The numbers behind our advice — and how they are produced
Every figure on this page is read live from the data files we publish, the same ones that drive the chart on our homepage. Nothing here is typed in by hand — if a file stops updating, the page says so instead of showing an old number.
A real account, next to the model and the market
This is the time-weighted return of a real brokerage account traded on our strategy, shown against the model portfolio and against the S&P 500 over exactly the same period. We show it as it is, including the gap between the live account and the model.
Returns are time-weighted, so deposits and withdrawals do not change the figure.
Sharpe and Sortino are computed on this page from the published daily series, annualised over 252 trading days with a risk-free rate of 0% — recompute them yourself from the file linked below.
Thirteen years of the strategy, at the downside limit you choose
The same series that drives the chart on our homepage: the strategy's daily portfolio value in EUR, rebased to €10,000 at the start of the window you select. A tighter downside limit is the same strategy held at a lower exposure, with the rest sitting in cash.
Past performance is not a reliable indicator of future results. The chart shows the actual daily portfolio value of our live EUR strategy, rebased to €10,000 at the start of the selected period; figures are before advisory fees and taxes. Tighter downside limits are the same strategy held at a reduced exposure, with the remainder assumed to earn nothing in cash. The benchmark is the S&P 500 (SXR8.DE, EUR-quoted, dividends reinvested). The value of investments can go down as well as up; capital at risk.
The series starts in 2013, well before the first client account was opened: the early years are the strategy computed on historical data, the recent part is the strategy running live. The section above is its honest counterpart — a real account, in real markets.
How the portfolio is built
A short description of what the model actually does. We keep the specific signals and their weights to ourselves — the shape of the process, and every number it produces, we publish.
The model runs on a fixed schedule, once per trading day, on rules written in advance. Nobody overrides it because a headline felt alarming — the same inputs always produce the same recommendation.
Three to five positions at a time, chosen from a shortlist of large, liquid, low-cost ETFs: broad equities (SPY, QQQ), gold (GLD), credit (HYG), sector exposure (XLE, XLU), treasuries (TLT) and short-duration cash (ICSH). No single stocks, no exotic instruments.
You set the downside limit. Position sizes are scaled to the volatility of the assets so that the portfolio stays inside your line, and the unused part stays in cash instead of being invested for the sake of it.
Inflation, industrial production and long-run equity valuation are tracked continuously. When they point to a recessionary or richly valued regime, the model reduces exposure before drawdowns build, rather than after.
Public attention to trade and geopolitical conflict is scored daily from open sources and feeds into the risk budget, so exposure comes down while tension is rising — not once it has already hit prices.
The output is a recommendation. Your assets stay in your own name at your own broker and you confirm every trade — which is also why the live account above can differ from the model.
Check the numbers yourself
Both charts on this page are drawn from public files, rebuilt daily from our production database. Open them and recompute anything you see here.
Last update: —
Last update: —
What these figures do not include: advisory fees and taxes are not deducted from the 2013 track record. Returns on the real account are reported by the broker after its commissions and fees. Past performance is not a reliable indicator of future results and the value of investments can go down as well as up; capital at risk.