Many methods online promise to "beat the market". We test them over years of data, with the real costs paid by a Belgian investor, and we publish every result, including the bad ones.
ℹ️This page is for information only and is not investment advice. It shows simulated results (backtests) and live tracking. Past performance is no guarantee of future results.
Methods tested
8
families of popular strategies
Beat the ETF after costs
0
in a lasting way
Best: 💎 Fundamental
+0.2 pt/yr
vs the ETF, 2011-2026 (audited)
Cost of fees and taxes
−1.6 pt/yr
measured on 💎 during the audit
In short: in our tests, no method has durably beaten a simple S&P 500 ETF after costs. Some did well for a few years, then fell behind. Our 💎 Fundamental method, the best of our tests, breaks even over 15 years once all Belgian costs and taxes are counted: it is indistinguishable from an ETF. 🏆 Monthly ranking and ⚖️ 50/50 only pull ahead thanks to the single year 2026.
Our testing rules
📏 A buyable benchmark
A real S&P 500 UCITS ETF (iShares CSPX) or, failing that, SPY, dividends reinvested.
🧾 Real Belgian costs
Commission + stock exchange tax (TOB) + currency: about 0.43% on every purchase and every sale.
🕰️ No hindsight
Companies that later disappeared are included; accounts are used only after their publication date.
🔪 Several periods
A method must hold up over several periods, not thanks to a few good months.
Results table
Method tested
Period
Method
Benchmark
Worst drop
Verdict
💎 Fundamental profitable, growing companies
2011 → 2026
14.0%/yr*
13.8%/yr UCITS ETF
−28.4% ETF −22.7%
Indistinguishable from an ETF
🏆 Monthly ranking 10 strongest stocks
2011 → 2026 without 2026
15.3%/yr* 11.4%/yr*
13.9%/yr 13.7%/yr UCITS ETF
−39% ETF −22.7%
Does not beat it without 2026
⚖️ 50/50 half 🏆, half 💎
2011 → 2026 without 2026
14.8%/yr* 12.7%/yr*
13.9%/yr 13.7%/yr UCITS ETF
−24.9% ETF −22.7%
Does not beat it without 2026
Crash filter exit below the 10-month average
2007 → 2026
7.2%/yr
10.8%/yr
−28% SPY −55%
Does not beat it
Sector rotation top 3 sectors each month
2000 → 2026
7.0%/yr
8.4%/yr
−47%
Does not beat it
Technical rebounds 20-day average, supports
Backtest + live 2026
About 1 signal in 3 positive after 20 days
Does not beat it
Insider purchases "Form 4" filings
Backtest
Did worse than the S&P 500
Does not beat it
Pick 5 stocks and wait until they beat the index
1,000 draws
28% of portfolios stay stuck for 5 years
Does not beat it
* Audited results (💎, 🏆, ⚖️): prices including companies that later disappeared, purchase on the trading day after the decision, all costs (including rebalancing to equal weights) and Belgian tax on dividends (15% withheld in the US + 30% Belgian withholding tax). Before these corrections: 15.6%/yr for 💎. The other rows are compared with the US-listed SPY, dividends reinvested.
💎 Fundamental: 1 € invested in January 2011
💎 audited (Belgian costs and taxes)S&P 500 UCITS ETF (CSPX)💎 before audit
💎 audited ×7.75UCITS ETF ×7.58💎 before audit ×9.64
The gap between the gold and blue lines is what real costs and Belgian tax take away. Once they are counted, 💎 and the ETF end up almost in the same place.
Recent period: annual return
💎 Fundamental2017 → 2026, audited, vs UCITS ETF
14.1%
15.0%
🏆 Monthly ranking05/2018 → 03/2026, audited, vs UCITS ETF
13.3%
14.5%
⚖️ 50/5005/2018 → 03/2026, audited, vs UCITS ETF
13.2%
14.5%
Sector rotation2018 → 2026
10.7%
13.6%
Crash filter2018 → 2026
6.5%
13.6%
Method testedBenchmark, same period
Details of each test
💎 FundamentalIndistinguishable from an ETF
Rule: each month, keep the S&P 500 companies that are highly profitable (ROE > 20%, ROA > 8%), whose earnings per share grow by more than 25%, at a reasonable price (P/E > 15, PEG < 1.5). Official SEC filings.
Audited result, all Belgian costs and taxes included: 14.0%/yr over 2011 → 2026, vs 13.8% for a real S&P 500 UCITS ETF (+0.2 point a year, within the margin of chance).
2011 → 2016: 13.7%/yr vs 11.9% (+1.8). 2017 → 2026: 14.1%/yr vs 15.0% (−0.9).
The filter thresholds were chosen among 162 combinations, on 2011-2016. Its lead comes from that period: in the following years it did worse than the ETF.
8 years won out of 16. Recent years: 2024 at −7.5 points, 2025 at −14.8 points vs the ETF.
What the audit corrections cost: about 1.6 points a year, including 0.6 for Belgian tax on dividends.
Weighting differently (less in volatile stocks, cap per sector) did not help: equal weights remain the best version.
