All Strategies / Horizon
Backtest Results

Horizon Strategy: Research & Backtest Results

A systematic, rules-based research model for Nasdaq exposure. This page presents historical backtest analysis, not live client account results. Live distribution follows a single uniform post-close publisher update.

Understanding This Research

What This Is

A rules-based research model that generates daily model-state allocations across QQQ, TQQQ (3× leveraged), and SQQQ (3× inverse) using systematic criteria on Nasdaq data.

What This Is Not

  • Not personalized investment advice
  • Not a trade instruction or execution service
  • Not a guarantee of future returns
  • Not a representation of live client trading results
  • Not suitable for all investors

Who This May Interest

Investors researching systematic strategies who understand leveraged ETF risks, are comfortable with high volatility, and want to review methodology and historical analysis before any further engagement. This research is generally not designed for investors prioritizing capital preservation, short investment horizons, or near-term liquidity needs.

About the Benchmark Comparison

We compare Horizon to QQQ (Nasdaq-100 ETF) as a common reference point. However, Horizon's actual holdings rotate among QQQ, TQQQ, and SQQQ. Because TQQQ/SQQQ are leveraged instruments that reset daily, direct comparison carries caveats: the strategy takes on higher risk than QQQ alone, and its returns reflect that leverage. This comparison illustrates relative performance patterns, not risk-adjusted equivalence.

Key Assumptions & Limitations (read before metrics):
  • All headline figures are hypothetical/backtested and use a same-day close execution reference basis (T+0).
  • Current publication is post-close; readers who choose to implement usually execute next trading day (T+1), which can materially change outcomes.
  • Displayed results are gross of transaction costs, slippage, taxes, and advisory/subscription fees.
  • This is general educational research, not personalized investment advice or a trading instruction.
Annualized Return (CAGR)
Hypothetical (T+0 basis)
32.9% vs 19.0%
Backtested/hypothetical · gross of costs & taxes
Maximum Drawdown
Hypothetical (T+0 basis)
-41.2% vs -35.1%
Peak-to-trough decline
Sharpe Ratio (rf = 0%)
Hypothetical (T+0 basis)
0.98 vs 0.95
Higher value indicates more return per unit of volatility; simplified for comparability (see methodology)
Backtest Period
Hypothetical (T+0 basis)
~16 yrs
Feb 10, 2010 - Feb 20, 2026
Worst Calendar Year
Hypothetical (T+0 basis)
-23.9% vs -32.6%
Based on annual returns shown (2010*–2026*)
Negative Years
Hypothetical (T+0 basis)
4/17 vs 3/17
Count of negative calendar periods (of 17)

Reminder: Everything on this page is hypothetical/backtested and presented for informational purposes only. Headline metrics assume T+0 close execution; current subscriber workflow is post-close generation with next-day (T+1) execution, which can materially diverge from these results. Hypothetical backtests do not predict future results. Results assume no transaction costs, slippage, taxes, advisory fees, or subscription fees. Real-world implementation can differ materially (see sensitivity and disclosures).

Critical Context: Leveraged ETF Mechanics

Horizon allocates to TQQQ (3× daily leveraged) and SQQQ (3× daily inverse), which are instruments that reset their leverage each trading day. Because of this daily reset, multi-day and multi-year returns diverge from a simple multiple of the index. Volatility decay erodes value in choppy or sideways markets, and path dependency means the order of daily returns matters as much as the total. In trending markets, daily compounding can amplify gains; in volatile or mean-reverting markets, it compounds losses. The CAGR, ending value, and drawdown figures above fully reflect these compounding dynamics, both favorable and adverse. TQQQ and SQQQ are not buy-and-hold index products; substantial losses are possible in short periods. See full disclosures.

Hypothetical Growth

Hypothetical Growth of $10,000

Simulated equity curve from TQQQ inception (Feb 10, 2010) through Feb 20, 2026. No pre-launch synthetic data used.

