This sheet holds our full P&L (profit and loss) record. We send an updated copy every week with our Monday update email. It covers the Commodities book. Figures here are as of November 2, 2026.
What should you read first?
Start with the Performance tab. It sets each book against SPY (the S&P 500 ETF) over the same sessions. On any figure, check three things.
How many positions. One position is an anecdote.
What they returned. Every figure is after costs.
How many are still open. Those figures move until we sell.
Which figures are live?
Live record: what we published at the time, unedited since.
Backtest: our rules run on past data. No money followed it.
We never add the two together. Each name's backtest ends the day it went live.
What prices and costs are in it?
We buy and sell at the open, after 5 basis points (0.05%) each way. We value open positions at the last close, so their figures move daily.
How did the backtest compare with SPY?
The Portfolio Backtest tab runs our rules over the ten years to July 31, 2026.
Commodities: 16.72% a year, against 15.36% for SPY. Worst fall from a peak: 16.12%.
Compare yearly rates (CAGR), not totals. A total grows simply by covering more years.
Past performance is not a reliable indicator of future results. Returns are uneven across market conditions, so read the calendar-year rows on the same tab. In 2022, Commodities rose 17.50% while SPY fell 17.99% in an equity bear market. In 2023, Commodities fell 6.52% while SPY rose 24.72%.
What does one row tell you?
Here is ExxonMobil on the Per Ticker Live tab, from the November 2 sheet.
Positions: 4, one still open. We have bought ExxonMobil four times.
Time in market: 73.3%. Share of sessions held since July 2.
Total return: 29.69%. Four positions, compounded in order.
Total return, 15% vol-adjusted: 15.54%. The same positions, rescaled to the S&P 500's typical swing of about 15% a year. Every name then sits on one yardstick, so a jumpy name cannot win on the size of its swings alone. No book earned this figure.
SPY, same sessions: minus 0.44%. SPY only on the days we held ExxonMobil.
Losers stay in. Uranium: minus 19.71% over three positions.
How should you use the signals?
We measure our performance on the Systematic Portfolio. The Multi-model Signals are the work under the hood. We built them to feed the portfolio, not to be traded alone, one position at a time.
Judge any model on its whole backtest, then year by year and across market conditions such as rallies and sell-offs. Three reasons:
One position is mostly noise. Any single long can be right or wrong by chance. We tested each name's rules across its full backtest, through every position it took. As a whole, those rules give the portfolio a sound base, built to limit overfitting.
No model can rescue a weak name. Some names have lagged the market for years. Others have soared, so the model can look like it held back. Nobody knows which is which until afterwards. Past laggards can lead next. Past leaders can lag.
We do not model earnings dates. Each name has too few past earnings to model them reliably. Our positions ride through them, so judge the history as a whole.
Every model has limits, ours included. The full record is where they show.
Count first, then returns, then what is still open. Never mix live and backtest.
You can see how the Systematic Portfolio and Multi-model Signals work at yxinsights.com/systematic.
Appendix: the nine tabs and the terms
The nine tabs
Performance. Each book against SPY: the past week, since launch, then by month. It also shows a 75% Macro & Megacaps, 25% Commodities blend, not traded separately.
Summary. The same months, split by group within the channel.
Portfolio Trades. Every book position, open and closed.
Portfolio Backtest. The books over ten years to July 31, 2026.
Signal longs. Every signal position, one row each.
Per Ticker Live. The same record, one row per name.
Signals Backtest. Each name's signals before it went live.
Per Ticker Backtest. That backtest, one row per name.
About. Column definitions and the date we locked each record.
Terms
Book. One portfolio we publish.
Signal. Our daily call on one name: hold it, or hold nothing.
Position. One continuous run of holding a name.
Time in market. Share of sessions we held the name since its first live signal.
Total return. A name's position returns, compounded in order.
15% vol-adjusted. Each position restated at 15% volatility, roughly the S&P 500's long-run level. It compares names only. No book earned it.
CAGR. The steady yearly rate that gives the same total. Only the backtest shows it, because no book has a full year of live record yet.
Mean excess vs compounded excess. The mean counts each position once. The compounded figure tracks the money, so one large loss can flip its sign. Both are correct.
Basis point. 0.01%.
Dashes and empty cells
Dash on Performance and the backtest tabs. Withheld on purpose. We withhold Sharpe (return per unit of risk) on weeks, months and losing periods, where it misleads.
Dash on the trade tabs. No result yet. The session has not closed.
Empty cell on Per Ticker Backtest. No position in that name that year.
What the sheet leaves out
Tax. We rebuild weekly. Work out tax for your own account.
Fills. We publish target weights, not the prices anyone traded at.
Per-name settings. The rules are on our website. The settings behind each name stay private.
DISCLAIMER: This newsletter is intended for educational purposes only. Any information or analysis in this note does not constitute an offer to sell or a solicitation of an offer to buy any securities. Nothing in this note is intended to be investment advice, nor should it be relied upon to make investment decisions. Any opinions, analyses, or probabilities expressed in this note are those of the author as of the note's date of publication and are subject to change without notice.
