Data to 12 November 2026
About

Tools

Trend, with global (first version, with the Nifty 50)

As Trend, India only, with a sixth sleeve for the S&P 500 in rupees held through LRS: six equal sleeves. Its latest target is 13% Nifty 50, 13% Smallcap 250, 11% Midcap 150, 6% G-secs, 6% S&P 500, 1% gold, 51% cash . From Apr 2006 it returned 8.5% a year after tax, against 13.1% for its benchmark; its worst fall was −16%.

What it holds now

Target : 13% Nifty 50, 13% Smallcap 250, 11% Midcap 150, 6% G-secs, 6% S&P 500, 1% gold, 51% cash.

Each sleeve is an equal slice of the model, invested in proportion to its market's trend score.

Each sleeve is an equal slice of the model, invested in proportion to its market's trend score.
MarketTrend scoreSleeveSinceMore
Nifty 507979% investedOct 2026
Midcap 1506767% investedNov 2026
Smallcap 2507979% investedSept 2026
Gold44% investedMay 2026
G-secs3333% investedNov 2026
S&P 5003333% investedOct 2026

Its record

Growth of one rupee after costs, before tax, on a log scale, beside its benchmark: equal parts Nifty 50, Midcap 150, Smallcap 250, S&P 500, gold and 5-year G-secs, rebalanced each January. The gap can reflect asset allocation, risk exposure, rebalancing and timing; this comparison alone does not isolate timing skill. The marks show when the rule was published and when it entered this library; the record after entry is tracked separately.

Return a year, after tax8.5%Benchmark 13.1%
Against its benchmark−4.6 pts90% range before tax −5.4 to −0.6: clear of zero
Worst fall−16%Benchmark −36%
Deflated Sharpe ratio0.00Against the benchmark; 0.95 is the bar

Trend, with global (first version, with the Nifty 50): growth of one rupee

How to read it. Log scale, so equal slopes are equal returns.

Trend, with global (first version, with the Nifty 50) Benchmark

Loading chart…

Source: tipsheet models, computed from NSE total-return indices, domestic gold (World Gold Council), G-secs, the 91-day T-bill, and foreign index or iShares ETF NAV total returns in INR (ECB rates), as specified for each model.
Method JSON

Trend, with global (first version, with the Nifty 50): fall from its previous peak

How to read it. How far each was below its own previous high, weekly: the model as the line, its benchmark as the grey area, so wherever the line sits above the grey the model was falling less. The model's worst fall was −16%, against −36% for its benchmark, on daily closes; the weekly line can look a little shallower.

Benchmark Trend, with global (first version, with the Nifty 50)

Loading chart…

Source: tipsheet models, computed from NSE total-return indices, domestic gold (World Gold Council), G-secs, the 91-day T-bill, and foreign index or iShares ETF NAV total returns in INR (ECB rates), as specified for each model.
Method JSON

Apr 2006 to Nov 2026. Percentages except the Sharpe ratio, turnover and tax cost. Deflated Sharpe ratio against the benchmark: 0.00 (0.95 is the usual bar), counting every rule tried. The older IVV dividend sensitivity returned 11.0% a year before investor tax. New ETF versions use the full-workbook input separately. With remittance and forex costs of 0.25% each way instead of 0.60%, 8.8%.

Apr 2006 to Nov 2026. Percentages except the Sharpe ratio, turnover and tax cost. Deflated Sharpe ratio against the benchmark: 0.00 (0.95 is the usual bar), counting every rule tried. The older IVV dividend sensitivity returned 11.0% a year before investor tax. New ETF versions use the full-workbook input separately. With remittance and forex costs of 0.25% each way instead of 0.60%, 8.8%.
This modelBenchmark
Return a year, before tax10.713.8
Return a year, after tax8.513.1
Volatility7.712.5
Worst daily fall, before tax−16.1−36.4
Sharpe ratio0.5–
Average share in equity41.1–
Turnover a year3.6–
Tax cost, points a year2.2–

Backtests and recorded returns

The frozen paper log holds 25 sessions, 25 reconstructed. The latest consecutive record made on each session's date covers 0 sessions. Its complete return differs from the recomputed daily backtest on exactly the same sessions by 0.11 points. These are paper records, not actual portfolio executions. Returns are after costs and before tax.

The frozen paper log holds 25 sessions, 25 reconstructed. The latest consecutive record made on each session's date covers 0 sessions. Its complete return differs from the recomputed daily backtest on exactly the same sessions by 0.11 points. These are paper records, not actual portfolio executions. Returns are after costs and before tax.
Period and basisDatesTotal return, before taxA yearMore
Before the source rule was published (backtest)10 Apr 2006 to 29 Dec 2017252.6%11.4%
After source publication, before library entry (backtest)29 Dec 2017 to 8 Oct 2026129.6%9.9%
Since library entry (recomputed backtest)8 Oct 2026 to 12 Nov 2026−0.2%–
Frozen paper log, including reconstructed sessions8 Oct 2026 to 12 Nov 2026−0.3%–

What it held

The share of the model in each asset at every month-end.

Trend, with global (first version, with the Nifty 50): holdings at each month-end

Cash (91-day T-bill, as a liquid fund) S&P 500 in rupees (price index) Nifty Midcap 150 Nifty 50 Gold (domestic, rupees) Everything else

Loading chart…

Source: tipsheet models, computed from NSE total-return indices, domestic gold (World Gold Council), G-secs, the 91-day T-bill, and foreign index or iShares ETF NAV total returns in INR (ECB rates), as specified for each model.
Method JSON

The rule

As Trend, India only, with a sixth sleeve for the S&P 500 in rupees held through LRS: six equal sleeves.

Hurst, Ooi and Pedersen (2017); long-only. Rule published 2017; entered this library 8 Oct 2026.

Sources

The research this rule comes from. The Indian version substitutes Indian indices for the original assets, as the rule above says.

  1. Brian Hurst, Yao Hua Ooi and Lasse Heje Pedersen (2017). A Century of Evidence on Trend-Following Investing. The Journal of Portfolio Management.
  2. Tobias Moskowitz, Yao Hua Ooi and Lasse Heje Pedersen (2012). Time Series Momentum. Journal of Financial Economics.

Graded with the deflated Sharpe ratio of Bailey and López de Prado (2014), counting every rule tried: the pre-registered specification.

The workings

Method

Data

Every chart has a table view, a CSV download and a link to its published JSON. Bundles are listed in data status.

Read with