Our variants

Insider buying: Conviction model, 500 largest

Backtest from 22 Feb 2013 to 2 Nov 2026 · weekly · hypothetical

Watch this strategy

Return a year

−2.5%

S&P 500 +14.4%

Last 12 months

+3.7%

S&P 500 +12.1%

Sharpe ratio

−0.01

S&P 500 0.81

Biggest drop

−63.3%

S&P 500 −31.8%

About

Scores every insider purchase with a model of the trade, the insider and the company, fitted each year on earlier years only, and holds the best-scoring companies. Costs of 10 basis points a trade are deducted.

Insider backtests →

Key statistics

Measure	Strategy	S&P 500
Return a year (compound)	−2.5%	+14.4%
Total return	−30%	+529%
Last 12 months	+3.7%	+12.1%
Last 3 months	−12.4%	—
Last month	−2.7%	—
This year so far	+9.5%	—
Volatility a year	28.7%	15.8%
Sharpe ratio	−0.01	0.81
Biggest drop	−63.3%	−31.8%
Beta to the S&P 500	1.05	1.00
Alpha a year (t-statistic)	−13.7% (−2.15)	—
Share of weeks up	54%	—

Sharpe ratio uses returns above the 3-month Treasury bill rate. Alpha is the yearly return left after allowing for the strategy's exposure to the S&P 500 (CAPM); a t-statistic below about 2 means it cannot be told apart from luck.

Insights

Holdings (1)Backtest resultsRecently openedRecently liquidated

Company	Weight	Held since	Return since
XYL	100.0%	—	—

Equal weight at the latest rebalance, rebuilt every month.

Period	Return a year	S&P 500	Sharpe	S&P 500	Biggest drop	Beta	Alpha (t)
Whole backtest22 Feb 2013 to 2 Nov 2026	−2.5%	+14.4%	−0.01	0.81	−63.3%	1.05	−13.7% (−2.15)
2018 onwards5 Jan 2018 to 2 Nov 2026	−5.3%	+13.9%	−0.11	0.66	−63.3%	0.97	−14.9% (−1.75)
Last three years3 Nov 2023 to 2 Nov 2026	+1.0%	+21.8%	−0.01	1.16	−32.4%	0.82	−13.5% (−1.01)
Last year7 Nov 2025 to 2 Nov 2026	+9.2%	+14.2%	0.31	0.84	−16.9%	0.65	+2.9% (0.10)

Date	Company	Action
2 Nov 2026	XYL	Bought
19 Oct 2026	BABA	Bought

Dated at the rebalance that made the change.

Date	Company	Action
2 Nov 2026	BABA	Sold

Dated at the rebalance that made the change.

Backtests are hypothetical and use today's listed companies only, which flatters results. Past performance does not predict future returns. This is research, not investment advice. How we backtest