Our research, in the open

Every strategy of Kakushadze and Serur's 151 Trading Strategies and the academic literature we could rebuild, 509 in all, tested on point-in-time data and charged the official Interactive Brokers fee grid at nine account sizes, from USD 10k to 3M. Around it: the walk-forward improvements of v2, the pre-registered combinations of v3, a microcap bench and a survey of retail-only alpha. Pick an account size: every view follows.

Data as of 2026-11-03 · research output, not investment advice
Verdicts across account sizes

The account size changes the verdict

Number of strategies in each verdict group at each of the nine account sizes: same strategies, same data, the fees of that account.

What survives

Robust strategies at this size

Pass every test after costs, across the whole family of 509. The line is their rolling 36-month net Sharpe; the right column compares it with the S&P 500 held since 2010.

What the results say

Each number is read from the result tables at the selected account size.

Explorer

All 509 strategies at the selected account size. Click a row for its full sheet: what it does, why it got its verdict, how it behaves through time, what leverage and Swiss tax do to it.

UltraView

The terminal holds everything the lake does, about 3,400 instruments: every strategy, the ETFs and mutual funds that were among the 300 largest at any year end, the largest stocks of 19 markets, 67 national stock markets, 343 futures markets (equity indexes, government bonds, currencies, energy, metals, farm goods, volatility), crypto, the Fama-French factors and the benchmarks. Each is measured the same way: monthly, in excess of cash, after IBKR's costs where it trades. Search, filter, pin, press play; the strategy views below follow the strategy you pick.

The strategies through time

Every strategy year by year: rolling 36-month statistics at the selected account size, animated through time. A three-year Sharpe ratio is noisy: read the drift across many windows, not one year. Pick a strategy in the search box, or click a point: every view follows.

Landscape3-D · every strategy × every year

The whole collection in relief

One bar per strategy and year: height and colour are the rolling 36-month net Sharpe, blue above zero and red below, strategies grouped by family from back to front. The gold plane is the year being played, the gold strip the selected strategy. Hover a bar to read it, click it to select its strategy.

Time machinethe collection through time

Every strategy, year by year, on the measure you pick

Each year, the spread of a rolling 36-month measure across all strategies: the line is the median, the dark band the middle half, the light band the middle 80%. Gold: the selected strategy; dashed: the S&P 500 held (Sharpe ratios). Hover for every percentile; drag on the strip below to zoom on a period, double-click to show them all.

Galaxygross vs net · 36 months

What costs take, year after year

Each point is a strategy. On the diagonal costs take nothing; the further below it, the more they eat. A point beyond the frame is drawn hollow on its edge. Press play to run the years; the trail follows the selected strategy over its last twelve. Drag to pan, click then scroll (or pinch) to zoom, double-click to reset.

Familiesmedian net Sharpe

Families through time

Median rolling net Sharpe of each family, against the S&P 500 held (dashed). Hover a family to isolate it.

Projection3-D · four measures · year by year

Every strategy in four dimensions

Each point is a strategy over the 36 months to the year shown: three measures on the axes, a fourth in the colour. Its shadows on the three walls are the flat views of each pair. Press play above to watch the cloud move through the years; the gold trail follows the selected strategy over its last twelve. Axes span the 5th to 95th percentile of every strategy and year; beyond, a point sits on the edge of the cube.

Heat mapfamilies × years

The whole collection through time

One row per family of strategies, its median rolling 36-month net Sharpe; one column per year; blue when positive, red when negative. The top band is the S&P 500 held. Click a family to open it into its strategies (click again to close), a strategy to select it; the cursor follows the galaxy.

Surface3-D · 4 dimensions

Year × account size × Sharpe, coloured by growth

The selected strategy across the nine account sizes and all its years; height is the rolling net Sharpe, colour the log growth. The gold ridge is the galaxy's year. Drag to turn, click then scroll to zoom, hover to read any year and size.

Decaypublication time

Before and after the paper

Every dated strategy aligned on its publication year (0): median and interquartile range of the rolling Sharpe ratio, gross and net, by years before or after the paper (at least ten strategies per point). The bands run from the first to the third quartile.

Curves

Every rolling measure of the selected strategy

The publication mark is the year of the original paper; 2019 is the start of the period v3 never saw before its single evaluation.

Compare

Two to four strategies side by side, against SPY, the S&P 500 index or 60/40: their net growth and drawdowns on one time axis, their statistics over the same months, how much they move together, their verdict at each account size and what costs take from each. Every view of this page has its own address: share it and it opens as you see it.

Costs & size

Below USD 100k the per-order minimum and whole contracts decide; above, market impact starts to. What leverage and Swiss tax do on top.

Programs

v1 tests each strategy as published. v2 lets the literature's improvements compete, chosen out of sample. v3 is a pre-registered test of cost-aware combinations on years it never saw. The microcap bench asks whether small accounts have an edge where institutions cannot go; the survey asks the literature the same question.

Methods

How a verdict is reached, and what the results do not say.