Hypothetical: This is a single historical backtest and assumes same-day close execution (T+0). In current operations, signals are generated shortly after close; readers who choose to implement usually execute on T+1 based on brokerage/order handling, so this curve will not represent their realized path. The curve is shown gross of taxes and trading frictions (spreads, slippage, auction dynamics, and broker cutoffs), and excludes advisory/subscription fees. See the sensitivity box and disclosures.

Scale:

Chart showing hypothetical growth of $10,000 from February 2010 to February 2026. In this backtest, Horizon ends at approximately $954,914 while QQQ ends at approximately $162,630. Hypothetical only; gross of costs and taxes.

Horizon (simulated)
QQQ (benchmark)
$954,914
Horizon (Hypothetical Ending Value, gross)
$162,630
QQQ (Benchmark Ending Value, gross)

Implementation Sensitivity (Illustrative)

Even small implementation frictions can materially reduce long-horizon compounded results, especially for strategies that may rotate into daily-reset leveraged/inverse ETFs. The table below applies an annual performance drag (in percentage points) to the headline CAGR for illustration. This is not a cost model and is not a prediction of realizable returns.

Assumed Annual Drag Horizon Ending Value QQQ Ending Value
1.0 pp$846,044$142,058
2.0 pp$748,896$123,945
4.0 pp$585,132$94,020
6.0 pp$455,418$70,976

Important: Real-world drag depends on turnover, spreads, order type availability, broker cutoffs, taxes, and market conditions. Daily-reset leveraged/inverse ETFs can behave unexpectedly over multi-day horizons. Read the full disclosures before making any decisions.

Methodology: CAGR calculated from Feb 10, 2010 to Feb 20, 2026 (~16.0 years). Values use adjusted closing prices (dividends and splits reflected). Max drawdown = largest peak-to-trough decline computed on daily returns. Sharpe ratio = annualized excess return ÷ annualized volatility (risk-free rate = 0% for simplicity). Annualization uses 252 trading days; volatility is the annualized standard deviation of daily returns. No transaction costs, taxes, management/advisory fees, or subscription fees included.

Year-by-Year

Annual Return Distribution

Calendar year returns 2010–2026. *2010 = partial year (Feb–Dec); *2026 = YTD through Feb 20.

Hypothetical: Annual returns shown are derived from the backtest and exclude taxes, slippage, and other implementation frictions. Descriptive of one sample, not a forecast or probability estimate.

Bar chart comparing annual returns in one historical backtest sample. In this sample, Horizon's annual return exceeded QQQ's in 12 of 17 calendar periods. Notable years: 2020 Horizon +151.4% vs QQQ +48.6%; 2022 Horizon -23.9% vs QQQ -32.6%. Hypothetical only; descriptive of one sample, not a forecast or probability estimate.

Horizon
QQQ

Backtest Window Comparison

Backtest Window Comparison (Single Sample)

Observed relative results within one historical backtest sample. These counts are descriptive only and are not probability estimates.

12 of 17
Calendar Year Periods
In this single backtest sample, Horizon's annual return exceeded QQQ's in 12 of 17 calendar periods (2010-2026). Descriptive only, not a forecast or probability estimate.
Periods counted: 2010* (Feb–Dec), 2011–2025 (full years), 2026* (Jan–Feb YTD). These are individual years, not rolling windows.
12 of 12
5-Year Rolling Windows
In this sample, Horizon's 5-year CAGR exceeded QQQ's in 12 of 12 overlapping windows (2010-2014 through 2021-2025). Descriptive only, not a forecast or probability estimate.
Note: Overlapping windows from a single backtest are not independent observations. This is a historical description of this dataset, not predictive certainty.
7 of 7
10-Year Rolling Windows
In this sample, Horizon's 10-year CAGR exceeded QQQ's in 7 of 7 overlapping windows (2010-2019 through 2016-2025). Descriptive only, not a forecast or probability estimate.
Note: With only 7 non-independent windows from a ~16-year backtest, this is a limited sample and not a statistical guarantee.

Stress Periods

Performance During Market Declines

How Horizon performed in the two full calendar years in this dataset when QQQ posted negative returns (2018, 2022).

