The best per-ticker result backed by more than one trade is AMD's 308 percent
Rank every published study by its own single best-performing ticker and one number towers over the rest, NVDA's 3300 percent under 52-week high momentum, built on exactly one trade. Set that aside and the best result left standing, AMD at 308 percent under Williams %R reversal, is backed by 46.
Take every one of Tenachine's 22 published studies and find each one's own single best-performing ticker. Sort the 22 numbers and one sits far above the rest: NVDA under 52-week high momentum, positive 3300.42 percent. Tenachine's guide on that result already covers the reason it towers over everything else on the list. It came from exactly one trade.
Set that one trade aside and ask the same question again: of the 22 best-ticker results, which is the strongest one actually backed by a real number of trades? The answer is AMD under Williams %R reversal, positive 308.45 percent across 46 trades, a 71.74 percent win rate on that ticker specifically.
Why the trade count changes what the number means
A result from 46 trades and a result from 6 trades are not the same kind of claim, even when the percentages sit close together. CAT's 174.94 percent under 52-week low reversion came with a 100 percent win rate, every one of its 6 trades closed in the green, and 6 trades is exactly the kind of sample size where one or two large winners can produce a number a hundred more trades would pull back toward the middle. AMD's 71.74 percent win rate on 46 trades is a smaller, less dramatic figure, and it is a sturdier one. Tenachine's guide on 13F ranking sample sizes makes the same point about investor track records, a 100 percent win rate on 12 disclosed positions and a 121.9 percent weighted return on 10. The mechanism is identical here, just moved from an investor's disclosed trades to a single ticker inside a backtest.
Williams %R reversal's own aggregate win rate across all 214 completed tickers is 69.13 percent, so AMD's 71.74 percent is not a fluke sitting far outside what the strategy normally does. That is worth stating plainly because it is the opposite of NVDA's case: NVDA's one-trade result was the entire story on that ticker, while AMD's 46-trade result is close to what the strategy does everywhere, applied to a stock that happened to trend hard for a decade.
What the buy-and-hold gap still says
AMD itself returned 4423.97 percent buy and hold over the same window, so Williams %R reversal's 308.45 percent captured a small share of what simply holding the stock would have returned. That gap is real and worth naming, and it is not this guide's main point. Tenachine's guide on golden cross and NVDA already makes the fuller case for reading a strategy's return next to buy and hold before treating either number as a verdict on the stock. The point here is narrower: among the best single-ticker results this site has published, the size of the number and the number of trades behind it do not move together, and the second-best headline is the one actually built on a real sample.
None of this is a reason to trade Williams %R reversal on AMD specifically. A study run across 220 large caps and ten years describes what happened to that universe in that window, not a forecast for any one ticker going forward. The strategy's own aggregate numbers carry a warning of their own: an average maximum drawdown of negative 31.88 percent, the second deepest of any study this site has published, next to an average return of positive 27.21 percent. AMD's 46 trades are a real sample, and they are still one ticker out of the 214 that produced a usable run. Past performance does not predict future results, and a strategy that trended well with one stock for a decade is not guaranteed to do it again with the next one.