Weak points: worst drop −28.4% (vs −22.7% for the ETF), up to 41 months in a row behind the ETF, a cumulative lag that reached −19.7%. Rules frozen and live tracking started on 10/12/2026: to be judged after at least 12 months. Every transaction of the backtest is published in a downloadable log (5,438 rows) so anyone can redo the calculation.
🏆 Monthly rankingDoes not beat it without 2026
Rule: each month, the 10 S&P 500 stocks that rose the most over 12 months (excluding the last month). A stock is sold only when it leaves the top 10; stocks kept are never rebalanced.
Audited result (companies that disappeared included, costs of 0.43% per purchase and sale, Belgian tax on dividends): 15.3%/yr over 2011 → 2026 vs 13.9% for the S&P 500 UCITS ETF.
But it loses in every period: 2011 → 2018: 10.9% vs 12.5% (−1.6); 2018 → 2026: 13.3% vs 14.5% (−1.2); without 2026: 11.4% vs 13.7% (−2.2).
All of its lead comes from 2026 (+86% in 9 months), driven by a single theme: memory and storage (SanDisk, Micron, Western Digital, Seagate…).
The former Comparator figure (1,031 € for 100 € from 2011) was inflated by missing companies that disappeared: with them, 821 € at equal costs.
Weak points: worst drop −39% (vs −22.7% for the ETF), 177 months behind the ETF before 2026, one month at −24.5 points vs the ETF (August 2026), about 31% of the portfolio replaced each month. Live tracking since October 1, 2026, rules frozen.
⚖️ 50/50 (half 🏆, half 💎)Does not beat it without 2026
Rule: put half in 🏆 Monthly ranking and the other half in 💎 Fundamental, then let each half evolve.
Audited result: 14.8%/yr over 2011 → 2026 vs 13.9% for the ETF; without 2026: 12.7% vs 13.7% (−0.9); 2018 → 2026: 13.2% vs 14.5%.
Worst drop −24.9%, close to the ETF (−22.7%) and smaller than either method alone: the mix smooths the ride.
It beats the ETF in only 32% of 5-year investments.
Crash filter (10-month average)Does not beat it
Rule: a well-known rule, published in 2006. Hold the S&P 500 when its price is above its average of the last 10 months, otherwise move to cash.
2007 → 2026 (after its publication): 7.2%/yr vs 10.8%. 10,000 € grow to about 40,000 € instead of 76,000 €.
It did halve the worst drop (−28% vs −55%), mostly thanks to 2008.
78% of exits were false alarms: on average you buy back 7.4% higher than you sold.
Since 2010, simply holding 73% stocks and 27% cash did better, with a comparable worst drop.
Weak points: useless against fast crashes (2020, 2025), and each exit can trigger capital gains tax.
Sector rotationDoes not beat it
Rule: each month, invest in the 3 sectors (technology, energy, health care…) that rose the most over 12 months.
2000 → 2026: 7.0%/yr (monthly) and 8.2%/yr (quarterly) vs 8.4% for the S&P 500.
The small lead before costs comes from the 2000s (energy boom). Since 2010, the method does worse even with zero costs.
Holding all 11 sectors at equal weights did about as well as the S&P 500: picking sectors added nothing.
Weak points: about 22% of the portfolio replaced each month. With ETFs subject to the 1.32% TOB, the lag can reach almost 7 points a year.
Technical reboundsDoes not beat it
Rule: buy a trending stock when it pulls back to its 20-day average, a pivot point or a support.
Backtest on the S&P 500, companies that disappeared included: did worse than the index.
Live tracking (July → October 2026): about 1 signal in 3 positive after 20 days, about 40% at −10% or worse.
On average, the stock falls right after the signal (−3.5% after 3 days, −4.8% after 5 days).
Insider purchasesDoes not beat it
Rule: buy stocks whose executives have just bought shares of their own company (official "Form 4" filings).
Result: did worse than the S&P 500 in our backtest. Same verdict for the IPOs of start-ups from the Y Combinator incubator.
Pick a few stocks and waitDoes not beat it
Rule: pick 1 to 5 stocks and hold them "until they beat the index".
An S&P 500 stock beats the index 45% of the time over 1 year, but only 37% of the time over 5 years.
Out of 1,000 random 5-stock portfolios, 72% eventually pull ahead, but 28% stay stuck far behind for 5 years.
A rare few "super-winners" drive the market's performance: holding the whole market guarantees you own them.
What these figures tell us
Our results match public studies: over 15 years, about 9 professional funds out of 10 do worse than the S&P 500. Costs, false alarms and impatience often cost more than what a method brings.
We keep tracking 💎 Fundamental, 🏆 Monthly ranking and ⚖️ 50/50 live, every month, with frozen rules, to check whether their results hold over time. A monthly method should not be judged before at least 12 months of tracking.
For educational purposes only. This is neither investment advice nor a recommendation to buy or sell. Past performance, real or simulated, is no guarantee of future results. Simulated results (backtests) have limits: historical data, assumptions about costs, taxes and execution. Investing in the stock market involves a risk of capital loss.