2018
Horizon
+11.4%
QQQ
-0.1%
Horizon: positive return
2022
Horizon
-23.9%
QQQ
-32.6%
Horizon: smaller decline
Context: Only two calendar years since TQQQ's Feb 2010 launch saw QQQ finish negative. This is a very small sample. Horizon's relative performance in these periods reflects the backtest; it does not guarantee similar behavior in future downturns.

Rolling Analysis

Rolling Window Analysis

CAGR comparison across 5-year and 10-year windows. All windows begin from TQQQ's Feb 2010 launch, with no pre-launch simulated data.

Period Horizon CAGR QQQ CAGR Difference
2010*–2014+33.2%+21.0%+12.2%
2011–2015+18.7%+16.5%+2.2%
2012–2016+26.4%+17.2%+9.2%
2013–2017+39.2%+19.7%+19.5%
2014–2018+25.4%+13.3%+12.1%
2015–2019+23.6%+16.7%+6.9%
2016–2020+53.4%+24.6%+28.8%
2017–2021+65.1%+28.2%+36.9%
2018–2022+34.3%+11.8%+22.5%
2019–2023+50.1%+22.4%+27.7%
2020–2024+46.5%+19.6%+26.9%
2021–2025+31.1%+15.4%+15.7%
Important: These are overlapping windows from a single backtest, not independent trials. These counts are descriptive of this historical sample and do not imply probability of future outperformance. All data uses TQQQ's actual Feb 2010 launch date; no synthetic pre-launch returns.

Sample Output

Sample Daily Signal Output

This section shows the output format only. The example below is illustrative (not a live signal, not a recommendation, and not intended to be actionable).

Reminder: Nothing on this page is investment advice. If you are evaluating systematic strategies, focus on the process, assumptions, and risks, not a single example output.

Example · Nasdaq Universe Illustrative (format only)
Model State Transitions (Research Record, Not Trade Instruction)
CASH
State C
0% -> 0% (Delta 0%)
QQQ
State C
100% -> 100% (Delta 0%)
SQQQ
State C
0% -> 0% (Delta 0%)
TQQQ
State C
0% -> 0% (Delta 0%)
About Signal Delivery: The backtest headline metrics use a same-day close execution assumption (T+0) as a research reference basis. In current operations, Horizon is published shortly after market close in a single uniform publisher lane; the same signal is broadcast to all subscribers without customization or segmentation. Readers who choose to implement usually execute on the next trading day (T+1), depending on brokerage/order handling, which can materially change realized outcomes versus the T+0 backtest figures. Model state transitions are a research record, not trade instructions. The model outputs target allocation weights for informational use only; it does not execute trades or manage accounts. The specific rules that produced any allocation are not disclosed. See full disclosures for execution and implementation risks.

Methodology

Methodology Overview

The structural framework for how Horizon generates daily signals. Specific rules, thresholds, and weighting logic are proprietary and not disclosed here.

Scope note: This section is designed to make assumptions explicit (data, timing, and execution). It does not provide trading instructions and does not disclose proprietary decision logic.

Signal Structure

Universe

Nasdaq-listed equities serve as the input universe for the model's systematic rules. The strategy's actual holdings are limited to four instruments: QQQ (Nasdaq-100), TQQQ (3× daily leveraged Nasdaq-100), SQQQ (3× daily inverse Nasdaq-100), and/or CASH.

Timing & Locking

For backtest construction, the model uses a close-of-day decision basis and assumes same-day close execution (T+0 reference). In current operations, the daily signal is generated shortly after market close and published in a single uniform publisher lane; readers who choose to implement are generally limited to next trading-day execution (T+1), depending on brokerage/order handling. Once published, a signal is immutable for that trading day and cannot be overwritten.

Allocation Structure

Each day the model outputs a 100% allocation across QQQ, TQQQ, SQQQ, and/or CASH. The model may hold a single instrument at full weight or split allocations across instruments. The strategy does not hold individual stocks.

