Tools
Adaptive asset allocation
Each month-end, the three of seven assets (Nifty 100, Midcap 150, Smallcap 250, gold, 5-year G-sec, 10-year G-sec, S&P 500 in rupees) with the best six-month return, weighted for minimum variance on the last 60 sessions' covariance (Ledoit-Wolf shrinkage). It holds 53% S&P 500, 40% Midcap 150, 6% Smallcap 250 since Sept 2026. From Oct 2005 it returned 8.7% a year after tax, against 12.9% for its benchmark; its worst fall was −33%.
What it holds now
Target since Sept 2026: 53% S&P 500, 40% Midcap 150, 6% Smallcap 250. If the month ended today it would hold 53% S&P 500, 35% Midcap 150, 12% Smallcap 250.
Its record
Growth of one rupee after costs, before tax, on a log scale, beside its benchmark: equal parts Nifty 100, Midcap 150, Smallcap 250, S&P 500, gold, 5-year G-secs and 10-year G-secs, rebalanced each January. The benchmark holds the assets the rule chooses from but never switches, so the gap is what the rule's timing added or cost. The marks show when the rule was published and when it entered this library; only the stretch after the second mark is free of hindsight. Deciding on a different day of the month would have changed its return before tax from 9.5% to 12.3% a year.
Adaptive asset allocation: growth of one rupee
How to read it. Log scale, so equal slopes are equal returns.
Loading chart…
Adaptive asset allocation: 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 −33%, against −30% for its benchmark, on daily closes; the weekly line can look a little shallower.
Loading chart…
Oct 2005 to Oct 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. With the S&P 500's dividends included (iShares IVV's total return), it would have returned 8.3% a year after tax. With remittance and forex costs of 0.25% each way instead of 0.60%, 9.1%.
| This model | Benchmark | |
|---|---|---|
| Return a year, before tax | 10.5 | 13.5 |
| Return a year, after tax | 8.7 | 12.9 |
| Volatility | 15.0 | 11.0 |
| Worst fall | −32.8 | −29.9 |
| Sharpe ratio | 0.3 | – |
| Average share in equity | 57.7 | – |
| Turnover a year | 4.7 | – |
| Tax cost, points a year | 1.9 | – |
The live log starts on the entry date.
| Period | Dates | Total return | A year | More |
|---|---|---|---|---|
| Before the rule was published | Oct 2005 to Dec 2012 | 127.2% | 12.0% | |
| Published, before it entered this library | Dec 2012 to Oct 2026 | 263.1% | 9.8% | |
| Since it entered this library (live) | – | – | – |
As an index: Tipsheet Adaptive Allocation
The same rule published as an index: total return, before any fund cost, trading cost or tax, set to 1,000 at the end of 2006 like every Tipsheet index. Its launch date is 14 Oct 2026, when the rule entered this library; the lighter line before it is back-calculated. Tipsheet indices are reference series for research only: not official indices, not investable, and not advice.
Tipsheet Adaptive Allocation, 1,000 at the end of 2006
Loading chart…
- Rule
- Each month-end, the three of seven assets (Nifty 100, Midcap 150, Smallcap 250, gold, 5-year G-sec, 10-year G-sec, S&P 500 in rupees) with the best six-month return, weighted for minimum variance on the last 60 sessions' covariance (Ledoit-Wolf shrinkage).
- Rebalancing
- Decided on each month's last session; traded at the next close.
- Basis
- Total return, before any fund cost, trading cost or tax; 1,000 on 31 December 2006.
- Launch
- 14 Oct 2026. Earlier levels are back-calculated with a rule chosen after that history was known; later ones chain returns written once to a hash-chained log and are never restated.
- Download
- Daily levels, CSV · JSON · every index
What it held
The share of the model in each asset at every month-end.
Adaptive asset allocation: holdings at each month-end
Loading chart…
The rule
Each month-end, the three of seven assets (Nifty 100, Midcap 150, Smallcap 250, gold, 5-year G-sec, 10-year G-sec, S&P 500 in rupees) with the best six-month return, weighted for minimum variance on the last 60 sessions' covariance (Ledoit-Wolf shrinkage).
Butler, Philbrick and Gordillo (2012), Adaptive Asset Allocation: A Primer. Midcap 150 and Smallcap 250 are back-calculated before their launch on 2016-04-01. Rule published 2012; entered this library 14 Oct 2026.
Sources
The research this rule comes from. The Indian version substitutes Indian indices for the original assets, as the rule above says.
- Adam Butler, Michael Philbrick and Rodrigo Gordillo (2012). Adaptive Asset Allocation: A Primer. SSRN working paper.
- Roger Clarke, Harindra de Silva and Steven Thorley (2006). Minimum-Variance Portfolios in the U.S. Equity Market. The Journal of Portfolio Management.
- Olivier Ledoit and Michael Wolf (2004). A Well-Conditioned Estimator for Large-Dimensional Covariance Matrices. Journal of Multivariate Analysis.
Graded with the deflated Sharpe ratio of Bailey and López de Prado (2014), counting every rule tried: the pre-registered specification.
The workings
Data
Every chart has a table view, a CSV download and a link to its published JSON. Bundles are listed in data status.