OLTA Backtest Research

Programme launched: 2025-09-11 Last updated: 2026-04-16 Author: OLTA Research Status: Internal research, pre-publication. Intended to seed the public Research section of the OLTA app in 2026-Q3.

About this research

The OLTA backtest programme is a complete rebuild of the index library's evidence base. Through Q1 and Q2 of 2026, the team built a backtest engine, backfilled two years of daily price data for the crypto constituents and five years of daily data for the underlying equities behind the tokenised-equity baskets (routed through Dinari for single stocks and Backed Finance for ETFs). The engine simulates each of the 74 indices under its configured rebalance strategy across the available window, computes the standard institutional performance metrics, and stress-tests the results against eight historical crisis windows.

This research collection is the consolidated output. The intended audience is institutional allocators, professional research desks, and the OLTA team itself when defending design decisions to product and engineering stakeholders.

To cite this work in external materials, use the format: OLTA Research, "[Paper title]" (publication date as listed in the paper). research/backtest/[filename].md.

For questions, follow-ups, or the methodology brief (which carries the detailed parameters, per-basket weights, per-strategy variants, and the rolling-window dispersion configuration), contact OLTA Research.

The research framework is reproducible. To re-run the pipeline against fresh data, see the methodology paper section 8.

Table of contents

1. Executive Summary - The headline finding in one page. Read first.
2. Methodology - How the backtest engine works. Data sources, NAV computation, rebalance strategies, metrics, stress windows.
3. Current-State Analysis - All 74 indices' performance over the 2-year window. Rankings, family-by-family discussion.
4. Diversified Family Paper - The flagship finding. Six cross-asset baskets, five Live, one Watchlist.
5. Stress Test Report - Crisis window behavior across families. Aug 2024, Feb 2025, April 2025, Sept 2025.
6. Sharpe Decomposition - Why the high-Sharpe baskets are high-Sharpe. Driver attribution at the framework level.
7. Outperformance vs BTC - The 20-of-74 finding. Cross-asset construction as the path to BTC outperformance in this regime.
8. Family Summaries - One concise section per family with metrics and framework reading.
9. Downside-Risk Framework - The VaR/CVaR/Sortino/tail-beta stack computed for every index. Holding-horizon Value-at-Risk at 7 and 30 days, the cross-book tail distribution, and how the framework reaches the order screen.

Headline numbers at a glance

• BTC reference (2024-05-21 to 2026-05-21): Sharpe 0.14, total return +11.2%, max drawdown -49.5%.
• Top Sharpe (multi-year basket): the Semis Equities member in the [2.2, 2.4] band over 980 days.
• Top Diversified: the Sharpe-Max member in the [1.5, 1.6] band over 729 days, total return in the high-double-digit range, max drawdown roughly half of BTC's.
• Indices beating BTC's Sharpe: 20 of 74.
• Indices with positive 2-year return: 30 of 74.
• Indices with max drawdown smaller than BTC's -49.5%: 12 of 74.

Note on disclosure

The public research surface presents the framework, the family taxonomy, the headline outcomes in banded form, and the qualitative reading. Per-basket constituent weights, the rolling-window dispersion configuration, the regime-thresholds detail, and the per-strategy variant tables are documented in a methodology brief issued to institutional counterparties and grant reviewers under NDA on request. This follows the standard institutional disclosure pattern used by S&P DJI, MSCI, and BlackRock iShares for their published index methodologies.

Note on the regime

Every conclusion in this research collection is conditioned on a specific 2-year window. That window was a BTC-sideways, US-equity-bull, gold-bull, alt-coin-crash regime. The Diversified family was constructed and tested against exactly that regime; its outperformance is not regime-agnostic. A 2017-2018 window, a 2020-2021 window, or a 2030-2031 window may produce materially different relative rankings.

The framework is honest about this. Each paper has a Caveats section. The collection should be read with the regime fingerprint in mind.

Repository state

This document represents the published state of OLTA's research programme as of 2026-04-16. Updates:

• 2026-Q4: re-run of the backtest after another quarter of data, especially for the post-launch RWA family.
• 2027-Q1: full revisit, including a fresh design cycle for the Strategy family.
• 2027-Q2: re-evaluation of the Watchlist Diversified member for promotion to Live.

Future research not yet covered:

• A "futures + spot" comparison for the Sector family (basis trade as a separate alpha source).
• A "structured product" wrapper analysis (e.g., a basket with downside protection via options).
• An optimisation framework that takes the OLTA constituent universe and produces a tangent portfolio explicitly.
• A pre-trade analytics module: tracking error to BTC, beta-corrected returns, and a regime-classifier overlay.

Methodology and analysis published under the OLTA Research disclosure framework. Detailed parameters and per-basket weights are available in the methodology brief on request to institutional counterparties and grant reviewers.