Data & Bias Controls

Price Data

End-of-day adjusted closing prices from a third-party market data provider. Adjusted prices (dividends and splits reflected) are a standard way to approximate total-return series, but they are vendor-constructed and may differ across providers. In current operations, signals are generated after close from end-of-day inputs; data latency and vendor revisions remain implementation limitations.

No Look-Ahead

At each historical date, signals are computed using only information that would have been available by that day's close. Future prices are not used. As with any backtest, subtle look-ahead can still creep in through data-vendor reconstructions (e.g., corporate-action adjustments or constituent histories), so treat results as an approximation rather than a guarantee of realizable execution.

No Pre-Launch Data

The backtest begins on February 10, 2010, the actual launch date of TQQQ and SQQQ. No simulated, synthetic, or extrapolated pre-launch leveraged or inverse ETF returns are included anywhere in the analysis.

Backtest Implementation Assumptions

Trade Price

The backtest uses a same-day close execution assumption (T+0 reference) based on information available by market close. This is optimistic relative to publication timing: in current operations, signals are published after close and any reader who chooses to implement is generally limited to next trading day (T+1), depending on brokerage/order handling. Real fills can differ due to opening gaps, bid/ask spreads, order timing, and liquidity conditions, so realized results may differ materially.

Costs & Cash Yield

Zero transaction costs, zero bid-ask spread, zero slippage, and zero taxes are assumed throughout. No advisory, management, or subscription fee is deducted. Cash receives 0% interest (no T-bill proxy). ETF expense ratios are embedded in the adjusted price series and are therefore captured implicitly. Tax impact can be material for frequently rebalanced strategies and is not modeled.

Rebalancing & Universe

The model rebalances daily to its full target allocation (or maintains the prior allocation when unchanged). Where available in the dataset, historical listings and delisting information are used to help reduce survivorship bias; the dataset may still be incomplete. Results can differ under alternative universe definitions, different execution assumptions, or if turnover is materially higher in practice.

What is not disclosed: The specific rules, indicator types, lookback periods, signal thresholds, weighting logic, and any proprietary criteria that produce the daily allocation are not disclosed. This section describes the structural framework and data handling only, not the strategy's decision logic or any source of differentiation. The backtest reflects systematic, rules-based decisions with no discretionary override.

Want to Understand the Process?

Before considering any strategy, understand how it works. Review the methodology, see sample outputs, and read all disclosures.

This page is for informational purposes only. Not investment advice. All results are backtested/hypothetical.

Important Disclosures

Important Disclosures & Risk Warnings

This presentation is for informational and educational purposes only and does not constitute investment advice, a solicitation, or an offer to buy or sell any security. Karmaniti is a publisher of impersonal financial research and is not a registered investment adviser. Past performance is not indicative of future results. Backtested results are hypothetical and have inherent limitations; they do not reflect actual trading, actual costs, or actual market conditions.

Publisher Status / No Advisory Relationship: Karmaniti is a publisher of impersonal financial research and is not registered as an investment adviser with the U.S. SEC or any state securities authority. Viewing this page or receiving these materials does not create an investment advisory, broker-dealer, fiduciary, or agency relationship. Information is general and not tailored to any person’s circumstances.

Leveraged & Inverse ETFs: The Horizon strategy may allocate to leveraged and/or inverse ETFs (such as TQQQ and SQQQ). These instruments reset daily and are designed for short-term use. Due to daily compounding, their multi-day and multi-year returns can differ materially, and often significantly, from the stated leverage multiple applied to the index. Volatility decay can erode returns even when the underlying index is flat or rising over time. These are not buy-and-hold instruments, and large or rapid losses are possible. All performance data shown in this report begins on February 10, 2010, the actual launch date of TQQQ and SQQQ. No simulated pre-launch or synthetic leveraged/inverse returns are included anywhere in this presentation.

Benchmark Comparison Limitations: Results are compared to QQQ (Nasdaq-100 ETF) as a common reference, but Horizon's holdings rotate among QQQ, TQQQ, and SQQQ. This is not a like-for-like risk comparison. Horizon takes on significantly more risk than a QQQ-only strategy due to leveraged instrument usage.

