Read the regime
OpenLL classifies completed daily data as BULL, NEUTRAL, or BEAR with a volatility-aware EMA band.
Corrected research · allocation level 50
A cross-asset strategy tested from 2000 to 2026.
Meridian uses OpenLL to follow established trends across stocks, crypto, ETFs, bonds, and commodities. It ranks what is working, builds a target portfolio, and can hold cash when trends disappear.
Research only. Crypto uses completed Binance USDT spot candles; stocks, ETFs, bonds, and commodities use Yahoo adjusted daily OHLC. Current-survivor universe. Includes a modeled 0.10% cost per traded dollar. Historical results are not a forecast. Read the OpenLL methodology.
Loading verified research data...
From the last research observation before January 2021, a modeled $10,000 grew to $62,896 in Meridian versus $21,661 in the S&P 500. Meridian's recent maximum drawdown was deeper.
January 2021 to latest
Start with the period everyone can compare.
Same dates. Same modeled $10,000 start. Meridian compounded faster, but its recent maximum drawdown was deeper than the index. Both facts belong in the decision.
| Measure | Meridian | S&P 500 |
|---|---|---|
| Growth of $10,000 | $62,896 | $21,661 |
| Total return | +528.96% | +116.61% |
| Annualized return | +39.42% | +14.99% |
| Maximum drawdownclose to close | -30.33% | -24.34% |
| Longest underwaterpeak to recovery, rounded | 25 months | 23 months |
| Latest recovery statusas of Jul 13, 2026 | Underwater: -5.07% | Underwater: -1.25% |
Yearly returns
Calendar boundaries use the nearest committed observation on or before each boundary. 2026 is year to date through Jul 13.
| Year | Meridian | S&P 500 |
|---|---|---|
| 2021 | +149.18% | +29.71% |
| 2022 | -16.97% | -18.38% |
| 2023 | +26.99% | +26.18% |
| 2024 | +55.20% | +26.79% |
| 2025 | +18.74% | +16.96% |
| 2026 YTD | +29.90% | +9.35% |
Corrected 2000-2026 research
The five-year window is not the whole case.
The default book was also tested through the dot-com decline, the global financial crisis, COVID, and 2022. The longer window is Meridian's central evidence, with its limitations left attached.
MeridianAllocation level 50
$1,096,917
S&P 500Buy and hold
$82,098
Meridian annualized at 19.38% with a -31.18% close drawdown and -33.27% daily-bar-low stress drawdown. SPY annualized at 8.26% with a -55.19% close drawdown.
The strategy did not win every rally or every year. Its research advantage came mainly from avoiding the deepest market losses, then re-entering after trends were established. That confirmation can lag sharp recoveries.
The universe grows over time, and the crypto sleeve is dormant before 2015. Current-survivor bias remains unresolved, so these values are evidence of historical behavior, not achievable-return estimates.
76.06%
of 259 rolling five-year monthly starts beat SPY
100%
of those five-year windows finished positive
4 of 5
non-overlapping five-year blocks beat SPY
Rolling windows overlap, so 76.06% is persistence evidence, not 259 independent samples and not a forward probability. The five non-overlapping blocks are fewer but independent in time.
The strategy underneath
OpenLL finds the trend. Meridian builds the book.
The system does not predict bottoms or forecast earnings. It waits for established price trends, ranks the strongest opportunities, and changes the target when the evidence changes.
Price is treated as the testable summary of fundamentals, news, and flows. The tradeoff is deliberate: Meridian can miss the first part of a recovery, but it avoids making untestable predictions about what should happen next.
OpenLL calculation, source, and license →Rank conviction
Band, structure, support, and momentum rank the eligible universe across five asset classes.
Compose the target
The portfolio applies adaptive class budgets, per-name limits, and subsector candidate limits. Unused weight can remain in cash.
Trade each market on its clock
The target refreshes daily. Crypto is checked daily and updates above 20% sleeve drift. Stocks and the other non-crypto assets update above 10% drift only after a 30-day lock.
15%hard single-position cap at default level 50
2non-ETF candidates at most per subsector
1ETF candidate at most per sector
5 classesadaptive budgets can redistribute toward qualified trends
Cashremains when no class can absorb its target responsibly
Follow it your way
The tools make the strategy easier to use.
The simplest path is Telegram. The full app adds allocation choice, imported holdings, goal planning, and the evidence behind every book.
Meridian Portfolio Bot
One choice, then only meaningful changes.
Choose the allocation level you want to follow. The bot remembers it, shows today's effective target, and sends a personal update only when crypto clears its 20% band or non-crypto clears its 10% band after 30 days.
/portfolioChoose one of 21 allocation levels/bookSee the current positions and cashalertReceive only eligible crypto or monthly non-crypto updates
Use the full decision workflow
Inspect the target, match it to your circumstances, and generate changes from the holdings you actually own.
01
Choose an allocation level21 books, each paired with its observed historical drawdown reference.
02
Bring your current portfolioImport holdings and apply a separate account-level 10% drift check before calculating orders.
03
Plan around a real goalThe wizard maps target, date, and contributions to an allocation using declared assumptions, not backtest CAGR.
Agent API · public testing
The same strategy, machine-readable.
Software agents can read the methodology, current signals, all 21 research choices, and today's target book as decision-ready JSON. Public testing currently requires no authentication.
Four endpoints. No hidden second model.
The app, Portfolio Bot, and API read the same daily target. The API exposes current decisions and evidence boundaries without publishing user holdings or hidden historical series.
More in Meridian
Depth when you need it.
The landing stays focused on the strategy. The supporting research and signal tools remain one step away.
Cross-asset signalsCurrent BULL, NEUTRAL, BEAR states and flips
Backtest replayInspect historical signal behavior
Orange BTC and ETHSpecialist OpenLL and anchored VWAP timing
Public daily channelOne read-only digest of market flips
OpenLL methodologyExact calculation, Python source, and MIT license
Two-minute product tourSee the real Meridian workflow
Forward shadowFrozen model record without historical backfill
The 100 Apps ProjectReturn to the studio
Follow the target, not another stream of market noise.
Choose an allocation level in Meridian Portfolio Bot. Check the current book any time, and receive a personal message only when the model target changes enough to act.