Backtest · AMD
Backtest AMD — Candle by Candle
AMD is where the most expensive mistake in equity backtesting shows up most clearly, and it is not about AMD at all. Seven of the seventeen names in this catalogue are semiconductors — NVDA, AMD, AVGO, QCOM, MU, INTC and ARM — and on most days they move together. Test a rule on all seven, get 350 trades, and you do not have a 350-trade sample. You have something closer to a 50-trade sample counted seven times.
Correlated instruments do not multiply your sample
This matters because sample size is the thing standing between a result and a story, and correlation destroys it quietly. A hundred trades taken across seven names that all rallied on the same three days is not a hundred independent observations; it is roughly three, dressed up. The drawdown you measure will be understated for exactly the same reason — correlated positions lose on the same day, which is the day that matters.
AMD is the sharpest illustration because it is the highest-beta of the group: it does what NVDA does, further. So a strategy that looks merely good on NVDA often looks spectacular here, and the temptation is to conclude that you found a better instrument rather than more leverage on the same move.
The correct test is diversification of condition, not of ticker. Run your rule on AMD, then on something that is not a chip and not in the same index behaviour — NFLX, COIN, or step outside equities entirely to gold or EURUSD. What survives that is an edge.
What the AMD data actually is
The archive runs from May 2018 to the most recent session we hold, on every timeframe from one minute to monthly. It is open from the first chart — there is no plan that unlocks more of it, and no shorter window for people who have not paid.
Equity data here is regular hours only — 09:30 to 16:00 New York, no pre-market and no after-hours. That is not a gap in the data, it is what an equity session is, and it has a consequence worth planning for: every trading day opens with a gap rather than continuing from the previous close, so any rule you carried over from an index future or an FX pair is about to meet a discontinuity it has never been tested against.
Prices are split-adjusted throughout, so a chart of 2019 shows what you would compare against today rather than the raw tape. Getting this wrong is the classic silent equity-backtest bug: an unadjusted series prints a 90% single-day crash on the split date, and a stop-loss rule tested through it produces numbers that mean nothing.
How to backtest AMD without fooling yourself
Test AMD on its own first and record the result. Then add NVDA and check whether the combined equity curve is smoother or merely longer — if the drawdowns line up on the same dates, you have your answer about independence.
Use the sample size calculator with the honest trade count, not the inflated one. Dividing your chip-sector sample by something close to the number of correlated names is a crude adjustment, but it is far closer to the truth than not adjusting at all.
Size for beta. A position size derived from a test on a slower name will be too large here, and the position size calculator is how you keep risk constant across instruments that are not.
Frequently asked questions
Is AMD backtesting free?
Yes, and there is no card and no trial timer. You get AMD back to May 2018, with every tool the product has, for 2,000 candles of replay — about two hours of stepping through a chart. After that you top up for $5, or pay $19.99 once and never think about it again. Nothing you buy expires. So are the other six semiconductors in the catalogue, which is what makes the correlation test on this page possible at all.
Why does testing several chip stocks not give me a bigger sample?
Because they largely move together. Trades taken across seven correlated names on the same days are not seven independent observations — the effective sample is far smaller than the trade count, and the measured drawdown is understated because correlated positions lose simultaneously.
Which stocks here are semiconductors?
Seven of the seventeen: NVDA, AMD, AVGO, QCOM, MU, INTC and ARM. Knowing that list is the point — it is what stops you reading a sector move as strategy confirmation.
How far back does AMD go?
May 2018, regular hours only, split-adjusted, on every timeframe from one minute to monthly.
Open AMD and start stepping
Real AMD history, split-adjusted, one candle at a time. No account, no card, 2,000 candles free.
Open the chart →