spreadsheet. Orders are validated against real brokerage margin rules and can be rejected, exactly as a broker would reject them. Leverage pays actual historical financing rates, day by day. Then we re-run everything with costs multiplied and every signal executed a full day late. If a strategy can't survive being traded badly, we don't trust it traded well.
Verified against the tape
Simulated fills are checked against exchange tick data: the actual national best bid and offer at the moment of each fill, including every fill during crisis sessions. We then deliberately re-price every fill at the worst side of the quoted market, plus a penalty, and require the strategy to survive. An independent accounting replay must match the engine within strict tolerance.
Tested against our own bias
We log every experiment, including the failures. Our flagship research reflects a registry of over 1,400 recorded trials, and that number isn't a confession: it's an input. Statistical significance must beat not just the market, but the luck of 1,400 attempts. Acceptance rules are written down before results come in.
Stress-tested where it hurts
Every strategy is decomposed across bear markets, corrections, melt-ups, and trendless grinds, and its weak regimes are disclosed, not averaged away. We bootstrap thousands of alternative histories to estimate how bad drawdowns could get, not just how bad they were, and replay 1987-scale crash scenarios against the strategy's worst historical posture.
Frozen, then watched
Once validated, a strategy's parameters are frozen under a versioned identifier. Its signals are recorded daily in an append-only ledger where every entry is cryptographically chained to the last: retroactive edits are detectable by construction. No parameter changes. No do-overs. Forward evidence accumulates before any capital decision, ever.
What we won't claim
Hypothetical results are hypothetical, no matter how carefully they're built. They benefit from hindsight and are not a live track record. We won't claim one until it exists, and when it does, it will be reported from inception, append-only, separately from any backtest. We'd rather tell you that here than have you discover it later.