Execution Basis & T+1 Divergence: The headline hypothetical metrics shown on this page assume a same-day close execution basis (T+0). In current operations, Horizon is published shortly after market close in a single uniform publisher lane; readers who choose to implement are generally limited to next trading-day execution (T+1), depending on brokerage/order handling. Because prices can move between close and next-day execution, realized outcomes may differ materially from the backtest figures shown. Practical constraints such as broker cutoffs, order handling, spread/slippage, and gaps can increase this divergence.

Fractional Shares: Performance results shown here assume fractional share execution. Not all brokerages support fractional shares. If your brokerage does not support fractional shares, your actual results may differ from those shown.

Dividends & Adjusted Prices: Results shown use adjusted closing prices, which account for dividends and splits in the price series. However, actual dividend receipt, reinvestment timing, and tax treatment may vary by investor and brokerage. Results do not account for taxes, transaction costs, management fees, or other expenses. The inclusion of these costs would reduce performance.

No Guarantee of Signal Delivery Timing: Signals are intended to be generated and delivered shortly after market close. Technical issues, data provider outages, or other factors may delay delivery beyond the expected window. Delayed delivery can further affect next-day (T+1) execution outcomes. Current distribution is publisher-channel delivery (email/web); SMS delivery is not currently enabled.

Universe & Signal Generation: "Nasdaq Universe" refers to the set of Nasdaq-listed equities used as the stock selection universe for signal generation. The strategy's actual holdings are limited to the ETFs QQQ (Nasdaq-100), TQQQ (3× daily leveraged Nasdaq-100), and SQQQ (3× daily inverse Nasdaq-100). The strategy does not hold individual stocks. Historical constituent data reflects the universe as available at each historical date in the dataset.

Cost, Fee & Slippage Assumptions: All backtested results assume zero transaction costs, zero bid-ask spread, zero slippage, and zero taxes. No management fee, advisory fee, or subscription fee is deducted in the displayed hypothetical returns. ETF expense ratios are embedded in ETF prices. Inclusion of realistic costs and subscriber fees would reduce reported performance.

Tax Considerations: Many investors face short-term capital gains rates and other tax impacts from frequent rebalancing. This report does not model taxes; after-tax outcomes may be materially lower.

Hypothetical Performance Limitations: Simulated backtested results have inherent limitations that materially differ from actual trading. All headline figures shown here assume same-day close execution (T+0). If a subscriber receives signals too late for close execution and trades on T+1, these hypothetical figures are not representative of that subscriber's realized outcomes. Simulated results are created with the benefit of hindsight and cannot fully account for the impact of financial risk, liquidity constraints, potential for losses, actual market impact, or behavioral factors. The model may have been constructed, selected, or presented in part based on its historical simulated performance - this constitutes a form of look-back optimization that may not replicate in future market conditions. Simulated results do not represent what any investor actually achieved, and actual future results may be materially lower or result in losses.

Conflicts / Personal & Firm Trading Disclosure: Karmaniti LLC and its principals continuously research and test multiple strategies. The firm and/or principals may invest or trade firm/personal capital in QQQ, TQQQ, SQQQ, and other tickers, and may use the same strategy discussed here or different strategies. Positions and strategy usage may change at any time without notice and may differ from any model allocation shown.

No Affiliation: Karmaniti Research is not affiliated with, sponsored by, or endorsed by ProShares Advisors LLC (issuer of TQQQ and SQQQ), Invesco Capital Management LLC (issuer of QQQ), or any other ETF issuer, index provider, or financial institution. All ETF names and tickers are used for descriptive purposes only.

Not Financial Advice: This is not personalized investment advice. You should consult a qualified financial advisor before making any investment decisions. Investing involves risk, including the possible loss of principal.

Suitability: This research is generally intended for readers who understand high-volatility, leveraged/inverse ETF behavior and can tolerate substantial drawdowns. It is generally not designed for investors prioritizing capital preservation, short investment horizons, or near-term liquidity needs.