Our Record
Every stock we recommend is written down, then measured against the market.
Each weekday we record what Daily Opportunities, Buy Ideas and Short Ideas are showing. Once a holding period has fully elapsed, every pick is measured from the close on the day it appeared against the same window in its benchmark. Results are published here as they mature, whether or not they are good. Nothing is held back for being unflattering, and there is no step where we decide whether to show you a quarter.
Nothing has matured yet
Recording began on 19 Aug 2026, and a pick cannot be scored until its holding period has actually elapsed — 21 trading days for the shortest. The first results appear here automatically on the day they mature. We are not waiting to see whether they are good.
How this is measured
Entry is the closing price on the day a pick appeared — not an intraday price nobody could have got. Exit is the close a set number of trading days later. The benchmark is measured over the identical calendar window: QQQ for picks drawn from our technology universe, SPY for the earlier S&P coverage. Short Ideas are scored as shorts, so a stock that falls is a gain, and their excess is measured against shorting the benchmark rather than holding it.
What counts as a win
Beating the benchmark, not merely going up. A pick that rose 3% while the index rose 6% cost you money against doing nothing, and is counted as a loss here. The headline figure is the median rather than the average, because on a small sample one outlier can carry an average and say more about luck than about method.
Limits we will not hide
No exit is modelled: each pick is held for the full period with no stop loss, which on a short is unbounded — a stock that triples reads as −200%, and one such position can distort a small sample. These are also gross figures, with no commissions, slippage or position sizing. Past results do not predict future ones.
Why universes are listed separately
Our coverage narrowed to technology on 2 Sep 2026. Picks made before and after came from different products measured against different benchmarks, so blending them would produce a number describing neither. Each is shown on its own, with its own sample size, rather than merged into a single flattering total.
This page updates on its own as picks mature. Our pre-launch backtest, and its limitations, are documented separately on the Track Record page. Xymetrical is an investor intelligence tool, not investment advice.